IPCS INCHEM Home


    UNITED NATIONS ENVIRONMENT PROGRAMME
    INTERNATIONAL LABOUR ORGANISATION
    WORLD HEALTH ORGANIZATION


    INTERNATIONAL PROGRAMME ON CHEMICAL SAFETY



    ENVIRONMENTAL HEALTH CRITERIA 171





    DIESEL FUEL AND EXHAUST EMISSIONS









    This report contains the collective views of an international group of
    experts and does not necessarily represent the decisions or the stated
    policy of the United Nations Environment Programme, the International
    Labour Organisation, or the World Health Organization.


    Environmental Health Criteria  171

    DIESEL FUEL AND EXHAUST EMISSIONS


    First draft prepared by the staff members of the Fraunhofer Institute
    of Toxicology and Aerosol Research, Germany, under the coordination of
    Dr. G. Rosner


    Published under the joint sponsorship of the United Nations
    Environment Programme, the International Labour Organisation, and the
    World Health Organization, and produced within the framework if the
    Inter-Organization Programme for the Sound Management of Chemicals.


    World Health Organization
    Geneva, 1996

         The International Programme on Chemical Safety (IPCS) is a joint
    venture of the United Nations Environment Programme, the International
    Labour Organisation, and the World Health Organization. The main
    objective of the IPCS is to carry out and disseminate evaluations of
    the effects of chemicals on human health and the quality of the
    environment. Supporting activities include the development of
    epidemiological, experimental laboratory, and risk-assessment methods
    that could produce internationally comparable results, and the
    development of manpower in the field of toxicology. Other activities
    carried out by the IPCS include the development of know-how for coping
    with chemical accidents, coordination of laboratory testing and
    epidemiological studies, and promotion of research on the mechanisms
    of the biological action of chemicals.

    WHO Library Cataloguing in Publication Data

    Diesel Fuel and exhaust emission

    (Environmental health criteria ; 171)

    1.Air pollutants, Enviromental  2.Air pollution  3.Fueloils
    I.International Programme on Chemical Safety  II.Series

    ISBN 92 4 157171 3                 (NLM Classification: WA 754)
    ISSN 0250-863X

         The World Health Organization welcomes requests for permission to
    reproduce or translate its publications, in part or in full.
    Applications and enquiries should be addressed to the Office of
    Publications, World Health Organization, Geneva, Switzerland, which
    will be glad to provide the latest information on any changes made to
    the text, plans for new editions, and reprints and translations
    already available.

    (c) World Health Organization 1996

         Publications of the World Health Organization enjoy copyright
    protection in accordance with the provisions of Protocol 2 of the
    Universal Copyright Convention. All rights reserved. The designations
    employed and the presentation of the material in this publication do
    not imply the expression of any opinion whatsoever on the part of the
    Secretariat of the World Health Organization concerning the legal
    status of any country, territory, city or area or of its authorities,
    or concerning the delimitation of its frontiers or boundaries. The
    mention of specific companies or of certain manufacturers' products
    does not imply that they are endorsed or recommended by the World
    Health Organization in preference to others of a similar nature that
    are not mentioned. Errors and omissions excepted, the names of
    proprietary products are distinguished by initial capital letters.

    CONTENTS

    NOTE TO READERS OF THE CRITERIA MONOGRAPHS

    PREAMBLE

    ENVIRONMENTAL HEALTH CRITERIA FOR DIESEL FUEL AND EXHAUST EMISSIONS

    WHO DRAFTING GROUP ON ENVIRONMENTAL HEALTH CRITERIA FOR DIESEL FUEL
    AND EXHAUST EMISSIONS

    WHO TASK GROUP ON ENVIRONMENTAL HEALTH CRITERIA FOR DIESEL FUEL AND
    EXHAUST EMISSIONS

    PART A: DIESEL FUEL

         A1.   SUMMARY
               A1.1   Identity, physical and chemical properties, and
                      analytical methods
               A1.2   Sources of human and environmental exposure
               A1.3   Environmental transport, distribution, and
                      transformation
               A1.4   Environmental levels and human exposure
               A1.5   Effects on laboratory mammals and in vitro test
                      systems
               A1.6   Effects on humans
               A1.7   Effects on other organisms in the laboratory and the
                      field
               A1.8   Evaluation of human health risks
               A1.9   Evaluation of effects on the environment
         A2.   IDENTITY, PHYSICAL AND CHEMICAL PROPERTIES, AND ANALYTICAL
               METHODS
               A2.1   Identity
                      A2.1.1   Fuel components
                               A2.1.1.1   Alkanes
                               A2.1.1.2   Alkenes
                               A2.1.1.3   Aromatic compounds
                               A2.1.1.4   Sulfur
                      A2.1.2   Fuel additives
                               A2.1.2.1   Cetane number improvers
                               A2.1.2.2   Smoke suppressors
                               A2.1.2.3   Flow improvers
                               A2.1.2.4   Cloud-point depressors
                               A2.1.2.5   Wax anti-settling additives
                               A2.1.2.6   Other additives

                      A2.1.3   Quality aspects of diesel fuels
                               A2.1.3.1   Ignition performance and cetane
                                          number
                               A2.1.3.2   Density
                               A2.1.3.3   Sulfur content
                               A2.1.3.4   Viscosity
                               A2.1.3.5   Cold-flow properties
               A2.2   Physical and chemical properties
               A2.3   Analytical methods
               A2.4   Conversion factors
         A3.   SOURCES OF HUMAN AND ENVIRONMENTAL EXPOSURE
               A3.1   Natural occurrence
               A3.2   Anthropogenic sources
                      A3.2.1   Production and use
                               A3.2.1.1   Production process
                               A3.2.1.2   Use
                               A3.2.1.3   Production and consumption
                                          levels
                      A3.2.2   Emissions during production and use
                               A3.2.2.1   Air
                               A3.2.2.2   Water
                               A3.2.2.3   Soil
                      A3.2.3   Accidental releases to the environment
         A4.   ENVIRONMENTAL TRANSPORT, DISTRIBUTION, AND TRANSFORMATION
               A4.1   Transport and distribution between media
                      A4.1.1   Evaporation from and dissolution in the
                               aqueous phase
                      A4.1.2   Transport in and adsorption onto soil and
                               sediment
                               A4.1.2.1   Soil
                               A4.1.2.2   Sediment
               A4.2   Transformation and removal
                      A4.2.1   Photooxidation
                      A4.2.2   Biodegradation
                               A4.2.2.1   Microbial degradation
                               A4.2.2.2   Phytoplankton and marine algae
                               A4.2.2.3   Invertebrates and vertebrates
                      A4.2.3   Bioaccumulation
                      A4.2.4   Tainting
                      A4.2.5   Entry into the food chain
               A4.3   Ultimate fate after use
                      A4.3.1   Use in motor vehicles
                      A4.3.2   Spills
                      A4.3.3   Disposal
         A5.   ENVIRONMENTAL LEVELS AND HUMAN EXPOSURE
               A5.1   Environmental levels
               A5.2   Exposure of the general population
               A5.3   Occupational exposure during manufacture,
                      formulation, or use
         A6.   KINETICS AND METABOLISM IN LABORATORY ANIMALS AND HUMANS

         A7.   EFFECTS ON LABORATORY MAMMALS AND IN VITRO TEST SYSTEMS
               A7.1   Single exposure
               A7.2   Short-term exposure
                      A7.2.1   Subacute exposure
                               A7.2.1.1   Dermal exposure
                               A7.2.1.2   Inhalation
                      A7.2.2   Subchronic exposure
                               A7.2.2.1   Dermal exposure
                               A7.2.2.2   Inhalation
               A7.3   Long-term exposure
                      A7.3.1   Dermal exposure
                      A7.3.2   Inhalation
               A7.4   Dermal and ocular irritation; dermal sensitization
                      A7.4.1   Dermal irritation
                      A7.4.2   Ocular irritation
                      A7.4.3   Sensitization
               A7.5   Reproductive toxicity, embryotoxicity, and
                      teratogenicity
               A7.6   Mutagenicity and related end-points
                      A7.6.1   In vitro
                      A7.6.2   In vivo
               A7.7   Carcinogenicity
                      A7.7.1   Dermal exposure
                      A7.7.2   Inhalation
         A8.   EFFECTS ON HUMANS
               A8.1   Exposure of the general population
               A8.2   Occupational exposure
         A9.   EFFECTS ON OTHER ORGANISMS IN THE LABORATORY AND THE FIELD
               A9.1   Laboratory experiments
                      A9.1.1   Microorganisms
                               A9.1.1.1   Water
                               A9.1.1.2   Soil
                      A9.1.2   Aquatic organisms
                               A9.1.2.1   Plants (phytoplankton)
                               A9.1.2.2   Invertebrates
                      A9.1.3   Terrestrial organisms
                               A9.1.3.1   Plants
                               A9.1.3.2   Invertebrates
                               A9.1.3.3   Vertebrates
               A9.2   Field observations
                      A9.2.1   Microorganisms.
                               A9.2.1.1   Water
                               A9.2.1.2   Soil
                      A9.2.2   Aquatic organism
                      A9.2.3   Terrestrial organisms
                               A9.2.3.1   Plants
                               A9.2.3.2   Invertebrates
                               A9.2.3.3   Vertebrates

         A10.  EVALUATION OF HUMAN HEALTH RISKS AND EFFECTS ON THE
               ENVIRONMENT
               A10.1  Evaluation of human health risks
                      A10.1.1  Exposure of the general population
                      A10.1.2  Occupational exposure
                      A10.1.3  Non-neoplastic effects
                      A10.1.4  Neoplastic effects
               A10.2  Evaluation of effects on the environment
         A11.  RECOMMENDATIONS
               A11.1  Recommendations for the protection of human health
               A11.2  Recommendation for the protection of the environment
               A11.3  Recommendations for further research
         A12.  PREVIOUS EVALUATIONS BY INTERNATIONAL BODIES

    PART B: DIESEL EXHAUST EMISSIONS
         B1.   SUMMARY
               B1.1   Identity, physical and chemical properties, and
                      analytical methods
               B1.2   Sources of human and environmental exposure
               B1.3   Environmental transport, distribution, and
                      transformation
               B1.4   Environmental levels and human exposure
               B1.5   Kinetics and metabolism in laboratory animals and
                      humans
                      B1.5.1   Deposition
                      B1.5.2   Retention and clearance of particles
                      B1.5.3   Retention and clearance of polycyclic
                               aromatic hydrocarbons adsorbed onto diesel
                               soot
                      B1.5.4   Metabolism
               B1.6   Effects on laboratory mammals and in vitro test
                      systems
               B1.7   Effects on humans
               B1.8   Effects on other organisms in the laboratory and the
                      field
               B1.9   Evaluation of human health risks
                      B1.9.1   Non-neoplastic effects
                      B1.9.2   Neoplastic effects
               B1.10  Evaluation of effects on the environment
         B2.   IDENTITY AND ANALYTICAL METHODS
               B2.1   Identity
                      B2.1.1   Chemical composition of diesel exhaust
                               gases
                      B2.1.2   Type and composition of emitted particulate
                               matter
               B2.2   Analytical methods
                      B2.2.1   Sampling
                               B2.2.1.1   Sampling from undiluted exhaust
                                          gas (raw gas sampling)

                               B2.2.1.2   Sampling from diluted exhaust
                                          (dilution tube sampling)
                      B2.2.2   Extraction from particles
                      B2.2.3   Clean-up and fractionation
                      B2.2.4   Chemical analysis
               B2.3   Conversion factors
         B3.   SOURCES OF HUMAN AND ENVIRONMENTAL EXPOSURE
               B3.1   Anthropogenic sources
                      B3.1.1   Diesel exhaust emissions
                               B3.1.1.1   Emission of chemical
                                          constituents with the gaseous
                                          portion of diesel exhaust
                               B3.1.1.2   Emission of particulate matter
                                          and adsorbed components in
                                          diesel exhaust gases
                      B3.1.2   Parameters that influence diesel exhaust
                               emissions
                               B3.1.2.1   Engine conditions
                               B3.1.2.2   Fuel specification
                               B3.1.2.3   Malfunction
                      B3.1.3   Total emissions by diesel engines
                      B3.1.4   Control of emissions
                               B3.1.4.1   Particle traps
                               B3.1.4.2   Catalytic converters
               B3.2   Regulatory approaches
         B4.   ENVIRONMENTAL TRANSPORT, DISTRIBUTION, AND TRANSFORMATION
               B4.1   Transport and distribution between media
               B4.2   Transformation and removal
         B5.   ENVIRONMENTAL LEVELS AND HUMAN EXPOSURE
               B5.1   Exposure of the general population
               B5.2   Occupational exposure
                      B5.2.1   Truck drivers and mechanics
                      B5.2.2   Bus garage and other bus workers
                      B5.2.3   Fork-lift truck operators
                      B5.2.4   Railroad workers
                      B5.2.5   Mine workers
                      B5.2.6   Fire fighters
               B5.3   Biomonitoring
                      B5.3.1   Urinary mutagenicity
                      B5.3.2   Other analyses
         B6.   KINETICS AND METABOLISM IN LABORATORY ANIMALS AND HUMANS
               B6.1   Deposition
               B6.2   Retention and clearance of particles
               B6.3   Retention and clearance of polycyclic aromatic
                      hydrocarbons adsorbed onto diesel soot
               B6.4   Metabolism
         B7.   EFFECTS ON LABORATORY MAMMALS AND IN VITRO TEST SYSTEMS
               B7.1   Single exposure
               B7.2   Short-term exposure

               B7.3   Long-term exposure and studies of carcinogenicity
                      B7.3.1   Non-neoplastic effects
                      B7.3.2   Carcinogenicity
                               B7.3.2.1   Inhalation
                               B7.3.2.2   Other routes of exposure
               B7.4   Dermal and ocular irritation; dermal sensitization
               B7.5   Reproductive toxicity, embryotoxicity, and
                      teratogenicity
                      B7.5.1   Reproductive toxicity
                      B7.5.2   Embryotoxicity
                      B7.5.3   Teratogenicity
               B7.6   Mutagenicity and related end-points
                      B7.6.1   In vitro
                      B7.6.2   In vivo
                      B7.6.3   DNA adduct formation
               B7.7   Special studies
                      B7.7.1   Immunotoxicity
                      B7.7.2   Behavioural effects
               B7.8   Factors that modify toxicity; toxicity of
                      metabolites
               B7.9   Mechanisms of toxicity; mode of action
                      B7.9.1   Carcinogenic effects
                               B7.9.1.1   DNA-reactive mechanisms
                               B7.9.1.2   Cytotoxicity with regenerative
                                          cell proliferation
                               B7.9.1.3   Effects of particles
                               B7.9.1.4   Effects of polycyclic aromatic
                                          hydrocarbons
                      B7.9.2   Noncarcinogenic effects
         B8.   EFFECTS ON HUMANS
               B8.1   General population
                      B8.1.1   Acute exposure: olfactory, nasal, and
                               ocular irritation
                      B8.1.2   Air pollution
               B8.2   Occupational exposure
                      B8.2.1   Effects on the respiratory system
                               B8.2.1.1   Symptoms
                               B8.2.1.2   Acute changes in pulmonary
                                          function
                               B8.2.1.3   Pulmonary effects
                      B8.2.2   Epidemiological studies (noncarcinogenic
                               effects)
                               B8.2.2.1   Effects on the respiratory
                                          system
                               B8.2.2.2   Effects on the circulatory
                                          system
                      B8.2.3   Epidemiological studies (carcinogenic
                               effects)
                               B8.2.3.1   Lung cancer
                               B8.2.3.2   Urinary bladder cancer

         B9.   EFFECTS ON OTHER ORGANISMS IN THE LABORATORY AND THE FIELD
         B10.  EVALUATION OF HUMAN HEALTH RISKS AND EFFECTS ON THE
               ENVIRONMENT
               B10.1  Exposure of the general population
               B10.2  Occupational exposure
               B10.3  Non-neoplastic effects
                      B10.3.1  Hazard identification
                               B10.3.1.1  Humans
                               B10.3.1.1  Experimental animals
                      B10.3.2  Dose-response assessment
                               B10.3.2.1  Epidemiological studies
                               B10.3.2.2  Studies in experimental animals
                      B10.3.3  Exposure assessment
                      B10.3.4  Risk characterization
                               B10.3.4.1  Humans
                               B10.3.4.2  Experimental animals
               B10.4  Neoplastic effects
                      B10.4.1  Hazard identification
                               B10.4.1.1  Lung cancer: occupational
                                          exposure
                               B10.4.1.2  Urinary bladder cancer:
                                          occupational exposure
                      B10.4.2  Dose-response assessment
                               B10.4.2.1  Lung cancer
                               B10.4.2.2  Urinary bladder cancer
                      B10.4.3  Exposure assessment
                      B10.4.4  Risk characterization
                               B10.4.4.1  Human lung cancer
                               B10.4.4.2  Human urinary bladder cancer
                               B10.4.4.3  Risk characterization based on
                                          studies in experimental animals
                      Appendix B10.1  Construction of a biologically
                                      based (alternative) model
                      Appendix B10.2  E-M algorithm
                      Appendix B10.3  A tumour growth model
         B11.  RECOMMENDATIONS
               B11.1  Recommendations for the protection of human health
               B11.2  Recommendation for the protection of the environment
               B11.3  Recommendations for further research
         B12.  PREVIOUS EVALUATIONS BY INTERNATIONAL BODIES

    REFERENCES

    RESUMÉ

    RESUMEN
    

    NOTE TO READERS OF THE CRITERIA MONOGRAPHS

         Every effort has been made to present information in the Criteria
    monographs as accurately as possible without unduly delaying their
    publication. In the interest of all users of the Environmental Health
    Criteria monographs, readers are requested to communicate any errors
    that may have occurred to the Director of the International Programme
    on Chemical Safety, World Health Organization, Geneva, Switzerland, in
    order that they may be included in corrigenda.

                                 * * *

         A detailed data profile and a legal file can be obtained from the
    International Register of Potentially Toxic Chemicals, Case postale
    356, 1219 Châtelaine, Geneva, Switzerland (Telephone No. 979 9111).

                                 * * *


         This publication was made possible by grant number 5 U01
    ES02617-15 from the National Institute of Environmental Health
    Sciences, National Institutes of Health, USA, and by financial support
    from the European Commission.

    Environmental Health Criteria

    PREAMBLE

    Objectives

         The WHO Environmental Health Criteria Programme was initiated in
    1973, with the following objectives:

    (i)    to assess information on the relationship between exposure to
           environmental pollutants and human health and to provide
           guidelines for setting exposure limits;

    (ii)   to identify new or potential pollutants;

    (iii)  to identify gaps in knowledge concerning the health effects of
           pollutants;

    (iv)   to promote the harmonization of toxicological and
           epidemiological methods in order to have internationally
           comparable results.

         The first Environmental Health Criteria (EHC) monograph, on
    mercury, was published in 1976; numerous assessments of chemicals and
    of physical effects have since been produced. Many EHC monographs have
    been devoted to toxicological methods, e.g. for genetic, neurotoxic,
    teratogenic, and nephrotoxic effects. Other publications have been
    concerned with e.g. epidemiological guidelines, evaluation of
    short-term tests for carcinogens, biomarkers, and effects on the
    elderly.

         Since the time of its inauguration, the EHC Programme has widened
    its scope, and the importance of environmental effects has been
    increasingly emphasized in the total evaluation of chemicals, in
    addition to their health effects.

         The original impetus for the Programme came from resolutions of
    the World Health Assembly and the recommendations of the 1972 United
    Nations Conference on the Human Environment. Subsequently, the work
    became an integral part of the International Programme on Chemical
    Safety (IPCS), a cooperative programme of UNEP, ILO, and WHO. In this
    manner, with the strong support of the new partners, the importance of
    occupational health and environmental effects was fully recognized.
    The EHC monographs have become widely established, used, and
    recognized throughout the world.

         The recommendations of the 1992 United Nations Conference on
    Environment and Development and the subsequent establishment of the
    Intergovernmental Forum on Chemical Safety, with priorities for action
    in the six programme areas of Chapter 19, Agenda 21, lend further
    weight to the need for EHC assessments of the risks of chemicals.

    Scope

         The Criteria monographs are intended to provide critical reviews
    of the effect on human health and the environment of chemicals,
    combinations of chemicals, and physical and biological agents. They
    include reviews of studies that are of direct relevance for the
    evaluation and do not describe every study that has been carried out.
    Data obtained worldwide are used, and results are quoted from original
    studies, not from abstracts or reviews. Both published and unpublished
    reports are considered, and the authors are responsible for assessing
    all of the articles cited; however, preference is always given to
    published data, and unpublished data are used only when relevant
    published data are absent or when the unpublished data are pivotal to
    the risk assessment. A detailed policy statement is available that
    describes the procedures used for citing unpublished proprietary data,
    so that this information can be used in the evaluation without
    compromising its confidential nature (WHO, 1990).

         In the evaluation of human health risks, sound data on humans,
    whenever available, are preferred to data on experimental animals.
    Studies of animals and in vitro systems provide support and are used
    mainly to supply evidence missing from human studies. It is mandatory
    that research on human subjects be conducted in full accord with
    ethical principles, including the provisions of the Helsinki
    Declaration.

         The EHC monographs are intended to assist national and
    international authorities in making risk assessments and subsequent
    risk management decisions. They represent a thorough evaluation of
    risks and are not in any sense recommendations for regulation or
    setting standards. The latter are the exclusive purview of national
    and regional governments.

    Content

         The layout of EHC monographs for chemicals is outlined below.

    *    Summary: a review of the salient facts and the risk evaluation of
         the chemical
    *    Identity: physical and chemical properties, analytical methods
    *    Sources of exposure
    *    Environmental transport, distribution, and transformation
    *    Environmental levels and human exposure
    *    Kinetics and metabolism in laboratory animals and humans
    *    Effects on laboratory mammals and in-vitro test systems
    *    Effects on humans
    *    Effects on other organisms in the laboratory and the field
    *    Evaluation of human health risks and effects on the environment
    *    Conclusions and recommendations for protection of human health
         and the environment

    *    Further research
    *    Previous evaluations by international bodies, e.g. the
         International Agency for Research on Cancer, the Joint FAO/WHO
         Expert Committee on Food Additives, and the Joint FAO/WHO Meeting
         on Pesticide Residues

    Selection of chemicals

         Since the inception of the EHC Programme, the IPCS has organized
    meetings of scientists to establish lists of chemicals that are of
    priority for subsequent evaluation. Such meetings have been held in
    Ispra, Italy (1980); Oxford, United Kingdom (1984); Berlin, Germany
    (1987); and North Carolina, United States of America (1995). The
    selection of chemicals is based on the following criteria: the
    existence of scientific evidence that the substance presents a hazard
    to human health and/or the environment; the existence of evidence that
    the possible use, persistence, accumulation, or degradation of the
    substance involves significant human or environmental exposure; the
    existence of evidence that the populations at risk (both human and
    other species) and the risks for the environment are of a significant
    size and nature; there is international concern, i.e. the substance is
    of major interest to several countries; adequate data are available on
    the hazards.

         If it is proposed to write an EHC monograph on a chemical that is
    not on the list of priorities, the IPCS Secretariat first consults
    with the cooperating organizations and the participating institutions.

    Procedures

         The order of procedures that result in the publication of an EHC
    monograph is shown in the following flow chart. A designated staff
    member of IPCS, responsible for the scientific quality of the
    document, serves as Responsible Officer (RO). The IPCS Editor is
    responsible for the layout and language. The first draft, prepared by
    consultants or, more usually, staff at an IPCS participating
    institution is based initially on data provided from the International
    Register of Potentially Toxic Chemicals and reference data bases such
    as Medline and Toxline.

         The draft document, when received by the RO, may require an
    initial review by a small panel of experts to determine its scientific
    quality and objectivity. Once the RO finds the first draft acceptable,
    it is distributed in its unedited form to over 150 EHC contact points
    throughout the world for comment on its completeness and accuracy and,
    where necessary, to provide additional material. The contact points,

    usually designated by governments, may be participating institutions,
    IPCS focal points, or individual scientists known for their particular
    expertise. Generally, about four months are allowed before the
    comments are considered by the RO and author(s). A second draft
    incorporating the comments received and approved by the Director,
    IPCS, is then distributed to Task Group members, who carry out a peer
    review at least six weeks before their meeting.

         The Task Group members serve as individual scientists, not as
    representatives of any organization, government, or industry. Their
    function is to evaluate the accuracy, significance, and relevance of
    the information in the document and to assess the risks to health and
    the environment from exposure to the chemical. A summary and
    recommendations for further research and improved safety are also
    drawn up. The composition of the Task Group is dictated by the range
    of expertise required for the subject of the meeting and by the need
    for a balanced geographical distribution.

         The three cooperating organizations of the IPCS recognize the
    important role played by nongovernmental organizations, so that
    representatives from relevant national and international associations
    may be invited to join the Task Group as observers. While observers
    may provide valuable contributions to the process, they can speak only
    at the invitation of the Chairperson. Observers do not participate in
    the final evaluation of the chemical, which is the sole responsibility
    of the Task Group members. The Task Group may meet  in camera when it
    considers that to be appropriate.

         All individuals who participate in the preparation of an EHC
    monograph as authors, consultants, or advisers must, in addition to
    serving in their personal capacity as scientists, inform the RO if at
    any time a conflict of interest, whether actual or potential, could be
    perceived in their work. They are required to sign a statement to that
    effect. This procedure ensures the transparency and probity of the
    process.

         When the Task Group has completed its review and the RO is
    satisfied as to the scientific correctness and completeness of the
    document, it is edited for language, the references are checked, and
    camera-ready copy is prepared. After approval by the Director, IPCS,
    the monograph is submitted to the WHO Office of Publications for
    printing. At this time, a copy of the final draft is also sent to the
    Chairperson and Rapporteur of the Task Group to check for any errors.

         It is accepted that the following criteria should initiate the
    updating of an EHC monograph: new data are available that would
    substantially change the evaluation; there is public concern about
    health or environmental effects of the agent because of greater
    exposure; an appreciable time has elapsed since the last evaluation.

    FIGURE 1

         All participating institutions are informed, through the EHC
    progress report, of the authors and institutions proposed for the
    drafting exhaust emissions  of the documents. A comprehensive file of
    all comments received on drafts of each EHC monograph is maintained
    and is available on request. The chairpersons of task groups are
    briefed before each meeting on their role and responsibility in
    ensuring that these rules are followed.

    WHO DRAFTING GROUP ON ENVIRONMENTAL HEALTH CRITERIA FOR DIESEL FUEL
    AND EXHAUST EMISSIONS

    WHO, Geneva, 6-9 December 1993

     Members

    Dr J.A. Bond, Chemical Industry Institute of Toxicology, Research
    Triangle Park, NC, United State

    Dr R.P. Bos, University of Nijmegen, Nijmegen, Netherlands

    Dr R. Brown, Medical Research Council Toxicology Unit, University of
    Leicester, Leicester, United Kingdom  (Joint Rapporteur)

    Dr Chao Chen, Human Health Assessment Group, United States
    Environmental Protection Agency, Washington DC, United States

    Dr I. Farkas, National Institute of Hygiene, Budapest, Hungary

    Dr E. Garshick, Pulmonary Section, Brockton/West Roxbury VA Medical
    Center, West Roxbury, MA, United States

    Dr P. Gustavsson, North Western Health Board, Stockholm, Sweden

    Dr D. Guth, United States Environmental Protection Agency, Research
    Triangle Park, NC, United States

    Dr U. Heinrich, Department of Experimental Hygiene, Fraunhofer
    Institute of Toxicology and Aerosol Research, Hanover, Germany

    Dr R.F. Hertel, Federal Health Office, Bundesgesundheitsamt, Berlin,
    Germany

    Professor G. Oberdörster, Department of Environmental Medicine,
    University of Rochester Medical Center, Rochester, NY, United States

    Dr W. Pepelko, United States Environmental Protection Agency,
    Washington DC, United States

    Dr P.J.A. Rombout, Laboratory of Toxicology, National Institute of
    Public Health and Environmental Protection, Bilthoven, Netherlands
     (Vice-Chairman)

    Dr G. Rosner, Hazardous Substances Documentation Group, Fraunhofer
    Institute of Toxicology and Aerosol Research, Hanover, Germany

    Dr J. Roycroft, National Institute of Environmental Health Sciences,
    Research Triangle Park, NC, United States

    Dr A. Sivak, Environmental Health Sciences, Saint Augustine, FL,
    United States  (Chairman)

    Dr B.H. Thomas, Environmental Health Directorate, Ottawa, Canada

    Dr L. Turrio, Istituto Superiore di Sanita, Laboratorio Tossicologia
    Comparata e Ecotossicologia, Rome, Italy

    Mr. R. Waller, Department of Health, London, United Kingdom

     Secretariat

    Dr P. Boffetta, International Agency for Research on Cancer, Lyon,
    France (6-7 December 1993)

    Dr E. Smith, International Programme on Chemical Safety, World Health
    Organization, Geneva, Switzerland

     Representatives/Observers

    CONCAWE
         Dr R.H. McKee, Exxon Biomedical Sciences, East Millstone, NJ,
         United States

    UNITED KINGDOM DEPARTMENT OF THE ENVIRONMENT
         Dr P.T.C. Harrison, MRC Institute of Environment & Health,
         Leicester, United Kingdom

    WHO TASK GROUP ON ENVIRONMENTAL HEALTH CRITERIA FOR DIESEL FUEL AND
    EXHAUST EMISSIONS

    Fraunhofer Institute of Toxicology & Aerosol Research, Hanover
    27 June-1 July 1994

     Members

    Dr J.A. Bond, Chemical Industry Institute of Toxicology, Research
    Triangle Park, NC, United States

    Dr R. Brown, Medical Research Council Toxicology Unit, University of
    Leicester, Leicester, United Kingdom

    Dr Chao Chen, Human Health Assessment Group, United States
    Environmental Protection Agency, Washington DC, United States

    Dr E. Garshick, Pulmonary Section, Brockton/West Roxbury VA Medical
    Center, West Roxbury, MA, United States

    Dr U. Heinrich, Department of Experimental Hygiene, Fraunhofer
    Institute of Toxicology and Aerosol Research, Hanover, Germany

    Dr R.F. Hertel, Federal Health Office, Bundesgesundheitsamt, Berlin,
    Germany

    Dr Jun Kagawa, Tokyo Women's Medical College, Tokyo, Japan

    Professor G. Oberdörster, Department of Environmental Medicine,
    University of Rochester Medical Center, Rochester, NY, United States

    Dr P.J.A. Rombout, Laboratory of Toxicology, National Institute of
    Public Health and Environmental Protection, Bilthoven, Netherlands

    Dr M. Roller, Medizinishes Institut für Umwelthyglene an der
    Universität Düsseldorf, Düsseldorf, Germany  Dr G. Rosner, Hazardous
    Substances Documentation Group, Fraunhofer Institute of Toxicology and
    Aerosol Research, Hanover, Germany

    Dr A. Sivak, Environmental Health Sciences, Saint Augustine, FL,
    United States

    Dr L. Turrio, Istituto Superiore di Sanita, Laboratorio Tossicologia
    Comparata e Ecotossicologia, Rome, Italy

    Mr R. Waller, Department of Health, London, United Kingdom

     Secretariat

    Dr H. Moller, International Agency for Research on Cancer, Lyon,
    France

    Dr E. Smith, International Programme on Chemical Safety, World Health
    Organization, Geneva, Switzerland

    Dr M. Younes, European Centre for Environment and Health, Bilthoven,
    Netherlands

     Representatives/Observers

    Dr Lutz Von Meyerinck, BP Oil Europe, Brussels, Belgium

    GERMAN AUTOMOBILE ASSOCIATION
         Dr N. Pelz, Mercedes-Benz AG, Stuttgart, Germany

    ILSI
         Dr I.T. Salmeen, Chemistry Department, Ford Motor Company,
         Dearborn, MI, United States

    IUTOX
         Dr P. Montuschi, Department of Pharmacology, School of Medicine,
         Catholic University of the Sacred Heart, Rome, Italy

    ENVIRONMENTAL HEALTH CRIETERIA FOR DIESEL FUEL AND EXHAUST EMISSIONS

         A WHO Task Group on Environmental Health Criteria for Diesel Fuel
    and Exhaust Emissions met at the Fraunhofer Institute of Toxicology
    and Aerosol Research, Hanover, Germany from 27 June to 1 July 1994.
    Dr G. Rosner, Fraunhofer Institute, welcomed the participants on
    behalf of the Institute and its Director, Professor U. Mohr, and
    Dr E.M. Smith, IPCS, welcomed the participants on behalf of
    Dr M. Mercier, Director of the IPCS, and on behalf of the heads of the
    three IPCS cooperating organizations (UNEP, ILO, and WHO). The Task
    Group reviewed and revised the draft and evaluated the risks for human
    health and the environment from exposure to diesel fuel and exhaust
    emissions.

         The first draft of the monograph was prepared at the Fraunhofer
    Institute. After international circulation for comment, this draft was
    extensively revised by a Working/Drafting Group, convened at WHO,
    Geneva, from 6 to 9 December 1993, and a second draft was prepared for
    further international circulation for comment. The membership of the
    drafting group is shown previously. A final draft, incorporating
    comments received from the IPCS contact points for Environmental
    Health Criteria monographs and new material, was completed at the
    Fraunhofer Institute under the coordination of Dr G. Rosner, with
    important contributions to the text from the following Institute staff
    members:

    Dr B. Bellman
    Dr A. Boehnke
    Dr O. Creutzenberg
    Dr J. Kielhorn

         Dr E.M. Smith of the IPCS Central Unit was responsible for the
    scientific content of the monograph and Mrs E. Heseltine, Lajarthe,
    France, for the editing.

         The efforts of all who helped in the preparation and finalization
    of the monograph are gratefully acknowledged.

    PART A  DIESEL FUEL

    A1.  SUMMARY

    A1.1  Identity, physical and chemical properties, and analytical
          methods

         Diesel fuel is a complex mixture of normal, branched, and cyclic
    alkanes (60 to > 90% by volume; hydrocarbon chain length, usually
    between C9 and C30); aromatic compounds, especially alkylbenzenes
    (5-40% by volume); and small amounts of alkenes (0-10% by volume)
    obtained from the middle-distillate, gas-oil fraction during petroleum
    separation. Benzene, toluene, ethylbenzene, and xylenes and polycyclic
    aromatic hydrocarbons (PAHs), especially naphthalene and its
    methyl-substituted derivatives, may be present at levels of parts per
    million in diesel fuel. The sulfur content of diesel fuels depends on
    the source of crude oil and the refinery process. It is regulated by
    law in a number of countries and is usually between 0.05 and 0.5
    weight percent. Additives are used to influence the flow, storage, and
    combustion of diesel fuel, to differentiate products, and to meet
    trademark specifications. At room temperature, diesel fuels are
    generally moderately volatile, slightly viscous, flammable, brown
    liquids with a kerosene-like odour. The boiling ranges are usually
    between 140 and 385°C (> 588°C for marine diesel fuel); at 20°C, the
    density is 0.87-1.0 g/cm3 and the water solubility is
    0.2-5 mg/litre. The quality and composition of diesel fuel influence
    the emissions of pollutants from diesel engines considerably.
    Important variables are ignition behaviour (expressed in terms of
    cetane number), density, viscosity, and sulfur content. The
    specifications of commercial diesel fuel differ considerably in
    different countries.

         Heating fuels and some kerosene jet fuels produced during the
    refining process may have a composition similar to that of diesel
    fuel, although with different additives. Biological data on these
    mixtures have therefore also been taken into account in the
    assessments of toxicity and ecotoxicity.

         Owing to the complexity of the diesel fuel mixture, there is no
    specific analytical method, and the analytical techniques used in most
    environmental assessments are suitable only for measuring the total
    petroleum hydrocarbon mixture. The methods consist of preliminary
    solvent extraction, a clean-up procedure to remove naturally occurring
    hydrocarbons, and subsequent detection by gravimetry, infrared
    spectroscopy or gas chromatography. Neither the gravimetric nor the
    infrared technique provides useful qualitative or quantitative
    information on contaminants and can thus be used only for screening.
    Gas chromatography combined with detection techniques such as flame
    ionization and mass spectrometry is the standard procedure for
    analysing environmental samples. Many other methods are available for
    the analysis of individual hydrocarbons in diesel fuels.

    A1.2  Sources of human and environmental exposure

         Diesel fuels are produced by refining crude oils. In order to
    meet technical specifications for performance, diesel fuels are
    generally blended; further formulation with additives improves their
    properties for specific uses. Diesel fuels are widely used as
    transportation fuels. The more volatile fuels, with low viscosity, are
    required for high-speed engines and the heavier grades for railroad
    and ship diesel engines. Much heavy-duty road transport is powered by
    diesel engines. Passenger cars powered by diesel engines are becoming
    increasingly more common in Europe and Japan (10-25%), whereas in
    North America the percentage of diesel-fuelled passenger cars is about
    1-2%, with a slightly decreasing tendency. Diesel fuel is used in
    stationary engines and in boilers, e.g. reciprocating engines, gas
    turbines, pipeline pumps, gas compressors, steam processing units in
    electric power plants, burner installations, and industrial space and
    water heating facilities.

         Over the last five years, the worldwide demand for diesel fuels
    has increased steadily. In 1985, the following amounts of diesel fuel
    were consumed: about 170 000 kt per year in North America; about
    160 000 kt per year, including gas oils, in the European Union; and
    about 46 000 kt per year in Australia, Japan, and New Zealand,
    equivalent to a total of 1062 kt per day. In 1990, world demand was
    reported to be about 1110 kt per day.

         No information is available on emissions during the production of
    diesel fuels; however, this source would seem to be of minor
    importance, because the refining process is carried out in closed
    systems. Emissions may occur principally during storage and
    transportation. Diesel fuels are released as a result of spills and at
    filling stations during the refuelling of vehicles. The atmosphere and
    the hydrosphere are the most heavily affected environmental
    compartments. Soil contamination with diesel fuels may occur during
    accidents and is also a problem in railroad yards. The numerous
    techniques for cleaning soils contaminated with diesel fuel include
    excavation, biological methods, and containment.

    A1.3  Environmental transport, distribution, and transformation

         Very few data are avilable on the environmental fate of diesel
    fuels, but the mechanisms of their distribution and transformation are
    considered to be comparable to those of heating fuels, such as No. 2
    fuel oil, which have been well studied. Spills of diesel fuel on water
    spread almost immediately to form a 'slick'. The polar and
    low-relative-molecular-mass components dissolve and leach out of the
    slick, and the volatile components evaporate from the surface;
    microbial degradation also begins. Chemical and biological weathering
    alter the composition of the spill. These processes are dependent on

    temperature; spills that occur in Arctic conditions are more
    persistent than those that occur in temperate climates. In marine
    environments, most of the low-relative-molecular-mass aromatic species
    are dissolved into the water phase, but the primary branched alkanes,
    cycloalkanes, and remaining aromatic compounds may remain in sediments
    for more than a year.

         Although no information is available on the photooxidation of
    diesel fuels in water and air, evaporated oil components are degraded
    photochemically. No. 2 fuel oil has been shown to be photooxidized
    rapidly in water under environmental conditions.

         The individual constituents of diesel fuel are inherently
    biodegradable, to varying degrees and at different rates. The
     n-alkane,  n-alkylaromatic, and simple aromatic molecules in the C10-C22
    range are the most readily degradable. Smaller molecules are generally
    rapidly metabolized. Long-chain  n-alkanes are more slowly degraded,
    owing to their hydrophobicity and because they are viscous or solid at
    ambient temperatures. Branched alkanes and cycloalkanes are relatively
    resistant to biological breakdown, and PAHs are resistant. The overall
    rates of degradation of hydrocarbons are limited by temperature, water
    content, oxygen, pH, inorganic nutrients, and microbial metabolic
    versatility.

         Unicellular algae can take up and metabolize both aliphatic and
    aromatic hydrocarbons, but the extent to which this actually occurs in
    nature is poorly understood. Unlike microorganisms that use petroleum
    carbons as a carbon source, animals generally oxidize and conjugate
    products, rendering end-products that are more soluble and therefore
    easier to excrete. All animal species tested can take up petroleum
    hydrocarbons. PAHs, crude oil, and refined petroleum products are
    known to induce cytochrome P450 enzymes and to increase the levels of
    hydrocarbon metabolism in numerous marine and freshwater fish species.

         Few data are available on the bioaccumulation of diesel fuel in
    the laboratory, but there is plentiful evidence from studies of spills
    and laboratory studies on other oils, particularly No. 2 fuel oil,
    that aquatic organisms bioconcentrate hydrocarbons. The
     n-octanol-water partition coefficient for diesel fuel is 3.3-7.06,
    which suggests high potential bioaccumulation; however, many of the
    lower-relative-molecular-mass compounds are readily metabolized, and
    the actual bioaccumulation of higher-relative-molecular-mass compounds
    is limited by their low water solubility and large molecular size.
    Thus, actual bioaccumulation may be low.

         Fish have been tainted by diesel fuel after spills. No data are
    available on the biomagnification of diesel fuel.

         No experimental data are available on the movement of diesel fuel
    through the soil, although a direct correlation between the movement
    and kinematic viscosity has been proposed. The movement of kerosene
    through soil depends on the moisture content and nature of the soil.

    A1.4  Environmental levels and human exposure

         As diesel fuels are complex mixtures, the environmental levels
    have not been measured. The individual constituents of diesel fuels
    can be detected in almost all compartments of the environment,
    although their source cannot be verified. The general population may
    be exposed to diesel fuel at filling stations and as a result of
    spills.

         Occupational exposure to diesel fuel occurs in a large number of
    activities. Because of their low volatility, diesel fuels should
    generate only low concentrations of vapours at normal temperatures,
    but high operating temperatures can result in significant
    concentrations.

    A1.5  Effects on laboratory mammals and in-vitro test systems

         The acute toxicity of diesel fuels is low after oral or dermal
    exposure or after inhalation. The oral LD50 value was > 5000 mg/kg
    body weight in all species tested (mouse, rabbit, rat, guinea-pig).
    Dermal application resulted in an LD50 value of > 5000 mg/kg body
    weight in mice and rabbits, although values of > 2000 mg/kg body
    weight were reported for some kerosenes and middle distillates, with
    different protocols and lower limit doses. The LC0 value in rats
    exposed by inhalation was about 5 mg/litre, except for one
    straight-run middle distillate for which a value of 1.8 mg/litre was
    seen.

         In rabbits treated dermally with up to 8000 µl/kg body weight per
    day and mice with up to 40 000 mg/kg body weight per day, acanthosis
    and hyperkeratosis due to severe irritation were seen. Rabbits were
    more sensitive than mice. Inhalation of diesel fuel was neuro-
    depressive in mice at concentrations up to 0.2 mg/litre but not
    in rats exposed to up to 6 mg/litre. Body and liver weights were
    reduced in rats.

         Mice, rats, and dogs did not show significant cumulative toxicity
    after inhalation of up to 1.5 mg/litre subchronically. The specific
    nephropathy syndrome seen in male rats is linked to an inherent
    accumulation of hyalin droplets in the renal tubules.

         The only effects of long-term exposures were ulceration after
    dermal application to mice (250 or 500 mg/kg body weight per day) and
    significant alterations in organ weight after inhalation of 1 or
    5 mg/litre by rats. In both studies the mean body weights were
    reduced.

         Various types of diesel fuels were slightly to severely
    irritating to the skin of rabbits. Diesel fuels do not irritate the
    eye, but some kerosenes have been reported to have a slight irritating
    effect. Diesel fuels do not cause skin sensitization.

         Diesel and jet fuels (kerosene) were neither embryotoxic nor
    teratogenic in two studies in rats exposed by inhalation to 100 or
    400 ppm and in one study in which rats were given up to 2000 mg/kg
    body weight per day by gavage. In the last study, reduced fetal weight
    was observed.

         Tests in  Salmonella typhimurium did not provide clear evidence
    of mutagenicity. Some positive findings in  S. typhimurium and in
    mouse lymphoma cells were considered to be equivocal owing to the
    inconsistency of the results. Tests for genotoxicity in mice  in vivo
    (induction of micronuclei or chromosomal aberrations) also gave
    equivocal or negative responses.

         Diesel fuels induced a low level of dermal carcinogenicity. In
    the present state of research, it cannot be concluded whether the
    carcinogenic potency of diesel fuels is mediated by a genotoxic
    mechanism or by chronic dermal damage.

    A1.6  Effects on humans

         Non-occupational exposure to diesel fuel can occur during manual
    filling of fuel tanks. The primary source of dermal exposure is
    accidental spills, which result in immediate high levels of exposure
    but are of short duration.

         After accidental dermal contact, anuria, renal failure, gastro-
    intestinal symptoms, and cutaneous hyperkeratosis have been reported.
    Toxic lung disease has been observed after accidental ingestion of
    diesel fuel and subsequent aspiration. Persistent productive cough has
    been reported after inhalation. In a case-control study of men exposed
    to diesel fuel, an increased risk for cancer of the lung other than   

    adenocarcinoma was found; a positive association was also seen with
    prostatic cancer, although a higher risk was noted for the group with
    'nonsubstantial' exposure than for that with 'substantial' exposure.
    In a cross-sectional study of factory workers exposed to kerosene jet
    fuels, dizziness, headache, nausea, palpitation, pressure in the
    chest, and eye irritation were found to be more prevalent than in
    unexposed controls. The time-weighted average concentration of 
    vapour from the fuel in the breathing zone was estimated to be
    128-423 mg/m3.

    A1.7  Effects on other organisms in the laboratory and the field

         Diesel fuel is more toxic than crude oil to aquatic organisms and
    plants. The ecotoxicity of diesel fuel is generally attributed to
    soluble aromatic compounds, but insoluble aliphatic hydrocarbons may
    also be implicated. Of the aromatic compounds, monoaromatics are the
    least toxic, their acute toxicity increasing with molecular mass up to
    the four- to five-ring compounds, although these are poorly soluble in
    seawater. In some animals, e.g. fish and birds, physical coating of
    the body surface by the fuel can produce toxicity and mortality.

         Laboratory experiments have been carried out on diesel fuel,
    water-soluble fractions, oil-water dispersions, and microencapsulated
    oil. Diesel fuel did not significantly reduce the growth in culture of
    the green alga  Euglena gracilis, whereas a low concentration (0.1%)
    almost completely inhibited the growth of  Scenedesmus quadricauda.
    Light diesel fuel (0.05%) stimulated the growth, photosynthesis, and
    chlorophyll asynthesis of  Chlorella salina but slightly inhibited
    respiration; at higher concentrations, the growth rate and
    photosynthesis were greatly reduced. Long-term exposure inhibited the
    growth of the benthic algae  Ascophyllum nodosum and  Laminaria
     digitata. In blue-green algae, photosynthesis was reduced by the
    aromatic and asphaltic fractions but not by the aliphatic fraction.

         Diesel fuel was acutely toxic to  Daphnia spp., chironomid
    larvae, and the mollusc  Viviparus bengalensis (Gastropoda). A
    concentration of 0.1 ml/litre caused the death of tidepool copepods,
     Tigriopus californicus, within five days.

          Mytilus edulismussels accumulate diesel fuel, have markedly
    reduced feeding and growth rates, and show reproductive toxicity after
    chronic exposure to diesel fuel. The EC50 for spawning in mussels
    exposed for 30 days was about 800 µg/litre. The LC50 of micro-
    encapsulated diesel oil after exposure of maturing mussels for
    30 days was about 5000 µg/litre. Diesel oil was more toxic to larvae
    than to juveniles: 10 µg/litre had adverse effects on the growth of
    larvae.

         Freshwater crabs  (Barytelphusa cunicularis) exposed to sublethal
    concentrations of diesel fuel for up to 96 h generally reduced their
    oxygen consumption, particularly at lower exposures up to 8 h. With
    longer exposures, the oxygen consumption was equal to or higher than
    that of the controls.

         In 96-h tests of acute toxicity in juvenile salmonids under
    static conditions, diesel fuel was more toxic to pink salmon,
     Onchorhychus gorbuscha (LC50: 32-123 mg/litre), than to cohosalmon,
     O. kisutch (LC50: 2186-3017 mg/litre), or rainbow trout,
     O. mykiss (LC50: 3333-33 216 mg/litre), irrespective of water type.

         The threshold for detection of behavioural responses of cod
    ( Gadus morhua L.) exposed to diesel fuel in seawater was within
    100-400 ng/litre. The Antarctic fish  Pagothenia borchgrevinki
    withstood an undiluted water-soluble fraction of diesel fuel oil for
    up to 72 h but showed signs of stress.

         Birds are affected externally and internally by oil
    contamination. Diesel fuel destroys the waterproof nature of the
    birds' plumage and is ingested during preening. Diesel and fuel oil
    fed by gavage at 2 ml/kg body weight to ducks caused lipid pneumonia,
    extreme inflammation of the lungs, fatty infiltration of the liver,
    and hepatic degeneration after 24 h. Administration of diesel or fuel
    oil at 1 ml/kg caused severe irritation of the digestive tract and
    toxic nephrosis. Higher doses resulted in adrenal enlargement (mainly
    due to hyperplasia of cortical tissue), depression of plasma
    cholinesterase levels, ataxia, and tremors. Doses up to 20 ml/kg body
    weight were not fatal to healthy birds, but the LD50 for diesel and
    fuel oil administered to birds under stress was 3-4 ml/kg body weight.

         After spills of diesel fuel, zooplankton appear to be highly
    vulnerable to dispersed and dissolved petroleum constituents but less
    so to floating oils. Aquatic organisms may be affected in a number of
    ways, including direct mortality (fish eggs, copepods, and mixed
    plankton), external contamination by oil (chorions of fish eggs and
    cuticles and feeding appendages of crustaceans), tissue contamination
    by aromatic constituents, abnormal development of fish embryos, and
    altered metabolic rates.

    A1.8  Evaluation of human health risks

         The general population can be exposed to diesel fuel and other
    middle distillates at filling stations, as a result of accidental
    spills, during the handling of such fuels, and during use of kerosene
    for domestic cooking or heating. Workers can be exposed to diesel fuel
    and other middle distillates while handling and discharging the fuel
    at terminals, storage tanks, and filling stations; during the
    manufacture, repair, maintenance, and testing of diesel engines and
    other equipment; during use of diesel fuel as a cleaning agent or
    solvent; and in handling and routine sampling of diesel fuel in the
    laboratory. Owing to the low volatility of diesel fuel, only low
    concentrations of vapour are likely to occur at room temperature,
    although in confined spaces at high temperatures significant levels
    may be found.

         Exposure to vapour is minimal during the normal handling of
    diesel fuel. The most likely effect on human health is dermatitis
    after skin contact. Diesel fuel is a skin irritant but does not appear
    to irritate the eye. Acute toxic effects on the kidney can occur after
    dermal exposure, but the effects of long-term dermal absorption of low
    concentrations are unknown.

         Diesel fuels are toxic when ingested, sometimes resulting in
    regurgitation and aspiration, which can cause chemical pneumonia; the
    same is true for any hydrocarbon in a particular range of viscosity.

         In rodents exposed by inhalation to diesel fuel at concentrations
    up to 0.2 mg/litre, a neurodepressive effect was seen in mice but not
    in rats at the higher concentrations. Subchronic exposure by
    inhalation to various distillate fuels induced specific
    alpha2-microglobulin nephropathy in male rats; this effect is
    considered irrelevant for humans.

         Diesel fuels were neither embryotoxic nor teratogenic in animals
    exposed orally or by inhalation.

         There is no clear evidence of mutagenic activity in bacteria, and
    the results of other tests for genotoxicity  in vitro and  in vivo were
    equivocal.

         A case-control study of workers exposed to diesel fuel suggested
    an increased risk for cancer of the lung other than adenocarcinoma and
    for prostatic cancer. In neither case was there an exposure-response
    relationship. In view of the small number of studies available, the
    small number of cases, and the correspondingly wide confidence
    intervals, no conclusion can be drawn about the carcinogenicity to
    humans of diesel fuel.

         In mice, dermally administered diesel fuels had weak carcinogenic
    potential. In view of the absence of clear genotoxicity, cancer could
    be induced by nongenotoxic mechanisms, e.g. by chronic dermal
    irritation characterized by repeated cycles of skin lesions, causing
    epidermal hyperplasia.

    A1.9  Evaluation of effects on the environment

         The environment can be polluted by accidental release of diesel
    fuel on a large scale, such as during tanker disasters and pipeline
    leaks, or on a smaller scale from contamination of soil around
    factories or garages. In water, diesel fuel spreads almost
    immediately, polar and low-relative-molecular-mass components dissolve
    and leach out, volatile components evaporate from the water surface,
    and microbial degradation begins. The extent to which 'weathering'
    takes place depends on the temperature and on climatic conditions. The
    chemical composition of spills changes with time: after spillage on
    water, some fractions evaporate, and the evaporated diesel components
    are degraded photochemically; in sediment, diesel fuel appears
    generally to be delivered to bottom sediments by settling particles;
    in soil, the components of diesel fuel migrate at different rates,
    depending on the soil type.

         The individual constituents of diesel fuel are inherently
    biodegradable, but the rates of biodegradation depend heavily on
    physical and climatic conditions and on microbial composition.

         Aquatic organisms, in particular molluscs, bioaccumulate
    hydro-carbons to varying extents, but the hydrocarbons are depurated
    on transfer to clean water. Diesel fuel may be bioaccumulated; no data
    are available about biomagnification.

         Spills of diesel fuel have an immediate detrimental effect on the
    environment, causing substantial mortality of biota. Recolonization
    may occur after about one year, depending on the animal or plant
    species and the chemical and physical content of the spill residues.
    Aquatic organisms that survive diesel fuel spills can be affected by
    external oil contamination and tissue accumulation: abnormal
    development and altered metabolic rates are signs of the resulting
    stress.

    A2.  IDENTITY, PHYSICAL AND CHEMICAL  PROPERTIES, AND ANALYTICAL
         METHODS

    A2.1  Identity

         Diesel fuels are a gas-oil fraction occurring during petroleum
    separation and commonly known as middle distillates (International
    Agency for Research on Cancer, 1989a). Gas oils are generally blended
    materials formulated to meet technical specifications (CONCAWE, in
    press). Commercial diesel fuels contain aliphatic, olefinic,
    cycloparaffinic, and aromatic hydrocarbons (see section A2.1.1), and 
    additives (see section A2.1.2) to improve their fuelling properties 
    (Sandmeyer, 1981).

         Four qualities of diesel fuel are available commercially: diesel
    fuel (general), diesel fuel No. 1, diesel fuel No. 2, and diesel fuel
    No. 4 (see Table 1). The composition of diesel fuels is comparable to
    that of heating oils (e.g. No. 1 and No. 2 fuel oils), except for the
    additives (International Agency for Research on Cancer, 1989a; Agency
    for Toxic Substances and Disease Registry, 1995).

         Diesel fuel No. 2 is used mainly as automobile fuel and
    corresponds to diesel fuel (general). In Europe (countries of the
    European Union (EU) and European Free Trade Area (EFTA)), the
    specifications for diesel fuel for transportation purposes are given
    in European standard EN 590 (European Committee for Standardization,
    1993), which also provides for changes in the specifications to meet
    the requirements of different climatic conditions. In Sweden, two
    further qualities of on-road diesel fuel are available: environmental
    class 1 and class 2 (city diesel), with sulfur contents of 0.05% and
    0.001%, respectively (Standardization Board in Sweden, 1991). The
    sulfur content of city diesel corresponds to that of kerosene. Diesel
    fuel for ship engines is covered by the International Standards

    Organization (ISO) standard 8217 (International Standards
    Organization, 1987). In the United States of America, three grades of
    diesel fuel are available: diesel fuel No. 1 (relatively high
    volatility) for road vehicle engines subject to frequent speed and
    load changes; diesel fuel No. 2 (lower volatility) for industrial or
    heavy-duty, high-load engines running at uniform speed; and diesel
    fuel No. 4 (viscous) for low- and medium-speed engines such as those
    used in ships (American Society for Testing and Materials, 1988,
    1992).

         Several types of kerosene, also derived from the gas-oil fraction
    of petroleum separation, are used as aviation turbine fuels (JP fuels,
    jet fuels). Their hydrocarbon composition is comparable to that of
    diesel fuels.

    A2.1.1  Fuel components

    A2.1.1.1  Alkanes

         Normal, branched, and cyclic alkanes (paraffins) are the most
    abundant components (about 65-85%) of diesel fuels. Pristane
    (2,6,10,14-tetramethylpentadecane) and phytane (2,6,10,14-
    tetramethyl-hexadecane) are of particular interest environmentally, as
    the ratios of pristane to heptadecane and of phytane to octadecane
    make it possible to identify the source of a fuel spill; furthermore,
    as these ratios increase during biological degradation, they can be
    used to estimate the age of an environmental contamination and the
    degree of elimination. Cycloalkanes and bicycloalkanes constitute a
    significant portion of the mixture, but individual compounds are
    present only at low levels and are difficult to analyse. Alkyl
    derivatives of cyclopentane, cyclohexane, and cycloheptane are common
    components (Block et al., 1991; Table 2).

    A2.1.1.2  Alkenes

         Alkenes are not common components  of crude oil but may be 
    present in diesel fuel if converted products are added after cracking.
    These alkenes have predominantly branched and cyclic structures (Block
    et al., 1991). The total alkene content of diesel fuels is up to 10%
    (CONCAWE, 1985) (see also Table 2).

    A2.1.1.3  Aromatic compounds

         Aromatic compounds constitute  5-30% of automotive diesel fuel, 
    5-40% of marine diesel fuel (CONCAWE, 1985), and 10-30% of diesel 
    fuel No. 2 (Block et al., 1991). Table 2 shows the specifications of 
    commercial diesel fuels in this regard.

        Table 1.  Synonyms and trade names of commercial diesel fuels
                                                                                                
    Name           CAS           CAS Registry       Range of       Synonyms
                   name          number             carbon
                                                    numbers
                                                                                                

    Diesel fuel    Diesel oil    68334-30-5         C9-C20a        Auto diesel, automotive
    (general)                                       C10-C28b       diesel oil, DERV, diesel,
                                                                   diesel fuel oil, diesel
                                                                   oil, gas oil

    Diesel fuel                  Not assigned       C9-C16c        Diesel fuel oil No. 1,
    No. 1                        (essentially       C4-C16 (for    diesel oil No. 1, No. 1
                                 equivalent to      wide-cut       dieseld
                                 kerosene,          aviation)c
                                 8008-20-6)

    Diesel fuel    No. 2         68476-34-6                        Diesel fuel, diesel fuel
    No. 2          diesel        (applicable for                   oil No. 2, diesel oil
                   fuel          specific                          No. 2, No. 2 diesele
                                 viscosity
                                 limits)

    Diesel fuel                  68476-31-3         C10-C30f       Marine diesel fuel,
    No. 4                                                          distillate marine diesel
                                                                   fuel
                                                                                                

    Adapted from International Agency for Research on Cancer (1989a) and supplemented
    a  From CONCAWE (in press)
    b  From CONCAWE (1985); automotive gas oil (automotive diesel fuel, DERV)
    c  From CONCAWE (1985, 1995)
    d  In Europe, fuels similar to US diesel No. 1 are commonly referred to as
       'kerosene' or 'Arctic diesel'
    e  Term uncommon in Europe. In the United Kingdom, distillate fuels are frequently
       categorized as Class A1 (road diesel) and A2 (off-highway diesel)
    f  From CONCAWE (1985); distillate marine diesel
    
         Only trace quantities of toxicologically relevant benzene,
    toluene, ethyl benzene, and xylene compounds (for physicochemical
    properties, see Table 6) are present in diesel fuel No. 2, but
    significant levels are found in diesel fuel No. 1 (Arctic diesel),
    which has lower flash-point specifications (Block et al., 1991); the
    International Agency for Research on Cancer (1989a) cites a
    flash-point of 0.25-0.5%. The concentration of benzene in kerosenes is
    < 0.01% by volume; wide-cut aviation kerosenes may have higher
    levels, but they are usually < 1% by volume (IPCS, 1986; CONCAWE,
    1995; IPCS, 1993).

        Table 2.  Hydrocarbon specifications of some commercial diesel fuel oils
                                                                                                

    Specification             Diesel      Keroseneb           Distillate       Diesel
                              fuela                           marine           fueld
                                                              dieselc

                                                                                                

    Paraffins/naphthenes      65-95e      78-96               60-90
    (volume %)                            (wide-cut
                                          aviation, < 1)

    n-Hexane (volume %)                   < 0.01
                                          (wide-cut
                                          aviation, < 1)

    Saturates (volume %)                                                       59.4-76.6

    Olefins (volume %)        0-10        0-5                 0-10             0.0-1.0

    Aromatics (volume %)      5-30        4-25                5-40             23.4-39.6
                                          (wide-cut
                                          aviation,
                                          6-25)

    Aromaticity (weight %)                                                     11-15
                                                                                                

    See also Table 1

    a  From CONCAWE (1985, in press); fuel oil similar to diesel fuel (general)
    b  From CONCAWE (1985, 1995); fuel oil similar to diesel fuel No. 1
    c  From CONCAWE (1985, in press); fuel oil similar to diesel fuel No. 4
    d  From German Scientific Association for Petroleum, Natural Gas, and Coal (1991);
       three samples of diesel fuel (general)
    e  Depending on origin of crude oil
    
         Alkyl benzenes (particularly C3 and C4) are common components of
    diesel fuel. Polycyclic aromatic hydrocarbons (PAHs), e.g.
    naphthalene, phenanthrene, acenaphthene, acenaphthylene, fluorene,
    fluoranthene, and pyrene, are also present, as are alkyl- and
    cycloalkyl-substituted homologues of these substances, the predominant
    ones being naphthalene and its methyl-substituted derivatives (see
    Table 6 for physicochemical properties) (Block et al., 1991).

         As some PAHs and the benzene, toluene, ethyl benzene, and xylene
    components have been shown to be toxic and ecotoxic, these classes of
    compounds are usually included in analytical procedures for
    environmental contamination by diesel fuel. The PAH content of diesel
    fuels varies widely, the highest levels being found in low-quality
    fuel blended for large users (frequently railroad companies). The
    mid-range aromatic (and PAH) content of diesel blends is limited by
    the cetane number specification (Block et al., 1991) (see section
    A2.1.3).

         The concentrations of total PAHs in diesel fuel are < 5% by
    volume, although some marine diesel fuels may contain > 10% by volume
    (CONCAWE, 1985; International Agency for Research on Cancer, 1989a).
    In straight-run gas-oil components, which are the major blending
    material of diesel fuel (see section A3.2.1.1), three-ring PAHs
    predominate; use of heavier atmospheric vacuum or cracker gas oils in
    a diesel-fuel mixture leads to an increasing content of four- to
    six-ring PAHs (CONCAWE, in press).

         The concentrations of these constituents in a commercial diesel
    fuel (unspecified) and, for comparison, in No. 2 fuel oil are shown in
    Table 3. It should be noted that the PAH content of No. 2 fuel oil is
    not limited by cetane number specification, so that it may include a
    larger proportion of these hydrocarbons (Block et al., 1991). Table 3
    also gives the composition of the water-soluble  fractions of diesel
    and No. 2 fuel oils and indicates how the solubility of a compound
    affects the composition of the fraction and of the whole oil.

    A2.1.1.4  Sulfur

         The sulfur content of middle distillates depends on the source of
    crude oil (Booth & Reglitzky, 1991; CONCAWE, 1995). The sulfur content
    of most diesel fuels is 0.1-0.5% by weight; it is higher than that of
    gasoline, which is about 0.02% by weight (Scheepers & Bos, 1992a).
    Only diesel fuel No. 4 (distillate marine diesel) has a sulfur level
    > 1% by weight (CONCAWE, 1985, in press). ISO standard 8217
    (International Standards Organization, 1987) specifies a sulfur
    content of marine diesel fuel of 1.0-2.0% by weight.

         The sulfur content of some diesel fuels ranges from 0.01 to > 3%
    by weight. In Europe (countries of EU and EFTA), the sulfur content is
    restricted to a maximum of 0.2-0.3% by weight, and, as of 1996, it
    will be further reduced to 0.05% by weight (European Commission,
    1993). In the United States in 1988, the maximal sulfur content
    permitted was 0.5% by weight for diesel fuels No. 1 and 2 and 2.0% by
    weight for diesel fuel No. 4 (American Society for Testing and


        Table 3.  Concentrations of toxicologically relevant aromatic hydrocarbons in diesel fuel and No. 2 fuel oil and in 10% water-soluble
              fractions prepared from them
                                                                                                                                                

    Compound                            Diesel fuela   Diesel fuelb   No. 2 fuelc     No. 2 fueld      Water-soluble fractions (µg/litre)
                                        (% by wt)      (% by wt)      (% by wt)       (% by wt)                                                 
                                                                                                       No. 2            Diesel
                                                                                                       fueld            fuela
                                                                                                                                                

    Benzene                             0.1            > 0.02         0.006-0.008     NR               550              344
    Toluene                             0.7            0.25-0.5       0.01-0.08       NR               1040             777
    Alkylbenzenese                                                                                     970
    Ethylbenzene                        0.2            0.25-0.5       0.01-0.08       NR                                139
    Xylene                              0.5            0.25-0.5       0.01-0.08       NR                                875
    Polycyclic aromatic hydrocarbons
      Naphthalene                       0.4                           0.273           0.4              840              6.6
      1-Methylnaphthalene               NR                            NR              0.82             340              66.2
      2-Methylnaphthalene               NR                            0.67            1.89             480              108
      Dimethylnaphalenese               NR                            NR              3.11             240              NR
      Trimethylnaphthalenese            NR                            NR              1.84             30               NR
      Fluorenese                        NR                            NR              0.36             20               NR
      Phenanthrenese                    NR                            0.15            0.53             20               NR
                                                                                                                                                
    NR, not reported
    a  From Dunlap & Beckmann (1988); the analytical method is not described in detail; the concentration of benzene seems to be
       unusually high.
    b  From International Agency for Research on Cancer (1989a); according to CONCAWE (1985), some marine diesel fuels may contain
       more than 10% polycyclic aromatic hydrocarbons.
    c  From Stone (1991); summary of several reports
    d  From Anderson et al. (1974); US National Research Council (1985)
    e  Total of several isomers
        Materials, 1988); in October 1993, the maximum was reduced to 0.05% by
    weight (US Environmental Protection Agency, 1992a; American Society
    for Testing and Materials, 1992). The permitted sulfur level in Brazil
    is > 3% by weight (A. Sivak, personal communication, 1993). In Japan,
    the sulfur content of diesel fuels was reduced to 0.2% by weight in
    1994, and a further reduction, to 0.05% by weight, is under discussion
    (CONCAWE, 1990a). Blended marine diesel fuel may also contain up to
    about 15% residual components, i.e. material with an initial
    boiling-point above about 350°C (CONCAWE, in press).

    A2.1.2  Fuel additives

         Only agents that are added to fuels at a concentration < 1% are
    described as 'additives'. A more appropriate term for substances
    present at higher concentrations is 'fuel components'. Fuels are
    treated with additives for a number of reasons (see Table 4 and
    below); they also differentiate products and determine the trademark
    quality of commercial fuels (Fabri et al., 1990).

    A2.1.2.1  Cetane number improvers

         Cetane number improvers upgrade the ignition characteristics of a
    base fuel more economically than refinery processes (Fabri et al.,
    1990) (see section A2.1.3). Primary alkyl nitrates (e.g. isooctyl
    nitrate) are often used to improve cetane number. Polyethyleneglycol
    dinitrates, although effective at much lower concentrations, have a
    number of disadvantages, including their price and the fact that they
    may not improve the performance of fuel in low-compression engines
    (Russell, 1989).

    A2.1.2.2  Smoke suppressors

         Organometallic compounds containing barium, calcium, manganese,
    or iron have been used to reduce diesel smoke. With barium-based
    products, 85-95% of the metal is emitted as particulates in the
    exhaust (Russell, 1989); however, barium and calcium compounds are no
    longer used.

    A2.1.2.3  Flow improvers

         Cold-flow improvers increase the fluidity of the fuel by
    modifying the growth of wax crystals formed by higher homologues of
    paraffins at low temperatures. The wax content of diesel fuels is
    influenced by the origin of the crude oil, the distillation range of
    the fuel, and the source of blend components (Coley, 1989).


        Table 4.  Diesel fuel additives
                                                                                                                                                

    Additive                    Material                           Concentration       Effect
                                                                   (ppm)
                                                                                                                                                

    Ignition improvers,         Organic nitratesa                                      Enhancement of self-ignition qualitiesa
    cetane enhancersa

    Smoke suppressors,          Organic compounds of Ca, Ba,                           Reduction of soot; increase in metal
    combustion enhancersa       or (sometimes)Mga                                      sulfate emissionsa

    Detergentsa                 Amines, imidazolines,                                  Prevention and removal of coke deposits
                                succinimides, etc.b                                    on fuel injector tips, etc.a

    Flow improvers              Olefin-ester copolymers            50-500              Interaction with wax crystals and
                                                                                       modification of their growth; prevention
                                                                                       of formation of agglomerates

    Cloud-point depressors      Olefin-ester copolymers            About 1000          Depression of cloud-point; prevention of
                                                                                       formation of agglomerates

    Wax anti-settlers           Modified ethylene-vinyl            100-500c            Reduction of crystal size and rate of
                                acetate copolymersc                                    settling; prevention of formation of agglomerates

    Anti-static agents          Not reported                       Not reported        Reduction of building up of charges of
                                                                                       static electricity

    Anti-corrosion chemicals    Alkenyl succinic acids and         5-50                Prevention of corrosion or rusting of
                                esters, dimer acids, amine                             storage tanks, pipelines, and metal fuel
                                salts                                                  system components
                                                                                                                                                

    Table 4 (contd)
                                                                                                                                                

    Additive                    Material                           Concentration       Effect
                                                                   (ppm)
                                                                                                                                                

    Antioxidants                Hindered phenols or amines         25-200              Prevention of aging processes

    Anti-foam agents            Silicones                          Up to 20            Reduction of foaming tendency; reduction
                                                                                       of risk of ground pollution from oil spills

    Dehazers                    Quaternary amine salts             5-50                Reduction of formation of hazes;
                                                                                       acceleration of haze clearance

    Biocides                    Imines, amines,                    About 200           Prevention of growth of bacteria and
                                imidazolines, etc.                                     fungi in storage tank bottoms

    Lubricants                  Surface-active agents such         50-500              Compensation of lower viscosity of fuels
                                as polyfunctional acids and                            in low-temperature regions
                                derivatives

    Odour maskers               Natural, identical substances,     10-100              Reduction or elimination of smell
                                such as vanillin and terpenesc
                                                                                                                                                

    From Coley (1989), except as noted
    a  From Organisation for Economic Co-operation and Development (1993)
    b  From Russell (1989)
    c  From Fabri et al. (1990)
        A2.1.2.4  Cloud-point depressors

         Diesel fuels must be easily filterable, as wax crystals formed at
    low temperatures can clog fuel filters. Cloud-point depressors are
    therefore added (Fabri et al., 1990) which consist of substances with
    lower cloud-points, e.g. kerosene. Addition of 10% kerosene lowers the
    cloud-point of diesel fuel by about 2°C. Olefin-ester copolymers
    depress the cloud-point by 3-4°C but are not currently in commercial
    use (Coley, 1989).

    A2.1.2.5  Wax anti-settling additives

         These additives inhibit the tendency of wax to settle by reducing
    the crystal size and slowing the settling rate. A five-fold reduction
    in wax crystal size slows the settling rate by one-twenty-fifth. With
    increasing temperature, small dispersed crystals redissolve more
    readily than settled wax (Coley, 1989).

    A2.1.2.6  Other additives

         Detergents, including amines, amides, imidazolines, and
    succinates, are used to reduce injector nozzle fouling. Detergents
    such as polyalkenyl succinimides also improve fuel stability,
    resistance to corrosion, and combustion efficiency (Russell, 1989).
    Antistatic agents lower the risk of building up a charge of static
    electricity during pumping at high rates at bulk terminals or in
    large-capacity truck fuel tanks.

         Other additives used are anti-oxidants (phenols, amines),
    anti-corrosion chemicals (alkenyl succinic acids, esters, dimer acids,
    amine salts), anti-foam agents (silicones), dehazers (anti-emulsion
    agents) (quaternary ammonium salts), biocides, lubricants for cold
    regions (surface-active polyfunctional acid derivatives), and odour
    maskers (vanillin, terpenes) (Coley, 1989; Fabri et al., 1990).

    A2.1.3  Quality of diesel fuels

    A2.1.3.1  Ignition performance and cetane number

         The cetane number determines the ignition performance of
    transport fuels relative to a scale on which methyl naphthalene
    corresponds to a combustion rate of 0 and cetane to one of 100
    (Scheepers & Bos, 1992a). The cetane number is calculated by comparing
    the ignition quality of a fuel with that of two reference fuel blends
    of known cetane numbers under standard operating conditions (American
    Society for Testing and Materials method D 613 CFR). A high cetane
    number improves cold starting and engine durability and reduces noise,
    fuel consumption, smoke emissions during warm-up, and exhaust
    emissions (Russell, 1989). In Europe (countries of the EU and EFTA),

    the minimal cetane number must be in the range 45-49, depending on the
    climatic conditions (European Committee for Standardization, 1993);
    cetane numbers are usually 49-53 (CONCAWE, 1987). In the United
    States, the cetane number must be at least 30 for diesel fuel No. 4
    and 40 for diesel fuels No. 1 and 2 (American Society for Testing and
    Materials, 1988, 1992).

    A2.1.3.2  Density

    The density of diesel fuel influences engine performance: higher
    density leads to enrichment of the fuel:air mixture, which results in
    greater engine power output. Enrichment may, however, increase the
    particulate content of exhaust gas emissions (Fabri et al., 1990) (see
    section B3.1.2.2). A density range is specified in fuel standards in
    some countries (American Society for Testing and Materials, 1992;
    European Committee for Standardization, 1993).

    A2.1.3.3  Sulfur content

         Gas and particle emissions in diesel engine exhaust are
    influenced by the sulfur content of the fuel; there is a direct
    relationship between particle production and sulfur content (Hare,
    1986) (see section B3.1.2.2).

    A2.1.3.4  Viscosity

         Too low a viscosity can lead to wear in the injection pump; too
    high a viscosity impairs fuel injection and mixture formation (Fabri
    et al., 1990). In Europe (countries of the EU and EFTA), the viscosity
    of commercial diesel fuels at a temperature of 40°C must be
    1.5-4.5 mm2/s (European Committee for Standardization, 1993). In the
    United States, the permitted ranges of viscosity at 40°C are
    1.3-2.4 mm2/s for diesel fuel No. 1, 1.9-4.1 mm2/s for diesel fuel
    No. 2, and 5.5-45.0 mm2/s for diesel fuel No. 4 (American Society
    for Testing and Materials, 1988, 1992).

    A2.1.3.5  Cold-flow properties

         Cloud-point and cold filter plugging point characterize the
    behaviour of diesel fuels at low temperatures (Fabri et al., 1990).
    These points are lowered by the addition of cloud-point depressors and
    by special blending techniques, e.g. increasing the kerosene content
    of the fuel (see section A2.1.2).

         Changes in diesel fuel quality have been assessed. Wade & Jones
    (1984) reported a deterioration in fuel quality with decreasing cetane
    number and found that a 90% increase in boiling-point led to greater
    emissions of particulates, nitrogen oxides and PAHs. A decline in
    diesel fuel quality on the European market was predicted, as the
    rising demand would lead to greater use of fuels from catalytic or
    thermal cracking processes (CONCAWE, 1987) (see section A3.2.1.1). A
    study by the Organisation for Economic Co-operation and Development
    (1993) indicated an improvement in the quality of diesel fuel in the
    United States due to a decline in aromaticity.

    A2.2  Physical and chemical properties

         Diesel fuel is a brown, slightly viscous, flammable liquid at
    room temperature (Sandmeyer, 1981). It generally has a kerosene-like
    odour (Agency for Toxic Substances and Disease Registry, 1995). Its
    physical and chemical properties are listed in Table 5.

         The water solubility of diesel fuels varies. The aqueous
    solubility of crude and fuel oils in the environment is clearly
    dependent on the salinity of the water and the age of the oil slick
    (see section A4.1.1) and is of the same order of magnitude as the
    solubility of fuel oils: at room temperature, 0.37-0.53 mg/litre in
    sea water (Boehm & Quinn, 1974) and 0.7-11 mg/litre in tap water
    (Lysyj & Russell, 1974). The solubility of a fuel slick decreases with
    its age as the concentration of long-chain hydrocarbons increases; the
    solubility of fresh crude oil is 29.3-32.3 mg/litre, whereas that of
    weathered crude oil is 0.06-23.2 mg/litre at 25°C (Mackay & Shiu,
    1976).

         The physicochemical properties of the toxicologically relevant
    benzene, toluene, ethyl benzene, xylene and PAH components are given
    in Table 6.

    A2.3  Analytical methods

         Exhaustive identification and quantification of the individual
    constituents of commercial diesel fuel (see section A2.1) is almost
    impossible owing to their number and complexity. In most environmental
    assessments, therefore, the mixture is analysed as total petroleum
    hydrocarbon (Block et al., 1991). All such methods involve preliminary
    solvent extraction of the matrix, with e.g. trichlorotrifluoroethane.
    This step may also extract naturally occurring hydrocarbons, which
    interfere with the analysis; however, some of these compounds are
    polar and can be removed on silica gel. Three analytical methods are
    available:


        Table 5.  Physicochemical properties of diesel fuels
                                                                                                                                                

    Property                             Diesel fuel               Diesel fuel              Diesel fuel            Diesel fuel
                                         (general)                 No. 1                    No. 2                  No. 4
                                                                                                                                                

    Melting-point (°C)                                             - 34a                    18a                    - 29-9a

    Boiling range (°C)                   160-190b                  145-300 (wide-cut        282-338a               170-420d
                                         143-384e                  aviation, 45-280)c                              101- > 588a
                                                                   193-293a

    Flash-point (°C)                     > 56b                     > 21 - < 55              52 (closed cup)a       > 56d
                                         58-66e                    (wide-cut aviation,                             > 54 (closed cup)a
                                         (Pensky-Martens)          < 21)c
                                                                   38 (closed cup)a

    Autoignition temperature (°C)                                  177-329a                 254-285a               263a

    Density (g/cm3)                      0.81-0.90 (15°C)b         0.805 (wide-cut          0.87-0.95              0.87-0.92 (15°C)d
                                         0.82-0.84 (15°C)e         aviation, 0.75-0.801)    (20°C)                 0.81-0.94 (15°C)
                                                                   (15°C)c 0.81-0.94                               1 (20°C)
                                                                   (15°C)

    Kinematic viscosity (mm2/s)          2-7.4 (40°C)b             1.5-2.5 (wide-cut                               2-7.4 (40°C)d
                                         2.20-3.25 (40°C)e         aviation, about 1.1
                                                                   (20°C)c

    Vapour pressure (kPa)                About 40 (40°C)b          About 10 (wide-cut       2.83-35.2              2.83-35.2 (21°C)a
                                         0.04f                     aviation, 140-210)       (21°C)a
                                                                   (Reid, 37.8°C)c
                                                                   2.83-35.2 (21°C)a
                                                                                                                                                

    Table 5 (contd)
                                                                                                                                                

    Prpoerty                             Diesel fuel               Diesel fuel              Diesel fuel            Diesel fuel
                                         (general)                 No. 1                    No. 2                  No. 4
                                                                                                                                                

    Water solubility (mg/litre)          1f                        About 5 (20°C)a          About 5 (20°C)a        About 5 (20°C)a
                                         0.2f

    Henry's law constant                 4.3 × 103f                6.03-7.5 × 105a          6.03-7.5 × 105a        6.03-7.5 × 105a
    (Pa × m3/mol) (20°C)

    n-Octanol-water partition                                      3.3-7.06a                3.3-7.06a              3.3-7.06a
    coefficient (log Kow)

    Soil sorption coefficient            3.04f                     3.0-6.7a                 3.0-6.7a               3.0-6.7a
    (log Koc)

    Diffusion coefficient in air         4.63 × 10-2f
    (cm2/s)

    Odour threshold (ppm)                                          0.7a                                            0.5g
                                                                                                                                                

    a   From Agency for Toxic Substances and Disease Registry (1995)
    b   From CONCAWE (in press); automotive gas oil
    c   From CONCAWE (1985, 1995); kerosenes
    d   From CONCAWE (in press); distillate marine diesel
    e   From German Scientific Association for Petroleum, Natural Gas, and Coal (1991); average of three samples
    f   From Custance et al. (1993); water solubility measured, other data from literature; no further details
    g   From Fraser Williams (Scientific Systems) Ltd (1985)
        --    Gravimetric detection: determination of the weight of residue
         remaining after solvent evaporation. Although this method is
         useful for measuring gross contamination, it cannot be used in
         trace analysis.

    --    Infrared detection: the absorbance of petroleum hydrocarbons is
         detected in a solvent matrix at the maximal value, near
         2930 cm-1. This carbon-hydrogen bond stretching absorbance is
         directly related to the hydrocarbon concentration in the extract.
         The results are dependent on the hydrocarbon standard used for
         quantification.

    --    Gas chromatographic detection: capillary gas chromatography
         combined with flame ionization or mass spectrometric detection.

         The gravimetric and infrared techniques are fast, relatively
    simple and widely accepted by regulatory authorities; however, they do
    not provide sufficient qualitative and quantitative information on
    composition. Gas chromatography is therefore the standard procedure
    used to identify and quantify fuel constituents in environmental
    samples. Different distribution and degradation processes in
    environmental compartments (see section A4) mean that the composition
    of petroleum hydrocarbons in air, water, soil, and biota may differ
    considerably from that of the original fuel. Newton et al. (1991)
    developed an analytical method for identifying and quantifying traces
    of diesel contamination in tinned fish products on the basis of the
     n-alkane pattern.

         Because of the suspected toxicity of aromatic components in
    diesel fuel, a detailed analysis is often necessary. Two groups of
    compounds are detected:

    --   volatile aromatic compounds: Analysis of these minor components
         of diesel fuel requires special enrichment techniques, such as
         purge-trap gas chromatography, and methods of detection including
         flame ionization and mass spectrometry.

    --   PAHs: Gas chromatography with flame ionization or mass
         spectrometric detection, or liquid chromatography, is used after
         solvent extraction. As PAHs are poorly resolved from the diesel
         fuel matrix, gas chromatography-mass spectrometry is necessary in
         most cases to quantify the components. The target analytes of
         commonly used analytical methods do not, however, include C3
         and C4 alkyl-substituted benzenes or most alkyl-substituted
         PAHs in diesel fuels.


        Table 6.  Physicochemical properties of aromatic components of diesel fuel
                                                                                                                                                

    Aromatic compound      Water solubility      Henry's law constant    Diffusion coefficient (cm2/s)           Octanol-water partition
                           (mg/litre)            (Pa × m3/mol)           In air           In water               coefficient (log Pow)
                                                                                                                                                

    Benzene                1791                  550                     0.087            9.8 × 10-6             1.99
    Toluene                534.8                 602                     0.083            8.6 × 10-6             2.52
    Ethylbenzene           161.0                 855                     0.076            7.8 × 10-6             2.94
    Xylene                 156.0                 778                     0.076            1 × 10-5               2.94
    Polycyclic aromatic    0.067                 0.007                   0.067            2.12 × 10-6            5.30
      hydrocarbons
                                                                                                                                                

    Adapted from Custance et al. (1993)
        A2.4  Conversion factors

         As diesel fuel vapour is a complex mixture of gases, it is
    impossible to give a conversion factor for converting parts per
    million in the gaseous phase to SI units.

    A3.  SOURCES OF HUMAN AND ENVIRONMENTAL EXPOSURE

    A3.1  Natural occurrence

         Diesel fuels are derived from crude oil, which can be considered
    a 'natural product'. Nevertheless, human and environmental exposure
    results almost exclusively from anthropogenic activities.

    A3.2  Anthropogenic sources

    A3.2.1  Production and use

    A3.2.1.1  Production process

         Diesel fuel is produced during the refining of crude oils but is
    generally blended to meet the specifications for technical
    performance. The blending components may be produced by atmospheric
    distillation of crude oil (straight-run atmospheric gas oil), vacuum
    distillation of atmospheric residue (vacuum gas oil), thermal cracking
    (thermally cracked gas oil), or catalytic cracking processes (e.g.
    light catalytically cracked gas oil; cycle oil). The main components
    are straight-run gas oils: 80% of European automotive diesel fuel is
    made up of these components (Booth & Reglitzky, 1991). Secondary
    processing of heavier fractions is increasingly necessary in order to
    meet product demand (CONCAWE, in press).

         Diesel fuel No. 1 is manufactured by a straight-run distillate
    process (Agency for Toxic Substances and Disease Registry, 1995);
    diesel fuel No. 2 is generally made by mixing straight-run and
    catalytically cracked distillates; and diesel fuel No. 4 is produced
    by adding blending stocks to distillation residues in order to meet
    viscosity specifications (International Agency for Research on Cancer,
    1989a). Further variations are introduced by formulation with
    additives to improve fuel properties (see section A2.1.2).

    A3.2.1.2  Use

         Diesel fuel is widely used as a transport fuel for light- and
    heavy-duty vehicles; the Organisation for Economic Co-operation and
    Development (1993) classifies vehicles weighing < 3.5 t as light-duty
    and those weighing > 3.5 or, occasionally, > 5 t as heavy-duty.
    Diesel fuel No. 1 is suitable for engines that undergo frequent
    changes in speed and load (Agency for Toxic Substances and Disease

    Registry, 1995). Heavier grades (diesel fuels No. 2 and 4) are used
    for trucks, railroad and marine diesel engines, and stationary engines
    in continuous high-load service (Sandmeyer, 1981; International Agency
    for Research on Cancer, 1989a). Diesel fuels are also used in
    stationary gas turbines, e.g. to generate electric power during
    peak-load periods. Residual fuel oils, such as diesel fuel No. 4, are
    used to generate steam in electric power plants (International Agency
    for Research on Cancer, 1989a), in commercial and industrial burner
    installations without preheating facilities, in plants and factories
    for space and water heating, for pipeline pumping, and in gas
    compression; they are also sprayed on unmade roads to compact the
    surface and are used in the manufacture of asphalt cement (Agency for
    Toxic Substances and Disease Registry, 1995).

    A3.2.1.3  Production and consumption levels

         The demand for diesel fuel has increased worldwide over time. In
    1990, world demand was about 1100 kt/day (Booth & Reglitzky, 1991).
    The production and consumption of diesel fuel in different regions
    over time are shown in Table 7.

         Diesel-fuelled passenger cars are relatively common in western
    Europe. The percentages of passenger cars with diesel engines in
    various European countries over time are shown in Table 8. In many
    European countries, taxi-cabs are equipped almost exclusively with
    diesel engines. In 1985, diesel-fuelled heavy-duty trucks comprised
    about 86% of the fleet in Norway, 89% in Denmark, 95% in Sweden, and
    100% in the former Federal Republic of Germany. In some countries of
    Europe, diesel fuel consumption has increased steadily over the last
    decade, and consumption for heavy-duty vehicles is predicted to more
    than double between 1980 and 2005 (Organisation for Economic
    Co-operation and Development, 1993). In western Germany, diesel fuel
    consumption was almost 60% higher in 1990 (about 5100 kt) than in 1984
    (about 3200 kt) (Federal Ministry for Transport, 1992).

         Diesel-fuelled passenger cars are less common in North America.
    In the United States in 1986, about 1.6% of passenger cars were
    diesel-fuelled, and the tendency was predicted to decrease slightly
    (US Department of Energy, 1988). In contrast, 82% of heavy-duty trucks
    were diesel-fuelled in 1985 (Organisation for Economic Co-operation
    and Development, 1993), and 59300 kt of diesel fuel were consumed by
    highway traffic in 1986, representing 14.8% of all the highway fuel
    used and about one-half of United States diesel fuel consumption.
    On-road use of diesel fuel was predicted to increase to 15.5% of all
    highway fuel (66600 kt) in 1995. A difference in fuel use patterns is
    seen between light- and heavy-duty vehicles. In 1986, about 4000 kt
    were used by passenger cars and light trucks, with an estimate of
    about 2300 kt for 1995, whereas in 1986 about 55300 kt were used by
    heavy-duty trucks, with an estimate of 64300 kt for 1995 (US

    Department of Energy, 1988). In Canada, only about 1% of passenger
    cars are equipped with diesel engines. In 1992, 14126 kt of diesel
    fuel were produced; 13508 kt were sold for domestic purposes (heating)
    and about 10849 kt for on-road use (Thomas, personal communication,
    1994).

         The percentage of diesel-fuelled vehicles is increasing in Japan
    (Table 9).

    Table 7.  Production and use of diesel fuel (including gas oils in
              Europe) in different regions, and development over time
                                                                        

    Region              Production/    1975 (kt)    1980 (kt)   1985 (kt)
                        Use
                                                                        

    United States       Production     134 967      138 323     135 181
                        Use            133 300      136 161     130 297

    Canada              Production                              19  187
                        Use                                     19  130

    OECD (all 24        Production     370 240      408 848     372 728
    countries)          Use            380 181      403 642     386 081

    OECD (Europe)       Production                              173 474
                        Use                                     190 960

    European Union      Production                              152 091
                        Use                                     163 132

    Australia, Japan,   Production                              44  886
    New Zealand         Use                                     45  694
                                                                        

    Adapted from International Agency for Research on Cancer (1989a);
    OECD, Organisation for Economic Co-operation and Development

    A3.2.2  Emissions during production and use

    A3.2.2.1  Air

         No data are available on emissions during the production and use
    of diesel fuel. Release to the atmosphere during production in the
    refining process is unlikely, as closed systems are used, but
    volatilization may occur during storage and transport. Diesel fuel may

    be released into the atmosphere as a result of spills during storage
    and transport and at filling stations during bulk storage and vehicle
    tank filling. The low-relative-molecular-mass constituents
    (short-chain alkanes, benzene, toluene, ethyl benzene, and xylene
    compounds) are the most likely to evaporate under environmental
    conditions.

    Table 8.  Percentage of diesel passenger cars in western Europe, and
              development over time
                                                                        

    Country                          Diesel passenger cars (%)
                                                                        

                                     1984      1988    1991   1992a
                                                                        

    France                           15.8      21.1    25.5   38
    Germany (without former          16.9      15.1    15.3   14.9 
      German Democratic Republic)                             (1993)b
    Germany (without former          About 8   13.6    13
      German Democratic Republic)c
    Italy                            18.7      11.9     7.4   NR
    Spain (without station wagons)   15.3      10.5    14.6   NR
    United Kingdom                    1.1       4.5    11.3   12.5
    United Kingdomd                   NR          5       9   19 (1993)
                                                                        

    Adapted from American Automobile Manufacturers Association (1993),
    unless otherwise stated; NR, not reported
    a  From Menne et al. (1994)
    b  With the former German Democratic Republic
    c  From German Institute for Scientific Research (1993)
    d  From Quality of Urban Air Review Group (1993)

    Table 9.  Percentage of diesel-fuelled vehicles in Japan and
              development over time
                                                                        

    Year     Passenger cars (%)    Trucks (%)     Buses (%)
                                                                        

    1984     3.8                   23.7           49.5
    1988     5.8                   28.5           63.2
    1991     8.3                   40.7           81.7
                                                                        

    Adapted from American Automobile Manufacturers Association (1993)

    A3.2.2.2  Water

         No data are available on effluents and emissions during diesel
    fuel production. Diesel fuel oils may be released to surface waters as
    a result of leakage from storage tanks or tankers (see section
    A3.2.3). In the United States, the total volumes of spills of diesel
    fuel oils in 1991 were (Agency for Toxic Substances and Disease
    Registry, 1993): diesel fuel No. 1, 10.6 t (20 notations); diesel fuel
    No. 2, 8.9 t (28 notations); and diesel fuel No. 4, 39.3 t (35
    notations).

         Groundwater can be contaminated with diesel fuel constituents by
    leakage from underground bulk storage tanks, but quantitative data are
    lacking. In the Province of New Brunswick, Canada (Thomas, personal
    communication, 1993), of a total of 671 gasoline, diesel, and fuel oil
    tanks examined in 1987, 11% leaked diesel or fuel oil; in 1989, 6% of
    1085 tanks had leaks; and in 1991, 13% of 539 tanks leaked fuel. No
    data were available about the amounts leaked, and diesel contamination
    was determined on the basis of a cluster of 'middle-range peaks'
    detected by gas chromatography.

         On the basis of chronic petroleum inputs to sediments of
    Narragansett Bay and Rhode Island Sound, United States, Van Vleet &
    Quinn (1978) estimated that about 200 kt of petroleum hydrocarbons may
    be released to surface waters worldwide from municipal treatment
    plants. The precise source of the inputs cannot be verified in most
    cases, but they are due, for example, to accidental discharge to sewer
    systems, disposal of used crankcase and lubricating oils, oil washed
    from roads, and atmospheric deposition of hydrocarbons.

         No data were available from other countries.

    A3.2.2.3  Soil

         Diesel fuels may be released to soil as a result of accidental
    spills (see section A3.2.3) and leakage from storage tanks or
    pipelines. Diesel-contaminated soil is a major problem in railroad
    yards, where diesel fuel is used in locomotives and as a solvent to
    clean moving metal parts (the remaining paraffins give a waxy
    anti-corrosive coat). Spillage during refuelling, engine maintenance,
    and steam cleaning, leakage from fuel storage tanks, and absorption of
    diesel fuel on sand used for traction are other possible sources of
    soil contamination (Dineen, 1991); however, no quantitative data are
    available.

    A3.2.3  Accidental releases to the environment

         Data on the accidental release of diesel fuels are summarized in
    Table 10. The effects on the environment depend on both the amount and
    environmental conditions.


        Table 10.  Diesel fuel spills and their effects on the environment (see also section A9.2)
                                                                                                                                                

    Year     Place                       Amount (t); type        Effects                                               Reference
                                         of diesel fuela
                                                                                                                                                

    Aquatic environment
    1972     North of Puget Sound,       > 600; diesel fuel      Substantial mortality of some intertidal taxa;        Woodin et al. (1972)
             Washington State, USA       No. 2                   larval recruitment within six months

    1973     East Lamma Channel          2000-3000;              Almost total kill of meiofauna within four days;      Wormald (1976);
             near Hong Kong              diesel fuel No. 4       heavy mortality among bivalves and molluscs,          Stirling (1977)
                                                                 Nerita albicilla and Monodonta labio, and in
                                                                 marine fish farms; almost no effect on
                                                                 Clypeomorus humilis or Planaxis sulcatus

    1978     Svea, Mijenfjord, Norway    111                     Most fuel trapped in ice; transport out of fjord      Carstens &
                                                                 during break-up; heavy mortality among shoreline      Sendstad (1979)
                                                                 invertebrates in some regions of fjord

    1983     Yaquina Bay, Oregon,        About 240 (with         Decline in population of Rhepoxynius abronius         Kemp et al. (1986)
             USA                         Bunker C oil)           by 75%; influence of spill not clear

    1984     Queen Charlotte             130 (plus 30 t          Low fuel levels in water during flood tide, high      McLaren (1985)
             Islands, Canada             gasoline)               levels during ebb as diesel was retained in
                                                                 sediment; high mortality among barnacles
                                                                                                                                                

    Table 10 (contd)
                                                                                                                                                

    Year     Place                       Amount (t); type        Effects                                               Reference
                                         of diesel fuela
                                                                                                                                                

    1987     Macquarie Island,           230; diesel fuel        High mortality among one species of holothuroid,      Pople et al. (1990)
             Australia                   No. 4                   one of isopod, one of limpet, and one of chiton;
                                                                 decreased populations of crab, two species of
                                                                 starfish, and gastropod; small number of dead
                                                                 algae; slow  recovery of organisms

    1989     Arthur Harbor,              510; diesel fuel        High mortality among limpets; partial recovery one    Kennicutt et al.
             Antarctica                  No. 1                   year after spill; small effect on macroalgae          (1991a,b,1992a,b)

    Terrestrial environment
    1983     San Bernadino County,       3                       Hydrocarbon contamination up to 1500 mg/kg;           Frankenberger et al.
             CA, USA                                             plume migrating towards surface water; in-situ        (1989)
                                                                 bioremediation (see also section A3.2.4)

    NR       Califonia, USA              193; diesel fuel        Contamination at 6-34 m below surface;                Peters et al. (1992)
                                         No. 2                   hydrocarbon levels at 18.3, about 1500 mg/kg;
                                                                 remediation by flooding with surfactants
                                                                 (see also section A3.2.4)
                                                                                                                                                

    aWhen available
        A4.  ENVIRONMENTAL TRANSPORT, DISTRIBUTION, AND TRANSFORMATION

         Few data are available on the  environmental fate of diesel
    fuels, but there are adequate data on the environmental behaviour of
    individual hydrocarbon components. The transport, distribution, and
    transformation of crude oils and some fuel oils (e.g. heating oils
    No. 1 and 2) have been studied, and the mechanisms of distribution and
    transformation are considered to be comparable to those of diesel
    fuel.

         The calculated half-lives for the toxicologically relevant PAH
    components are listed in Table 11.

    A4.1  Transport and distribution between media

         Diesel fuels are released into the environment mainly during
    storage, transport, and use (see section A3.2.3). The hydrosphere and
    geosphere are the affected compartments.

    A4.1.1  Evaporation from and dissolution in the aqueous phase

         Oils and fuels spilled on water spread out almost immediately;
    polar and low-relative-molecular-mass components dissolve and leach
    out of the slick; and volatile components evaporate from the water
    surface. With the start of microbial degradation the pattern of
    substances changes, as was observed, for example, after a major oil
    spill off the French coast (Atlas et al., 1981). The sum of these
    processes is known as chemical and biological 'weathering'. At the
    same time, the oil is emulsified to oil-in-water and water-in-oil
    ('mousse') emulsions (Atlas & Bartha, 1973). In laboratory experiments
    with several diesel fuels, however, stable mousses were not formed at
    any temperature (US National Research Council, 1985).  Experiments
    with South Louisiana crude oil showed that the evaporation rate was
    more than twice the dissolution rate (Harrison et al., 1975). The loss
    by evaporation of oil spread on water, which in oil spills amounts to
    about 25% of the total released mass, is generally dependent on the
    following factors (US National Research Council, 1985):

    --   the exposed area, which increases with time as the film spreads;

    --   the oil-phase vapour pressures, which decrease with time as the 
         low-relative-molecular-mass components evaporate. Components 
         with vapour pressures higher than that of  n-octane evaporate 
         rapidly from the surface of spilled fuel oil, while those with
         vapour pressures lower than that of  n-octadecane persist and
         form a viscous residue that retards the volatilization of other
         constituents (Regnier & Scott, 1975);

    --   the oil-air mass transfer coefficient, which depends on wind
         speed and hydrocarbon diffusivity (for the diffusion coefficient
         in air, see section A2.2, Table 5);

    --   diffusion barriers, such as emulsions, or formation of a 'skin'
         on the surface. Many surface waters have a thin layer of organic
         material of natural origin on the surface, and these films may
         reduce spread and volatilization, altering the persistence and
         concentration of the oil in water (Edgerton et al., 1987).

    Table 11.  Calculated half-lives of polycyclic aromatic hydrocarbon
               components in environmental compartments
                                                                        

    Component                Air       Water       Soil        Sediment
                                                                        

    Naphthalene              1 day     1 week      2 months    8 months
    Acenaphthalene           2 days    3 weeks     8 months    2 years
    Fluorene                 2 days    3 weeks     8 months    2 years
    Phenanthrene             2 days    3 weeks     8 months    2 years
    Anthracene               2 days    3 weeks     8 months    2 years
    Pyrene                   1 week    2 months    2 years     6 years
    Fluoranthene             1 week    2 months    2 years     6 years
    Chrysene                 1 week    2 months    2 years     6 years
    Benz[a]anthracene        1 week    2 months    2 years     6 years
    Benzo[a]fluoranthene     1 week    2 months    2 years     6 years
    Benzo[a]pyrene           1 week    2 months    2 years     6 years
    Perylene                 1 week    2 months    2 years     6 years
    Dibenz[a,h]anthracene    1 week    2 months    2 years     6 years
                                                                        

    Adapted from Mackay et al. (1992)

         On the basis of the constant of Henry's law (see section A2.2,
    Table 5) and the classification of Thomas (1990), diesel fuels can be
    considered to volatilize significantly from the aqueous phase. The
    constant may change during volatilization, as the composition of the
    mixture changes with time. Lockhart et al. (1987) observed a clear
    decrease in the hydrocarbon composition of the 50% water-soluble
    fraction of diesel fuel in open-aerated test cylinders. Within 48 h,
    the concentration of hydrocarbons with a very low boiling-point
    (< 115°C) decreased from 0.73 to 0.04 mg/litre and that of
    hydrocarbons with a low boiling-point (115-270°C) from 3.2 to
    0.17 mg/litre. Under sealed or open non-aerated conditions, there were
    only minor changes. Regnier & Scott (1975) showed that  n-decane
    volatilizes from diesel fuel with a half-life of 1.65 h at 30°C,
    whereas the half-life was 9.67 h at 5°C in a darkened chamber with a
    constant wind speed of 21 km/h. Edgerton et al. (1987) reported that
    diesel fuel spilled on ice had lost only 2-4% of its lightest fraction
    after 10 days.

         Dissolution of spilled oils and fuels in water is less important
    from the viewpoint of mass loss than the ecotoxic effects of the
    mixture. Lysyj & Russell (1974) conducted long-term experiments (42
    days) on the dissolution of different fuel oils and gasoline in water.
    After a period of stable concentration, during which physical
    dissolution was probably the main mechanism for oil-to-water transfer,
    the dissolution of organic compounds accelerated, probably because of
    chemical modification by oxidation or microbial action of the
    water-insoluble petroleum fraction.

         Under equilibrium mass-transfer conditions, the levels of PAHs
    (naphthalene and its methyl derivatives, acenaphthene, fluorene,
    phenanthrene, anthracene, and fluoranthene) in the water phase in
    contact with commercial automobile diesel fuel were about four to six
    orders of magnitude lower than those in the diesel fuel itself, i.e.
    in the low microgram per litre range. The concentrations of these
    constituents were likely to be smaller under non-equilibrium
    conditions but might be higher in the presence of surfactants,
    emulsifiers, or co-solvents (Lee et al., 1992). When No. 2 fuel oil
    and kerosene were placed in contact with drinking-water for 17 h,
    although each fuel contained about 50% by weight of aliphatic
    hydrocarbons, the water-soluble fractions contained primarily aromatic
    compounds: total, > 93% by weight; proportion of benzene and
    substituted derivatives, 19.4% by weight in No. 2 fuel oil and 53.2%
    by weight in kerosene (Coleman et al., 1984). The chemical composition
    of the water-soluble fractions of No. 2 fuel oil is compared with that
    of the whole oil in Table 3.

         During chemical weathering, other changes in composition (e.g.
    autoxidation of  n-alkanes) may occur that are not necessarily
    associated with a net mass loss (Davis & Gibbs, 1975).

    A4.1.2  Transport in and adsorption onto soil and sediment

    A4.1.2.1  Soil

         The movement of diesel fuel through soil is directly correlated
    with its kinematic viscosity (see section A2.2, Table 5). As the
    viscosity is about one-half that of water and about one-fourth to
    one-fifth that of gasoline, the rate of percolation of water into soil
    is about twice and that of gasoline about four to five times that of
    diesel fuel (Stone, 1991).  The retardation coefficient (Rd) is an
    indicator of the migration of substances into groundwater. In the
    aqueous phase, the maximal rate of movement of dissolved compounds is
    that of groundwater, which is represented by an Rd of 1. Stone (1991)
    calculated the Rd from the adsorption isotherms of individual diesel
    constituents, and derived the following classification:

    --   low mobility (Rd > 100): fluorene, phenanthrene, pyrene, 
         benzanthracene, benzo[ a]pyrene, fluoranthene;

    --   medium mobility (100 > Rd > 10): naphthalene, dimethylbenzenes, 
         ethylbenzene, toluene;

    --   high mobility (Rd < 10): benzene, quinoline, cresols, phenol. 

         In effect, soil contaminated with diesel fuel acts like a
    chromatographic column, separating individual constituents by their
    adsorption. Experimental data on the adsorption of diesel fuels on
    soils and their transport are not available; however, the movement of
    a synthetic kerosene through a clay-sand soil was dependent on the
    moisture content. With increasing moisture content, the penetration
    rate of nonvolatile components through the soil increased and the
    adsorption of more volatile components decreased. In contrast, upward
    mobility of the individual constituents (both volatile and
    nonvolatile) of the fuel decreased with increasing moisture content.
    In fully water-saturated soil, upward transport of kerosene was
    completely inhibited (Acher et al., 1989). Desorption of a simulated
    kerosene applied to three types of soil with the same moisture content
    but different contents of organic matter was found to be complete
    after 30 days under environmental conditions. The slowest desorption
    was from the soil with the highest level of organic matter (Yaron et
    al., 1989). Transport of kerosene through three types of sand was
    influenced mainly by volatilization of C9-C13 hydrocarbon
    constituents; the increase in the viscosity of the kerosene residue
    was followed by a decrease in the infiltration rate (Galin et al.,
    1990).

    A4.1.2.2  Sediment

         In marine environments, oils and fuels are generally delivered to
    bottom sediments on settling particles. In areas with low
    concentrations of suspended material (i.e. terrigenous minerals,
    plankton, detrital particles, resuspended bottom sediments), the rates
    of hydrocarbon sedimentation would be low (US National Research
    Council, 1985).

         No experimental data are available on diesel fuels in sediments.
    In laboratory experiments with fuel oil, binding of the hydrocarbon
    components to clay minerals of marine sediments was shown to be due to
    physical adsorption of the van der Waals type. The compounds are
    associated firmly because, once they are incorporated into sediments,
    desorption to the overlying water is slow. Saturated hydrocarbon
    constituents of fuel oils were transported slowly to the sediment
    (Meyers & Quinn, 1973). In studies of a laboratory marine ecosystem to
    which No. 2 fuel oil was added for 24 weeks, the concentrations in the
    sediment remained low until 135 days after the start of the
    experiment. Then, a steep increase was seen, leading to a level of 9%
    by weight of the total fuel added (i.e. 12% by weight of the total
    saturated hydrocarbons). The highest concentrations of saturated

    hydrocarbons were found in the surface flocculent layer; 2-3 cm below
    the sediment surface, the levels were below the detection limit. The
    smaller particles (< 45 µm) of water-suspended material seemed to
    adsorb about twice as much fuel oil as larger particles (> 45 µm)
    (Wade & Quinn, 1980). In a comparable study in a model marine
    ecosystem with No. 2 fuel oil, the particulate matter contained 40-50%
    of the total amount of aliphatic hydrocarbons added and only 3-21% of
    the aromatic fraction, indicating that most of the aromatic
    hydrocarbons were dissolved in the water phase (see section A4.1.1)
    (Gearing et al., 1980). Although biodegradation may remove many of the
    soluble aromatic hydrocarbons (see section A4.2.2), 10-20% by weight
    of the fuel oil, consisting primarily of branched alkanes,
    cycloalkanes, and aromatics, may remain in the upper 2 cm of sediment
    for over a year (Gearing et al., 1980; Oviatt et al., 1982).

    A4.2  Transformation and removal

    A4.2.1  Photooxidation

         No data were available on the photooxidation of diesel fuel in
    air or water. In the atmosphere, all evaporated oil components are
    degraded photochemically by hydroxy radicals and other species within
    hours or days to carbon monoxide, carbon dioxide, and oxygenated
    compounds (US National Research Council, 1985; IPCS, 1986, 1993). 
    Under laboratory conditions, No. 2 fuel oil was rapidly photo-oxidized
    in water. Two photooxidant pools were identified: hydrogen peroxide
    and a mixture of indane and tetralin hydrogen peroxides, the first
    being the faster acting (Herbes & Whitley, 1983). Peroxides are formed
    by two mechanisms: the first is the reaction of free radicals with
    triplet oxygen to produce peroxy radicals and hydroperoxides, which
    continue the chain reaction; the second is addition of excited singlet
    oxygen to reactive acceptors such as olefins. The photodegradation
    products of fuel oils are phenols, naphthols, and carboxylic acids
    (Larson et al., 1977, 1979).

    A4.2.2  Biodegradation

    A4.2.2.1  Microbial degradation

         The individual constituents of spilled oil are inherently
    degradable to varying degrees and at different rates. The  n-alkane,
     n-alkylaromatic, and simple aromatic molecules in the C10-C22
    and the most readily degradable. Smaller molecules are generally range
    are the least toxic rapidly metabolized, although they tend to be more
    toxic. Long-chain  n-alkanes are more slowly degraded because of
    their hydrophobicity and also because they are viscous or solid at
    ambient temperatures. Branched alkanes and cycloalkanes are relatively
    resistant to attack. PAHs are the compounds most recalcitrant to
    microbial degradation (Morgan & Watkinson, 1988).

         A study of the breakdown of heating oils and diesel fuels in a
    sub-soil demonstrated breakdown rates of 0.014-0.022 g/kg per day,
    which were slower than the rates of topsoil degradation (Flowers et
    al., 1984). The mineralization rates for heavy oils, sludges, and
    crude oils were 0.02-0.6 gram of hydrocarbon per kilogram of soil per
    day. Under Arctic conditions, the rates were considerably slower.

         The relative contribution of bacteria and fungi to hydrocarbon
    degradation is unclear. At acid pH, fungi are clearly more important,
    but closer to neutrality hydrocarbon contamination tends to enhance
    bacterial growth.

         The overall rates of hydrocarbon degradation are limited by:

    --   temperature (optimal temperature for biodegradation, 30-40°C);

    --   water content (in soils, water contents of 50-80% capacity are 
         optimal for microbial activity);

    --   oxygen (the metabolism of both aliphatic and aromatic
         hydro-carbons normally requires oxygen);

    --   pH (hydrocarbons are mineralized most rapidly at 6.5-8.0);

    --   inorganic nutrients (as petroleum contaminants are extremely 
         deficient in nitrogen and phosphorus, a large quantity of carbon 
         sources tends to result in rapid depletion of the available pools
         of major nutrients, i.e. enhancement of degradation in fertilized
         soils);

    --   microbial metabolic versatility (topsoils generally contain many 
         hydrogen-degrading bacteria, typically 105 - 106 per gram, but the 
         content of subsoils is unclear) (Morgan & Watkinson, 1988).

         A remedial treatment consisting of liming, fertilization, and
    tilling was evaluated on a laboratory scale for its effectiveness in
    cleaning up a sand, a loam, and a clay-loam soil contaminated at
    50-135 mg/g of soil by diesel and other oils. The disappearance of
    hydrocarbons was maximal at 27°C (Song et al., 1990).

         In sandy soils, 50-85% of the hydrocarbons were degraded within a
    few weeks, depending on the experimental conditions. Degradation
    efficiency increased with additional soil inoculation (Grundmann &
    Rehm, 1991).

         Frehland et al. (1992) examined the substrate specificity and
    growth conditions of nine bacterial strains on samples of
    diesel-contaminated soil under different conditions. The
    biodegradative properties of Pseudomonas spp. were superior (59, 60,
    and 80% in three samples) than those of Micrococcus, Bacillus,
    Acinetobacter, and other species tested.

    A4.2.2.2  Phytoplankton and marine algae

         Unicellular algae can take up and metabolize both aliphatic and
    aromatic hydrocarbons, but the extent to which this occurs in nature
    is poorly understood. The ability to metabolize naphthalene is
    widespread (US National Research Council, 1985).

    A4.2.2.3  Invertebrates and vertebrates

         Unlike microorganisms, which use petroleum as a source of carbon,
    animals generally oxidize and conjugate products to render them more
    soluble and therefore easier to excrete. All animal groups tested can
    take up petroleum hydrocarbons, either through the body surface,
    across respiratory surfaces, or via the gut, the route varying with
    the organism and its feeding habits. In copepods  (Calanus
     helgolandicus), dietary uptake was more important than that from
    water (Corner et al., 1976), but aromatic hydrocarbons accumulated by
    the polychaete  Neanthes arenaceodentata were derived from water and
    not from sediment (Rossi, 1977). Hydrocarbons are stored in lipid-rich
    tissues. The principal route of metabolism of hydrocarbons by
    metazoans is conversion of lipophilic compounds by cytochrome
    P450-dependent monooxygenases or mixed-function oxidases to
    derivatives that are more water-soluble. Mixed-function oxygenase
    activity has been found in all fish species studied. Lee (1981)
    reported 18 marine invertebrate species belonging to four phyla
    (Annelida, Arthropoda, Echinodermata, and Mollusca) that are known to
    contain mixed-function oxidase activity in the hepatopancreas,
    digestive gland, and other tissues. Sea birds and some other waterfowl
    also have cytochrome P450 systems, but few estimates have been made of
    their capacity for hydrocarbon metabolism.  Some crude oils, PAHs, and
    refined petroleum products are known to induce cytochrome P450 enzymes
    and to increase the levels of hydrocarbon metabolism in marine and
    freshwater fish, including embryonic and larval forms and polychaetes
    (US National Research Council, 1985). Induction is usually described
    in terms of the rate of metabolism of benzo[a]pyrene  in vitro.

    A4.2.3  Bioaccumulation

         Few data are available on the bioaccumulation of diesel fuel
    itself, although there is adequate evidence from studies on other oils
    that aquatic organisms bioconcentrate hydrocarbons. The
     n-octanol-water partition coefficient (log Pow) for diesel fuel is
    3.3-7.1 (see section A2.2, Table 5). Although a higher bioaccumulation
    potential might be expected from these log Pow values, the measured
    bioconcentration factors are lower than predicted because many
    compounds of low relative molecular mass, such as substituted
    benzenes, are readily metabolized, and the actual bioaccumulation of
    compounds of higher relative molecular mass is limited by their low
    water solubility and large molecular size.

         In a continuous-flow system, diesel fuel was taken up by
     Mytilus edulis mussels exposed for 41 days to a concentration of
    200-400 µg/litre. The hydrocarbon levels in tissue were about 1000
    times higher after exposure than before. Depuration was rapid during
    the first 15-20 days and then decreased rapidly to a minimum (Fossato
    & Canzonier, 1976).

         The accumulation of No. 2 fuel oil has been studied in a number
    of organisms (Lee, 1977). When No. 2 fuel oil (38% aromatics) was
    added to a commercial shrimp pond, the concentration of naphthalenes
    in shrimps, clams, and oysters exceeded the level in ambient water
    after 38 days. Depuration in clean water was complete after 10 days
    for shrimps and after 47-96 days for oysters (Cox et al., 1975). In a
    flow-through system to study accumulation of a No. 2 fuel oil-water
    dispersion, aromatic petroleum hydrocarbons were accumulated at a rate
    of 89.6 mg/kg tissue by  Rangia cuneata (clams) and 311.7 mg/kg by
     Crassostrea virginica (oysters) during 8 h of exposure. The
    bioaccumulation factor for methyl naphthalenes in clams was 2.3-26.7
    and increased with increasing alkylation of the naphthalene ring.
    Depuration in clean water was complete after 28 days (Neff et al.,
    1976a). The fuel components of a small spill of No. 2 fuel oil had the
    following biological half-lives in mussels  (Mytilus edulis): aliphatic
    hydrocarbons (C16 and C23), 0.2 and 0.8 days; alkyl naphthalenes
    (C2 and C3), 0.9 and 1.5 days; phenanthrene, 2.1 days (Farrington
    et al., 1982). Lobsters exposed to a simulated small spill of No. 2.
    fuel oil bioconcentrated PAH constituents in the hepato-pancreas and
    muscle within three to four days after exposure. The levels remained
    elevated for up to 10-11 days, and depuration was complete after 20-21
    days.

         The polychaetous annelid  Neanthes arenaceodentata rapidly
    accumulated 14C-naphthalene (to a biomagnification of 40 times) from
    solution during exposure for 24 h and slowly released the accumulated
    hydrocarbons within 300 h after return to clean sea water (Rossi,
    1977).

         Oral administration of South Louisiana crude oil to mallard ducks
     (Anas platyrhynchus) at a dose of 5 ml daily led to detectable
    aromatic petroleum hydrocarbon levels in each of the tissues examined:
    liver, breast muscle, heart muscle, skin, brain, uropygial gland, and
    blood. The skin and the underlying adipose tissue had accumulated far
    more aromatic hydrocarbons than other tissues. The individual
    components did not accumulate in the same relative concentrations as
    in the crude oil, suggesting different uptake and/or metabolism; two-
    and three-membered, condensed ring aromatic compounds were accumulated
    to a greater extent than alkylbenzenes (Lawler et al., 1978).

    A4.2.4  Tainting

         Tainting (unpleasant off-flavours) of various aquatic organisms
    by petroleum hydrocarbons has been reported, and diesel fuel has been
    suspected of tainting fish in freshwater rivers. Mackie et al. (1972)
    reported tainted rainbow trout up to at least nine days after a spill
    of diesel fuel into water (Six Mile Water, Ireland). Ackman & Noble
    (1973) described tainting of whitefish in the Athabaska River, Canada,
    which was attributed to diesel fuel from a railroad yard.

    A4.2.5  Entry into the food chain

         No data are available on the effect of diesel fuel on the food
    chain. The effects of petroleum compounds on the marine food web have
    been studied, e.g. on zooplankton  Calanus helgolandicus (Corner et
    al., 1976), on phytoplankton (Mahoney & Haskin, 1980), and on
    freshwater crayfish, one of the primary foods of waterfowl (Tarshis,
    1981). In a study in situ on the distribution, biotransformation, and
    flux of PAHs in the aquatic food chain (seston --> mussel [ Mytilus
     edulis] --> eider duck [ Somateria molissima]), the PAH composition
    changed significantly (Broman et al., 1990). The theoretical
    inhalation of atmospheric PAHs was insignificant in comparison with
    exposure from food. PAHs were biotransformed relatively rapidly. Their
    concentrations did not increase with increasing trophic level.

    A4.3  Ultimate fate after use

    A4.3.1  Use in motor vehicles

         Diesel fuel is combusted in motor vehicles and stationary engines
    (see section A3.2.1.2). The resulting exhaust is discussed in Part B.

    A4.3.2  Spills

         The fate of diesel fuel after accidental spills is eventual
    degradation by abiotic and biotic processes (see section A3.2.3)

    A4.3.3  Disposal

         In general, incineration is recommended for fuel oils (Fraser
    Williams (Scientific Systems) Ltd, 1985). This procedure is also
    useful for disposing of fuel production wastes.

         Because diesel fuel is released during use, mainly accidentally
    into water or onto soil (see sections A3.2.2.2, A3.2.2.3, and A3.2.3),
    the affected compartments may have to be remediated. In general, oil
    spills on beaches and into surface waters can be treated by burning

    off; absorption onto straw, polyurethane foam, activated carbon, or
    peat; or addition of sinking agents, gelling agents, or dispersants.
    Mechanical systems can also be used (Fraser Williams (Scientific
    Systems) Ltd, 1985). Various operations for treating soil affected by
    diesel fuel have been reviewed (Dineen, 1991) and are as follows:

    --    Disposal: excavation of contaminated soil and disposal at a
         regulated landfill; the treatment used most commonly during the
         last decade; advantageous for small volumes.

    --    Biodegradation: The natural process of soil microorganisms,
         which is stopped by a spill as a result of depletion of oxygen or
         nutrients (mostly nitrogen and phosphorus), is restored by
         providing sufficient amounts of the depleted substances. The
         process can be carried out  in situ (Frankenberger et al., 1989)
         (see section A3.2.3, Table 10) or after excavation.
         Biodegradation is the most widely applicable alternative to
         landfill disposal, although its duration is uncertain and may be
         relatively long and a large space is required (see sections
         A9.1.1.2 and A9.1.3.1).

    --    Thermal treatment: incineration of contaminated soil e.g. in
         cement kilns or in specially designed thermal units, where the
         soil is baked to remove the fuel, which is subsequently oxidized
         (thermal desorption). The thermal desorption process is also
         feasible  in situ. In cement kilns, the desorbed diesel fuel is
         used for heating the kiln, and the soil is incorporated into the
         cement mixture.

    --    Road base: The contaminated soil is spread on roads under 
         construction and is compacted and oiled before paving. If the 
         affected soil has a high proportion of expandable clays, it may
         not be suitable as a road base.

    --    Asphalt batching: The contaminated soil replaces part of the
         sand and gravel component of asphalt; this is possible only if
         the content of expandable clays is not too high.

    --    Chemical oxidation: A strong oxidant, usually hydrogen
         peroxide, in combination with a silica polymer stabilizer to
         control the reaction is used to oxidize diesel hydrocarbons to
         carbon dioxide and water. The reaction is highly exothermic, and
         uncontrolled volatilization and leaching of some constituents may
         occur.

    --    Fixation: addition of a strong base to convert diesel
         hydrocarbons to organic acids and subsequent treatment with a
         dissolved silica polymer to effectively encapsulate the acids in
         amorphous silica; immobilizes diesel fuel  in situ and prevents
         migration into surface or groundwater. The long-term stability of
         the immobilized chemicals is, however, unknown.

    --    Soil washing: Diesel fuel is washed out of soil by a
         biodegradable, non-toxic surfactant (Peters et al., 1992). The
         operation is feasible  in situ and after excavation and is
         particularly effective on materials with a low surface:volume
         ratio (e.g. gravels and coarse sands). The major disadvantage is
         the need to collect and treat or dispose of the 
         surfactant-diesel mixture after washing.

    --    Containment: construction of a low-permeability surface
         structure ('cap') that prevents infiltration of rain, snow, and
         other moisture that could lead to leaching of diesel constituents
         into groundwater. Long-term monitoring of the area of the spill
         is necessary.

    A5.  ENVIRONMENTAL LEVELS AND HUMAN EXPOSURE

    A5.1  Environmental levels

         As diesel fuels are complex mixtures of a great variety of
    hydrocarbons (see section A2.1), general 'environmental levels' cannot
    be described. The individual constituents of diesel fuels should be
    detectable in almost all compartments of the environment, but the
    sources cannot be verified.

    A5.2  Exposure of the general population

         Nonoccupational exposure to diesel fuel occurs when fuel tanks
    are filled manually. The primary source of dermal exposure is
    accidental spills, although these exposures to high levels are of
    short duration. Drinking-water can be contaminated by components of
    petroleum products, including diesel fuel, after leakage into
    groundwater from underground storage tanks. Specific data are not
    available on the contribution of diesel fuel components to groundwater
    contamination. Owing to the low volatility of diesel fuels at normal
    temperatures, there is little exposure to vapours (International
    Agency for Research on Cancer, 1989a).

    A5.3  Occupational exposure during  manufacture, formulation, or use

         Workers are exposed to diesel fuel in a number of occupations,
    including manual handling and discharge of fuels; loading and
    unloading of diesel fuel; retailing at filling stations; manufacture,
    repair, maintenance, and testing of diesel engines and other
    equipment; use of diesel fuel as a cleaning agent or solvent; and
    laboratory handling and routine sampling of diesel fuel (CONCAWE,
    1985).

         At normal temperatures, low concentrations of vapours are likely
    to be generated from diesel fuels, owing to their low volatility. Only
    in confined spaces or at high operating temperatures do significant
    levels of vapour occur. Hydrogen sulfide gas may evolve from residual
    fuel oils (CONCAWE, 1985).

    A6.  KINETICS AND METABOLISM IN LABORATORY ANIMALS AND HUMANS

         No quantitative data are available; however, the systemic and
    local effects observed after exposure to diesel fuels by various
    routes indicate that they are absorbed dermally, orally, and by
    inhalation.

    A7.  EFFECTS ON LABORATORY MAMMALS AND IN-VITRO TEST SYSTEMS

    A7.1  Single exposure

         Diesel fuels have a low order of acute toxicity after
    administration orally, dermally, or by inhalation. Acute LD50 Table
    12. In rats and mice, the values after oral or dermal exposure were
    values are summarized in generally > 5000 mg/kg body weight, and the
    values in rabbits after dermal exposure were > 2000 mg/kg body
    weight. Since the physicochemical properties of kerosenes are similar
    to those of diesel fuels (see also section A2.1), the LD50 values
    for various types of kerosene are included.

         Beck et al. (1984) reported an oral LD50 value of 7500 mg/kg
    body weight (9 ml/kg body weight) for a market-place sample of diesel
    fuel in male and female rats. Gross necropsy of animals that died
    during the test showed haemorrhagic gastroenteritis, gastrointestinal
    tympani, and pneumonia with abscess formation. No deaths were seen
    among rabbits treated dermally with 5000 mg/kg body weight under an
    occlusive covering for 24 h.

         After oral administration of 12 000 mg/kg body weight kerosene to
    rats, no change was seen in heart weight or in the relative weights of
    lung, spleen, and kidney. There were no histopathological changes
    (Muralidhara et al., 1982). Exposure of rats for 8 h by inhalation to
    about 0.1 mg/litre deodorized kerosene did not induce adverse effects.
    Four male cats that inhaled 6.4 mg/litre for 6 h showed no toxic
    effects on the central nervous system or body weight during the
    exposure or during a subsequent 14-day observation period (Carpenter
    et al., 1976).

         Male Sprague-Dawley rats given JP-5 jet fuel at a dose of
    18 912 mg/kg body weight (24 ml/kg body weight) orally had
    vacuolization of periportal hepatocytes and serum enzyme alterations,
    indicating hepatic damage, within two days (Parker et al., 1981).  All
    male and female B6C3F1 mice treated dermally with 5000-40000 mg/kg
    body weight marine diesel fuel survived a 14-day observation period
    (US National Toxicology Program, 1986).


        Table 12.  Acute toxicity (LD50 and LC50 values) of diesel and other fuels
                                                                                                                                                

    Fuel                                      Species        Route            LD50 or      Dose                       Reference
                                                                              LC50
                                                                                                                                                

    Diesel fuels
    Diesel fuel (market-place sample)         Rat            Oral             LD50         7500 mg/kg bw              Beck et al. (1984)
    Diesel fuel (market-place sample)         Rabbit         Dermal           LD50         > 5000 mg/kg bw            Beck et al. (1984)
    Home heating oil No. 2a                   Rat            Oral             LD50         14 500-21 100 µg/kg bw     American Petroleum
                                                                                                                      Institute (1980a,b,c)
    Home heating oil No. 2a                   Rabbit         Dermal           LD50         > 5000 µg/kg bw            American Petroleum
                                                                                                                      Institute (1980a,b,c)
    Marine diesel fuel, JP-5, JP-8            Mouse          Oral             LD50         > 16 000 mg/kg bw          Schultz et al. (1981)
      shale, or petroleum oila
    Marine diesel fuel, shale, or             Mouse          Dermal           LD50         > 16 000 mg/kg bw          Schultz et al. (1981)
      petroleum oila

    Other fuels
    Straight-run middle distillateb           Rat            Oral             LD50         > 5000 mg/kg bw            American Petroleum
                                                                                                                      Institute (1985a)
    Hydrodesulfurized middle distillateb      Rat            Oral             LD50         > 5000 mg/kg bw            American Petroleum
                                                                                                                      Institute (1982a,b)
    Straight-run middle distillateb           Rabbit         Dermal           LD50         > 2000 mg/kg bw            American Petroleum
                                                                                                                      Institute (1985a)
    Hydrodesulfurized middle distillateb      Rabbit         Dermal           LD50         > 2000 mg/kg bw            American Petroleum
                                                                                                                      Institute (1982a,b)
    Straight-run middle distillateb           Rat            Inhalation       LC50         1.8 mg/litre               American Petroleum
                                                                                                                      Institute (1987a)
                                                                                                                                                

    Table 12 (contd)
                                                                                                                                                

    Fuel                                      Species        Route            LD50 or      Dose                       Reference
                                                                              LC50
                                                                                                                                                

    Hydrodesulfurized middle distillateb      Rat            Inhalation       LC50         4.6 mg/litre               American Petroleum
                                                                                                                      Institute (1983a)
    Light catalytically cracked distillatec   Rat            Oral             LD50         4660 mg/kg bw (m)          American Petroleum
                                                                                           3200 mg/kg bw (f)          Institute (1985b)
    Light catalytically cracked distillatec   Rabbit         Dermal           LD50         > 2000 mg/kg bw            American Petroleum
                                                                                                                      Institute (1985b)
    Light catalytically cracked distillatec   Rat            Inhalation       LC50         4.7-5.4 mg/litre           American Petroleum
                                                                                                                      Institute (1986a,b)
    Jet fuel Aa                               Rat            Oral             LD50         > 25 000 µl/kg bw          American Petroleum
                                                                                                                      Institute (1980d)
    JP-5 shale oild                           Mouse          Dermal           LD50         8000 mg/kg bw              Schultz et al. (1981)
    JP-5 petroleum oild                       Mouse          Dermal           LD50         11 200 mg/kg bw            Schultz et al. (1981)
    JP-8 shale oila                           Mouse          Dermal           LD50         6400 mg/kg bw              Schultz et al. (1981)
    JP-8 petroleum oila                       Mouse          Dermal           LD50         8000 mg/kg bw              Schultz et al. (1981)
    Jet fuel Aa                               Rabbit         Dermal           LD50         > 5000 µl/kg bw            American Petroleum
                                                                                                                      Institute (1980d)
    Kerosenea                                 Rat            Oral             LD50         50 000 mg/kg bw            Parker et al. (1981)
    Kerosenea                                 Guinea-pig     Oral             LD50         16 320 mg/kg bw            Deichmann et al. (1944)
    Kerosenea                                 Rabbit         Oral             LD50         22 720 mg/kg bw            Deichmann et al. (1944)
    Straight-run kerosenea                    Rat            Oral             LD50         > 5000 mg/kg bw            American Petroleum
                                                                                                                      Institute (1985c)
    Hydrodesulfurized kerosenea               Rat            Oral             LD50         > 5000 mg/kg bw            American Petroleum
                                                                                                                      Institute (1982b)
                                                                                                                                                

    Table 12 (contd)
                                                                                                                                                

    Fuel                                      Species        Route            LD50 or      Dose                       Reference
                                                                              LC50
                                                                                                                                                

    Straight-run kerosenea                    Rabbit         Dermal           LD50         > 2000 mg/kg bw            American Petroleum
                                                                                                                      Institute (1985c)
    Hydrodesulfurized kerosenea               Rabbit         Dermal           LD50         > 2000 mg/kg bw            American Petroleum
                                                                                                                      Institute (1982c)
    Petroleum-derived hydrocarbona            Rat            Inhalation       LC50         > 5.3 mg/litre             American Petroleum
                                                                                                                      Institute (1987b)
    Petroleum-derived hydrocarbona            Rat            Inhalation       LC50         > 5.2 mg/litre             American Petroleum
                                                                                                                      Institute (1983c)
    Deodorized kerosenee                      Rat            Inhalation       LC50         > 0.1 mg/litre             Carpenter et al. (1976)
                                                                                           (saturation)
                                                                                                                                                

    a   Hydrocarbon fraction similar to that of diesel fuel (see section A2.1)
    b   Hydrocarbon fractions from which also diesel fuels are gained during petroleum separation (see section A3.2.1.1)
    c   Blending component of diesel fuel
    d   Narrow boiling range
    e   Used as a solvent, not a fuel; very low aromaticity
             Exposure of Sprague-Dawley rats to an aerosol of diesel fuel at a
    concentration × time product of 8 mg.h/litre caused less than 1%
    mortality and was estimated to represent the maximal tolerated
    exposure. On this basis, 12 mg.h/litre was chosen as the highest
    concentration in a study of repeated exposure (see section A7.2).
    After exposure to 0.5 mg/litre for 2 h, a transient depression of the
    startle reflex was observed. Further findings were a dose-related
    decrease in respiratory frequency and influx of polymorphonuclear
    cells into the lungs for several days. In a 6-h inhalation test, the
    no-observed-adverse-effect level (NOAEL) was 2.7 mg/litre (Dalbey et
    al., 1982, 1987).

    A7.2  Short-term exposure

    A7.2.1  Subacute exposure

    A7.2.1.1  Dermal exposure

         Dermal exposure to 0.1 ml/kg body weight per day kerosene for one
    week was not lethal to male mice. Significant decreases in the
    relative weights of thymus, spleen, adrenals, and abdominal lymph
    nodes were observed (Upreti et al., 1989).

         Eight rabbits (weighing 2.5-3.5 kg) were exposed dermally
    (occlusive covering, 24 h/day, five days per week, two weeks) to 4 or
    8 ml/kg body weight per day of a market-place sample of diesel fuel.
    None of the animals given the low dose died, although there was a
    weight loss of 0.3 kg, while five animals given the high dose died
    with a weight loss of 0.5 kg; the control group had a weight gain of
    0.2 kg. Severe chronic dermal irritation and anorexia were seen, which
    were followed by cachexia and death (Beck et al., 1984).

         Three doses between 1 and 10 ml/kg body weight per day of 10, 30,
    or 50% fuel oil No. 2 were applied to the skin of New Zealand white
    rabbits. After a 12-day treatment including a two-day rest, acute
    dermal corrosion was seen in all animals treated with the 10% oil and
    in animals given 10 ml/kg body weight per day of the 30 or 50% oil.
    There was histopathological evidence of dermal toxicity in all treated
    groups (American Petroleum Institute, 1980a,b,c).

         Application of 8000 µl/kg body weight per day of jet fuel A to
    four male and four female New Zealand white rabbits for 12 days,
    including a two-day rest, induced acute dermal corrosion.
    Histopathological evidence of dermal and hepatic toxicity and
    hyperplastic changes in the urinary bladder transitional epithelium
    were seen (American Petroleum Institute, 1980d).

         Groups of five male and five female New Zealand white rabbits
    were treated with 200, 1000, or 2000 mg/kg body weight per day of a
    straight-run distillate, a light catalytically cracked distillate, or
    two hydrodesulfurized distillates, three times per week for four
    weeks. Moderate to severe irritation was seen, the degree of skin
    irritation being directly related to the number of doses applied.
    Microscopic examination of the skin at the application sites revealed
    the presence of acanthosis and hyperkeratosis (American Petroleum
    Institute, 1983a, 1984, 1985a,b).

         In a study of marine diesel fuel containing 12.7% paraffins,
    43.7% naphthalenes, and 43.6% aromatic compounds, groups of male and
    female B6C3F1 mice received dermal applications (with an occlusive
    covering) of 2000, 8000, 20000, or 40000 mg/kg body weight per day for
    14 consecutive days. None of the animals at 20000 or 40000 mg/kg body
    weight per day survived. Skin lesions were observed in all treated
    groups, which consisted of moderate acanthosis, parakeratosis, and
    hyperkeratosis. These findings were accompanied by a mixed cellular
    inflammatory infiltrate in the upper dermis. The NOAEL was 8000 mg/kg
    body weight per day. In the same study, JP-5 navy fuel containing
    52.8% cycloparaffins, 30.8% paraffins, 15.9% aromatics, and 0.5%
    olefins (see also Table 2) was applied at doses of 5000-40000 mg/kg
    body weight per day for 14 days. All animals at 40000 mg/kg body
    weight per day and females at 30000 mg/kg body weight per day group
    died before the end of the study. Body weight loss was observed at
    doses > 10 000 mg/kg body weight per day. The dermal findings were
    similar to those in the study of marine diesel fuel. The NOAEL was
    5000 mg/kg body weight per day (US National Toxicology Program, 1986).

    A7.2.1.2  Inhalation

         Rats (strain, age, and sex not specified) were exposed for
    6 h/day to a droplet aerosol of deodorized kerosene at a mean
    concentration of 7.7 mg/litre for four days. After a day with no
    treatment, the skin of the extremities was dry and flakes had formed.
    The body weight remained unchanged (Carpenter et al., 1976).

         CD-1 mice were exposed for 8 h/day for five days to uncombusted
    diesel vapour at concentrations of 0.065, 0.135, or 0.204 mg/litre.
    The dose-response for neurotoxicity (measured by square-box activity)
    was 50% of the control level at the high dose, 100% at the medium
    dose, and 150% at the low dose. A drastic deterioration of motor
    coordination was seen at the high dose in the rotarod test. In the hot
    plate test, increased sensitivity followed by tolerance was seen at
    the high dose, a slight increase at the medium dose, and a substantial
    increase at the low dose; normalization occurred in all groups within
    24 h. No effects were seen in the inclined plane test or the corneal
    reflex test. Vasodilatation, poor grooming, and tremors were
    described. In general, the most obvious effects were on the nervous
    system, diesel fuel acting as a neurodepressant (Kainz & White, 1982,
    1984).

         Dalbey et al. (1982, 1987) investigated the toxic effects of
    diesel fuel aerosols to assess the health risks of their military use
    as smoke screens. Groups of male and female Sprague-Dawley rats were
    exposed to concentrations of 1.3-6 mg/litre for 2 or 6 h, one or three
    times per week, for a total of nine exposures; 12 mg.h/litre was
    chosen as the highest concentration-time product that caused 6.25%
    mortality. A significant depression of body weight and a slower growth
    rate were seen initially, and liver weights were significantly
    decreased. No neurotoxicity was seen. Histologically, there was focal
    accumulation of free pulmonary cells and increased numbers of
    leukocytes in bronchial lavage fluid. In pulmonary function tests, a
    non-dose-dependent change in lung volume was seen, involving a
    decrease in total and vital capacity and increases in functional
    residual capacity and residual volume. No other organs were affected.

         Groups of 20 male and 20 female Sprague-Dawley rats exposed by
    inhalation to 0.025 mg/litre of one of three hydrodesulfurized middle
    distillates for 6 h/day on five days per week for four weeks showed no
    treatment-related changes. Mild subacute inflammatory changes were
    observed in the nasal respiratory mucosa of the animals in one group
    that had been exposed to distillate API 81-09 (American Petroleum
    Institute, 1986c).

    A7.2.2  Subchronic exposure

    A7.2.2.1  Dermal exposure

         Groups of 10 male and 10 female B6C3F1 mice received dermal
    applications under an occlusive covering of 250 or 4000 mg/kg body
    weight per day marine diesel fuel or 500-8000 mg/kg body weight per
    day JP-5 navy fuel for five days per week for 13 weeks. The marine
    diesel fuel caused no treatment-related deaths. Males had reduced body
    weight gain; mild chronic active dermatitis at the site of application
    was observed at the highest dose. At 1000 mg/kg body weight per day,
    all animals survived and had weight gains similar to those of the
    controls. Exposure to JP-5 navy fuel killed five males at the
    high-dose group and caused dermatosis, splenic extramedullary
    haematopoiesis, and liver karyomegaly in all groups (US National
    Toxicology Program, 1986).

    A7.2.2.2  Inhalation

         The results of studies of subchronic inhalation of various
    distillate fuels are summarized in Table 13.


        Table 13.  Studies of subchronic exposure to distillate fuels by inhalation
                                                                                                                                                

    Fuel                   Species         Exposure (high dose)                Results                             Reference
                                                                                                                                                

    Diesel fuels
    Diesel fuel            Rat             4 h/d, 2 d/week, 13 weeks +         Reversible loss of body weight;     Lock et al. (1984)
                                           8 weeks' recovery (1.5 mg/litre)    no other apparent effect

    Marine diesel fuela    Mouse, rat      90 d, continuous (0.05 and          Nephrotoxic effects specific to     Bruner (1984)
                                           0.3 mg/litre)                       male rats

    Other fuels
    JP-8 jet fuelb         Mouse, rat      24 h/d, 7 d/week,90 d + 24          Nephrotoxic effects specific to     Mattie et al. (1991)
                                           months' recovery (1 mg/litre)       male rats

    JP-5 jet fuelb         Rat, dog        24 h/d, 7 d/week,90 d               Nephrotoxic effects specific to     Gaworski et al. (1984)
                                           (0.75 mg/litre)                     male rats

    JP-4 jet fuelb         Mouse, rat      6 h/d, 5 d/week,12 months           Nephrotoxic effects specific to     Bruner et al. (1993)
                                           (5 mg/litre)                        male rats; increased incidence
                                                                               of hepatocellular adenomas in
                                                                               female mice

    Deodorized             Rat, dog        6 h/d, 5 d/week,13 weeks            No apparent effects                 Carpenter et al. (1976)
    kerosenec                              (0.1 mg/litre)
                                                                                                                                                

    a  Hydrocarbon fraction similar to that of diesel fuel
    b  Narrow boiling range
    c  Not a fuel; very low aromaticity
             Groups of 25 male rats and four male dogs were exposed to
    deodorized kerosene at concentrations up to 0.1 mg/litre,
    approximately the saturation limit at 25°C, for 6 h/day on five days
    per week for about 13 weeks. Although dose-related changes were
    observed in survival, body weight gain, organ weights, and
    haematological and clinical chemical parameters, none of the findings
    was statistically significantly different from those in controls. No
    pathological change was seen in any of the organs examined (Carpenter
    et al., 1976).

         Groups of 12 male and 12 female Sprague-Dawley rats were exposed
    in whole-body chambers for 4 h/day on two days per week for 13 weeks
    to 0.25, 0.75, or 1.5 mg/litre aerosolized diesel fuel, followed by an
    eight-week recovery period. There were no deaths. All treated animals
    lost weight, indicating slight toxicity, during the first four weeks
    of exposure, but the effect was reversed during the recovery period in
    most cases. A slight depression of the startle reflex was seen at the
    start of the study, and the number of macrophages in bronchial lavage
    fluid was slightly elevated in all treated animals immediately after
    cessation of exposure. Further tests of pulmonary function, gas
    exchange, and dynamic lung function did not reveal any significant,
    dose-related change. Overall, no significant cumulative toxicity was
    seen (Lock et al., 1984).

         A specific effect of diesel fuel in male rats, the hyaline
    droplet nephropathy syndrome, was seen in several studies. Gaworski et
    al. (1984) studied the effects of a 90-day exposure to 0.15 or
    0.75 mg/litre JP-5 jet fuel derived from either petroleum or shale oil
    for 24 h/day, in groups of three male and three female beagle dogs, 75
    male and 75 female Fischer 344 rats, and 111-150 female C57Bl/6 mice.
    The toxicity of the two types of fuel was not significantly different.
    The only substantial effect at both concentrations was a renal tubular
    lesion in male rats, probably caused by hyaline droplet accumulation
    and unique to male rats. Other signs of toxicity were mild liver-cell
    changes and mild nasal inflammation in rats. Cowan & Jenkins (1980)
    studied petroleum and shale-oil marine diesel fuel using an identical
    study design. Reduced body weight gain and hyaline droplet nephropathy
    were reported in male rats. In another study, groups of 95 male and 75
    female Fischer 344 rats and 100 male and 100 female C57Bl/6 mice were
    exposed for 24 h/day on seven days per week to JP-8 jet fuel at
    concentrations up to 1 mg/litre for 90 days, and then held for 20-21
    months in order to assess the long-term consequences of exposure. The
    main finding was accumulation of hyaline droplets in the kidneys of
    male rats (Mattie et al., 1991). In a 90-day study, groups of 75 male
    and 75 female Fischer 344 rats were exposed continuously to 0.05 or
    0.3 mg/litre of marine diesel fuel. The main finding was again hyaline
    droplet nephropathy in males at both levels (Bruner, 1984). In a study

    of the carcinogenicity of JP-4 jet fuel vapour, hyaline droplet
    nephropathy was observed in groups of 100 male Fischer 344 rats
    exposed for 6 h/day, five days per week for 12 months to
    concentrations up to 5 mg/litre (Bruner et al., 1993) (see also
    section A7.7).

         The hyaline droplet nephropathy syndrome is not considered to be
    relevant for humans. It is linked to an inherent peculiarity of renal
    protein. The histopathological sequence is: accumulation in the renal
    proximal tubules of an excessive amount of hyaline droplets containing
    alpha2-microglobulin, a special protein of male rats which is
    synthesized in the liver; cytotoxicity and single-cell necrosis of the
    tubular epithelium; sustained regenerative tubule-cell proliferation;
    development of intraluminar granular casts from sloughed cell debris;
    tubular dilatation and papillary mineralization; foci of tubule
    hyperplasia; and renal tubule tumours (US Environmental Protection
    Agency, 1991). Mechanistically, a globulin-substance complex that is
    resistant to hydrolytic degradation is responsible for the protein
    overload in kidneys.

    A7.3  Long-term exposure

    A7.3.1  Dermal exposure

         Groups of 50 male and 50 female B6C3F1 mice received dermal
    administrations of 250 or 500 mg/kg body weight per day marine diesel
    fuel or JP-5 navy fuel on the clipped dorsal interscapular region on
    five days per week for 103 weeks. (The doses were derived from studies
    reported in section A7.2.) The high dose of marine diesel fuel caused
    a dose- and time-dependent reduction in mean body weight in animals of
    each sex, by up to 23% after week 40 in males and up to 20% after week
    72 in females. Survival of treated females was decreased, and exposure
    to the high dose was stopped at week 84 owing to severe ulceration.
    Application of 500 mg/kg body weight per day JP-5 navy fuel resulted
    in a time-dependent reduction in mean body weight in animals of each
    sex: by up to 22% after week 76 in males and up to 25% after week 60
    in females. Treatment of females with the high dose had to be
    terminated at week 90. Overall, JP-5 navy fuel caused a lesser chronic
    inflammatory response than marine diesel fuel (US National Toxicology
    Program, 1986).

    A7.3.2  Inhalation

         Groups of 100 male and 100 female Fischer 344 rats and C57Bl/6
    mice were exposed to JP-4 jet fuel vapour for 6 h/day on five days per
    week for 12 months at concentrations of 1 or 5 mg/litre (see section
    A7.7). The sex-specific alpha2-microglobulin nephropathy syndrome was
    observed in 21% of male controls and in 20% of males at the low dose,
    and 77% at the high dose (see also section A7.2). No significant,
    dose-related alteration was seen in clinical findings, mortality,

    haematological parameters, or clinical chemistry. All treated male
    rats had a significant reduction in mean body weight. At the high
    dose, male rats had significantly increased relative liver and kidney
    weights and females had a significant decrease in relative spleen
    weight. These alterations were not significant in comparison with
    absolute organ weights (Bruner et al., 1993).

    A7.4  Dermal and ocular irritation; dermal sensitization

    A7.4.1  Dermal irritation

         The results of studies of dermal irritation (Table 14) must be
    interpreted critically, as different protocols were used, e.g. in
    Europe and the United States. The main difference is in exposure time:
    24 h in the American protocols and 4 h in those of the European Union
    and the Organisation for Economic Co-operation and Development, which
    may pose a problem when only mild irritation is observed, leading to a
    classification of 'non-irritant' according to protocols of the latter
    organizations.

         A market-place sample of diesel fuel, tested at a dose of 0.5 ml
    for 24 h in rabbits, was assessed as 'extremely irritating',
    corresponding to a Draize score of 6.81 (Beck et al., 1984).

         Topical applications three times a week to mice of various types
    of kerosene and gas oil with boiling ranges similar to that of diesel
    fuel produced degenerative changes in the skin surface, including
    necrosis and hyperplasia (CONCAWE, 1991).

         Four types of American military fuels were investigated:
    petroleum JP-5 fuel, shale JP-5 fuel, petroleum marine diesel fuel,
    and shale marine diesel fuel. None was active in the Draize test, in
    which six female New Zealand white rabbits received 0.5 ml of
    undiluted fuel on intact and abraded skin under an occlusive covering
    for 24 h (Cowan & Jenkins, 1980). Applications to rabbit skin of
    0.5 ml of various shale oil- and petroleum-derived fuels under
    occlusive dressings for 4 h were not irritating (Schultz et al.,
    1981).

         Various modifications of fuel oil No. 2 (10, 30, and 50%) were
    found to be moderately irritating to rabbit skin (American Petroleum
    Institute, 1980a,b,c), and two hydrodesulfurized middle distillates
    resulted in primary irritation indices of 4.3 and 5.9 (American
    Petroleum Institute, 1982a,b). A straight-run middle distillate had a
    moderately irritating effect, with an irritation index of 3.2
    (American Petroleum Institute, 1985d). Jet fuel A was classified as
    mildly irritating, with an irritation score of 1.96 (American
    Petroleum Institute, 1980d). A hydrosulfurized kerosene was moderately
    irritating, with an index of 4 (American Petroleum Institute, 1982c). 

    A light catalytically cracked distillate (American Petroleum
    Institute, 1985d) and a straight-run kerosene (American Petroleum
    Institute, 1985e) had severely irritating effects, with irritation
    indexes of 5.5 and 5.6.

        Table 14.  Studies of dermal irritation in rabbits exposed for 24 h to various fuels
                                                                                             

    Fuel                              Classification of         Reference
                                      irritationa
                                                                                             

    Diesel fuel                       Extremely irritating      Beck et al.
                                                                (1984)

    Marine diesel fuel or marine      Not irritating            Cowan & Jenkins
      shale fuel                                                (1980)

    Hydrodesulfurized middle          Moderately irritating     American Petroleum
      distillate                                                Institute (1982a,b)

    Straight-run middle distillate    Moderately irritating     American Petroleum
                                                                Institute (1985d)

    Light catalytically cracked       Severely irritating       American Petroleum
      distillate                                                Institute (1985d)

    Heating oil No. 2                 Moderately irritating     American Petroleum
                                                                Institute (1980a,b,c)

    Jet fuel A                        Mildly irritating         American Petroleum
                                                                Institute (1980d)

    Various kerosenes                 Slightly to severely      CONCAWE (1995)
                                      irritating

    Shale oil- and                    Not irritating            Schultz et al. (1981)
      petroleum-derived fuels

    Hydrodesulfurized kerosene        Moderately irritating     American Petroleum
                                                                Institute (1982c)

    Straight-run kerosene             Severely irritating       American Petroleum
                                                                Institute (1985e)
                                                                                             

    a   Note differences in protocols between regulations of the United States and
        the Organisation for Economic Co-operation and Development (see section A7.4.1)
    
    A7.4.2  Ocular irritation

         All of the available reports indicate no or, at the most, a
    slight irritating effect of diesel fuels and kerosenes on the rabbit
    eye.

         Four types of American military fuels (petroleum JP-5 fuel, shale
    JP-5 fuel, petroleum marine diesel fuel, and shale marine diesel fuel)
    gave generally negative responses in the Draize test, in which nine
    female New Zealand white rabbits were given 0.1 ml of undiluted
    substance (Cowan & Jenkins, 1980). A market-place sample of diesel
    fuel (0.1 ml, undiluted) was also not irritating in the Draize test
    (Beck et al., 1984). None of several kerosenes tested was more than
    slightly irritating to the eye (CONCAWE, 1995), and none of the shale
    oil- and petroleum-derived fuels tested caused visible irritation in
    the rabbit eye (0.1 ml, undiluted) (Schultz et al., 1981).

         Three modifications of fuel oil No. 2 (10, 30, and 50%) were
    classified as mildly irritating in rabbits (American Petroleum
    Institute, 1980a,b,c). Two hydrodesulfurized middle distillates did
    not cause primary irritation in the Draize test (American Petroleum
    Institute, 1982a,b); a straight-run middle distillate (American
    Petroleum Institute, 1985d) and a straight-run kerosene also gave
    negative results (American Petroleum Institute, 1985e). Jet fuel A was
    classified as mildly irritating, with an average score of 2.67 after
    one day and 0 after seven days (American Petroleum Institute, 1980d).
    A hydrodesulfurized kerosene was not irritating (American Petroleum
    Institute, 1982c).

    A7.4.3  Sensitization

         Three modifications of fuel oil No. 2 (10, 30, and 50%) were were
    classified as non-sensitizing for the skin of guinea-pigs (American
    Petroleum Institute, 1980a,b,c). A straight-run middle distillate
    (American Petroleum Institute, 1985d), a straight-run kerosene
    (American Petroleum Institute, 1985e), jet fuel A (American Petroleum
    Institute, 1980d), a hydrodesulfurized middle distillate (American
    Petroleum Institute, 1984), and a light catalytically cracked
    distillate (American Petroleum Institute, 1985c) were also not
    sensitizing.

         Four types of fuels used in the United States Army, petroleum
    JP-5 fuel, shale JP-5c fuel, petroleum marine diesel fuel, and shale
    marine diesel fuel, were not sensitizing in a modified Landsteiner
    test in groups of 24 guinea-pigs (Cowan & Jenkins, 1980). A
    straight-run kerosene, hydrosulfurized kerosene, jet fuel A (CONCAWE,
    1995), and shale oil- and petroleum-derived jet and diesel fuels
    (Schultz et al., 1981) were also not sensitizing.

    A7.5  Reproductive toxicity, embryotoxicity, and teratogenicity

         Groups of 20 pregnant CRL:COBS rats were exposed by inhalation on
    days 6-15 of gestation to 100 or 400 ppm of diesel fuel for 6 h/day
    (American Petroleum Institute, 1979). A significant reduction in
    maternal food consumption was seen at the high concentration, and
    maternal body weight gain and fetal weights were reduced, although not
    statistically significantly. Overall, no significant, dose-related
    effects were observed in dams, and the compound had no teratogenic or
    embryotoxic effect and no effect on fetal growth or development.

         No signs of developmental or maternal toxicity were seen when
    groups of 20 pregnant rats were exposed to jet fuel A or fuel oil for
    6 h/day on days 6-15 of gestation (Beliles & Mecler, 1982). 
    Sprague-Dawley rats were fed 500, 1000, 1500, or 2000 mg/kg body
    weight per day of JP-8 aviation fuel by gavage on days 6-15 of
    gestation, and fetuses were examined on day 20. Significantly reduced
    body weights were seen in dams at the two higher doses and in fetuses
    at 1000-2000 mg/kg body weight per day, but the number of
    malformations was not altered. The investigators concluded that JP-8
    aviation fuel is a developmental toxin at doses > 1000 mg/kg body
    weight per day but is not teratogenic (Cooper & Mattie, 1993).

    A7.6  Mutagenicity and related end-points

    A7.6.1  In vitro

         Various samples of diesel and related fuels have been tested for
    mutagenicity in Salmonella typhimurium (Henderson et al., 1981;
    Conaway et al., 1984; Vandermeulen et al., 1985; US National
    Toxicology Program, 1986; McKee et al., 1994). Negative results were
    obtained in direct, unmodified Ames' assays (Conaway et al., 1984;
    Vandermeulen et al., 1985; US National Toxicology Program, 1986). In
    one study, refined shale oil, jet fuel, and shale oil- and
    petroleum-derived marine diesel fuel were not mutagenic (Schultz et
    al., 1981). Negative results were also found in studies of aliphatic
    and aromatic fractions of diesel fuel in  S. typhimurium TA98, TA100,
    and TA1535 in the presence of an exogenous metabolic system (Henderson
    et al., 1981).

         The number of revertants of TA98 increased to 2.3-2.5 times the
    background levels in a suspension assay (Conaway et al., 1984), but
    increases were found at a limited number of doses. In a modified
     Salmonella assay with an extraction step to concentrate the
    mutagenic aromatic compounds, addition of an exogenous metabolic
    system from Aroclor-induced hamster liver, and use of TA98 as the most
    responsive strain, increased numbers of revertants were seen (McKee et
    al., 1994); but the number never exceeded twice the background
    response, and the results were considered to be equivocal.

         A dimethyl sulfoxide extract of an intermediate, catalytically
    cracked distillate (CAS No. 644741-60-2) was mutagenic in
     S. typhimurium TA98 with an increased concentration of exogenous
    metabolic system to optimize detection. Of two dimethyl sulfoxide
    extracts of light hydrocracked distillate (CAS-No. 64741-77-7) tested
    in the same system, one induced a weak, non-dose-dependent response,
    while the other induced a clear dose-dependent positive response
    (Blackburn et al., 1984). Under the same assay conditions, five
    samples of cracked middle distillates also gave positive results
    (German Scientific Association for Petroleum, Natural Gas, and Coal,
    1991). These results in a series of assays in Salmonella with a number
    of variations provide no clear evidence of mutagenic activity.  In the
    unicellular alga  Chlamydomonas reinhardtii (streptomycin resistant)
    tested in a natural aquatic environment, water-soluble fractions of
    two crude and two refined oils did not induce forward mutations
    (Vandermeulen & Lee, 1986).  Neither aliphatic nor aromatic dimethyl
    sulfoxide fractions of diesel fuel No. 7911 induced mutation in
     S. typhimurium. After reaction with nitrogen dioxide, which may
    activate the mutagenicity of diesel exhaust, it was mutagenic in
    strain TA100, the aromatic fraction being 40-fold more active than the
    aliphatic fraction (Henderson et al., 1981).

         A market-place sample of diesel fuel did not induce mutation in
    L5178Y tk+/- mouse lymphoma cells with or without metabolic
    activation (Conaway et al., 1984). A hydrodesulfurized middle
    distillate (CAS 64742-80-9) was active in the presence of metabolic
    activation, but the concentrations were cytotoxic (American Petroleum
    Institute, 1987c).

    A7.6.2  In vivo

         Undiluted diesel fuel No. 2 increased the frequency of
    chromosomal aberrations in the bone marrow of Sprague-Dawley rats
    given single or repeated (five days; 125, 417, or 1250 mg/kg body
    weight) oral doses. Diesel fuel given intraperitoneally for one or
    five days at 0.6, 2, or 6ml/kg body weight significantly increased the
    number of aberrant cells at the highest dose (Conaway et al., 1984).
    In order to assess these findings further, McKee et al. (1994)
    evaluated micronucleus induction by diesel fuel No. 2 and related
    materials in male and female CD-1 mice treated by gavage at up to
    5 g/kg body weight. No effect was seen.

         No effect was seen on the frequency of dominant lethal mutations
    in male CD-1 mice exposed to 100 or 400 ppm of either diesel fuel or
    jet fuel A for 6 h/day, five days per week for eight weeks (American
    Petroleum Institute, 1980e, 1981).

    A7.7  Carcinogenicity

    A7.7.1  Dermal exposure

         Diesel fuel has been shown to have at least weak carcinogenic
    potency after dermal administration in a number of studies. As some of
    the test samples contained low concentrations of carcinogenic PAHs, it
    has been suggested (Biles et al., 1988) that the tumours induced may
    have been a consequence of the dermal damage produced by repeated
    treatment with these materials. The available studies are summarized
    in Table 15.

         Doses of 25 mg per animal of petroleum and shale fuels were
    applied to groups of 25 male and 25 female C3H/HeN mice three times a
    week, for 62 weeks for jet fuel A with subsequent observation up to
    105 weeks, and for 104 weeks for JP-4 fuel. Controls received mineral
    oil. The shale and petroleum diesel fuels were characterized for
    boiling range (jet A, 171-271°C; JP-4, 60-249°C) and sulfur content
    (jet A: petroleum, 3510 ppm; shale, 89 ppm; JP-4: petroleum, 1830 ppm;
    shale, 199 ppm). After 60 weeks, increased incidences of squamous-cell
    carcinomas and fibrosarcomas were reported (jet A: petroleum, 26%;
    shale, 28%; JP-4: petroleum, 26%; shale, 50%; controls, 0-2%), but no
    statistical analysis was reported (Clark et al., 1988).

         Groups of 50 male and 50 female B6C3F1 mice were given 250 or
    500 mg/kg body weight per day marine diesel fuel dermally, on five
    days per week for 103 or 84 weeks (see section A7.3.). A control group
    received the vehicle, acetone, only. The survival rates were 26/50
    males and 29/50 females at the high dose at 84 weeks, 20/49 males and
    12/50 females at the low dose at 104 weeks, and 30/50 male and 40/50
    female controls. Males at the high dose had a significantly increased
    number of hepatocellular adenomas or carcinomas (29%; control, 18%).
    The total incidences of squamous-cell papillomas and carcinomas at the
    application site and the adjacent inguinal site were: control, 1/50;
    low dose, 2/49; high dose, 3/50 in males; and control, 0/50; low dose,
    1/45; and high dose, 2/48 in females. No information was available on
    historical frequencies in acetone-treated mice; the incidence of
    papillomas and carcinomas in 3500 observations in untreated mice was
    0.3-0.4%. In a parallel study with JP-5 navy fuel, 1/50 papillomas
    were seen at the low dose and 4/49 carcinomas at the high dose in
    males, and 0/49 carcinomas were seen at the low dose and 1/47 at the
    high dose in females. The investigators concluded that there was
    equivocal evidence for the carcinogenicity of marine diesel fuel and
    no evidence for the carcinogenicity of JP-5 navy fuel (US National
    Toxicology Program, 1986).

         Two middle-distillate fractions of each of two crude oils
    (boiling range, 170-370°C) were painted at a dose of 50 mg onto the
    skin of male C3H mice twice a week for 18 months. The yield of benign
    and malignant tumours (details not specified) was 2-19 % (Lewis,
    1983).


        Table 15.  Studies of chronic dermal exposure of mice (usually 50 mice/group) to distillate fuels
                                                                                                                                                

    Fuel                      Exposure                      Results                                                    Reference
                                                                                                                                                

    Marine diesel fuel        0, 250, and 500 mg/kg bw,     1/50, 2/49, and 3/50 (m) and 0/50, 1/45, and 2/48 (f)      US National Toxicology
                              5 d/week, 103 weeks           with squamous-cell papillomas and carcinomas after         Program (1986)
                                                            104 weeks

    Straight-run gas oil      25 mg/animal, twice/week,     Occasional ulcerations, scarring, epidermal dysplasia;     American Petroleum
    (laboratory sample)a      up to 2 years                 after 2 years, 17/50 with tumours; controls, 0%            Institute (1985c)

    Straight-run middle       50 µl/animal, twice/week,     Mild to moderate irritation (mild desquamation,            American Petroleum
    distillatea               up to 2 years                 occasional scabbing); after 2 years, 10/50 with            Institute (1989a)
                                                            squamous-cell carcinomas or fibrosarcomas;
                                                            controls, 0%

    Straight-run              50 µl/animal, twice/week,     29/50 with malignant and 1/50 with benign tumours;         American Petroleum
    keroseneb                 > 2 years                     latency, 76 weeks; controls, 0%                            Institute (1989a)

    Two cracked               50 µl/animal, twice/week,     Mild to moderate irritation; after 24 months, 39/50        American Petroleum
    distillatesc              up to 2 years                 and 26/50 with tumours                                     Institute (1989a)

    Intermediate              50 mg/animal, twice/week,     42/43 with tumours; mean latency, 17 weeks;                Blackburn et al. (1984)
    catalytically cracked     80 weeks                      controls, 0/50                                             
    distillatec

    Two hydrodesulfurized     50 µl/animal, twice/week,     Acanthosis, fibrosis, hyperkeratosis, crusting; at end     American Petroleum
    middle distillatesa       lifetime                      of study, 24/50 and 27/50 with tumours (no controls        Institute (1989b)
                                                            reported)
                                                                                                                                                

    Table 15 (contd)
                                                                                                                                                

    Fuel                      Exposure                      Results                                                    Reference
                                                                                                                                                

    Hydrodesulfurized         50 µl/animal, twice/week,     26 malignant and 1 benign tumours; latency, 77             American Petroleum
    keroseneb                 > 2 years                     weeks (no controls reported)                               Institute (1989b)

    Jet A and JP-4 fuels      25 mg/animal, three           After 60 weeks, increased skin tumour rate; jet A,         Clark et al. (1988)
    from shale or             times/ week, 62 weeks         26% petroleum, 28% shale; JP-4, 26% petroleum,
    petroleum oilsb           + observation (jet A)         50% shale
                              or 104 weeks (JP-4)

    Two middle-distillate     50 mg/animal, twice/week,     2-19% tumour rate (no further information)                 Lewis (1983)
    fractions                 18 months

    No. 2 fuel oilb           12, 25, or 50 µl/animal,      Overall tumour rate, 10%; control, 0%                      Witschi et al. (1987)
                              three times/week, up to
                              100 weeks

    Two heating oilsa         25 µl/animal, three           7/50 and 9/50 with tumours; control, 0/50                  Biles et al. (1988)
                              times/week, lifetime
                                                                                                                                                

    a  Hydrocarbon fractions similar to that of diesel fuel (see section A2.1)
    b  Hydrocarbon fractions from which also diesel fuels are gained during petroleum separation (see section A3.2.1.1)
    c  Blending component of diesel fuel
             Application of American Petroleum Institute No. 2 fuel oil
    dermally to C3H mice at 12, 25, or 50 µl per animal, three times
    weekly for up to 100 weeks induced an overall tumour incidence (type
    not specified) of 10% (15/150 in the three treatment groups; controls,
    0/100) (Witschi et al., 1987).

         Ten samples of middle-distillate fuels were applied dermally to
    C3H/HeJ mice at a dose of 25 µl, three times a week for life. A virgin
    heating oil blending base (142-307°C) and a commercial No. 2 heating
    oil (156-352°C) had aromatic contents closest to that of diesel fuel.
    The blending base induced five skin carcinomas and two papillomas in
    50 animals, and the commercial heating oil produced eight skin
    carcinomas and one papilloma in 50 animals. There were no tumours in
    the respective control groups (Biles et al., 1988).

    A7.7.2  Inhalation

         Fischer 344 rats and C57Bl/6 mice were exposed for 6 h/day on
    five days per week for 12 months to 1 or 5 mg/litre of fuel vapours
    (see section A7.3), followed by a 12-month recovery period. No
    significant pulmonary neoplastic changes were seen. Female mice at the
    high dose had a slightly but significantly increased incidence of
    benign hepatocellular adenomas (10%; controls, 2%); however, the trend
    was reversed in male mice. No conclusions about carcinogenic potency
    were drawn from this study (Bruner et al., 1993).

    A8.  EFFECTS ON HUMANS

    A8.1  Exposure of the general population

         Like similar petroleum distillates, such as kerosene, diesel fuel
    can cause dermal irritation and defatting after dermal exposure
    (Sandmeyer, 1981). Considerable exposure and subsequent absorption may
    also result in acute renal failure secondary to acute renal tubular
    necrosis. A 28-year-old man developed progressive oliguria one day
    after washing his hair with diesel fuel and had acute tubular necrosis
    on renal biopsy (Barrientos et al., 1977). A 47-year-old man developed
    acute tubular necrosis after cleaning his arms and hands with diesel
    oil for several weeks (Crisp et al., 1979). Both recovered renal
    function.

         Accidental aspiration of diesel fuel resulted in immediate
    coughing, followed by dyspnoea, cyanosis, and loss of consciousness
    (Perez Rodriguez et al., 1976). X-Ray examination revealed diffuse
    alveolar infiltrates. Mildly elevated levels of alkaline phosphatase,
    serum glutamic-oxaloacetic transaminase, and serum glutamic-pyruvic
    transaminase were also seen. After 37 days, residual infiltrates
    remained, but the patient had improved. In another case, abdominal
    pain and vomiting followed ingestion of a reported 1.5 litres of
    diesel fuel. Pulmonary infiltrates were noted, presumably from
    subsequent aspiration of diesel fuel (Boudet et al., 1983).

    A8.2  Occupational exposure

         The odour threshold values for diesel fuels No. 1 and 4 are
    reported to be 0.7 and 0.5 ppm, respectively (see section A2.2, Table
    5). Six volunteers exposed to 140 mg/m3 of deodorized kerosene for
    15 min did not experience throat irritation. Olfactory fatigue was
    induced in three subjects, and one reported taste sensation. The
    investigators suggested that this concentration was the sensory
    threshold for kerosene and estimated the odour threshold at
    0.6 mg/m3 (Carpenter et al., 1976).

         Acne and folliculitis developed on the arms and thighs of a man
    who had worked in various automobile workshops for 15 years. He had
    handled diesel oil with his bare hands and then wiped them on his
    trousers (Das & Misra, 1988). Hyperkeratosis was noted in 320 drivers
    working in Russian oil fields with regular dermal contact with diesel
    fuel (Gusein-Zade, 1974). A 33-year-old man was exposed to diesel fuel
    vapour (concentration not reported) for 10 days while driving a truck
    with a fuel injector leak. He developed abdominal cramps, nausea,
    vomiting, and then acute renal failure with anaemia and
    thrombocytopenia (Reidenberg et al., 1964). Two aviators who were
    exposed to JP-5 fuel vapour for 1 h while flying a small airplane
    experienced eye irritation, difficulties in coordination and
    concentration, and fatigue (Porter, 1990).

         Siemiatycki et al. (1987) conducted a case-control study in
    Montreal, Canada, and found an increased risk for squamous-cell
    carcinoma of the lung in men exposed to diesel fuel. When all types of
    lung cancer except adenocarcinoma were combined, an odds ratio
    (adjusted for smoking) of 1.6 (90% confidence interval (CI), 1.0-2.4;
     n = 39) was found for 'any' exposure. In four subcategories, the odds
    ratios were 1.9 (short-low), 2.0 (short-high), 1.1 (long-low), and 2.0
    (long-high). The possible effects of exposure to combustion products
    were not taken into consideration. An association with prostatic
    cancer was also observed, although there was no dose-response
    relationship, with odds ratios of 2.3 (CI, 1.3-4.0;  n = 17) for the
    subgroup with nonsubstantial exposure and 1.4 (0.7-2.9;  n = 8) in
    the subgroup with substantial exposure.

         In another case-control study, no association was found between
    the occurrence of renal-cell cancer and occupational exposure to fuel
    oils, including diesel fuel (levels and duration not quantified)
    (Partanen et al., 1991).

         In a cross-sectional epidemiological study, the effects of
    long-term exposure of factory workers to jet fuels (not specified) was
    investigated. Dizziness, headache, nausea, palpitation, pressure in
    the chest, and eye irritation were found to be more prevalent than in
    unexposed controls. The time-weighted average concentration of jet
    fuel vapour in the breathing zone was estimated to be 128-423 mg/m3
    (Knave et al., 1978).

    A9.  EFFECTS ON OTHER ORGANISMS IN THE LABORATORY AND THE FIELD

         Comparison and evaluation of studies of the ecotoxicological
    effects of diesel fuel are complicated by a number of factors:

         (1) There are comparatively few ecotoxicological studies on
    diesel fuel, and most work in the 1970s was concentrated on the
    effects of No.2 fuel oil, which has a similar but not identical
    chemical composition to diesel fuel No. 2 (see section A2). The
    composition of marine diesel fuel differs greatly from both (see
    Table 1). As described in the same section, diesel fuel itself has
    various specifications in different countries; in particular, the
    composition of non-hydrocarbon compounds from e.g. catalytic processes
    or additives may vary widely. There may also be differences between
    batches of the same oil (Hedtke & Puglisi, 1982) (see Table 20). The
    amount of the chemical actually in solution is also pertinent to
    ecotoxicological studies. Table 3 gives the concentration ranges of
    toxicologically relevant constituents of diesel and No. 2 fuel oil as
    a whole and as 10% water-soluble fractions.

         (2) The chemical composition of fuel oils and, as a consequence,
    their toxicity change with time, as oil can be modified by
    biodegradation, photooxidation, and volatilization (see section A4).
    Their toxicity also depends on whether they are mixed with water (sea
    water or freshwater), sediment, or soil, the temperature (tropical or
    Arctic), and physical conditions (e.g. storm conditions with continual
    mixing and breaking up of oil). Oil modified by these processes is
    usually less toxic than fresh oil (Baker, 1970; Morrow, 1973).

         (3) A variety of methods have been used in ecotoxicological
    studies on fuel oils. There is an obvious difference between
    conditions in the laboratory and the field. Most of the field studies
    that have been reported are related to fuel spills. In laboratory
    studies, either water-soluble fractions (see below) or oil-water
    dispersions have been used, and the methods of preparing them vary
    among laboratories. Oil films and, more recently, microencapsulated
    diesel oil have also been tested. In none of the laboratory studies
    reported were test media prepared with 'water accommodated fractions',
    as described and recommended by CONCAWE (1993) in laboratory protocols
    for petroleum products. Thus, the amount of product that must be added
    to a given volume of medium to produce the effects reported cannot be
    derived from the results of these studies, making comparison of data
    on ecotoxicity from different studies practically impossible.

         Hedtke & Puglisi (1982) tested the lethality of No. 2 fuel oil in
    two species of freshwater fish and two species of amphibians and found
    variable LC50 values (see Table 20). Oil-water dispersions were very
    toxic to fish. It is thought that these tiny oil droplets make contact
    with the gills and directly damage the respiratory epithelium
    (Engelhardt et al., 1981; Poirier et al., 1986). With floating oils,

    the relationship between toxicity and oil-water contact is a function
    of the equilibrium time for the soluble components of the oil and
    water, death occurring more rapidly when the oil is added 96 h before
    the organisms are exposed. Flow-through tests were more sensitive than
    static tests, as, in the latter, the chemical nature and toxicity of
    the oil change with time (Hedtke & Puglisi, 1982). Lockhart et al.
    (1987) showed that the results of static tests in fish varied
    considerably with aeration of test containers. Although flow-through
    techniques give a better estimate of real exposure conditions, many of
    the results reported in the literature are of static tests.

         (4) The susceptibility of organisms to fuel oils and their
    components depends not only on the species and strain but also on the
    biological stage and the time of development within stages (Kühnhold,
    1977; Winters et al., 1977). As the effects of No. 2 fuel oil and
    diesel oil have been tested in a wide variety of organisms and life
    stages under various conditions, direct comparison is difficult.

         In spite of these difficulties, it can be concluded that diesel
    oil is generally more toxic than crude oil, as seen, for example, in
    barnacle nauplii (Winters et al., 1977),  Daphnia (Cladocera),
    chironomid larvae, and the mollusc  Viviparus bengalensis (Gastropoda)
    (Das & Konar, 1988). Crude oils themselves differ in toxicity,
    depending on their source (Anderson, 1977a; Winters et al., 1977), as
    do their corresponding products.

         Rice et al. (1977a) concluded that the toxicity of fuel oil is
    due to the chemical toxicity of soluble aromatic compounds rather than
    to the physical toxicity of dispersed oil droplets. Physical coating
    by oil of the gills of fish and the feathers of birds is also a cause
    of toxicity. Monoaromatic compounds seem to be the least toxic, their
    acute toxicity increasing with molecular size up to the four- to
    five-ring compounds (e.g. chrysene and benzo[ a]pyrene) (Neff et al.,
    1976b; Rice et al., 1977a), but they are not very soluble in sea water
    (see Table 6). The toxicity of one-, two-, and three-ring compounds
    (benzenes, naphthalenes, and phenanthrenes, respectively), seemed to
    increase with increasing degrees of alkylation of the aromatic
    nucleus. Neff et al. (1976b) reported 96-h LC50 values of 0.3-0.6 ppm
    for 1-methylphenanthrene, fluoranthrene, and phenanthrene, indicating
    the high toxicity of these polynuclear aromatic compounds for the
    polychaete,  Neanthes arenaceodentata. Of the pure compounds tested
    by Winters et al. (1977), the most toxic to barnacle nauplii were
    indan, naphthalene, xylene, and substituted benzenes and naphthalenes.
    In four freshwater species, the water-soluble fraction was again
    associated with the substituted benzenes, and naphthalenes were found
    to be associated with the observed toxicity (Lockhart et al., 1987).
    The concentrations of naphthalenes, methylnaphthalenes, and
    dimethylnaphthalenes are particularly high in fuel oils (see Table 3),
    and these compounds have been used as analytical markers in several
    studies.

         Experiments involving microencapsulated oil fractions have shown
    that insoluble fractions also inhibit growth, e.g. of juvenile mussels
    (Stromgren & Nielsen, 1991). Anderson (1979) indicated that the larval
    stage might be the most sensitive to No. 2 fuel oil. Eggs of fish and
    invertebrates were often tolerant, and adults were sometimes more
    sensitive than juveniles.

         Extensive studies have been carried out on lethal toxicity to
    marine organisms under the same conditions. The 96-h LC50 values for
    a wide range of invertebrates in static exposure conditions
    (temperature, 18-22°C) were 1-20 ppm for crude oil in water and
    0.4-6 ppm for No. 2 fuel oil (Anderson et al., 1974; Neff et al.,
    1976b; Rossi et al., 1976). At 4-10°C, the 96-h LC50 values appeared
    to be lower (Rice et al., 1976, 1977a). Table 16 gives the results of
    lethality studies reported from several laboratories.

    Table 16.  Lethality (median lethal concentrations) to marine
               zooplankton of physically dispersed and water-soluble
               fractions of No. 2 fuel oil
                                                                        

    Zooplankton            Oil-water dispersions   Water-soluble
                           (mg/litre)              fractions (mg/litre)
                                                                        

    Ctenophora                                     0.59 (1 day)
    Mollusca
     Clams
      Embryos                                      0.43 (2 days)
      Larvae                                       1.3 (2 days)
      Larvae                                       0.53 (10 days)
     Pteropods             < 0.2 (2 days)
    Crustacea
     Barnacles                                     2.6 (1-h LC50)
     Copepods                                      1.0 (3 days)
     Amphipods             0.3 (2 days)            2.5 (3 days)
     Decapods
      Shrimp post-larvae   1.7; 9.4
      Lobsters                                     1.2-6.6
    Teleosts                                       1.5
                                                                        

    Adapted from US National Research Council (1985). Values are for total
    measured (extractable) hydrocarbons, calculated from initial
    concentrations  measured by spectroscopy and chromatography

         No. 2 fuel oil (mostly the water-soluble fraction) has been found
    to have sublethal effects in all phyla examined (Anderson, 1977a,b,
    1979; US National Research Council, 1985). With increasing
    concentrations, alterations were seen in the respiratory rate of the
    fish  Cyprinodon variegatus (Anderson et al., 1974) and  Oncorhynchusm
     gorbuscha (pink salmon) (Rice et al., 1977b), the crustaceans
     Mysidopsis almyra and  Penaeus aztecus, and the glass shrimp
     Palaemoetes pugio (Anderson, 1979); and changes were found in the
    pulse rate in embryonic estuarine fish,  C. variegatus and  Fundulus
     heteroclitus (Anderson, 1977a) These effects have been used as a
    measure of stress caused by exposure. Dysregulation of the ability to
    regulate internal concentrations of solutes (osmotic and ionic
    regulation) was demonstrated in the brown shrimp  Penaeus aztecus
    (Anderson et al., 1974; Cox, 1974; Anderson, 1979). The behaviour of
     F. similus exposed to a 100% water-soluble fraction of No. 2 fuel
    oil was correlated with the content of naphthalene in four organs
    (Dixit & Anderson, 1977). Sublethal studies on No. 2. fuel oil showed
    reduced growth rates, survival, and reproduction in a range of
    organisms (Anderson, 1979; Table 17), including the annelid Neanthes
    arenaceodentata (Rossi & Anderson, 1978) and the estuarine fish
     F. grandis (Ernst et al., 1977).

    A9.1  Laboratory experiments

         Diesel fuel is a complex mixture of substances with a range of
    solubilities in water (see section A2.2). In testing for aquatic
    toxicity, it is important that the constituents in the medium be
    characterized (Bennett et al., 1990), but these are seldom reported.
    The results described in this section refer to diesel fuel, unless
    otherwise stated.

         As the solubility of diesel fuel in sea and freshwater is low,
    investigations of the effects of diesel fuel on organisms in the
    laboratory often involve preparation of a water-soluble fraction or an
    oil-water dispersion. Water-soluble fractions can be prepared by
    stirring one part of diesel fuel into nine parts of (sea) water for
    20 h (Anderson et al., 1974) or one part of diesel fuel into eight
    parts of sea water for 24 h (Pulich et al., 1974), and then extracting
    the water phase. The resulting water phases contain 6-20 µg/ml of
    total hydrocarbons; the 100% water-soluble fraction of No. 2 fuel oil
    contains about 7 µg/ml of total hydrocarbons. Results are expressed as
    the concentration (percent dilution of the water-soluble fraction)
    and/or as the measured hydrocarbon concentration that causes the
    specified level of effect. The results cannot, however, be correlated
    with the original amount of fuel oil that was added to produce the
    water-soluble fraction or oil-water dispersion (CONCAWE, 1993).


        Table 17.  Effects of No. 2 fuel oil on the growth and reproduction of marine animals
                                                                                                                                     

    Species                    Exposure    Concentration (ppm)       Growth or reproduction          Reference
                               (days)                        
                                           TH      TN      TA
                                                                                                                                     

    Fish
    Cyprinodon variegatus      7           2.0     0.6     1.7       0% of eggs hatched              Anderson et al. (1976)

    Decapods
    Rithropanopeus harrasii    27          1.0     0.3     0.9       Reduced survival and            Neff et al. (1976b)
                                                                     extended development to
                                                                     megalopa
    Palaemonetes pugio         12          0.9     0.3     0.8       Reduced growth rate of          Tatem (1977)
                                                                     larvae
                               3           1.4     0.6               Reduced viability of eggs
                                                                     from exposed gravid females

    Polychaetes
    Neanthes arenaceodentata   22          1.0     0.3     0.9       Reduced growth of larvae        Rossi & Anderson (1976)
                               28          0.3     0.1               Reduced growth of               Anderson (1977a)
                                                                     juveniles by 30%
    Ctenodrilus serratus       28          2.2     0.5     1.4       Reduced survival and            Carr & Reish (1977)
                                                                     reproduction
    Ophryotrocha sp.           28          1.3     0.3     0.9       Reduced survival and            Carr & Reish (1977)
                                                                     reproduction
                                                                                                                                     

    From Anderson (1977b)
    TH, total hydrocarbons; TN, total naphthalenes; TA, total aromatic compounds
             Oil-water dispersions are produced by shaking a measured quantity
    of oil in water for 5 min (Anderson et al., 1974) or 30 min (Pulich et
    al., 1974). In some cases, the oil-water dispersion is maintained by
    continuous mixing, but more usually it is allowed to settle, and the
    aquatic organisms are exposed to the material in the water and to a
    surface film. The effects seen in such tests can be due to both
    physical action and toxicity, especially in small invertebrates such
    as  Daphnia.

         Insoluble fractions of crude oil and other hydrocarbon
    formulations strongly inhibited growth when they were
    microencapsulated and ingested by juvenile mussels (Stromgren et al.,
    1986; Stromgren & Reiersen, 1988) and larvae (Stromgren & Nielsen,
    1991). In microcapsules with a thin acacia-gelatine structure, the
    total oil is dispersed into particles of 1-10 µm, which are suitable
    for ingestion by  Mytilus larvae. The water-soluble components leak
    through the wall, providing a low concentration, and are ingested by
    the molluscs.

         None of the studies involved use of the 'water accommodated
    fractions' recommended by CONCAWE (1993), which allows the results to
    be expressed as 'loading rate', defined as the amount of product that
    must be equilibrated with the aqueous test medium in order to produce
    a specified level of effect.

    A9.1.1  Microorganisms

    A9.1.1.1  Water

         Marcus & Scott (1989) tested the effects of 0, 10, 20, and 40%
    water-soluble fractions of a shale-diesel fuel mixture and a
    petroleum-diesel fuel mixture on the growth of faecal coliform
    bacteria. The 20 and 40% fractions of the petroleum-diesel fuel
    mixture resulted in significantly lower bacterial densities than the
    shale-diesel fuel mixture. The latter appeared to biostimulate or
    mediate against toxicity, suggesting a difference in the chemical
    characteristics of the two fuels.

         The effects of diesel and furnace oils, petrol, kerosene, and
    Assam crude oil and their paraffinic, aromatic, and asphaltic
    fractions (higher relative molecular mass, poorly defined) on the
    photosynthesis and respiration of blue-green algae  (Cyanobacteria), in
    particular Anabaena doliolum were investigated by Singh & Gaur (1990).
    The chemical composition of the diesel and furnace oils was almost
    identical, with 37 and 36% paraffins, 51 and 54% aromatic compounds,
    and 6 and 0.6% asphaltic compounds, respectively. The aromatic
    fractions of these oils were the most toxic, followed by the asphaltic
    fractions; photosynthesis was reduced to 40% by the aromatic fractions
    and to 50% by the asphaltic fractions of both oils at a concentration

    of 5.0 mg/litre. The effects were dependent on concentration up to
    10 mg/litre. Photosynthesis was stimulated 160% by 5 mg/litre of the
    paraffinic fractions of crude oil , but only 100% by those of diesel
    oil and 110% by those of furnace oil. In general, crude oil inhibited
    photosynthesis and respiration by  A. doliolum least and diesel and
    furnace oils the most.

    A9.1.1.2  Soil

         In outdoor lysimeters, the toxicity of spills of 2.3 ml/cm2
    diesel and heating fuels was investigated in soil microbes, using
    Microtox(R) measurements (Wang & Bartha, 1990). (For parallel assays
    of seed germination and plant growth, see section A9.1.3.1.) At time
    0, both fuels were moderately toxic (EC50 = 80-90 mg of contaminated
    soil); this was followed by a period of increased toxicity, which
    started to decrease after 6-12 weeks. The values in soils treated by
    biodegradation returned to background within 20 weeks, although diesel
    fuel had significant residual toxicity in untreated soil. Seed
    germination and plant growth were inhibited in a similar manner, and
    diesel oil was more toxic and persistent than heating oil (section
    A9.1.3.1). Biodegradation treatment, in which the conditions for
    hydrocarbon degradation by microbes were optimized by mixing,
    aeration, pH control, and addition of mineral nutrients (fertilizers),
    strongly decreased fuel persistence and toxicity.

    A9.1.2  Aquatic organisms

    A9.1.2.1  Plants (phytoplankton)

         Experiments in both the laboratory and the field have shown that
    hydrocarbons can inhibit algal growth, although enhancement is
    occasionally noted at lower concentrations of oil (US National
    Research Council, 1985).

         After 12 days of exposure to 10% diesel oil, the growth of cells
    of  Euglena gracilis was not significantly reduced, whereas a
    concentration of 0.1% almost completely inhibited the growth of
    Scenedesmus quadricauda. Cells of  S. quadricauda grown in culture
    media containing diesel fuel (concentration not given) became
    chlorotic, suggesting damage to the photosynthesizing system
    (Dennington et al., 1975).

         A concentration of 0.05% light diesel fuel was found to stimulate
    growth rate, growth yield, photosynthesis, and chlorophyll a
    synthesis, while at the same time slightly inhibiting the respiration
    of  Chlorella salina CU-1 (Chan & Chiu, 1985). At higher
    concentrations (0.5 or 5%), the growth rate, growth yield, and

    photosynthesis were greatly reduced (to 50, 50, and 62% and 43, 32,
    and 41% of the control levels, respectively), but the effect on algal
    respiration was less severe (81% with 0.5% diesel and 86% with 5%).
    Diesel fuel containing dispersants caused greater inhibition than did
    the compounds separately.

         The lengthwise growth of the benthic algae  Ascophyllum nodosum
    and  Laminaria digitata and other rocky-shore communities was
    measured while they were kept in 50-m3 concrete basins and exposed
    continuously to diesel fuel for two years (Bokn, 1987). An average
    hydrocarbon concentration of 130 µg/litre continuously inhibited
    growth in both species, while a concentration of 30 µg/litre caused
    periodic inhibition. Under oil-free conditions during the subsequent
    season, the plants recovered completely.

         Boiler fuel was more toxic than diesel fuel to the kelp
    Macrocystis; both fuels reduced photosynthesis (Baker, 1970).

    A9.1.2.2  Invertebrates

    (a)  Several species

         The acute toxicity of diesel fuel to  Daphnia (Cladocera),
    chironomid (insect) larvae, and the mollusc  Viviparus bengalensis
    (Gastropoda) was tested after the organisms were acclimatized to
    laboratory conditions for four, two, and four days, respectively. The
    bioassays were conducted at 28 ± 2°C with unchlorinated borehole water
    (pH 7.0 ± 0.2; oxygen, 8 mg/litre; free carbon dioxide, 1.2 mg/litre;
    total alkalinity as calcium carbonate, 240 mg/litre; hardness,
    260 mg/litre). The results are shown in Table 18. Plankton and insect
    larvae exhibited erratic, uncoordinated movements before becoming
    lethargic. In the mollusc, active avoidance and heavy secretion of
    mucus were observed. The concentrations at which these behavioural
    changes occurred were not specified (Das & Konar, 1988).

         Fourteen species of five phyla (Echinodermeta, Mollusca,
    Annelida, Arthropoda, and Urochordata) were exposed to 0.5% No. 2
    diesel fuel in sea water. Only the larvae of the echinoderm
    Crossaster, which were also the largest, survived up to eight days;
    all of the other larvae died 3-72 h after being placed in the
    fuel-water mixture (Chia, 1973).

         The periwinkles  Pachymelania aurita (Muller) and  Tympanotonus
     fuscatus (Linne) were exposed to refined diesel fuel films for 24
    and 48h at concentrations of 2.5, 5.0, 7.0, or 10.0% and to emulsions
    for 24 h at concentrations of 0.5, 1.5, 2.5, or 3.5%; they were then
    shaken for 2 min with brackish water. After exposure to the emulsion,
    the periwinkles were washed and reintroduced into oil-free, brackish
    water for 24 h and their activity recorded. Both surface films and
    emulsions were harmful to both gastropods, the emulsions being the
    more toxic.  T. fuscatus was less susceptible than  P. aurita,

    possibly because the former can retract further into its shell. The
    mean percentage survival in the presence of oil film (7.5% diesel
    fuel) was 100% at 24 h and 80% at 48 h for T. fuscatus and 80% at 24 h
    and 50% at 48 h for P. aurita. The mean percentage survival in the
    presence of 3.5% diesel fuel-water emulsion was 90% for  T. fuscatus
    and 80% for  P. aurita (Dambo, 1993).

    Table 18.  Acute toxicity of diesel fuel to Daphnia (Cladocera),
               chironomid larvae (insects), and the mollusc Viviparus
               bengalensis (Gastropoda)
                                                                        

    Species         LC5 (mg/litre)   LC50 (mg/litre)   LC95 (mg/litre)

                                                                        

    Daphnia magna   1.5 (0.8-2.5)    20 (19.2-21.0)    38.5 (37.5-39.5)

    Chironomid      5.0              346.0 (238-455)   865.0 (750-1018)
    larvae

    Viviparus       2.0              254.0 (185-320)   637.0 (575-700)
    bengalensis
                                                                        

    From Das & Konar (1988)

    (b)  Molluscs

         The feeding rate and growth of mussels ( Mytilus edilus L.) was
    markedly reduced when they were exposed to 30 or 130 µg/litre diesel
    fuel for eight months (Widdows et al., 1985). Recovery of
    physiological responses (clearance rate, respiration, food absorption
    efficiency, and ammonia excretion) after transfer to unpolluted water
    was concomitant with the depuration of hydrocarbons from the tissues.
    Mussels exposed to either dose recovered completely within about 55
    days.

         The effects on the nutrient storage system and reproductive
    cell systems of the mussel  M. edilus of exposure to a diesel
    hydrocarbon-sea water emulsion and subsequent depuration were studied
    in order to assess the capacity of these systems to recover after
    discontinuation of exposure to the oil. After 144 days of exposure to
    hydrocarbon concentrations of 27 and 128 units, the volume of storage
    cell types was significantly reduced and there was a reduction in the
    volume of ripe gametes. After 53 days of depuration, the volume of the
    storage cells and of developing gametes increased. Exposure to either
    concentration also increased the volume of atretic (degenerating)

    gametes, but depuration allowed a return to normal (Lowe & Pipe,
    1986). The authors suggested that exposure to the hydrocarbons reduced
    the storage reserves and increased gamete atresia and resorption, so
    that the storage pool was partially replenished, enabling the organism
    to tolerate better the hydrocarbon insult.

         The sensitivity of  M. edulis to pollution with diesel fuel
    depends on the salinity of the water (Tedengren & Kautsky, 1987). Low
    salinity and diesel fuel acted synergistically to a greater degree in
    Baltic than in North Sea mussels.

         Adults and larvae of  M. edulis were exposed to microencapsulated
    diesel fuel at concentrations of 200, 600, 1000, 1300, or
    5000 µg/litre and the effects on mortality, larval growth, and
    spawning frequency were recorded (Stromgren & Nielsen, 1991). The EC50
    for spawning in mussels exposed for 30 days was about 800 µg/litre,
    and the LC50 (30days) for maturing mussels was about 5000 µg/litre.
    The longitudinal growth of larvae was significantly reduced at
    10 µg/litre; the EC50 for growth (10 days) corresponded to about
    25-30 µg/litre (Stromgren & Nielsen, 1991), which is lower than the
    EC50 of about 1000 µg/litre for growth of juvenile mussels reported
    by Stromgren & Reiersen (1988).

         Diesel fuel is therefore more toxic to mussel larvae than to
    juveniles.

         The EC50 values (10 days) for growth of larvae of the Quahog
    clam,  Mercenaria spp., exposed to water-soluble fractions of various
    crude and refined fuels were 220-4200 µg/litre, indicating that the
    insoluble fractions of hydrocarbons, ingested as microcapsules, are
    far more toxic than the water-soluble fractions alone (Byrne & Calder,
    1977). Stromgren et al. (1986) drew the same conclusion from their
    results for the growth of juvenile mussels. Larval mortality during
    exposure to diesel fuel increased steeply up to 50 µg/litre, at which
    dose only 20-30% survived 10 days of exposure. At 500 µg/litre, there
    was 100% mortality. Byrne & Calder (1977) reported LC50 values (10
    days) of 50-2100 µg/litre for larvae of  Mercenaria spp.

    (c)  Crustaceans

         Freshwater crabs  (Barytelphusa cunicularis) were exposed to
    sublethal concentrations of diesel fuel (4.5, 3.7, 3.2, and 2.6 ppm)
    for 24, 48, 72, and 96 h, respectively (Sarojini et al., 1989). The
    oxygen consumption of the crabs was measured as an index of stress
    caused by the fuel and compared with that of normal controls. The
    crabs responded in general by lowering their oxygen consumption up to
    8 h (down to 50% at 4 h), particularly at the lower doses; with longer
    exposures, the oxygen consumption was higher than (approx. 120% at
    24 h) or equal to (at 96 h) that of the controls.

          Gammarus spp. (Crustacea, Amphipoda) were obtained from marine
    and brackish water in the North and Baltic Seas and tested in two
    laboratories for their sensitivity to diesel-fuel pollution (Tedengren
    et al., 1988).  G. oceanicus was obtained from the Fucus belt at a
    depth of 1-3 m and  G. duebeni from rock pools of various salinities.
    The respiration, excretion, and atomic O:N and O:P ratios of the two
    species were compared after a 6-h exposure to 10 mg/litre diesel fuel
    emulsified in a syringe and injected into experimental aquaria; the
    resulting hydrocarbons were not measured. Exposure to diesel fuel
    generally resulted in decreased oxygen consumption and a rise in
    nutrient excretion, leading to lowered O:N and O:P ratios.
     G. oceanicus was the most sensitive, and the response was aggravated
    by simultaneous lowering or raising of the salinity. No signs of
    impaired swimming or other physical or mechanical effects were
    detected in either species. The investigators concluded that
     G. duebenihas a greater tolerance to pollutants and changes in
    salinity, probably because it has broad physiological niches and has
    evolved in and become adapted to more variable environments.

         In a comparison of the effects of an oil-water dispersion of
    diesel fuel, kerosene, gasoline, and benzene on the tidepool copepod
     Tigriopus californicus, diesel fuel was the most detrimental: a
    concentration of 0.10 ml/litre sea water caused 100% mortality within
    five days (Barnett & Kontogiannis, 1975).

    (d)  Fish

    Freshwater

         Table 19 shows the results of a series of 96-h static tests for
    acute toxicity with diesel fuel in freshwater on several species of
    juvenile salmonids,  Oncorhynchus kisutch (coho),  O. gorbuscha
    (pink), and  O. mykiss (rainbow trout), exposed at 14°C in 16 h of
    light and 8 h of darkness (Wan et al., 1990). The average loading
    density was 250 mg/litre (range, 100-420 mg/litre) of Canadian diesel
    fuel with a pour-point of -17.8°C. Three types of water were used for
    dilution: soft acid city tapwater, hard alkaline lakewater, and
    intermediate reconstituted deionized city tapwater. The LC50 values
    varied between the species from 32 to 33216 mg/litre. Diesel fuel was
    more toxic to pink salmon than to coho salmon or rainbow trout,
    irrespective of the water type.

         The LC values for diesel fuel in golden orfe  (Leuciscus idus
     melanotus) in a static test were: LC0 = 40-120 mg/litre; LC50 =
    120-160 mg/litre; LC100 = 160-205 mg/litre (Juhnke & Lüdemann,
    1978).

         Various freshwater fish  (Fundulus diaphanus, Roccus saxatilis,
     Lepomis gibbosus, Roccus americanus, Anguilla rostrata, and  Cyrinus
     carpio) were exposed to No. 2 or No. 4 fuel oil (both dispersed) in
    a static test at 19°C, pH 7.1, and 60 mg/litre hardness (Rehwoldt et
    al., 1974). The 96-h LC50 values for No. 2 fuel oil ranged from
    22.2 mg/litre for striped bass  (Roccus saxatilis) to 49.1 mg/litre
    for carp  (Cyrinus carpio); those for No.4 fuel oil ranged from
    21 mg/litre for banded killifish  (Fundulus diaphanus) to 48.1 mg/litre
    for carp. These concentrations represent total oil added and not the
    oil dissolved in the water column.

        Table 19.  Acute toxicity of diesel fuel to juvenile Pacific salmonids in different
               types of freshwater
                                                                                                

    Water type             Salmonid      LC50 (mg/litre
                                                                                                
                                         24-h         48-h         72-h        96-h
                                                                                                

    Soft                   Coho            28 972       28 972     12 345      10 299
                           Pink             1 972          276         84          74
                           Rainbow       > 32 000        5 525      4 600       3 017

    Intermediate           Coho          > 55 560     > 55 560     33 216      33 216
    (reconstituted)        Pink             1 829          376        133         123
                           Rainbow        168 363      168 363     28 787       2 186

    Hard (lake)            Coho          > 26 743       26 743      4 845       3 333
                           Pink             1 404          302         48          32
                           Rainbow       > 23 108       23 108      5 102       2 447
                                                                                                

    From Wan et al. (1990). Coho, Oncorhynchus kisutch; pink, O. gorbuscha; rainbow
    trout, O. mykiss
             The LC50 values of water-soluble fractions and emulsions and
    the floating layer of No. 2 fuel oil for four freshwater species under
    static and flow-through conditions are shown in Table 20 (Hedtke &
    Puglisi, 1982).

    Marine

         In an investigation of the threshold doses at which cod  (Gadus
     morhua L.) detected samples of Ekofisk diesel fuel, sea water was
    allowed to gravitate through the diesel fuel. The total content of
    dissolved materials, estimated by gas chromatography, were 15%
    benzenes, 2% toluenes, 6% xylenes, 15% naphthalenes, 3% phenols, 42%
     n-alkanes (C10-C25), and 17% unidentified compounds. There were
    no visible surface films or microdroplets in the samples, although the
    authors suspected the presence of undissolved microdroplets. Behaviour
    was observed with a sea-water olfactometer, and reaction patterns
    after injection of samples (snapping, increased activity, coughing,
    darting, backing) were noted. The most frequent reaction was snapping
    of jaws followed by a short period of activity. None of the doses
    produced alarm reactions. The mean activity scores increased with
    increasing doses of oil compounds. In a set of four experiments, the
    threshold for detection of diesel oil compounds was 100-400 ng/litre,
    showing that cod detect fuel compounds at doses lower than those
    hitherto observed (Hellstrom & Doving, 1983).

         The effect of acute exposure to the water-soluble fraction of
    Arctic diesel fuel (1:9 with sea water) on survival and metabolic rate
    was studied in an Antarctic fish  (Pagothenia borchgrevinki). The
    fish proved to be extremely tolerant and withstood an undiluted
    water-soluble fraction for at least 72 h. None died as a direct
    consequence of contact with the water-soluble fraction, but they
    showed signs of stress. In unpolluted water, little spontaneous
    swimming activity was seen, but on transfer into water containing
    diesel fuel they became agitated and increased their swimming activity
    for a short time. Their ventilation rates increased but later
    decreased; the depth of each ventilation increased considerably.
    Coughing became apparent shortly after transfer and persisted for up
    to 72 h. The haematocrit increased and remained elevated throughout
    the experiment. Oxygen consumption was increased to about twice that
    of the controls (Davison et al., 1992).

         In further studies on  P. borchgrevinki (Davison et al., 1993),
    a water-soluble fraction was prepared (1:2 with sea water) and then
    diluted to 33%; the fish were kept for seven days. On initial
    placement, behaviour similar to that with the higher concentration in
    the previous experiment was noted. After seven days, the only
    noticeable differences between exposed and control fish of similar
    weight and length were large amounts of mucus streaming from opercula,
    increased coughing, and slightly deeper ventilation. Plasma chloride
    and osmolarity were similar in the two groups. Haematocrit and
    haemoglobin concentration were significantly higher in the treated
    fish; spleen weights and oxygen consumption were similar. This
    Antarctic fish can thus survive prolonged periods of exposure to low
    levels of water-soluble hydrocarbons.


        Table 20.  96-h LC50 values for No. 2 fuel oil on freshwater organisms under various conditions
                                                                                                                                                

    Species                                 Life stage     Test conditions                     Effect             Concentration
                                                                                                                  (µl/litre water)
                                                                                                                                                

    Jordanella floridae (flagfish)          Adult          Water-soluble fraction of 10%       No mortality       No dilution
                                                           oil-water mixture; static
                                                           Emulsion; flow-through              LC50               60.5

    Pimephales promelas (fathead minnow)    Adult          Emulsion; flow-through              LC50               38.6
                                                           Water-soluble fraction of 10%       No mortality       No dilution
                                                           oil-water mixture; static
                                                           Floating layer; static              LC50               > 160 000
                                                                                                                  48 300 (2nd
                                                                                                                  sample)

    Rana sylvatica (frog)                   Larvae         Water-soluble fraction of           LC50               413 000
                                                           10% oil-water mixture; static
                                                           Emulsion; static                    LC50               26.4
                                                           Emulsion; flow-through              LC50               4.9
                                                           Floating layer; static              LC50               < 5000

    Ambystoma maculatum (salamander)        Larvae         Emulsion; flow-through              LC50               86.4
                                                                                                                                                

    From Hedtke & Puglisi (1982)
        (e)  Amphibians

         Larvae of the frog Rana sylvatica were tested in 96-h static and
    flow-through tests with No. 2 fuel oil as a water-soluble fraction, an
    emulsion, or a floating layer (Hedtke & Puglisi, 1982) (Table 20).
    Flow-through exposure to the emulsion was the most toxic (LC50 =
    4.9 µl/litre), and static exposure to the water-soluble fraction was
    the least toxic (LC50 = 413 000 µl/litre). The LC50 of No. 2 fuel
    oil emulsion in larvae of the salamander Ambystoma maculatum in a
    flow-through test was 86.4 µl/litre).

    A9.1.3  Terrestrial organisms

    A9.1.3.1  Plants

         Seed germination and growth of soya beans and ryegrass were
    inhibited by a diesel fuel spill of 2.3 ml/m2 (Wang & Bartha, 1990;
    see sections A3.2.4 and A9.1.1.2). Four weeks after the spill,
    biodegradation had at least partially restored the ability of the soil
    to support seed germination and plant growth. Morphological effects,
    such as reverse geotropism, were noted on some partially emerged
    seedlings. In untreated diesel-contaminated soil, there was no
    evidence of seed germination or plant growth. Parallel experiments
    showed that heating oil was less toxic.

    A9.1.3.2  Invertebrates

         No data were available.

    A9.1.3.3  Vertebrates

         Birds are affected externally and internally by oil
    contamination. The fuel rapidly destroys the waterproofing of the
    birds' plumage, and as the birds attempt to preen the contaminant off
    their feathers they ingest fuel. When No. 1 fuel oil was administered
    by gavage at doses of 1-4 ml/kg body weight to ducks, a dose of
    2 ml/kg caused lipid pneumonia, extreme inflammation of the lungs, and
    fatty infiltration and degeneration of the liver after 24 h (Croxall,
    1977). Administration of 1 ml of diesel or No. 1 fuel oil/kg produced
    severe irritation of the digestive tract (enteritis), with diarrhoea
    and bile pigments in faeces. Signs of toxic nephrosis were also noted.
    Adrenal enlargement (mainly hyperplasia of cortical tissue),
    depression of plasma cholinesterase levels, and incoordination,
    ataxia, and tremors were reported at higher doses. The treatment was
    not fatal to healthy birds, even at doses up to 20 ml/kg body weight;
    however, when No. 1 fuel and diesel fuel were administered to birds
    under stress, the LD50 was 3-4 ml/kg body weight (Hartung & Hunt,
    1966).

    A9.2  Field observations

         The following information is derived mostly from observations
    made after diesel spills (see Table 10).

    A9.2.1  Microorganisms

    A9.2.1.1  Water

         No data were available.

    A9.2.1.2  Soil

         Six weeks after the ship  Bahia Paraiso spilled Arctic diesel
    fuel into the Antarctic Sea, the impact of the spill on microbial
    ecology was investigated (Karl, 1992). The acute effects on the
    metabolic activities of sedimentary microorganisms appeared to be
    negligible, even in saturated sea water.

    A9.2.2  Aquatic organisms

         In field studies conducted after fuel spills to investigate the
    effects on populations and ecosystems, zooplankton appeared to be
    highly vulnerable to dispersed and dissolved petroleum constituents
    but less so to floating oils. Field observations on oil spills in
    general showed that individual organisms were affected in many ways,
    including direct mortality (fish eggs, copepods, mixed plankton),
    external contamination by fuel (chorion of fish eggs, cuticles, and
    feeding appendages of crustacea), tissue contamination by aromatic
    constituents, abnormal development of fish embryos, and altered
    metabolic rates (US National Research Council, 1985).  A spill of
    No. 2 diesel fuel was extremely toxic: faunal repopulation of the
    affected areas did not occur during the subsequent eight months
    (Blumer et al., 1970). A spill of 750000 litres of No. 2 diesel fuel
    caused substantial mortality in some taxa of the intertidal
    population, whereas others showed little or no effect. There was
    substantial recovery within six months, and there was no observable
    mortality among the subtidal fauna or flora (Woodin et al., 1972).    
    After a spill of marine diesel fuel (2000-3000 t) off Hong Kong, much
    of the oil was carried to land by onshore winds. Wormald (1976)
    described the effects of the oil spill on intertidal meiofauna. Within
    four days of the spill, there was almost total mortality: nematodes
    and harpacticoid copepods were observed in an advanced state of decay,
    and other meiofauna were not detected. The aromatic fractions probably
    evaporated within four to six weeks but persisted longer in deeper
    sediment. Harpacticoids and nematodes did not recolonize the beach
    until nearly eight months after the spill. The density of meiofauna at
    all stations observed reached a maximum 10 months after the spill and

    subsequently showed large fluctuations, which were probably the result
    of ecological successions during recovery of the beach fauna. One year
    later, an increase in the number of macrofauna, especially polychaete
    worms, was noted, which may have been responsible for the decrease in
    meiofaunal populations. High rainfall 11 months after the spill may
    have brought the soluble aromatic compounds nearer to the surface and
    decreased the harpacticoids and nematode populations. After 15 months,
    the meiofauna were again well established. Recovery at high shore
    stations began only when the oil content of the sediment had decreased
    by at least 50%, primarily through physical dispersion. Recolonization
    of the lower shore had occurred earlier. Stirling (1977) studied the
    effects of this spill on rocky shore fauna (see also Table 10),
    concentrating on large, common species of gastropods  (Monodonta labio,
     Thais clavigera, Nerita albicilla, N. polita, Lunella coronata,
     Clypeomorus humilis, and Planaxis sulcatus), bivalve molluscs  (Donax
     semignosus, Atactodea striata, Tapes philippinarum, and  Circe
     stutzeri), and crustaceans (shore and hermit crabs). Acute mortality
    of gastropods was greatest at moderately contaminated sites where
    oil-dispersing chemicals had been applied to floating oil slicks.
    Long-term disturbances were most significant at the heavily polluted
    sites where dispersants had not been used. Recovery of animals taken
    from the oiled beaches was studied in clean salt water in the
    laboratory. Bivalve molluscs and the gastropods  Monodonta labioand
     Thais clavigera were the most sensitive and  Clypeomorus humilisand
     Planaxis sulcatus the least sensitive. Field observations of acute
    mortality were consistent with this order of sensitivity. Population
    studies showed the greatest reductions in  Monodonta labio and  Nerita
     albicilla, which were eliminated for at least 13 months from the
    site of greatest oil pollution.

         The release of 600 000 litres of Arctic diesel fuel from the
     Bahia Paraiso in 1989 coated intertidal macroalgae, limpets, and
    birds as well as sediments and shores (Kennicutt et al., 1991b).
    Macroalgae  (Leptosomia simplex and  Monostroma spp.) were only
    moderately affected. Intertidal limpet  (Nacella concinna) populations
    were reduced by 50% within the first few weeks after the spill
    (Kennicutt et al., 1990), and a year later had still not fully
    recovered (Kennicutt & Sweet, 1992). Bottom-feeding organisms, such as
    clams  (Laternula elliptica) and fish  (Harpagifer anarcticus and
     Notothenia coriiceps neglecta) were found to have PAHs in their gut
    contents and muscle tissues (Kennicutt et al., 1991b). After the
    spill, all of the chicks of the local population of South Polar skuas
     (Catharacta maccormicki) died, although few of the adults or the
    chicks died from direct contact with oil (Eppley, 1992). Within less
    than two years, and at many locations within a few weeks, the PAH
    content of sediment and intertidal organisms returned to background
    levels (Kennicutt & Sweet, 1992), owing partly to the high volatility
    of the spilled fluid and the high energy environment. Sampling of
    intertidal limpets showed that the tissue contaminants were primarily
    naphthalenes, but fluorenes and phenanthrenes were often the most

    abundant aromatic contaminants in tissues with higher concentrations
    of PAHs (Kennicutt & Sweet, 1992; Kennicutt et al., 1992b)  After a
    spill of 8000 litres of diesel fuel into an unspoiled freshwater
    stream in California, United States, most animals were affected within
    one to four days. Thousands of aquatic insects perished; some
    crayfish, aquatic leeches, and freshwater planarians were killed; and
    over 2500 fish, tadpoles (but no frogs), snakes, a turtle, and common
    merganser  (Mergus merganser) were found dead (Bury, 1972).

    A9.2.3  Terrestrial organisms

    A9.2.3.1  Plants

         Diesel fuel was applied at 10 litres/m2 to the soil of five
    common plant communities in north-east Greenland, comprising wet
    marsh, grassland, and three types of dwarf-shrub heath. The plant
    species were observed for up to three years after the application.
    Within a few days, most species had lost chlorophyll or had abscissing
    leaves. Species with xeromorphic leaf characteristics reacted more
    slowly than species with orthophyllous leaves. During three successive
    growing seasons, shrubs and forbs showed no signs of recovery.
    Graminoids showed slight resistance and recovered in all mesic and wet
    plots but were killed in dry plots. Three species of sedge  (Carex
     bigelowii, C. saxatilis, and  C. stans) and mosses recovered to some
    extent in all plots. Recovery of mosses was excellent in the wet
    plots, good in the mesic plots, and poor in the dry plots. In general,
    the phytotoxic effects of diesel fuel were much more pronounced than
    those of crude oil in similar plots (Holt, 1987).

    A9.2.3.2  Invertebrates

         No data were available.

    A9.2.3.3  Vertebrates

         No data were available.

    A10.  EVALUATION OF HUMAN HEALTH RISKS AND EFFECTS ON THE ENVIRONMENT

    A10.1  Evaluation of human health risks

         Diesel fuel is produced commercially in various qualities with
    regard to volatility, aromaticity, cetane number, and sulfur content.
    The composition of diesel fuel, which influences the type and amount
    of compounds emitted in the exhaust, has not changed greatly during
    the last few decades, although the cetane number has been slightly
    increased, resulting in better ignition, and the sulfur content, which

    influences the release of particulate matter, has been reduced in some
    countries. Owing to the lack of data on diesel fuel, some data on
    heating oils and kerosenes (jet fuels) with compositions similar to
    that of diesel fuel are included in the evaluations of toxicological
    and environmental effects.

    A10.1.1  Exposure of the general population

         The general population can be exposed to diesel fuel and other
    middle distillates at filling stations, as a result of accidental
    spills, when handling such fuels, or when using kerosene for domestic
    cooking or heating. No data on exposure were available.

    A10.1.2  Occupational exposure

         Workers can be exposed to diesel fuel and other middle
    distillates during manual handling and discharge of the fuel, i.e.
    during retailing at filling stations; manufacture, repair,
    maintenance, and testing of diesel engines and other equipment; in
    jobs where diesel fuel is used as a cleaning agent or solvent; and in
    the handling and routine sampling of diesel fuel in the laboratory. At
    room temperature, very low concentrations of vapours are likely to be
    generated from diesel fuel because of its low volatility; significant
    levels of vapour are likely to occur only in confined spaces and at
    high temperatures.

    A10.1.3  Non-neoplastic effects

         Diesel fuels are toxic when ingested, usually accidentally.
    Ingestion may result in regurgitation and aspiration, which can cause
    chemically induced pneumonia. The latter effect is, however, not
    specific for diesel fuel and can occur with all hydrocarbons of a
    particular viscosity range.

         Exposure to vapours is minimal during normal handling of diesel
    fuel. The most likely effect on human health is dermatitis as a result
    of skin contact. Dermal absorption takes place and can result in acute
    toxic effects on the kidney. The health effects of long-term
    absorption of low levels are unknown, but the available data on acute
    human toxicity indicate that practices such as washing hands in diesel
    fuel should be avoided. Although groundwater contamination and entry
    into drinking-water are potential sources of adverse health effects,
    such contamination would be noticeable and affect palatability so that
    inadvertent ingestion of contaminated drinking-water is unlikely.     
    In experimental animals, diesel fuel has little acute toxicity after
    exposure orally, dermally, or by inhalation. The oral LD50 values in
    rats were > 5000 mg/kg body weight and those in mice, rabbits, and
    guinea-pigs even higher. After short-term dermal exposure of mice for
    14 consecutive days, the NOAEL for two middle distillates (marine

    diesel and JP-5 navy fuel) was 5000-8000 mg/kg body weight per day.
    Inhalation of up to 0.2 mg/litre had a neurodepressive effect in mice
    but not in rats. Subchronic exposure to various distillate fuels
    induced mainly alpha2-microglobulin nephropathy in male rats, which is
    not considered relevant for humans.

         Female mice that received dermal applications of 250 or 500 mg/kg
    body weight per day of marine diesel fuel or JP-5 navy fuel on five
    days per week for 103 weeks showed decreased survival due to severe
    ulceration. Rats and mice exposed by inhalation to 1 or 5 mg/litre had
    significant alterations in organ weight.

         Diesel fuel irritates the skin but not the eye.

         Exposure to diesel or jet fuels orally or by inhalation was
    neither embryotoxic nor teratogenic in rats; although doses toxic to
    the dams reduced fetal weight, no effects on viability were seen.     
    No clear evidence of mutagenic activity was seen in a series of tests
    in  Salmonella typhimurium. A positive response achieved only under
    special conditions appeared to be equivocal. Other tests for
    genotoxicity  in vitro or  in vivo did not show clearly positive
    responses.

         Owing to lack of data, a quantitative risk assessment is not
    possible.

    A10.1.4  Neoplastic effects

         A case-control study of cancers at several sites in relation to
    exposure to diesel fuel suggested an increased risk for lung cancer
    other than adenocarcinoma and for prostatic cancer. In neither case
    was there an exposure-response relationship. As few studies were
    available, the number of cases was small, and the confidence intervals
    were correspondingly wide, no conclusion can be drawn at present about
    the carcinogenicity to humans of diesel fuel.

         Twelve studies of dermal carcinogenicity in animals demonstrated
    that diesel fuels have weak carcinogenic potency in mice. As no clear
    genotoxicity is seen, cancer may be induced by nongenotoxic
    mechanisms, such as chronic dermal irritation characterized by
    repeated cycles of skin lesions and epidermal hyperplasia.

    A10.2  Evaluation of effects on the environment

         Although there are numerous data on the environmental behaviour
    of the individual components of diesel fuel, few data are available
    on diesel fuel as a whole. The environment can be polluted by
    accidental release of diesel fuel on a large scale, such as during
    tanker disasters and pipeline leaks, or on a smaller scale, from
    contamination of soil around factories and garages.

         In water, diesel fuel spreads out almost immediately, polar and
    low-relative-molecular-mass components dissolve and leach out of the
    slick, volatile components evaporate from the water surface, and
    microbial degradation begins. The extent to which such 'weathering'
    takes place depends on the temperature and climatic conditions, and
    the chemical composition of spills changes with time. After spillage
    on water, some fractions evaporate, and the evaporated diesel
    components are degraded photochemically. In sediment, diesel fuel is
    generally delivered to bottom sediments by settling particles. In
    soil, the various components of diesel fuel migrate at different
    rates, depending on the soil type.

         The individual constituents of diesel fuel are inherently
    biodegradable, but the rates depend greatly on the physical and
    climatic conditions and microbial composition. PAHs are the most
    recalcitrant molecules.

         Aquatic organisms, in particular molluscs, bioaccumulate
    hydrocarbons to various extents, but the hydrocarbons are depurated on
    transfer to clean water. Bioaccumulation of diesel fuel may occur, but
    there are few data to indicate biomagnification.

         Although there have been a number of spills of diesel fuel, few
    reports are available on their environmental effects. Generally,
    diesel fuel is more toxic than crude oil to aquatic and plant species.
    Spills of diesel fuel have an immediate detrimental effect on the
    environment, causing substantial mortality of biota; recolonization
    occurs after about one year, depending on the animal and plant species
    involved and the chemical and physical content of the spill residues.
    Aquatic organisms that survive diesel fuel spills can still be
    affected by external contamination and tissue accumulation; abnormal
    development and altered metabolic rates are signs of such stress.

         There are few data on the effects of diesel fuel on aquatic and
    terrestrial organisms in experimental situations. In studies of
    water-soluble fractions and emulsions of diesel fuel, the actual
    chemical composition has rarely been analysed, although this is
    crucial to an understanding of the toxicity of diesel fuel. No data
    are available about flow-through conditions or mesocosms, which better
    reflect environmental conditions

    A11.  RECOMMENDATIONS

    A11.1  Recommendations for the protection of human health

         Exposure to diesel fuel may cause irritation and dermatitis;
    therefore, skin contact should be avoided, and diesel fuel should not
    be used for cleaning purposes.

    A11.2  Recommendation for the protection of the environment

         As for all petroleum products, accidental releases to the
    environment should be avoided and should be cleaned up as soon as
    possible if they occur.

    A11.3  Recommendations for further research

         A basic data set on the aquatic toxicity of diesel fuels,
    including complete analysis of test substances, should be generated.  
    The fate and environmental impact of diesel fuel spills and leakages
    should be investigated in laboratory and controlled field studies.    
    The exposure of humans to diesel fuel in various situations should be
    quantified.

         The mechanism of the carcinogenic action of diesel fuel on the
    skin of experimental animals should be clarified.

    A12.  PREVIOUS EVALUATIONS BY INTERNATIONAL BODIES

         The carcinogenic risks of diesel fuels for human beings were
    evaluated by a working group convened by the International Agency for
    Research on Cancer in 1988 (International Agency for Research on
    Cancer, 1989a). It concluded that there was inadequate evidence for
    the carcinogenicity of diesel fuels in humans, and there was limited
    evidence for the carcinogenicity of marine diesel fuel in experimental
    animals.

    PART B

    B1.  SUMMARY

    B1.1  Identity, physical and chemical properties, and analytical
          methods

         Diesel engine exhaust emissions contain hundreds of chemical
    compounds, which are emitted partly in the gaseous phase and partly in
    the particulate phase of the exhaust. The major gaseous products of
    combustion are carbon dioxide, oxygen, nitrogen, and water vapour;
    carbon monoxide, sulfur dioxide, nitrogen oxides, and hydrocarbons and
    their derivatives are also present. Benzene and toluene are present in
    the lower weight percent range in the gaseous part of the hydrocarbon
    fraction. Other gaseous exhaust components are low-relative-molecular-
    mass polycyclic aromatic hydrocarbons (PAHs).

         A main characteristic of diesel exhaust is the release of
    particles at a rate about 20 times greater than that from
    gasoline-fuelled vehicles. The particles are composed of elemental
    carbon, organic compounds adsorbed from fuel and lubricating oil,
    sulfates from fuel sulfur, and traces of metallic components. Most of
    the total particulate matter appears to occur in the submicrometre
    range, between 0.02 and 0.5 µm. Agglomeration may occur during aging,
    up to a maximal diameter of 30 µm. The emitted particles have a large
    surface area. Organic compounds generally contribute 10-30% of the
    total particulate matter, but poorly designed and maintained engines
    may result in as much as 90%. Higher-relative-molecular-mass,
    oxygenated and nitro-PAHs occur at concentrations of parts per million
    in this fraction.

         Specific transient or steady-state driving cycles are used in
    measuring vehicle emissions. The exhaust can be sampled from undiluted
    or diluted exhaust gas. It is difficult to obtain artefact-free
    samples, as exhaust constituents can undergo chemical reactions,
    adsorption and desorption processes, and condensation and diffusion.
    The toxicologically relevant PAHs in diesel particulate matter are
    usually determined by Soxhlet extraction, clean-up, and fractionation,
    with subsequent analysis by high-performance liquid chromatography or
    gas chromatography coupled with mass spectrometry.

    B1.2  Sources of human and environmental exposure

         Diesel engine exhausts are emitted mainly from motor vehicles;
    other sources are stationary, railway locomotive, and ship diesel
    engines. The emissions from diesel motor vehicles have been well
    described, but the individual results are often not comparable owing
    to differences in parameters such as driving cycle, engine type, and
    fuel composition. The individual components are emitted in the

    following quantities: carbon dioxide, about 1 kg/km; carbon monoxide,
    nitrogen oxides, total gaseous hydrocarbons, and particulate matter,
    0.1-20 g/km; and aliphatic compounds, alcohols, aldehydes, light
    aromatics, and PAHs, micrograms per kilometer. The emissions of carbon
    monoxide, nitrogen oxides, total gaseous hydrocarbons, and particulate
    matter are regulated by law in a number of countries.

         In principle, there is no difference between the quality and
    quantity of exhaust emissions from light- and heavy-duty engines,
    although heavy-duty vehicles release larger relative amounts of
    particulate matter. Exhaust emissions depend on driving cycle
    (transient or steady state), engine conditions (injection and
    aspiration techniques, maintenance, total mileage), and fuel
    composition (sulfur content, aromaticity, volatility); adjustment of
    the engine plays a major role.

         The release of particulate matter increases with decreasing
    air:fuel ratio, increasing load, and increasing temperature. More
    particulate is released from older, intensively used engines than from
    new, low-mileage engines, probably because of a greater consumption of
    lubricating oil. The emission of particles from light-duty diesel
    vehicles is also correlated with the sulfur content of the fuel, as
    the formation of metal sulfates increases the particle mass; particle
    emissions from heavy-duty vehicles have not yet been established.
    Increasing fuel aromaticity also increases particle emissions.

         PAHs and oxygenated PAHs from diesel and spark-ignition engines
    are qualitatively similar. Oxygenated and nitrated PAHs are emitted in
    the low microgram per kilometer range, but the actual concentrations
    of these compounds are uncertain, as decomposition and formation can
    occur during sampling. PAH emissions increase with increasing load and
    temperature and with the age of the engine, probably owing to
    increased consumption of lubricating oil. PAH emissions also depend on
    the injection technique of the engine: they increase with increasing
    air:fuel ratio in engines with direct injection, whereas they decrease
    in engines with indirect injection. The aromaticity and volatility of
    the fuel are directly correlated with the emission of PAHs.
    Malfunction of engine devices, especially the fuel injection system,
    increases the emission of the main exhaust components. There are few
    data on the contribution of diesel motor emissions to the total
    man-made release of combustion products.

         Diesel exhaust emissions can be reduced by improving engine
    design and by use of particle traps (trap oxidizers) and catalytic
    converters. While particle traps remove both the soot and soluble
    organic compounds adsorbed onto the particles, catalytic converters
    reduce the levels mainly of carbon monoxide and gaseous hydrocarbons.
    In practice, it is dificult to regenerate particulate traps. Catalytic
    converters require fuels with a low sulfur content, as sulfur poisons
    the active centres of the catalyst.

    B1.3  Environmental transport, distribution, and transformation

         The compartment first affected by diesel exhaust emissions is the
    atmosphere. The hydrosphere and geosphere are contaminated indirectly
    by dry and wet deposition. The environmental fate of the individual
    constituents of diesel exhaust is generally well known: Particles
    behave like (non-reacting) gas molecules with regard to their
    mechanical transport in the atmosphere; they may be transported over
    long distances and even penetrate the stratosphere. The overall
    removal rate of diesel particles is estimated to be low, resulting in
    an atmospheric lifetime of several days. During aging, particles may
    coagulate, with higher fall-out rates, thus reducing the total
    airborne level. The elemental carbon of diesel particulates may act as
    a catalyst in the formation of sulfuric acid by oxidation of sulfuric
    dioxide. The organic components adsorbed on elemental carbon may
    undergo a number of physical and chemical reactions with other
    atmospheric compounds and during exposure to sunlight.

    B1.4  Environmental levels and human exposure

         As diesel exhaust is a complex mixture of a great variety of
    compounds, general 'environmental levels' cannot be given. The
    individual components of diesel exhaust should be detectable in all
    compartments of the environment, although their source cannot usually
    be verified. The environmental levels of most of the individual
    constituents are known.

         The general population is most likely to be exposed to diesel
    exhaust in busy streets or parking areas, particularly underground.
    The identification of sources is difficult, and the contribution of
    diesel exhaust to total pollution by traffic combustion products is
    generally calculated on the basis of emission factors and the
    percentage of diesel-fuelled vehicles. The levels of exposure of the
    general population and workers to diesel particulate matter are
    toxicologically relevant. The daily average ambient concentrations of
    diesel particulates near roads are 8-42 µg/m3. The estimated annual
    average concentrations were 5-10 µg/m3 in urban areas and
    < 1.5 µg/m3 in rural areas in Germany and 1-2 µg/m3 in urban
    areas and 0.6-1 µg/m3 in rural areas in the United States of America.
    The concentrations are directly correlated with traffic density and
    decrease with increasing distance from roads. Identification of
    sources may also be difficult in work-places, especially in mines,
    where the total dust burden is high. Carbon core analysis has been
    used to determine specific exposure to diesel exhaust in the
    work-place, and the levels of particulate matter to which workers are
    exposed have been reported to be 0.04-0.134 mg/m3 for truck drivers
    and 0.004-0.192 mg/m3 for railroad workers. Total respirable
    particles and total suspended particles have also been used as
    measures of occupational exposure.

    B1.5  Kinetics and metabolism in laboratory animals and humans

    B1.5.1  Deposition

         Diesel exhaust particles, with a mass median aerodynamic diameter
    of about 0.2 µm, undergo some filtration in the nose, and the
    deposition efficiency in the lung is only slightly greater than the
    minimal value found for particles with a mass median aerodynamic
    diameter of about 0.5 µm. Thus, 10-15% of inhaled diesel soot
    particles are deposited in the alveolar region of the lung of rats and
    guinea-pigs; in humans, about 10% is deposited in the alveolar region.

    B1.5.2  Retention and clearance of particles

         Mucociliary clearance of particles from ciliated airways is
    almost complete within 24 h. The long-term clearance of diesel
    particles from the alveolar region was measured in several studies in
    rats exposed by inhalation to labelled diesel or test particles.
    Half-times of 60-100 days were reported for controls with low lung
    burdens (< 1 mg/lung), whereas the half-times increased to 100-600
    days in rats with lung burdens increasing from 1 to 60 mg/lung. In
    several studies, effects were seen to be caused by particle overload,
    which has been described in a variety of species and with a number of
    particulate materials. The phenomenon is generally observed when the
    deposition rate of particles of low solubility and low acute toxicity
    exceeds their clearance rate for a considerable time. The half-time of
    unimpaired alveolar clearance in humans is several hundred days, which
    is longer than that in rats.

         A dosimetric lung model was developed on the basis of data on the
    deposition and retention of diesel particles in rats after long-term
    inhalation and of data on particle deposition and retention in humans.
    The model can be used to predict retention of diesel particles and
    adsorbed organic material in the lungs of people of different ages. No
    data are available on changes in the retention of individual compounds
    after prolonged exposure to diesel exhaust.

     B1.5.2  Retention and clearance of polycyclic aromatic hydrocarbons
             adsorbed onto diesel soot

         PAHs in diesel soot adhere strongly to the surface of particles.
    About 50% of the PAHs adsorbed onto diesel particles is cleared from
    the lung within one day, but the retention half-times for the
    remaining PAHs were 18-36 days. Studies with 3 H-benzo[ a]pyrene and
    14C-nitropyrene show that when PAHs are associated with particulate
    matter, their clearance from the lungs is significantly delayed in
    comparison with the clearance of inhaled PAHs not associated with
    particulate matter.

    B1.5.4  Metabolism

         Benzo[ a]pyrene coated on diesel exhaust particles was metabolized
    by oxidation to benzo[ a]pyrene phenols, diols, and quinones in the
    lung and in cell cultures of pulmonary macrophages. Nitropyrene
    adsorbed to diesel particles was metabolized to acetylaminopyrene-
    phenol after inhalation. More DNA adducts were found in the lungs and
    type II cells of rats exposed to diesel exhaust than in controls. The
    oxidative metabolism of some organic compounds to epoxides may be
    responsible for the formation of these adducts, but adducts were found
    only after exposure to particles. Certain organic substances in diesel
    exhaust have been shown to be responsible for the formation of DNA
    adducts; however, the carbonaceous core itself (without extractable
    organic material) can also induce adducts, by chronic damage of
    epithelial cells.

    B1.6  Effects on laboratory mammals and in vitro test systems

         The few data available suggest that diesel exhaust has little
    acute toxicity. Mice treated intratracheally with diesel exhaust
    particles died due to lung oedema; the LD50 was about 20 mg/kg body
    weight. Methanol-extracted diesel exhaust particles were not lethal a
    concentrations up to about 33 mg/kg body weight. In hamsters, the LD50
    after intraperitoneal administration was 1280 mg/kg body weight. No
    data are available for exposure by inhalation.

         Exposure of rats, guinea-pigs, and cats to diesel exhaust with a
    particle content of 6 mg/m3 for about four weeks altered lung
    function, including a 35% increase in pulmonary flow resistance in
    guinea-pigs and a 10% decrease in vital capacity (expiratory flow) in
    cats. Histopathologically, focal thickening of alveolar walls, a
    significantly increased type-II cell labelling index, and
    accumulations of particleladen macrophages were found. The
    accumulations were located near the terminal bronchioles and became
    larger, due to macrophage attachment (sequestration), during a
    subsequent recovery period.

         Mice, rats, hamsters, cats, and monkeys did not have drastic
    decreases in body weight or reduced survival times after long-term
    inhalation of up to 4 mg/m3 . Dose-related toxic effects seen in all
    species after long-term inhalation of diesel exhaust were: increases
    up to 400% in lung weight; pulmonary inflammation measurable by
    biochemical (cytoplasmic marker enzymes, collagen) and cytological
    (increase in polymorphonuclear neutrophils) parameters; impairment of
    lung mechanics; increasing numbers of particle-laden macrophages with
    focal accumulations (sequestration) under overload conditions; and
    subsequent proliferative alterations of epithelial cells and onset of
    fibrosis.

         Limited information on reproductive and developmental toxicity
    suggests that inhalation of diesel exhaust is not critical. In most
    experiments, no effects were seen in mice, rats, hamsters, rabbits, or
    monkeys; however, after intraperitoneal injection, sperm abnormalities
    were seen in mice given diesel exhaust particles and embryotoxicity
    occurred in hamsters given diesel exhaust extract.

         Most tests for genotoxicity  in vitro have been performed with
    diesel exhaust extracts rather than with total exhaust; positive
    responses were found in the absence of metabolic activation, i.e. the
    genotoxic effects appeared to be PAH-independent. About one-half of
    investigations  in vivo gave negative responses; the only positive
    responses were sister chromatid exchange induction with total diesel
    exhaust and with organic extracts and micronucleus induction with
    organic extracts.

         Immunotoxic effects were generally not seen after inhalation of
    diesel exhaust; however, an enhanced anti-ovalbumin immunoglobulin E
    antibody titre was seen in one study and an increased susceptibility
    to infections in mice in two experiments. Studies in rats suggested
    that inhalation of diesel exhaust affects behavioural and
    neurophysiological status.

         In studies of the carcinogenicity of diesel exhaust administered
    by inhalation to rats, the gaseous phase (without the particulate
    fraction) was not carcinogenic. In all valid studies in rats, diesel
    exhaust was found to be carcinogenic at particle concentrations of >
    2 mg/m3, corresponding to an equivalent continuous exposure of about
    1mg/m3. No effect was seen in hamsters or mice. In studies by
    intratracheal instillation, both diesel exhaust particles and carbon
    black induced tumours, and the surface area of the carbonaceous
    particles appeared to be correlated with the tumorigenic potency.
    Long-term inhalation of carbon black virtually devoid of PAHs at
    similar concentrations also resulted in lung tumours in rats.

         It is not clear whether the carcinogenicity of diesel exhaust
    involves DNA-reactive or non-DNA-reactive mechanisms (or a
    combination). Various models have been used to elucidate the
    carcinogenicity of diesel exhaust.

    B1.7  Effects on humans

         Diesel exhaust contributes to air pollution in general. Although
    the role of diesel particles cannot be singled out in acute or chronic
    studies, they may be partly responsible for the range of health
    effects found to be associated with air pollution.

         Typical diesel exhaust has a characteristic odour, which some
    people find offensive, particularly at high concentrations. The
    symptoms seen after acute and chronic exposure to diesel exhaust have
    been described in studies and anecdotal reports of occupationally
    exposed subjects. Acute exposure to diesel exhaust has been associated
    with irritation of the ocular and nasal mucous membranes, and an
    increased frequency of respiratory symptoms has been observed in
    occupational cohorts; however, the contribution of diesel particulates
    is not known. No consistent short-term effect on pulmonary functions
    has been found, but asthma attacks have been reported.

         In a controlled study in which eight healthy non-smoking
    volunteers were exposed to diluted diesel exhaust in a chamber for
    60 min, the phagocytosis rate of alveolar macrophages in broncho-
    alveolar lavage fluid was reduced.

         Some cross-sectional and longitudinal studies on workers with
    long-term occupational exposure to diesel exhaust show decrements in
    lung function and an increased prevalence of respiratory symptoms, but
    these studies are limited by short exposure. Cohort studies in which
    deaths from cardiovascular and/or cerebrovascular disease due to
    diesel emissions were investigated did not show a significant excess.

         The relationships between cancers of the lung and of the urinary
    bladder and occupational exposure to diesel exhaust have been
    evaluated in a number of epidemiological studies. Only those studies
    that were considered relevant for evaluating the carcinogenic effects
    of diesel exhaust are included in this monograph. The most relevant
    studies with regard to lung cancer are those of railroad workers, bus
    garage workers, and stevedores, who have definite exposure to diesel
    exhaust. The four most informative studies all reported an increased
    risk for lung cancer, with relative risks ranging from 1.4 for
    railroad workers and 1.3-2.4 for bus garage workers (depending on
    exposure category), to a three- to sixfold increase in risk for
    stevedores (depending on the exposure assessment used, but with wide
    confidence intervals). Adjustment for smoking was possible in one
    case-control study of railroad workers and in the study of stevedores,
    and in both cases, the effect of diesel exposure was not materially
    influenced. In the three studies in which smoking could not be
    adjusted for, the analysis was based on comparisons of subgroups of
    the cohorts, so that confounding by tobacco smoking was less likely
    than when external comparison groups were used.

         Several case-control studies have been conducted to examine the
    relationship between urinary bladder cancer and presumed exposure to
    diesel exhaust. An increased risk was found, especially for truck
    drivers; however, all of these studies are limited by poor
    characterization of exposure. Hence, a causative association between
    exposure to diesel exhaust and an increased risk for urinary bladder
    cancer cannot be established.

    B1.8  Effects on other organisms in the laboratory and the field

         The effect of diesel emissions as such have been addressed in
    only one study, of green algae.

    B1.9  Evaluation of human health risks

         The risk assessment paradigm of the United States National
    Academy of Sciences (US National Research Council, 1983) was used to
    assess the risks for both cancer and non-cancer end-points. The four
    steps in this process comprise: (1) hazard identification, (2)
    dose-response assessment, (3) exposure assessment, and (4) risk
    characterization.

         Epidemiological studies of long duration with well-defined
    exposure and follow-up (> 20 years) are considered to be the most
    informative. Four studies of lung cancer in occupationally exposed
    individuals met these criteria. The relative risks for lung cancer
    associated with exposure to diesel exhaust were generally low and were
    susceptible to chance, to the effects of unmeasured confounding
    factours, and to the difficulty in accurately adjusting for known
    confounding factors. Studies with less precise definitions of exposure
    support the conclusions of these studies. Overall, it is considered
    that diesel exhaust is probably carcinogenic to humans; however, no
    quantitative data are available for estimating human risk.

    B1.9.1  Non-neoplastic effects

         Two general approaches were used for risk characterization: a
    no-observed-adverse effect level (NOAEL) divided by an uncertainty
    factor; and a benchmark concentration. In both approaches, a
    sophisticated dosimetric model was used which decreases any
    uncertainty in interspecies extrapolation of dose.

         The no-effect level of diesel exhaust particles in humans was
    calculated to be 0.139 mg/m3. The guidance value for the general
    population calculated from the dosimetric model was 5.6 µg/m3, and
    that calculated without the model was 2.3 µg/m3.

         The benchmark concentration approach takes into account the
    entire exposure-response relationship rather than relying on a single
    data point from studies by inhalation, as in the NOAEL approach. Three
    sensitive end-points were identified: chronic alveolar inflammation,
    impaired lung clearance, and hyperplastic lung lesions. The benchmark
    concentrations, calculated from the same dosimetric model used in the
    NOAEL approach, were 0.9-2 µg/m3for inflammation, 1.2-3 µg/m3 for
    impaired lung clearance, and 6.3-14 µg/m3 for hyperplastic lesions.

    B1.9.2  Neoplastic effects

         A linearized multistage model was used to estimate unit risks due
    to exposure to diesel exhaust. Because the results of the available
    epidemiological studies were considered inadequate for a quantitative
    estimate of unit risk, data from several studies of long-term
    inhalation in rats were used in which carcinogenesis occurred at
    concentrations >2 mg/m3. A unit risk of 3.4 × 10-5 µg/m3 (geometric
    mean of four risk estimates) diesel exhaust particles was calculated.
    An alternative biologically based model yielded a similar unit risk,
    under the assumption that diesel particles affect cell initiation
    and/or proliferation at low concentrations.

    B1.10  Evaluation of effects on the environment

         Insufficient information was available to evaluate the specific
    effects of diesel exhaust emissions. The effects of combusted diesel
    fuel should be similar to those of other fossil fuels and are related
    to the consumption of diesel fuel.

    B2.  IDENTITY AND ANALYTICAL METHODS

    B2.1  Identity

         Diesel engine emissions contain hundreds of chemical compounds,
    which are emitted partly in the gaseous phase and partly in the
    particulate phase of the exhaust. These substances form particles or
    contribute to the gaseous phase of the emissions, depending on vapour
    pressure, temperature, and the concentration of individual species.
    The amount and composition of diesel exhaust are related to engine
    conditions and fuel specifications (see sections B3.1.2.1 and
    B3.1.2.2). A total of 445 compounds has been identified or tentatively
    identified as constituents of diesel exhaust emissions (Westerholm,
    1987).

         Diesel exhaust has an offensive odour (odour threshold, 320 ppm).
    Compounds with a relative molecular mass > 80 mainly determine the
    odour (Scheepers & Bos, 1992b). Partridge et al. (1987) concluded that
    benzaldehyde and a methylbenzaldehyde isomer are the major substances
    responsible; unburnt aromatic hydrocarbons may also contribute.
    Aldehyde levels and odour emissions are directly correlated with the
    total hydrocarbon content of the exhaust; decreased hydrocarbon levels
    lead to diesel exhaust with less odour (Organisation for Economic
    Co-operation and Development, 1993).

    B2.1.1  Chemical composition of diesel exhaust gases

         The main gaseous products of diesel exhaust are carbon dioxide,
    oxygen, nitrogen, and water vapour; carbon monoxide, sulfur dioxide,
    nitrogen oxides, and hydrocarbons and their derivatives are also
    present. A representative composition of diesel exhaust gas
    (light-duty engine) is shown in Table 21.

         The levels of carbon monoxide, nitrogen oxides, total
    hydrocarbons, and particulates emitted by diesel engines are regulated
    by law in a number of countries (see section B3.2). The composition of
    diesel exhaust gases is similar to that of gasoline engine gases, but
    because of the relatively higher fuel:air ratio, carbon monoxide and
    hydrocarbons occur in lower concentrations in diesel exhaust, and the
    emission of nitrogen oxides, particulate matter, and sulfur compounds
    is higher (the latter being due mainly to the higher sulfur content of
    diesel fuels). The relative concentrations in the aldehyde fraction
    are as follows: about 45% by weight formaldehyde, 17% by weight
    acetaldehyde, 14% by weight acetone and acrolein, < 10% by weight
    crotonaldehyde, propionaldehyde, isobutyraldehyde, benzaldehyde, and
    hexanaldehyde; methylethylketone is also present (Volkswagen AG,
    1989).

    Table 21.  Composition of light-duty diesel engine exhaust
                                                                        

    Component                               Concentration
                                            (% by weight)
                                                                        

    Carbon dioxide                              7.1
    Water vapour                                2.6
    Oxygen                                     15.0
    Nitrogen                                   75.2
    Carbon monoxide                             0.03
    Hydrocarbons                                0.0007
    Nitrogen oxides                             0.03
    Hydrogen                                    0.002
    Sulfur dioxide                              0.01
    Sulfates                                    0.00016
    Aldehydes                                   0.0014
    Ammonia                                     0.00005
    Particulates                                0.006
                                                                        

    Adapted from Volkswagen AG (1989)

         Diesel exhaust contains toxicologically relevant compounds such
    as benzene, toluene, ethylbenzene, xylene, and PAHs. In the
    hydrocarbon fraction, benzene and toluene are present in relative
    concentrations of 11 and 4% by weight, respectively (Volkswagen AG,
    1989).

         Possible sources of PAHs in diesel exhaust are unburnt PAHs from
    the fuel, pyrosynthesis of PAHs during combustion, crankcase oils, and
    engine and/or exhaust system deposits (Scheepers & Bos, 1992a). These
    constituents occur partly in the gaseous fraction of total
    hydrocarbons (benzene, toluene, ethylbenzene, and xylene components,
    PAHs of lower relative molecular mass) and partly in the particulate
    fraction due to adsorption (PAHs of higher relative molecular mass).
    The partitioning of constituents between the particulate and gas
    phases has been measured by several investigators (Hampton et al.,
    1983; Schuetzle, 1983; Schuetzle & Frazier, 1986). Most PAHs with five
    or more rings and aliphatic hydrocarbons with more than 18 carbon
    atoms are expected to be primarily adsorbed onto particles (Schuetzle,
    1983). There is a direct relationship between adsorption behaviour and
    vapour pressure (Hampton et al., 1983); compounds with vapour
    pressures < 10-5 mbar (7.5 × 10-6 mm Hg; 2.09 kPa) at room
    temperature occur predominantly in the particulate phase. An empirical
    relationship between boiling-point and particle- to gas-phase
    partition coefficients has been derived for PAHs and aliphatic
    hydrocarbons in order to estimate gas-phase emission rates (Schuetzle
    & Frazier, 1986). According to this equation, the concentration of
    e.g. anthracene in the gaseous phase is about 20 times higher than
    that in the particle phase, whereas the levels of e.g. pyrene and
    fluoranthene are nearly equal in the two phases.  Oxygenated PAHs are
    found in diesel engine exhaust at concentrations as high (Volkswagen
    AG, 1989) or twice as high (Jensen & Hites, 1983) as those of their
    parent PAHs, perhaps because these engines have a larger air:fuel
    ratio than gasoline motors (Scheepers & Bos, 1992a). 9-Fluorenone and
    9,10-anthracenequinone are especially abundant in diesel exhaust
    (Schuetzle & Frazier, 1986). In tests of diesel passenger cars by
    Volkswagen AG (1989), phenalene-1-one, naphthalene dicarboxylic acid
    anhydride, cyclopenta[ def]phenanthrene-4-one, monoaldehydes of
    phenanthrene and anthracene, mono- and diketones of benzanthracenes
    and benzo[cd]pyrene, and benzofluorenones were also found (for
    artefact formation during sampling, see section B2.2.1; for emission
    factors, see section B3.1.1.1).

         Nitrogen-substituted PAHs are formed during combustion as a
    result of either nitration of PAHs by nitric acid or addition of
    nitrogen monoxide or dioxide to PAH free radicals generated during
    combustion (Schuetzle, 1983). In the exhaust of a four-cylinder diesel
    engine, nitrofluorene, dinitrofluorenone, nitroanthracene,
    nitrofluoranthene, mono- and dinitropyrenes, and nitrobenzo[ a]pyrene
    were detected at concentrations about one order of magnitude lower
    than those of oxygenated PAHs (Volkswagen AG, 1989).

         A number of investigators have found nitrated PAHs in extracts of
    diesel engine exhaust particles (Handa et al., 1983; Hartung et al.
    1984). Many three-, four-, and five-ring structures are involved
    (Schuetzle et al., 1982). The concentration of 1-nitropyrene, one of
    the prevalent species, has been reported to be 15-25 mg/kg particulate
    matter, whereas benzo[ a]pyrene has been found at concentrations up
    to 50 mg/kg. The nitrated PAHs are important in determining the health
    effects of exposure to diesel exhaust, since they are effective
    mutagens in microbial and human cell systems (Pederson & Siak, 1981;
    Pitts et al., 1982; Patton et al., 1986). Some nitrated PAHs are
    carcinogenic to animals (Ohgaki et al., 1985; Imaida et al., 1991).

         Combustion products originating from fuel and lubricating oil
    additives or from corrosion may also occur. For example, nitrate-based
    cetane improvers (see also section A2.1.2) are assumed to form nitro-
    PAHs (Organisation for Economic Co-operation and Development, 1993).

    B2.1.2  Type and composition of emitted particulate matter

         Diesel exhaust is characterized by a higher content (10-20-fold)
    of particulate matter than that of gasoline-fuelled passenger cars
    (Egebäck & Bertilsson, 1983; Volkswagen AG, 1989; Williams et al.,
    1989; CONCAWE, 1990b; Hammerle et al., 1994a). How soot particles are
    formed is still under discussion, but the precursor is probably
    acetylene generated during combustion. As a result of the numerous
    cracking, dehydration, and polymerization processes, quite large
    amounts of carbon, hydrogen, hydroxyl, and oxygen radicals are
    available in the combustion chamber, which can produce cyclic and
    polycyclic aromatic hydrocarbons from acetylene by polymerization and
    ring closure. With increasing polymerization, the carbon content of
    the macromolecule increases, finally generating graphite-like soot
    particles (Klingenberg et al., 1991).

         Investigations of two four-cylinder engines with direct and
    indirect injection (see also section B3.1.2.1) showed that the emitted
    particles range in shape and size from spherical particles of
    < 0.01 µm to cluster and chain agglomerates with maximal dimensions
    of 30 µm (Dolan et al., 1980; Amann & Siegla, 1982; Klingenberg et
    al., 1991). The formation of agglomerates (coagulation, 'aging') is
    dependent not only on engine and exhaust conditions but also on
    environmental factors, such as photochemical processes and humidity.
    Between 50 and 80% of the total particulate emission occurs in the
    range of 0.02-0.5 µm (Israel et al., 1982). According to Amann &
    Siegla (1982), the emitted particles have a concentric, lamellar
    structure arranged around the centre, similar to the structure of
    carbon black.

         Emitted soot particles are composed of elemental carbon, adsorbed
    organic compounds (soluble organic fraction) from fuel and lubricating
    oil, traces of metallic compounds (iron, calcium, and zinc at < 0.6%
    by weight; Volkswagen AG, 1989), and sulfates. The organic components
    of emitted particles are distributed over the surface. Numerous data
    on the relative proportions of organic compounds are available
    (Horvath et al., 1987; Williams et al., 1987; Volkswagen AG, 1989;
    Williams et al., 1989; CONCAWE, 1992). Depending on the engine
    conditions and testing cycle, the organic compounds comprise 10-90% of
    the total particulate mass. The ranges of various constituents of
    particulate matter in nine light-duty and four heavy-duty vehicles
    (see also section B3.1.2.1) are shown in Table 22. The driving cycles
    used were combined European Union tests (ECE-15 and 'high speed'
    extra-urban driving cycle [EUDC]) for the light-duty vehicles and
    modified ECE-R49 tests for the heavy-duty engines (see also section
    B2.2.1); four fuels with different sulfur contents and an aromatic
    content of about 25% by volume were used (CONCAWE, 1992). (For a
    description of the influence of fuel composition on exhaust emissions,
    see section B3.1.2.2.)

         This study showed considerable variation, and no obvious
    influence of fuel composition or engine conditions was discernible,
    although a difference was seen between fuel and lubricating oil
    hydrocarbons derived from light- and heavy-duty vehicles,
    respectively. This is probably due to the different driving cycles:
    transient in the case of light vehicles, steady state in the case of
    heavy-duty engines. In the individual data, adjustment of the engine
    appeared to play a major role. In the same study, PAHs were determined
    in the soluble organic fraction of the exhausts of three light-duty
    vehicles representing different engine conditions and of the four
    heavy-duty engines. The data are shown in Table 23. The constituents
    of the particulate matter of the exhaust are clearly dependent on the
    driving cycle used. Under steady-state conditions (heavy-duty
    engines), the PAH content is about one order of magnitude lower than
    that under transient conditions (light-duty vehicles). In addition,
    transient conditions lead to considerable variations in levels. In the
    study of Volkswagen AG (1989) on a four-cylinder diesel engine in the
    US-72-test, a hot start was used, and higher levels of fluoranthene,
    pyrene, benzo[ ghi]fluoranthene and benzo[ c]phenanthrene were
    found, perhaps due to either engine condition or driving cycle.
    According to Volkswagen AG (1989), some of the PAHs, and especially
    pyrene, benz[ a]anthracene and benzo[ a]pyrene, seemed to decompose
    during sampling (see section B2.2.1).

        Table 22.  Particulate matter content of exhausts from light- and heavy-duty vehicles
                                                                                                

    Constitutent         Particulate matter (% by weight)                  Reference
                                                                      
                         Light-duty vehiclesa    Heavy-duty vehiclesa
                                                                                                

    Carbon               32.6-85.1               39.7-81.7                 CONCAWE (1992)
                         82-88 (passenger        68-89                     Williams et al.
                         cars)                                             (1989)b

                         76-83 (light vans)
                         60-75                                             Hammerle et al.
                                                                           (1994a)c

    Fuel-derived         1.1-55.0                9.8-32.5                  CONCAWE (1992)
    hydrocarbons         6.0-18                                            Hammerle et al.
                                                                           (1994a)c

    Lubricating          4.2-54.4                4.0-25.9                  CONCAWE (1992)
    oil-derived          5.7-15                                            Hammerle et al.
    hydrocarbons                                                           (1994a)c

    Soluble organic      10.2-64.7               14.0-58.4                 CONCAWE (1992)
    fraction             14-26 (passenger        22-90                     Williams et al.
                         cars)                                             (1989)b

                         36-48 (light vans)
    Sulfate (excluding   0.6-12.2                1.4-7.5                   CONCAWE (1992)
    bound water)         3.3 (passenger          0.3-6.9                   Williams et al.
                         cars)                                             (1989)b

                         2.5 (light vans)        1.5-2.0                   Hammerle et al.
                                                                           (1994a)c
                                                                                                

    a  According to the Organisation for Economic Co-operation and Development (1993),
       light-duty vehicles are generally classified as vehicles with a weight < 3.5 t,
       occasionally < 5 t; heavy-duty vehicles are those weighing > 5 t.

    b  Measurements for four passenger cars and 15 light vans in different Australian
       driving cycles corresponding roughly to US-FTP-72 and -75 tests for heavy-duty
       vehicles, and 10 trucks and two buses in one Australian urban traffic driving cycle

    c  Five passenger cars and light vans in European Motor Vehicle Emissions group cycle
       and US-FTP-75 test

    Table 23.  Polycyclic aromatic hydrocarbon content of the soluble organic fraction of
               particulate matter in diesel exhaust
                                                                                                

    Constituent                Soluble organic fraction            Referenceb
                               (mg/kg particulate matter
                                                        
                               Light-duty      Heavy-duty
                               vehiclesa       vehicles
                                                                                                

    Fluoranthene               8-238           4-52                CONCAWE (1992)
                               553-560                             Volkswagen AG (1989)
    Pyrene                     11-287          7-90                CONCAWE (1992)
                               579-668                             Volkswagen AG (1989)
    Benzo[ghi]fluoranthene     36-61           3-6                 CONCAWE (1992)
    + benzo[c]phenanthrene     163-179                             Volkswagen AG (1989)
    Benz[a]anthracene          9-285           1-16                CONCAWE (1992)
                               61-82                               Volkswagen AG (1989)
    Triphenylene and           14-138          4-13                CONCAWE (1992)
    chrysene                   117-138                             Volkswagen AG (1989)
    Benzo[b,j]fluoranthene     4-87            1-10                CONCAWE (1992)
                               180-193c                            Volkswagen AG (1989)
    Benzo[a]fluoranthen        6-69                                CONCAWE (1992)
    Benzo[e]pyrene             8-40            1-2                 CONCAWE (1992)
                               85-90                               Volkswagen AG (1989)
    Benzo[a]pyrene             2-48            1-3                 CONCAWE (1992)
                               38-54                               Volkswagen AG (1989)
    Perylene                   5-32            2                   CONCAWE (1992)
                               4-8                                 Volkswagen AG (1989)
    Dibenz[a,j]anthracene      10-78                               CONCAWE (1992)
    Indeno[123-cd]pyrene       6-54                                CONCAWE (1992)
                               55-73                               Volkswagen AG (1989)
    Dinbenz[a,h]- or           3-40                                CONCAWE (1992)
    -[a,c]anthracene
    Benzo[ghi]perylene         25-65           1-5                 CONCAWE (1992)
                               75-82                               Volkswagen AG (1989)
    Anthracene                 3-7                                 CONCAWE (1992)
    Dibenz[a,i]pyrene          8-31                                CONCAWE (1992)
    Coronene and               9-52                                CONCAWE (1992)
    dibenzo[a,e]pyrene         33-48d                              Volkswagen AG (1989)
                                                                                                

    a  According to the Organisation for Economic Co-operation and Development (1993),
       light-duty vehicles are generally classified as vehicles with a weight < 3.5 t,
       occasionally < 5 t; heavy-duty vehicles are those weighing > 5 t.
    b  From Volkswagen AG (1989): US-72 tests with one four-cylinder engine
    c  Including benzo[k]fluoranthene
    d  Excluding dibenz[a,e]pyrene
    
    B2.2  Analytical methods

    B2.2.1  Sampling

         Driving cycles are specified to simulate on-road conditions for
    measuring vehicle emissions (Grimmer et al., 1988). The test
    procedures can be divided in two basic types: steady-state and
    transient cycle tests. In the European Union and Japan, mainly
    steady-state tests are used, in which emissions from engines are
    measured at specified combinations of speed and load. The most widely
    used official tests for light- and heavy-duty vehicles are as follows:

     Light-duty vehicles:

    *    US FTP-75: transient cycle; cold start; driving distance, 17.8
         km; 10-min break after 28.7 min (United States) (Volkswagen AG,
         1989)

    *    US SET: transient cycle; 'sulfate emission test'; hot start;
         simulating rush-hour freeway driving; driving distance, 21.7 km
         (United States) (Volkswagen AG, 1989)

    *    ECE 15: transient cycle; cold start; simulating city-centre
         driving; driving distance, 4.052 km (European Union) (CONCAWE,
         1990a)

    *    US HDC: steady-state cycle; 'highway driving cycle'; hot start;  
         driving distance, 16.5 km; maximal speed, 96.4 km/h (United      
         States) (Volkswagen AG, 1989)

    *    EUDC: steady-state cycle; 'extra-urban driving cycle'; hot start
         (after ECE 15); driving distance, 6.755 km; maximal speed,
         120km/h (European Union) (CONCAWE, 1990a)

     Heavy-duty vehicles:

    *    HD-FTP: transient cycle; 'heavy-duty federal test procedure';
         simulating metropolitan driving (United States) (Organisation for
         Economic Co-operation and Development, 1993);

    *    ECE 13: steady-state cycle at 13 different modes; identical to
         ECE R49 (European Union) (Organisation for Economic Co-operation
         and Development, 1993). Various bus cycles are also used to test
         heavy-duty vehicles that are characterized by frequent
         stop-and-go driving (Westerholm & Egebäck, 1991; Prakash et al.,
         1992).

         It is difficult to avoid artefacts in sampling. Exhaust gas
    constituents can undergo chemical reactions, adsorption and desorption
    processes, and condensation or diffusion. For example, oxygenated and
    nitrated PAHs are chemically unstable, and artefacts may form on the
    sampling device, especially if large sampling volumes are needed (Lee
    & Schuetzle, 1983). Both decreases and increases in the concentrations
    of PAHs and oxy-PAHs were observed during sampling in a study by
    Volkswagen AG (1989); in particular, degradation of pyrene and
    nitropyrene and formation of naphthalene dicarboxylic acid anhydride
    on the filter occurred. PAHs can be converted into nitroarenes (Pitts
    et al., 1978; Lee et al., 1980; Gibson et al., 1981; Lee & Schuetzle,
    1983; Schuetzle & Perez, 1983; Gaddo et al., 1984; Levsen et al.,
    1988).

         The sampling techniques used for the collection of the gaseous
    and particulate phases of diesel exhaust comprise dilution tube
    sampling and raw gas sampling.

    B2.2.1.1  Sampling from undiluted exhaust gas (raw gas sampling)

         The device for sampling motor vehicle exhaust (total flow method;
    Kraft & Lies, 1981; Grimmer et al., 1973) consists of a glass cooler,
    a condensate separator, and a special micron filter of
    paraffin-impregnated glass-fibre material providing separation of
    > 99.9% of particles of 0.3-0.5 µm. The cooler is fed with cold tap
    water, and the filter is placed on top of the cooler to remove any
    uncondensed liquid or solid exhaust gas constituents that may pass
    through the cooler coil. The particulate filter retains undissolved
    particles of the exhaust gas, while the condensate consists of the
    dissolved components (Volkswagen AG, 1989).

         The US Environmental Protection Agency (Lies et al., 1986)
    defines the 'particulate phase' as all materials (with the exception
    of condensed water) which, at a temperature of < 51.7°C, are retained
    on a special filter after dilution with air. Large volumes are
    collected with this type of sampling, making analysis of trace
    components difficult. The sample volumes are reduced by diverting the
    exhaust emission flow and sampling only a part. It is difficult to
    ensure a representative partial flow, as the amount of exhaust gas
    produced varies considerably over time (Volkswagen AG, 1989).

    B2.2.1.2  Sampling from diluted exhaust (dilution tube sampling)

         The exhaust is mixed with ambient air in a dilution tunnel in a
    constant volume sampler method (Behn et al., 1985). The total flow of
    air and exhaust is kept constant with a suitable fan in a constant
    volume sampler unit. This sampling technique more realistically

    reflects the immediate dilution of exhaust leaving the tailpipe of
    vehicles on the road. The concentrations of exhaust components are,
    however, reduced in accordance with the degree of dilution (magnitude,
    1:10), and components that are present only at low concentrations
    cannot be determined.

         There are various techniques for sampling gaseous constituents.
    Substances present at sufficiently high concentrations are commonly
    sampled in a plastic bag; adsorption onto solid materials, consisting
    of extremely porous, tiny plastic pellets (e.g. Tenax; diameter,
    0.1-0.2 mm; specific surface area, 20-50 m2/g), or absorption in
    liquids are also used. Particulate constituents are sampled by
    filtration (Lee & Schuetzle, 1983; Volkswagen AG, 1989).

    B2.2.2  Extraction from particles

         PAHs adsorbed onto diesel exhaust particles are of toxicological
    relevance, and there are numerous methods for extracting these
    constituents. The most frequently used method is Soxhlet extraction,
    which permits extraction of 99% of the soluble organic fraction within
    8 h; an additional 16 h are necessary to extract the remaining 1%.
    Extraction is frequently carried out by sequentially changing the
    solvents, from polar to nonpolar, e.g. toluene, toluene or  n-propanol,
    dichloromethane, and acetonitrile. PAHs may be further extracted by
    refluxing the exhaust particulates in solvent for 1 h. Sublimation and
    extraction by ultrasonic agitation are also used (Lee & Schuetzle,
    1983; Levsen, 1988), and supercritical fluids such as carbon dioxide
    are effective (Hawthorne & Miller, 1986). The extraction efficiency of
    various solvents is still under discussion (Köhler & Eichhoff, 1967;
    Swarin & Williams, 1980; Schuetzle & Perez, 1981; Grimmer et al.,
    1982; Lee & Schuetzle, 1983; Schuetzle et al., 1985; Levsen, 1988).
    Aromatic solvents and their combinations seem to be best suited for
    extracting PAHs from diesel particulates (Levsen, 1988).

    B2.2.3  Clean-up and fractionation

         Complex fractionation schemes for separating the exhaust extract
    into aliphatic compounds, aromatic compounds, and moderately and
    highly polar fractions after the removal of acidic and basic fractions
    have been described by Petersen & Chuang (1982) and Levsen (1992).
    During fractionation, further artefacts may be produced by the
    components of a complex sampling mixture like diesel exhaust,
    especially with acidic and basic extraction and aggressive column
    separation procedures.

    B2.2.4  Chemical analysis

         Chemical analysis of separated, cleaned compounds of diesel
    exhaust involves numerous techniques, including thin-layer
    chromatography, gas chromatography, gas chromatography with mass
    spectrometric detection, and high-performance liquid chromatography
    (US National Institute for Occupational Safety and Health, 1991).
    Jinno et al. (1986) determined PAHs by supercritical fluid
    chromatography with a photodiode array detector.

         The detection methods vary widely and include ultra-violet,
    visible, and fluorescence spectrophotometry, mass spectrometry, and
    more or less specific devices, such as flame ionization, nitrogen
    flame ionization, and sulfur-specific and electron-capture detectors
    (US National Institute for Occupational Safety and Health, 1991). A
    thermal optical method for determining elemental carbon is currently
    being used as a surrogate for measurement of worker exposure in almost
    all internal investigations at the US National Institute for
    Occupational Safety and Health (Niemeier, personal communication,
    1994).

         Analytical methods are available for determining the organic
    constituents of diesel particulate matter, with special emphasis on
    the determination of PAHs and their nitrated and oxygenated
    derivatives (Levsen, 1988).

    B2.3  Conversion factors

         Because diesel exhaust emissions are complex mixtures of
    different gases and particles, it is impossible to give a conversion
    factor for converting parts per million to SI units.

    B3.  SOURCES OF HUMAN AND ENVIRONMENTAL EXPOSURE

    B3.1  Anthropogenic sources

         Diesel exhaust emissions are exclusively man-made and are
    distributed primarily in the atmosphere. The hydrosphere and geosphere
    may also be affected owing to dry and wet deposition of the exhaust
    constituents; however, determination of the source of deposited
    compounds is difficult. No data are available on the individual
    components of diesel exhaust and their deposition.

    B3.1.1  Diesel exhaust emissions

         Diesel fuels are widely used as transport fuels, and exhaust
    emissions are released mainly by road traffic. Although many
    investigations of these emissions have been reported, the individual
    results are frequently not comparable because the experimental
    parameters, e.g. driving cycles, are different, or different
    dimensions were chosen. For example, the results of tests on emissions

    from light-duty vehicles refer mainly to kilometres driven, whereas
    those for heavy-duty vehicles are related to the engine power in
    kilowatt-hours. Furthermore, the statistical power of studies of
    emission factors is often limited by the small number of vehicles or
    engines examined. Some investigators change more than one parameter
    (e.g. sulfur content or aromaticity of the fuel) during their studies,
    so that the specific cause of changes in exhaust composition is
    difficult to assess. Adjustment of individual engine and fuel
    parameters plays a major role in determining the quality and quantity
    of exhaust emissions (see sections B3.1.2.1 and B3.1.2.2).

         Further sources of diesel exhaust emissions include off-road and
    stationary engines (reciprocating or gas turbine), electricity
    generating plants, pipeline pumping, gas compression, commercial and
    industrial furnace installations, and space and water heaters in
    plants and factories (see section A3.2.1.2). Diesel exhaust is also
    emitted from marine and inland ships' engines; residual fuels and
    other heavy grades (for example Bunker C) are used in ocean shipping
    (von Meyerinck, personal communication, 1994).

    B3.1.1.1  Emission of chemical constituents with the gaseous portion
              of diesel exhaust

         The exhaust emissions from diesel-fuelled vehicles are generally
    given as milligrams of component emitted per kilometre of driving
    distance (Egebäck & Bertilsson, 1983; Westerholm et al., 1986;
    Westerholm, 1987; Volkswagen AG, 1989; CONCAWE, 1990b; Westerholm &
    Egebäck, 1991; Prakash et al., 1992; Scheepers & Bos, 1992a). Data for
    verifying tests are issued by national regulatory bodies, e.g. the
    German Federal Office for Motor Vehicles (1993, 1994). Depending on
    the testing cycle, diesel engine, and fuel (see sections B3.1.1.1 and
    B3.1.1.2), the data show considerable differences even though the
    values are of the same order of magnitude. Emission factors used in
    different regions to estimate the effects of regulated chemical
    constituents (carbon monoxide, nitrogen oxides, and hydrocarbons) in
    air are shown in Table 24.

         Like all hydrocarbon fuel-burning engines, diesel engines release
    significant amounts of carbon dioxide. Westerholm & Egebäck (1991)
    measured emission factors of about 1 kg/km from a bus and a truck in
    the US transient cycle and the bus cycle. In measurements of emissions
    from six buses in four driving cycles (US transient cycle and three
    bus cycles), Prakash et al. (1992) found emission factors up to about
    4 kg/km, especially in the bus cycles. For five European diesel
    passenger cars and one light van, Hammerle et al. (1994a) determined
    emission factors of 162-252 g/km carbon dioxide in the European Motor
    Vehicle Emissions cycle and the US FTP-75 test. Diesel engines are the
    most efficient of all common types of combustion engines, with peak
    thermal efficiencies typically > 40%, owing to the relatively large
    air:fuel ratio during combustion (see section B3.1.2.1) (Organisation
    for Economic Co-operation and Development, 1993).


        Table 24.  Emission factors for diesel vehicles in 1985 used as standards for estimating total impact
                                                                                                                           

    Region                 Passenger cars       Commercial vehicles
                                                                                                                           
                           g/km      g/kg       Average                  < 5 t                  > 5 t
                                                                                                                  
                                                g/km        g/kg         g/km        g/kg       g/km          g/kg
                                                                                                                           

    Carbon monoxide
    Denmark                1.0                                                                  2.0
    Germanya               1.5       20                     16
    Netherlands            2.7       35         3.9         14           2.0         20         4.6           14
    European Union                   19                                              19                       16
    Austria                          22.8                                            22.8                     20
    Sweden                                                  10.5
    Switzerland                                                          4.0
    United Statesb                                                                              3.1-6.2

    Nitrogen oxides
    Germanya               1.0       14                     58
    Netherlands            0.8       10         18.3        66           0.6         6          22.8          69
    European Union                   14                                              18                       62
    Sweden                                                  39
    Switzerland                                                                                 14.0
    United Statesb                                                                              3.1

    Hydrocarbons
    Germanya               0.5       7                      12           
    Netherlands            0.6       10         4.6         17           0.9         9          6.3           19
                                                                                                                           

    Table 24 (contd)
                                                                                                                           

    Region                 Passenger cars       Commercial vehicles
                                                                                                                           
                           g/km      g/kg       Average                  < 5 t                  > 5 t
                                                                                                                  
                                                g/km        g/kg         g/km        g/kg       g/km          g/kg
                                                                                                                           

    Hydrocarbons (contd)
    European Union                   10                                              13                       9
    Sweden                                      3.9
    Switzerland                                                                                 3.0
    United Statesb                                                                              1.2
                                                                                                                           

    Adapted from Organisation for Economic Co-operation and Development (1993)
    a  Excluding former German Democratic Republic
    b  1975-83, trucks tested in the US transient cycle
             Cyanides, ammonia, and sulfur dioxide are detectable in the
    gaseous portion of diesel exhaust. The average emission factors given
    by Volkswagen AG (1989) for seven passenger diesel cars with four- and
    five-cylinder engines in three driving cycles (US 75 test with cold
    start, US sulfate emission test, US highway driving test) were
    0.6 mg/km cyanides, 1.1 mg/km ammonia, and 233.0 mg/km sulfur dioxide.

         The emission of individual organic components and classes of
    compounds, summarized as total hydrocarbons, are shown in Table 25,
    which provides an overview of the quantities of individual substances
    emitted in diesel exhausts. Heavy-duty vehicles release more
    low-boiling hydrocarbons than passenger cars; however, the difference
    could be due to the driving cycle used, which involves frequent
    changes of speed and load (see section B2.2.1). The emissions of
    adsorbed organic components, such as PAHs (see section B3.1.1.2), are
    correlated with the release of particulate matter.

    B3.1.1.2  Emission of particulate matter and adsorbed components in
              diesel exhaust gases

         The emission of total particulate matter from diesel exhaust is
    shown in Table 26. Analysis of the particulate matter released from
    heavy-duty vehicles gave the following emission factors (Westerholm &
    Egebäck, 1991): 170-370 mg/km carbon, 8-71 mg/km soluble organic
    compounds derived from fuel, 56-167 mg/km soluble organic compounds
    derived from lubricating oil, 1.4-37 mg/km soluble sulfates, and
    0.6-3.7 mg/km nitrate. Comparable values were not available for
    light-duty vehicles.

         As PAHs and oxygenated PAHs from diesel and spark-ignition
    engines are qualitatively similar, their differentiation in the
    environment is difficult (Behymer & Hites, 1984); however,
    benzo[ b]naphtho-[2,1-d]thiophene is specific for diesel exhaust and
    can therefore be used as a marker (Grimmer et al., 1979). Emissions of
    PAH components are shown in Table 27.

         In experiments with one bus and one truck, Westerholm & Egebäck
    (1991) detected 30-220 µg/km of particulate-associated PAHs. The
    compounds measured included dibenzothiophene and its 4- and 3-methyl
    derivatives, 2-methylfluorene, phenanthrene and its methylated
    derivatives, anthracene and its 2-methyl derivative, fluoranthene,
    pyrene and its 1- and 2-methyl derivatives, retene, benzo[ a]fluorene,
    benzo[ ghi]fluoranthene, cyclopenta[ cd]pyrene, benzo[ a]anthracene,
    chrysene or triphenylene, benzo[ b,k]fluoranthene, benzo[ e]pyrene,
    benzo[ a]pyrene, perylene, indeno[1,2,3-cd]fluoranthene,
    indeno[1,2,3-cd]pyrene, picene, benzo[ ghi]perylene, and coronene.
    The emission factors for 1-nitropyrene were 0.13-7.3 µg/km.

    Table 25.  Emission factors for individual organic constituents and
               classes of compounds in the gaseous portion of diesel
               exhaust
                                                                        

    Class of compound   Emission factor (mg/km)
    or constituent                                                      

                        Light-duty vehiclesa   Heavy-duty vehiclesa
                                                                        

    Straight-chain aliphatic hydrocarbons

    Sum of C1-C4        5.3-6.6 (C1 + C2)b
                        < 1 - < 143c
    Sum of > C4         < 1-5c
    Olefins
    Ethylene            25.6-90.7d             32-135e
                        13.5-16.9b             19.0-82.9f
                        < 1-31c
    Propylene                                  7.3-22e
                                               8.1-44.0f
    1,3-Butadiene       0.2g
    Acetylene           5.7-7.4b               27.9-106.9 (+ ethane)f
    Alcohols            Methanol               1.6d
    Ethanol             < 1.2-22.2d
    Aldehydes
    Formaldehyde        8.1b                   60-255e
                        7.9-20.7d              140-150h
                        4.2-9.1g               19.0-207.4 f*
    Acetaldehyde        3.1b                   18-137e
                        < 22-36d               40-50h
                        2.0-4.4g
    Crotonaldehyde      1.2b
    Acrolein            < 1-1.8d               8.5-36e
                                               4h
    Benzaldehyde        0.6b                   10-14h

    Light aromatic compounds

    Benzene             2.6-4.3b               4.5-13.1e
                        ND-6c                  6.9-30.5f
                        9.5-18.1d
                        0.15-3.47g
                                                                        

    Table 25. (cont'd)
                                                                        

    Class of compound   Emission factor (mg/km)
    or constituent                                                      

                        Light-duty vehiclesa   Heavy-duty vehiclesa
                                                                          


    Toluene             1.1-1.9b               2-13.7e
                        ND-2c                  7.4-32.5f
                        < 1.0d
    Ethylbenzene        NDc
    Xylenes             NDc                    7.8-41.7f

    Phenols             0.7-1.5b
                                                                        

    ND, not determined
    a  According to the Organisation for Economic Co-operation and
       Development (1993), light-duty vehicles are generally classified as
       vehicles with a weight < 3.5 t, occasionally < 5 t; heavy-duty
       vehicles are those weighing > 5 t.

    b  From Volkswagen AG (1989): various numbers of diesel passenger cars
       in three driving cycles (US-75 test with cold start, US sulfate
       emission test, US highway driving test); determination of aldehydes
       only in the US-75 test

    c  From Scheepers & Bos (1992a): two diesel passenger cars (< 50 000 and
       > 100 000 km on odometer) in ECE test, 15 min at 90 km/h and 15 min
       at 120 km/h

    d  From Egebäck & Bertilsson (1983): two passenger cars in US-78 test

    e  From Westerholm & Egebäck (1991): a bus and a truck in bus cycle with
       different diesel fuels

    f  From Prakash et al. (1992): one bus in four driving cycles (US
       transient cycle, three bus cycles) with the exception (*) of eight
       buses in the same driving cycles

    g  From Hammerle et al. (1994a): five passenger cars and one light van
       in European Motor Vehicle Emissions Group cycle

    h  From Westerholm et al. (1986): one heavy-duty vehicle on a chassis
       dynamometer with two diesel fuels

    i  Mainly phenols, cresols and xylenols


        Table 26.  Emission factors for total particulate matter from diesel exhaust
                                                                                                                                                

    Measurement conditions                                     Emission factor (mg/km)        Reference
                                                                                                                                                

    Light-duty vehiclesa

    Seven passenger cars in three driving cycles               166.4-238.7                    Volkswagen AG (1989)
    (US-FTP-75 test with cold start; US sulfate emission
    test; US highway driving test)

    Four passenger cars, 15 light vans in three                180-250 (passenger cars)       Williams et al. (1989)
    Australian driving cycles (corresponding roughly to        360-440 (light vans)
    US-72 and US-75 tests)

    Seven passenger cars, two light vans in combined           70-365 (passenger cars)        CONCAWE (1990b)
    ECE 15/EUDC test with four diesel fuels                    233-609 (light vans)

    Various                                                    Germanyb, 300                  Organisation for Economic
                                                               Netherlands, 600               Co-operation and Development
                                                               Sweden, 400                    (1993)

    61 passenger cars and 30 light vans without catalyst       39-188 (mean, 104)             Federal Office for Motor Vehicles
    (optimally tuned) incombined ECE 15/EUDC test                                             (1993)

    73 passenger cars and 45 light vans with catalyst          48-127 (mean, 90)              Federal Office for Motor Vehicles
    (optimally tuned) in combined ECE 15/EUDC test                                            (1993)
                                                                                                                                                

    Table 26 (contd)
                                                                                                                                                

    Measurement conditions                                     Emission factor (mg/km)        Reference
                                                                                                                                                

    44 passenger cars and 8 light vans without catalyst        52-117 (mean, 95)              Federal Office for Motor Vehicles
    (optimally tuned) inUS FTP-75 test                                                        (1993)

    118 passenger cars and 34 light vans with catalyst         50-118 (mean, 71)              Federal Office for Motor Vehicles
    (optimally tuned) inUS FTP-75 test                                                        (1993)

    Five passenger cars and 1 light van in European Motor      52-117                         Hammerle et al. (1994a)
    Vehicle Emissions Group cycle and US FTP-75 test

    Heavy-duty vehiclesa

    One vehicle on chassis dynamometer (bus cycle);            680-1000                       Westerholm et al. (1986)
    two diesel fuels with different contents of sulfur and
    aromatic compounds

    10 trucks, 2 buses in one Australian driving cycle         800-7150                       Williams et al. (1989)
    simulating urban traffic

    One bus, one truck in two driving cycles (bus cycle;       230-600                        Westerholm & Egebäck (1991)
    US transient cycle) with two diesel fuels

    106 trucks and buses in ECE-R49 driving cycle              0.056-0.606 g/kWh              Federal Office for Motor Vehicles
                                                                                              (1994)
                                                                                                                                                

    Table 26 (contd)
                                                                                                                                                

    Measurement conditions                                     Emission factor (mg/km)        Reference
                                                                                                                                                

    Various                                                    Netherlands, 800 (< 5 t)       Organisation for Economic
                                                                           3300 (> 5 t)       Co-operation and Development
                                                               Sweden,      400 (< 5 t)       (1993)
                                                                           1700 (> 5 t)

    13 trucks and buses (three with particulate trap,          584-804 (without trap)         Lowenthal et al. (1994)
    10 without) in bus cycle with two diesel fuels             67-206 (with trap)
                                                                                                                                                

    a  According to the Organisation for Economic Co-operation and Development (1993), light-duty vehicles are generally classified as
       vehicles with a weight <  3.5 t, occasionally < 5 t; heavy-duty vehicles are those weighing > 5 t.
    b  Without the former German Democratic Republic
        Table 27.  Emission factors for polycyclic aromatic hydrocarbons in
               diesel exhaust
                                                                        

    Constituent              Emission factor (µg/km)
                                                                        

                             Light-duty vehiclesa   Heavy-duty vehiclesa
                                                                        

    Anthracene               17-63b                 0.9-3.2c; 9-26d;
                                                    3-14d+; 1.6e
    Phenanthrene             140f                   4.2-25c
                             295-524b               163-308d; 79-186d+;
                                                    12.2e
    Fluoranthene             172f                   10-39c
                             58-200g                27-34d; 14-17d+
                             60.4-79.7g             13.0e
    Pyrene                   186                    15-102c
                             < 0.9-22               28-31d; 9-21d+
                             66.6-87.1              22.6e
    Chrysene                 40                     9.5-26c
                             14-67                  8-32d; 4-12d+
                             10.7-22.3              Not measurede
    Benzo[b,k]fluoranthene   40                     1.6-8.4c
                             2.6-47                 9-12d,h; 2-8d,h+
                             10.3-15.2              5.6e
    Benzo[ghi]fluoranthene   40f                    1.4-13c; 6.9e
    Cyclopenta[cd]pyrene     2f                     2.5-11c
                             0.81ag*                1.4e
    Benzo[e]pyrene           3-38b                  1.6-23c
                             13.8-16.9              5-9d; 1-7d+; 2.6e
    Benzo[a]pyrene           10                     0.8-12c
                             < 1-19                 2-3d; 0.4-2d+
                             4.0-5.3                1.3e
    Benzo[a]anthracene       25                     3.0-12c
                             8-43                   5-12d; 1-4d+
                             2.7-3.9                3.6e
    Indeno[123-cd]pyrene     15f                    0.7-15c
                             3.5-4.4g
    Benzo[ghi]perylene       20                     0.6-30c
                             < 1-18                 2-4d; 0.5-1d+
                             8.7-11.7               Not detected
                                                    (detection limit
                                                    not given)e
                                                                        

    Table 27 (contd)
                                                                        

    Constituent              Emission factor (µg/km)
                                                                        

                             Light-duty vehiclesa   Heavy-duty vehiclesa
                                                                        

    Perylene                 < 1-2b                 Not measuredc
                             0.5-0.6g               1.0e
    Coronene                 10f                    < 0.1-12c
                             5.43g*                 Not detected
                                                    (detection limit
                                                    not given)e
                                                                        

    a  According to the Organisation for Economic Co-operation and
       Development (1993), light-duty vehicles are generally classified as
       vehicles with a weight < 3.5 t, occasionally < 5 t; heavy-duty
       vehicles are those weighing > 5 t.
    b  From Scheepers & Bos (1992a): two diesel passenger cars (< 50 000
       and > 100 000 km on odometer) in ECE test
    c  From Westerholm et al. (1986): one heavy-duty vehicle on a chassis
       dynamometer (bus cycle)
    d  From Lowenthal et al. (1994): 13 trucks and buses (three with
       particulate trap [d+], 10 without) in bus cycle
    e  From Rogge et al. (1993): two heavy-duty vehicles in special
       driving cycle
    f  From Egebäck & Bertilsson (1983): average of two passenger cars in
       US-78 test
    g  From Volkswagen AG (1989): seven passenger cars in three driving
       cycles (US-75 test with cold start, US sulfate emission test, US
       highway driving test) with the exception (*) of one four-cylinder
       engine, US-72 test with hot start
    h  Isomers not specified

         The emission of oxygenated PAHs by two diesel passenger cars (one
    four-cylinder and one five-cylinder engine, US 72 test with hot start)
    was 4-40 µg/km. In a comparable test (one four-cylinder engine, same
    driving cycle), the emission of nitro-PAHs was 0.1-3.5 µg/km. As a
    result of possible artefact formation during sampling (see section
    B2.2.1), these measurements are uncertain (Volkswagen AG, 1989). It is
    suspected that nitrate-based cetane improvers (see section A2.1.2)
    could increase nitro-PAH emissions in exhaust (Organisation for
    Economic Co-operation and Development, 1993).

         There are no significant differences in the emissions of adsorbed
    single substances between light- and heavy-duty vehicles, although
    heavy-duty vehicles release larger relative amounts of particulate
    matter (Salmeen et al., 1985).

         The chlorine content of diesel fuel is typically < 1 ppm
    (Oehme et al., 1991). Thus, no dioxins were found in the exhaust of
    heavy vehicles, at a detection limit of 100 pg/litre of fuel (Marklund
    et al., 1990). The dioxin emissions from one light-duty (indirect
    injection) and one heavy-duty (direct injection) diesel engine under
    different load conditions on a chassis dynamometer were 0.009-0.141 ng
    of toxicity equivalents per litre of fuel consumed. The highest value
    was reported in emissions from the light-duty engine at full load.
    Overall, the emissions were of the same order of magnitude as the
    dioxin emissions from a petrol-fuelled engine with a catalytic
    converter (Schwind et al., 1991).

    B3.1.2  Parameters that influence diesel exhaust emissions

    B3.1.2.1  Engine conditions

         In general, diesel engines are operated with excess air; the
    air:fuel ratio is about 1.5, and at lower air:fuel ratios, much more
    smoke is emitted (Organisation for Economic Co-operation and
    Development, 1993). Wide local variations in the equivalence ratio and
    the temperature of the mixture are caused by the heterogeneous nature
    of the combustion process. To avoid over-rich mixtures and to improve
    the air:fuel mixing processes, wall wetting and delay of fuel
    injection are used (CONCAWE, 1986). In naturally aspirated engines,
    the amount of air in the cylinder is independent of the power output.
    For these engines, the maximal power output is limited by the amount
    of fuel that can be injected without smoke formation. In turbocharged
    engines, however, increasing the amount of fuel injected increases the
    energy in the exhaust gas, and the turbocharger pumps more air into
    the combustion chamber. For this reason, the power output of
    turbocharged engines is not, in general, limited by smoke. Low
    air:fuel ratios can occur only during transient acceleration. Most
    turbocharged engines are further equipped with a cooling system
    (intercooler) in which compressed air reduces adverse thermal effects.
    In both types of engines, fuel is injected either directly into the
    combustion chamber or indirectly into a separate pre-chamber where it
    is mixed and partly burnt before reaching the main combustion chamber
    (Organisation for Economic Co-operation and Development, 1993).
    Scheepers & Bos (1992a) reported that most heavy-duty vehicles have
    direct injection motors, and indirect injection is used for light-duty
    vehicles.

         The most important influences of engine conditions on diesel
    exhaust emissions are summarized in Table 28. Contrasting results are
    sometimes obtained with regard to the effect of engine conditions on
    diesel exhaust emissions. Adjustment of the engine plays a major role
    (CONCAWE, 1990b), which was slightly less in the US FTP-75 test cycle
    than in the European cycle.


        Table 28.  Influence of engine conditions on diesel exhaust emissions
                                                                                                                                                

    Condition          Change       Influence on exhaust emissions                                        Reference
                                                                                                                                                

    Air:fuel ratio     Increase     Increased emissions of hydrocarbons and particulate soluble           Organisation for Economic
                                    fraction; decreased particulate and nitrogen oxide emissions          Co-operation and
                                    from turbocharged and intercooled engines                             Development (1993)

                                    Increased PAH emissions, decreased particulate emissions from         Barbella et al. (1988)
                                    direct injection engines; decreased PAH and particulate emissions
                                    from indirect injection engines

                                    Increased nitrogen oxide emissions, decreased particulate             Iida et al. (1986)
                                    emissions from direct injection engines

    Injection timing   Advance      Increased particulate emissions from indirect injection engines       Du et al. (1984); Williams et
                                                                                                          al. (1989)

                                    Decreased particulate and hydrocarbon emissions; increased            Organisation for Economic
                                    nitrogen oxide emissions                                              Co-operation and Development (1993)

                                    No influence on concentrations of PAHs and oxy-PAHs adsorbed          Jensen & Hites (1983)
                                    on particulate matter

                       Delay        Decreased nitrogen oxide emissions, increased volumetric fuel         Organisation for Economic
                                    consumption                                                           Co-operation and Development (1993)
                                                                                                                                                

    Table 28.  Influence of engine conditions on diesel exhaust emissions
                                                                                                                                                

    Condition          Change       Influence on exhaust emissions                                        Reference
                                                                                                                                                

    Injection timing   Delay        Decreased particulate emissions from indirect injection engines;      Williams et al. (1989)
    (contd)                         increased particulate emissions from direct injection engines
                                    Small effect on PAH emissions from indirect injection engines         Schuetzle & Frazier (1986)

                                    Small effect on particulate emissions from direct injection engines   Campbell et al. (1981a)

                                    No influence on concentrations of PAHs and oxy-PAHs adsorbed          Jensen & Hites (1983)
                                    on particulate matter

    Load and           Increase     Increased particulate and PAH emissions at high loads; decreased      Pipho et al. (1986)
    temperature                     particulate emissions at very high load from indirect injection
                                    engines

                                    Decreased concentrations of PAHs and oxy-PAHs adsorbed on             Jensen & Hites (1983)
                                    particulate matter

                       Decrease     Decreased particulate and nitrogen oxide emissions; extremely         Organisation for Economic
                                    cold charge air may increase hydrocarbon emissions                    Co-operation and
                                                                                                          Development (1993)

                                    Lowest PAH emissions at medium load; enrichment of three-             Pipho et al. (1986)
                                    and four-ring PAHs during idling (indirect injection engines)
                                                                                                                                                

    PAH, polycyclic aromatic hydrocarbons
        (a)  Engine aging

         Increased emissions of hydrocarbons, particulates, and PAHs have
    been reported from older, more intensively used light- and heavy-duty
    vehicle engines in comparison with newer ones (Stenberg et al., 1983;
    Braddock & Perry, 1986; Scheepers & Bos, 1992a). Stenberg et al.
    (1983) proposed that the combustion of more lubricating oil in older
    vehicles contributes to the enhanced emissions. Prakash et al. (1992)
    measured the emission characteristics of several buses and reported
    that a properly maintained engine would not emit significantly more
    hydrocarbons and nitrogen oxides during its lifetime, whereas carbon
    monoxide and particulate emissions could increase with age. These
    findings are similar to those of Volkswagen AG (1989) on the influence
    of distance travelled on exhaust emissions from one diesel passenger
    car over a total distance of about 30 000 km in three driving cycles
    (US FTP test, US sulfate emission test, US highway driving test). The
    average deterioration factors, relative to the emissions at about
    2500 km, were carbon monoxide, 0.97; nitrogen oxides, 0.96;
    hydrocarbons, 0.91; particulates, 1.07; fluoranthene, 0.96; pyrene,
    1.06; chrysene, 0.62; benzo[ b,k]fluoranthene, 1.11; benzo[ e]pyrene,
    1.51; benzo[ a]pyrene, 0.68; perylene, 0.96; indeno[1,2,3-cd]pyrene,
    2.91; and benzo[ ghi]perylene, 3.43.

    (b)  Lubricating oil

         As a result of emission control techniques such as turbocharging,
    intercooling, and injection timing, the relative importance of
    emissions caused by lubricating oils has increased (Organisation for
    Economic Co-operation and Development, 1993). Lubricating oils have
    been hypothesized to be a sink for incomplete combustion products such
    as PAHs and their derivatives, which are scrubbed from the exhaust
    gas. When this 'enriched' lubricating oil leaks into the combustion
    chamber, the constituents may be emitted in the exhaust (Scheepers &
    Bos, 1992a; see section B2.1.1). McKee & Plutnick (1989) found,
    however, that carcinogenic PAHs accumulate in gasoline- but not in
    diesel-fuelled engine oils.

         Oil consumption can be reduced by improving engine manufacturing
    specifications and engine seals (Organisation for Economic
    Co-operation and Development, 1993). The particle-bound hydrocarbons
    derived from lubricating oil contribute up to about 50% by weight in
    light-duty vehicles and to a maximum of about 26% by weight of the
    total emitted particulate matter (see section B2.1.2, Table 22).
    Experiments with 13C radiolabelled lubricating oil in three
    light-duty vehicles showed that addition of up to 3% lubricating oil
    increases the emitted particulate matter by about 50%, and the
    particulates consist mainly of the soluble organic fraction adsorbed
    onto the particles (Williams et al., 1989).

    B3.1.2.2  Fuel specification

         Fuel parameters influence the particulate emissions of diesel
    engines but seem to be less important than other parameters like
    engine adjustment, load, and maintenance (Weidmann et al., 1988).     
    The actual trend in legislation of the quality of diesel fuel, i.e. to
    a lower sulfur and aromatic content, will improve emissions,
    especially of particles. In addition, the more stringent emission
    standards in many countries will decrease diesel-related emissions.

    (a)  Sulfur content

         There have been numerous investigations on the influence of
    sulfur content in diesel fuel on the release of particulate matter
    (CONCAWE, 1990b; Van Beckhoven, 1991; Westerholm & Egebäck, 1991;
    Scheepers & Bos, 1992a). The sulfur content is presumed to be directly
    correlated with the release of particulate matter. About 3-5% of the
    fuel sulfur in light-duty vehicles and 2-3% of that in heavy-duty
    vehicles reacts with metal sulfates which are adsorbed onto the
    particulate matter, increasing its mass and its hygroscopic
    properties, as metal sulfates tend to absorb significant amounts of
    water from the atmosphere (Organisation for Economic Co-operation and
    Development, 1993).

         A reduced sulfur content leads to decreased levels of
    particulates in the exhaust of light-duty vehicles. For example,
    investigations on seven passenger cars and two light vans in a
    combined ECE 15/EUDC testing cycle with four diesel fuels of different
    sulfur contents (0.055, 0.12, 0.22, and 0.31% by weight) showed a
    statistically significant reduction in the emission of particulates up
    to a maximum of 7.4%. The investigators noted variations in
    particulate levels between the different vehicles, however, that were
    about two orders of magnitude higher than could be attributed directly
    to the influence of the sulfur content (CONCAWE, 1990b).

         Reduction of the emission rates of heavy-duty vehicles by
    lowering the fuel sulfur content is still under discussion. In a study
    of four heavy-duty engines in the ECE R49-13 mode driving cycle with
    the same four fuels as described above, no statistically significant
    reduction in emitted particulate matte was seen (CONCAWE, 1990b). In
    experiments with one heavy-duty vehicle on a chassis dynamometer bus
    cycle with two different fuels, however, a considerable reduction in
    the release of particles was found (Westerholm et al., 1986). This
    result was confirmed by Westerholm & Egebäck (1991) who investigated
    one bus and one truck in three driving cycles (ECE 13 test, bus cycle,
    US transient test) with eight diesel fuels of different sulfur and
    aromatic contents. Comparable results were found by Gairing et al.
    (1994) in tests with one heavy-duty engine (direct injection) in ECE
    R49 with 12 different fuels. Adjustment of the engines again appeared
    to be a decisive factor.

    (b)  Aromaticity (density)

         Aromaticity can be directly associated with density, as a higher
    density is frequently achieved by a larger content of aromatic
    compounds. Obviously, increasing the aromaticity leads to increased
    emissions of particulate matter, although volumetric fuel consumption
    is decreased (CONCAWE, 1986). This finding was confirmed by Hare
    (1986) for light- and heavy-duty vehicles on the basis of data from
    the automobile industry. The presumed increase in the release of
    particulates is 0.5-1.9% for a 1% by volume change in aromatic
    content. In a study of four light-duty vehicles in three Australian
    driving cycles (resembling roughly the US 72 and 75 tests), Williams
    et al. (1989) found reductions in particulate emissions of 22-27% at a
    10% difference in fuel aromaticity. Unfortunately, the two fuels also
    differed in sulfur content and volatility. Van Beckhoven (1991)
    achieved comparable results with passenger cars with indirect
    injection but found no significant influence for heavy-duty direct
    injection engines. Experiments by Westerholm & Egebäck (1991) on one
    bus and one truck showed a reduction in the emission of particulates
    when fuel with low levels of sulfur and aromatics was used. The
    increased aromatic content is probably also the source of higher
    levels of PAHs and nitro-PAHs in the exhaust (Schuetzle & Frazier,
    1986; Westerholm, 1987; Westerholm & Egebäck, 1991; Organisation for
    Economic Co-operation and Development, 1993).

    (c)  Volatility

         The volatility of diesel fuels is described by the boiling range,
    and especially by the 10% boiling-point (low boiling end fraction) and
    the 90% boiling-point (high boiling end fraction) of the mixture. The
    influence of the boiling range, particularly the 90% boiling-point, on
    the release of particulate matter is still under discussion. While
    some investigators (Hare, 1986; Williams et al., 1989) assume that
    emissions of particulate matter increase with an increased 90%
    boiling-point, others have not been able to confirm this effect,
    suggesting that other parameters, such as sulfur content and
    aromaticity, are responsible for the increase (Westerholm & Egebäck,
    1991; Organisation for Economic Co-operation and Development, 1993).

    (d)  Cetane number

         An increase in cetane number lowers the emission of diesel
    particulate matter. A slight decrease in cetane number, combined with
    a small increase in the density of diesel fuels (see section A2.1.3),
    was predicted in 1987 to result in an increase in emissions from 3 to
    12% for carbon monoxide, and to maxima of 2% for nitrogen oxides, 7%
    for hydrocarbons, and 6% for particles (CONCAWE, 1987).

    B3.1.2.3  Malfunction

         Malfunctions or defective components of the engine or exhaust gas
    system may result in considerably higher emissions than those from a
    properly maintained engine. Up to twofold increases in emissions of
    carbon monoxide, hydrocarbons, and particulate matter have been
    measured even in the absence of a significant effect on power or fuel
    consumption (Ullman et al., 1984). Malfunction of the fuel injection
    system may contributed markedly to higher emissions (Organisation for
    Economic Co-operation and Development, 1993). Volkswagen AG (1989)
    tested a diesel passenger car with defective injection nozzles in
    three driving cycles (US FTP test, US sulfate emission test, US
    highway driving cycle) and found statistically significant increases
    in the emissions of carbon monoxide (up to about 25%), hydrocarbons
    (up to about 45%), and particulate matter (up to about 15%), whereas
    the release of nitrogen oxides was almost unaffected. The amounts of
    benzene and PAHs emitted were slightly increased. Comparable results
    were obtained by Williams et al. (1989) for the release of particles
    by several light-duty vehicles equipped with defective injectors.
    Malfunction of US medium- to heavy-duty vehicle diesel engines
    increased the amount of nitrogen oxide emitted to about 4%, that of
    hydrocarbons to about 75%, and that of particulates to about 140%
    (Organisation for Economic Co-operation and Development, 1993).

    B3.1.3  Total emissions by diesel engines

         Few data are available on the emissions of different constituents
    of diesel exhaust into the atmosphere and their contribution to the
    total release by traffic and industry. The contribution of heavy-duty
    vehicles seems to be much greater than that of passenger cars, despite
    lower percent kilometres. For example, in westen Germany, only 8% of
    the total annual urban mileage was attributable to heavy-duty vehicles
    (Fromme, 1990). In contrast, in 1987, power plants contributed 15.3%
    (85 000 t), industry 63.5% (360 000 t), domestic and minor consumers
    8.1% (44 000 t), and total traffic 13% (72 000 t) to the particles
    released (Metz, 1989). The contribution of diesel particulate matter
    to the total particle burden of the atmosphere is highly dependent on
    local conditions (Table 29).

         Diesel vehicles, particularly in the light-duty category,
    represent a smaller proportion of motor traffic in the United States
    than in most European countries. Measurements from two heavy-duty
    vehicles in 1982 indicated that light- and heavy-duty diesel vehicles
    contributed about 6% to the total fine aerosol, organic carbon
    emissions in the Los Angeles basin, about 1200 kg/day being emitted by
    heavy-duty vehicles and 600-870 kg/day by light-duty vehicles
    (Hildemann et al., 1991; Rogge et al., 1993).

    B3.1.4  Control of emissions

         The automotive industry has made efforts to reduce diesel exhaust
    emissions (Obländer, 1991; Fortnagel, 1992; Hammerle et al., 1994b),
    and technology is advancing.

    B3.1.4.1  Particle traps

         Particle traps are placed in the exhaust system to collect
    particulate matter and to clean the filter by burning the trapped
    particles (trap oxidizers). Most traps are based on cellular ceramic
    monolith material or ceramic silica fibre coils. Many regeneration
    techniques have been developed for trap oxidizers (Organisation for
    Economic Co-operation and Development, 1993), including thermal
    regeneration, which is the burning of collected particulate matter,
    and catalytic regeneration, in which fuel catalysts, e.g. iron and
    copper, containing organic compounds form finely dispersed metal
    particles during combustion to reduce the burning temperature of the
    diesel particles so that there is more efficient regeneration
    (Volkswagen AG, 1989; Hammerle et al., 1994b). Traps remove the
    soluble organic fraction adsorbed onto particles and the dry
    carbonaceous soot, while catalysts can remove substances only from the
    soluble organic fraction. In catalytic regeneration, over 95% of the
    metal used as fuel catalyst is retained in the trap (Hammerle et al.,
    1994b).

         Studies of the emissions of one light-duty diesel passenger car
    by Volkswagen AG (1989) showed that the release of particulates was
    reduced by a thermal trap by at least 80% during the collection phase
    and by about 60% during the regeneration phase; the reduction was 80%
    for the whole cycle. Sulfate emissions decreased to about 50%;
    however, emission of carbon monoxide increased by about 10%, probably
    due to additional burning and/or incomplete oxidation of particulate
    matter on the trap. Comparable experiments with a catalytically
    regenerated trap, with addition of a manganese compound, also showed a
    trap efficiency of about 80% for particulate matter. Hydrocarbon and
    carbon monoxide emissions were slightly increased, probably due to
    incomplete oxidation on the trap. Sulfate emissions were reduced to
    less than half, emissions of PAHs were reduced to about 25%, and those
    of phenol and aldehydes to about 25-33%. About 10% of the manganese
    catalyst was detected in the exhaust.

         Measurements with a heavy-duty vehicle on a chassis dynamometer
    to test two particle traps showed decreases of 79 and 86% in the
    release of particles, depending on the trap and fuel characteristics.
    The emissions of hydrocarbon, carbon monoxide, aldehydes, and PAHs
    were slightly reduced (Westerholm et al., 1986).


        Table 29.  Diesel exhaust emissions and their contribution to total anthropogenic emissions
                                                                                                                                                

    Region                 Year      Annual emissions (kt/year)                                               Reference
                                                                                                        
                                     NOx            CO            HC            PM              SO2

                                                                                                                                                

    Light-duty vehicles
    Germanya               1986      52 (3.3)b      76 (1.2)b     23 (2.2)b     15 (26.3)b      NR            Fromme (1990)
                           1987      NR             NR            NR            12 (2.2)c       NR            Metz (1989)

    United Kingdom         1991      (0.7)c         (0.4)c        (1.1)c        (2)c,d          (0.2)c        Quality of Urban Air Review
                                                                                                              Group (1993)

    Light- and heavy-duty vehicles
    Europe                 NR        NR             NR            NR            NR              (1.6)c        OECD (1993)

    United States          1984      NR             NR            NR            NR              620 (3)c      OECD (1993)
                           1986      NR             NR            NR            54.8 (20)b      NR            Carey (1987)
                                                                                (4)c

    Germanya               1989      640            240           150           55              NR            Federal Environment
                                                                                                              Agency(1991)

    Heavy-duty vehicles
    OECD (Europe)          1980      1548 (13)c     NR            393 (4)c      NR              361 (2)c      OECD (1993)

    Germanya               1986      502 (32.2)b    130 (2.1)b    104 (10.0)b   37 (64.7)b      35 (2)c       OECD (1993); Fromme (1990)
                                                    (17)c         (2)c          (4)c            (7)c
                           1987      NR             NR            NR            32.5 (5.9)c     NR            Metz (1989)
                                                                                                                                                

    Table 29 (cont'd).
                                                                                                                                                

    Region                 Year      Annual emissions (kt/year)                                               Reference
                                                                                                      
                                     NOx            CO            HC            PM              SO2

                                                                                                                                                

    Heavy-duty vehicles (contd)
    Germany                1988      614            351           1.31          NR              38            Federal Environment Agency
                                                                                                              (1992)
    United Kingdom         1991      (19.9)c        (1.9)c        (5.4)c        (14)c,d         (1.0)c        Quality of Urban Air
                                                                                                              Review Group (1993)
    Canadae                1985      265            104           37            20              21            Environment Canada
                                                                                                              (1990)
                                                                                                                                                

    NOx, nitrogen oxides; CO, carbon monoxide; HC, hydrocarbons; PM, particulate matter; SO2, sulfur dioxide; NR, not reported; OECD,
    Organsiation for Economic Co-operation and Development
    a  Excluding the former German Democratic Republic
    b  Percentage of total traffic emissions
    c  Percentage of total man-made emissions
    d  PM10
    e  Sum of light- and heavy-duty trucks
             The effect of a ceramic wall-flow monolith particulate trap was
    studied on the emissions of a 1979 heavy-duty diesel engine with
    direct injection in two steady-state modes. Reductions of 70% (50% at
    full load) and 86% (75% at full load) in particle release were found.
    There was also a decrease in the soluble organic fraction, whereas the
    emitted portion of volatile organic carbons was only slightly reduced.
    The trap had almost no effect on the release of nitrogen oxides (Dorie
    et al., 1987).

         The measurements of Lowenthal et al. (1994) on 13 trucks and
    buses (three with particulate traps, 10 without) in a bus cycle showed
    a similar reduction in the release of particles (Table 26). There was
    a significant decrease in the emission of particle-bound PAHs.

    B3.1.4.2  Catalytic converters

         Diesel catalytic converters reduce gaseous and particle-bound
    hydrocarbon constituents of the exhaust, e.g. aldehydes, PAHs, nitro-
    PAHs, and carbon monoxide. Regeneration systems are not needed, as
    particulate matter is not collected by common flow-through catalytic
    converters. The efficiency in controlling particulates is, however,
    much lower than that of traps. Disadvantages of catalytic converters
    are further conversion of nitrogen monoxide to the more toxic nitrogen
    dioxide and of sulfur dioxide to particulate sulfates (Organisation
    for Economic Co-operation and Development, 1993). Catalysts that
    selectively oxidize organic compounds and oxidize almost no sulfur
    dioxide are currently in use. Removal of sulfur from diesel fuels
    would allow the use of more active catalysts; use of catalysts in
    diesel vehicles run on sulfur-containing fuel is limited because the
    strong adsorbing tendency of sulfur partly inactivates the catalyst
    (Hammerle et al., 1994b; see section A2.1.1).

    B3.2  Regulatory approaches

         The levels of carbon monoxide, nitrogen oxides, total
    hydrocarbons, and particulates emitted by diesel engines are regulated
    by law in a number of countries (Table 30). Most other industrialized
    countries have adapted United States or European regulations (for
    details, see CONCAWE, 1990a and Mercedes-Benz AG, 1994a).

    B4.  ENVIRONMENTAL TRANSPORT, DISTRIBUTION, AND TRANSFORMATION

         Few data are available on the fate  of diesel exhaust in the 
    atmosphere, which is the first environmental compartment affected. In
    general, the effects on the overall environment of diesel fuel
    combustion are typical of those of fossil fuel burning. The
    contribution of diesel fuel emissions is proportional to diesel fuel
    consumption both locally and globally.


        Table 30.  Limit values for components of diesel exhaust
                                                                                                                                                

    Region                 Carbon          Nitrogen             Hydrocarbons       Particulates      Comments
                           monoxide        oxides
                                                                                                                                                

    Light-duty vehicles (g/km)

    Austria                2.1             0.62                 0.25               0.124             < 3.5 t; since 1991; from 1995,
                                                                                                     adoption of European Union
                                                                                                     standards planned
    Canada                 2.1             0.62                 0.25               0.12              Since 1987
    European Union         2.72            0.97 (with           0.14               Since 1992
                                           hydrocarbons)
                           1.0             0.7                                     0.08              From 1996
    Finland                                                                                          Since 1993
    Japan                  2.1             0.7                  0.62               None              Since 1986
                           2.1             0.5                  0.4                0.2               Since 1994
    Sweden, Norway         2.1             0.62 (city)          0.25               0.124             < 3.5 t; from motor year 1992
                                           0.76 (highway)
    Switzerland            2.1             0.62 (city)          0.25               0.124             < 3.5 t; since 1988; from 1995,
                                           0.76 (highway)                                            adoption of European Union
                                                                                                     standards planned
    USA (California)       2.1-5.2         0.2-0.6              0.2-0.3 (except    0.05 (up to       Depending on mileage
                                                                methane)           31 000 km)
    US Environmental       2.1-2.6         0.6-0.8              0.2                0.05-0.12         Depending on mileage
    Protection Agency
                                                                                                                                                

    Table 30 (contd)
                                                                                                                                                

    Region                 Carbon          Nitrogen             Hydrocarbons       Particulates      Comments
                           monoxide        oxides
                                                                                                                                                

    Heavy-duty vehicles (g/kWh)

    Austria                4.9             9.0                  1.23               0.4
    Canada                 15.5            5.0                  1.3                0.25              g/bhp-h
                           15.5            5.0                  1.3                0.1               g/bhp-h; from 1995-97
    European Union         4.5             8.0                  1.1                0.36              Since 1992
                           4.0             7.0                  1.1                0.15              From 1995-96
    Japan                  7.4             5.0                  2.9                0.7               Indirect injection engines
                           7.4             6.0                  2.9                0.7               Direct injection engines
    Sweden                 4.9             9.0                  1.23               0.4
    USA                    15.5            5.0                  1.3                0.07              g/bhp-h; bus
                           15.5            4.0                  1.3                0.1               g/bhp-h; truck
                           15.5            5.0                  1.3                0.05              g/bhp-h; bus; from 1998
                           15.5            4.0                  1.3                0.1               g/bhp-h; truck; from 1998
                                                                                                                                                

    Adapted from Mercedes-Benz AG (1994b); reference year, 1994, unless otherwise stated; bhp-h, brake horsepower-hour
             Thus, the major components of diesel exhaust contribute to acid
    deposition (nitrogen and sulfur oxides), tropospheric ozone formation
    (carbon monoxide, nitrogen oxides, and hydrocarbons), and global
    warming (carbon dioxide and particulates; carbon monoxide, nitrogen
    oxides and hydrocarbons by ozone formation).

         The environmental fate of released particulate matter and of
    organic compounds, particularly the PAHs, is important environmentally.
    The mechanisms of the distribution and transformation of individual
    hydrocarbons in diesel exhaust have been well described (IPCS, 1986,
    1993).

    B4.1  Transport and distribution between media

         The major fraction (50-80%) of the particulate emissions of
    diesel engines is in the submicron size, ranging from 0.02 to 0.5 µm
    (see section B2.1.2). Once particles have been emitted, their
    mechanical transport in the atmosphere is like that of gas molecules
    (nonreactive). Together with carbon particles from other combustion
    processes, they may be transported over long distances and even
    penetrate the stratosphere (Muhlbaier Dasch & Cadle, 1989).

         Only one study was available on the agglomeration behaviour of
    the particles in the atmosphere. Albrechcinski et al. (1985) measured
    the aging behaviour of particulate matter from a 1978 diesel passenger
    car in a 600-m3 model chamber with a nominal exhaust gas dilution of
    150:1 and found peak mean aerosol diameters, based on volume size
    distributions, of 0.1-0.18 µm for 'fresh' aerosol. The concentrations
    of total particulate decreased during aging of the aerosol due to
    coagulation, fall-out, and perhaps losses to the walls of the chamber.
    The average decrease was 20% after 8 h and 60% after 24 h. Excess
    nitrogen dioxide seemed to inhibit particle growth. Particles with an
    average diameter of 1-10 µm constituted approximately 10% of the fresh
    aerosol, but this percentage decreased during aging as a result of
    apparently rapid fall-out of larger particles. It is not clear whether
    these results are also representative of the particle emission pattern
    of modern diesel passenger cars.

         Model experiments of the gas-phase chemistry of nitrogen oxides,
    total hydrocarbons, and sulfur dioxide from diesel exhaust were
    performed by varying parameters such as irradiation, nitrogen dioxide
    concentration, and ozone concentration (Albrechcinski et al., 1985).
    Irradiation alone had only a minimal effect, but excess nitrogen
    dioxide led to higher rates of decay of nitrogen monoxide in the dark
    than under irradiation, probably because of photodissociation of
    nitrogen dioxide to nitrogen monoxide. As nitrogen monoxide is rapidly
    converted to nitrogen dioxide in the presence of excess ozone, the

    nitrogen oxide concentration decreased as the ozone concentration
    reached a maximum, presumably partly because of the formation of
    nitric acid which was deposited onto the chamber walls. Since reactive
    organic compounds were not added, the authors considered that the
    experiments were not representative of urban atmospheres.

         The hydrosphere and geosphere may be affected indirectly by
    diesel exhaust emissions after dry or wet deposition of particulate
    matter or individual constituents (see section B4.2). In experiments
    with dysprosium-traced diesel fuel, Horvath et al. (1988) (see section
    B5.1) found 0.012-0.07 g diesel particulate matter per gram of road
    dust.

    B4.2  Transformation and removal

         Atmospheric removal of airborne carbon particles consists mainly
    of dry deposition and scavenging by precipitation (wet deposition).
    The rate of wet removal is directly correlated to the ratio of organic
    to elemental carbon and is low for small ratios (Muhlbaier Dasch &
    Cadle, 1989). As the overall removal rate of diesel particulates is
    estimated to be low, the atmospheric life-time is several days
    (Jaenicke, 1986).

         The elemental carbon of diesel particulates may act as a catalyst
    in the formation of sulfuric acid by oxidation of sulfur dioxide. The
    oxidation reaction is assumed to occur on the surface of solid soot
    particles or in liquid drops containing soot particles (Novakov, 1984;
    Mészáros & Mészáros, 1989). Furthermore, diesel particles themselves
    contain sulfate, adsorbed in the form of metal sulfates, at a
    concentration of about 1-12% by weight, depending on the sulfur
    content of the fuel (see section B2.1.2), which may be washed off if
    the particles come into contact with precipitation. The organic
    compounds adsorbed on the emitted particulate matter may undergo a
    number of physical and chemical reactions (volatilization of
    low-relative-molecular-mass compounds, formation of secondary
    aerosols, oxidation, nitration) with other atmospheric compounds or
    during exposure to sunlight (Pitts et al., 1985). The estimated
    half-lives for unbound PAHs in the atmosphere range from one day
    (naphthalene) to one week (e.g. pyrene, chrysene, benzofluoranthenes,
    benzo[a]pyrene) (Mackay et al., 1992; see Table 11); however, PAHs
    adsorbed onto particles may be more stable to nitrogen dioxide and
    ozone than airborne PAHs (Gibson et al., 1981).

    B5.  ENVIRONMENTAL LEVELS AND HUMAN EXPOSURE

         Because diesel exhaust is a complex mixture of a great variety of
    compounds (see also section B2.1), general 'environmental levels'
    cannot be described. The individual constituents of diesel exhaust
    should be detectable in all compartments of the environment. Since
    diesel exhaust is only one of numerous other sources, the levels of
    individual compounds cannot be correlated directly to diesel exhaust.

         Particulate matter appears to be the most relevant fraction of
    diesel exhaust for the general population and workers exposed to
    diesel exhaust emissions from a toxicological point of view. The
    concentration of particles is about 10- to 20-fold higher in diesel
    exhaust than in the exhaust from gasoline-fuelled vehicles, and the
    particle diameters are in the respirable range.

    B5.1  Exposure of the general population

         The general population is exposed to high levels of diesel
    exhaust in certain microenvironments, such as busy streets and parking
    areas. Exposure to the main gaseous components of diesel exhaust --
    carbon dioxide, carbon monoxide, nitrogen oxides, and total
    hydrocarbons -- can be estimated from the emission factors given in
    section B3, the kilometrage of diesel vehicles in relation to total
    traffic, and atmospheric dispersion models (Cuddihy et al., 1984;
    Volkswagen AG, 1989). In view of the fact that measured emission
    factors vary widely (see section B3.1.1) and are mainly dependent on
    adjustment of individual engines, the validity of these estimations is
    limited.

         On a global basis, there are wide variations in the contribution
    of diesel emissions to total airborne particulate matter, which
    depends on the following factors: percentage of diesel vehicles in the
    total volume of traffic; type, age, and maintenance of individual
    engines (see section B3.1.2.1); fuel quality (e.g. sulfur content and
    aromaticity; see section B3.1.2.2); and emission control techniques.
    Particle concentrations in environmental air samples are usually given
    as total particle concentrations, with no differentiation by source.
    The contribution of diesel particulate matter is then estimated on the
    basis of the percentage of diesel vehicle emissions in the total
    emissions from traffic or total man-made emissions. Emissions from
    running stationary diesel engines have up to now not been taken into
    account. Data on emissions from light- and heavy-duty vehicles in some
    industrialized countries are shown in Table 30 (section B3.1.3). The
    proportion of the concentrations to which people are exposed that is
    attributable to diesel vehicles varies widely, depending on traffic
    flow and composition in the vicinity and the magnitude of background
    pollution from other sources.

         In a study in the United Kingdom (Waller et al., 1965),
    measurements were made at a sampling site at the centre of a busy
    street canyon, where traffic flows were typically 1200 gasoline and
    600 diesel vehicles per hour, in order to assess the maximal
    concentrations of traffic pollutants in urban areas. The contributions
    of airborne particulate matter, expressed as seasonal means for
    daytime working hours, were 200-421 µg/m3, after allowance for the
    urban background measured at nearby control sites. The additional
    particulate matter was attributed primarily to the diesel vehicles,
    while the corresponding contribution to carbon monoxide, averaging
    15 ppm (17 mg/m3), was associated mainly with the gasoline vehicles.

         In France, Joumard & Perrin (1988) measured atmospheric
    particulate matter (diameter, 0.01-25 µm) in a 30-40-m wide street
    open to pedestrians, with two central lanes reserved for use by bus
    services involving weekday traffic of more than 1000 vehicles per day
    in each direction. The measurement point was in the immediate vicinity
    of two bus stops, 3 m from the edge of the pavement and at a height of
    2.5 m. The particle concentrations, determined by measuring ß
    radiation, were 34 µg/m3 during the period 23.00-05.00 h and
    67 µg/m3 during 07.00-19.30 h.

         Airborne particle concentrations were calculated for an
    eight-lane American freeway and for four- and six-lane European
    autoroutes. The one-dimensional Gaussian model used neglects chemical
    reactions during transport and possible deposition on the ground and
    assumes turbulent mixing in the air. The calculated maximal
    concentrations were 7-13 µg/m3 close to the road, and these
    decreased with distance from the measuring point. About 100 m away,
    the estimated particle concentration at a height of 1.5 m was three to
    four times lower than that next to the motorway (Volkswagen AG, 1989).

         Particle emissions, especially from diesel vehicles, were
    estimated roughly for various locations by the lead surrogate method,
    in which historical ambient lead concentrations in urban areas are
    used as indices of mobile source pollutant levels. Particle emission
    factors for light- and heavy-duty vehicles and the distribution
    characteristics of lead and diesel particulate are also taken into
    consideration (Rykowski & Brochu, 1986). The results are shown in
    Table 31.

         Concentrations of particles from diesel engines in the
    environment were determined in Vienna, Austria, at several measuring
    points during a four-week period. As there was only one distributor of
    diesel fuel for the whole network of service stations in Vienna, the
    rare earth element dysprosium was added as an indicator to the fuel at
    this central point, which was distributed to diesel vehicles at
    service stations and then emitted with the exhaust. Dust samples
    collected from the air were subjected to neutron activation analysis
    and the level of dysprosium was determined. The proportion of
    dysprosium in diesel soot was calculated on the basis of emission
    factors for diesel engines in the literature and from laboratory
    experiments, assuming that the marker is totally transferred from fuel
    to exhaust during combustion. At seven of the 11 measuring points for
    different traffic densities, the concentrations were almost linearly
    correlated to traffic density. Ambient particulate levels increased by
    5.5 µg/m3 per 500 diesel vehicles per hour, from 10.3 in a residential
    area to 23.6 µg/m3 near an urban freeway; the mean value was
    14.4 µg/m3. The diesel particulate concentration in the centre of
    Vienna was calculated to be 11 µg/m3 by correlating the measured
    values and traffic density. At four measuring stations, significant
    deviations from linearity were found which were due to local

    conditions. The concentration of the marker was below the detection
    limit before the testing period and one year later (Horvath et al.,
    1988). A deficiency of the study is that the dysprosium concentrations
    of diesel fuel before the addition were not taken into account, so
    that the emission factors used may not reflect the true situation
    (Holländer, personal communication, 1994). Nevertheless, the measured
    ambient particle concentrations fit well with estimated data for
    diesel exhaust emissions by other investigators.

    Table 31.  Estimated particle concentrations emitted from diesel
               engine exhaust
                                                                        

    Location and year                                    Concentration
                                                           (µg/m3)
                                                                        

    United Statesa, cities with > 1 million                 1.3-3
      residents (1984)
    United Statesa, cities with 100 000-250 000 residents   0.7-1.7
    Naples, Italy, urban area (1982)                        7.2-35.1
    Birmingham, United Kingdom, urban area (1982)           3.2-20.3
    Birmingham, United Kingdom, university area (1982)      0.8-5.5
    Stockholm, Sweden, inner city (1982)                    2.5-16.4
                                                                        

    From Rykowski & Brochu (1986)
    a  According to Cuddihy et al. (1984), the proportion of
       diesel-fuelled light-duty vehicles in the United States was
       about 2%.

         In a programme to measure airborne diesel soot levels in
    Düsseldorf, Germany, the total carbon content of fine dust (particle
    diameter, < 7µm) at a busy crossing was 8-42 µm3 as a daily
    average, 18 µg/m3 as an annual average, 21 µg/m3 on workdays, and
    14 µg/m3 on weekends. The carbon content after solvent extraction
    (elemental carbon) was 6-32 µg/m3 as a daily average, 16 µg/m3 on
    workdays, and 10 µg/m3 on weekends. In a residential area in
    Duisburg, Germany, the concentration of elemental carbon was 3 µg/m3.
    The method of determination was as follows: a defined volume of air
    was sucked through a fibreglass filter, and the respirable particle
    fraction was separated. This fine dust was further separated into an
    incinerable portion (according to German standard method TRGS 900),
    total carbon, and the carbon content after liquid extraction
    (elemental carbon). Total and elemental carbon were determined by
    coulometry. By correlation with diurnal changes in carbon and nitrogen
    monoxide concentrations, the main particle source was shown to be
    automobile traffic at the sampling location. Correlation of a traffic
    census with working day and weekend particle levels showed that the
    airborne particulate matter was due primarily to diesel vehicles
    (Elbers & Muratyan (1991). Further measurements at the same place in

    1991 and 1992 confirmed the results (Elbers & Richter, 1994).
    Validation of the analytical method gave detection limits of 0.39 for
    elemental carbon (relative standard deviation, 14% or 4.4 µg/m3) and
    0.65 µg/m3 for total carbon (relative standard deviation, 13% or
    7.8 µg/m3).

         The weekly average ambient concentrations of diesel particulate
    in Germany were estimated to be 5-10 µg/m3 in urban areas, 15-25 µg/m3
    in urban areas close to busy streets, and < 1.5 µg/m3 in rural
    areas; however, the method used for the estimations was not reported
    (State Committee for Immission Protection, 1992).

         Estimates of the US Environmental Protection Agency for 1986 gave
    annual mean ambient concentrations of diesel particle in the United
    States of 2.6 µg/m3 in urban areas and 2.4 µg/m3 in rural areas. A
    prediction for 1995 gave values of 1.5 µg/m3 for urban and 1.2 µg/m3
    for rural areas (Mauderly, 1992). Higher concentrations were estimated
    by the Motor Vehicles Manufacturing Association (Sienicki & Mago,
    1992), which predicted a mean value of 2.7 µg/m3 for the city of Los
    Angeles in 1995. Lower levels were estimated in a study by the US
    Environmental Protection Agency (1993) in which national average
    emission factors for diesel vehicles for 1988 were multiplied by urban
    and rural mileage. For 1990 and 1995, the annual exposure levels were
    estimated to be 2.0 and 1.2 µg/m3 in urban air and 1.1 and 0.6 µg/m3
    in rural air, respectively. These estimates were confirmed by ambient
    monitoring. Using data from the early 1980s, when the proportion of
    diesel vehicles, the number of kilometres travelled, and emission
    standards differed from present characteristics, McClellan (1986)
    derived ambient diesel particulate concentrations of 10-30 µg/m3 on
    American urban freeways and in urban street canyons.

    B5.2  Occupational exposure

         The levels of particulate matter to which workers are exposed
    when their predominant exposure to exhaust is that from diesel engines
    are described in this section. It is difficult to determine the
    contribution of diesel exhaust to respirable particles, as they
    include not only those from diesel engines but also others generated
    in the worker's environment, such as sand, dirt, coal dust, and
    cigarette smoke particles. Thus, the levels of total and respirable
    particulates reported in jobs with potential exposure to diesel
    exhaust may be significant overestimates of the actual exposure. Only
    two occupational groups have been studied in this regard: railroad
    workers (Woskie et al., 1988a,b) and mechanics and drivers in the
    trucking industry (Zaebst et al., 1991). The specific exposures of
    these groups to diesel particles range from 0.026 to 0.192 mg/m3. In
    the railroad workers, cigarette smoke was found to account for a
    substantial proportion of total respirable particulate matter, and the

    geometric mean exposure to respirable particulates of workers not
    exposed to diesel exhaust was reduced by over 80% when the
    contribution of cigarette particulate was considered (Woskie et al.,
    1988a).

         The principal marker chosen to measure exposure of workers in the
    trucking industry (including road drivers, local drivers, stevedores,
    and mechanics) to diesel exhaust was elemental carbon, which is
    derived from the core of the diesel exhaust particle and represents
    about 20% of respirable particulate; cigarette smoke does not
    contribute appreciable amounts of elemental carbon. Mechanics and
    stevedores had the highest exposure to diesel exhaust, and truck
    drivers had lower levels (Zaebst et al., 1991). The degree to which a
    marker of diesel exhaust exposure actually represents true exposure
    depends on the nature of all sources of respirable particles in the
    work-place. Occupational exposure to particulates is summarized in
    Table 32.

    Table 32.  Reported occupational exposure to particulates among
               workers exposed to diesel exhaust
                                                                        

    Occupational group     Particles (mg/m3)
                                                                        

                           Total           Respirable      Diesel
                                                                        

    Railroad workers       0.01-1.99       0.06-0.41       0.051-0.192a
    Mine workers           0-23.0          0-16.1
    Truck drivers          0.1-1.0         0.13-1.2        0.025b
    Bus garage and bus     0.15-1.2        0.01-0.61
    workers
    Stevedores                                             0.157b
    Truck mechanics                                        0.133b
                                                                        

    a  Estimated using nicotine to account for exposure to cigarette smoke
    b  Based on carbon core analysis, with an estimated 20% of respirable
       particulate due to carbon (Steenland et al., 1992)

         Because 1-nitropyrene is associated with the particle fraction of
    diesel exhaust, this compound was used as an indicator of the
    particulate fraction of diesel exhaust derived from various sources
    (e.g. ships, railroad engines, heavy-duty road vehicles, fork-lift
    trucks) (Scheepers et al., 1994a,b). The 1-nitropyrene content of
    fresh and aged diesel exhaust was measured in small air samples

    extracted with acetone by sonication and detected by gas
    chromatography-mass spectrometry. The following concentrations of
    respirable diesel exhaust were determined: 0.044-0.33 mg/m3 in
    various work-places with fork-lift trucks, 0.10-0.37 mg/m3 in
    railroad engine repair shops, 0.080mg/m3 in a river vessel, and 0.67
    or 3.33 mg/m3 in the breathing zone of railroad workers.

    B5.2.1  Truck drivers and mechanics

         In a study of 15 truck drivers in Geneva, Switzerland, exposure
    to particulate matter was measured during one working day with a
    direct-reading instrument and by sampling beside the drivers in the
    truck cabin (Guillemin et al., 1992). Local drivers had greater
    exposure to particulate matter (0.3 mg/m3) than long-distance
    drivers (0.1 mg/m3) and spent more time in polluted areas, such as
    streets with heavy traffic and construction sites. Smoking did not
    influence the exposure of professional truck drivers to particulate
    matter, probably because the ventilation rate of the truck cabins was
    relatively high, even on cold days.

         Industrial hygiene surveys were conducted during warm and cold
    weather at eight terminals and truck repair shops to determine
    exposure to submicrometre particles of elemental carbon among American
    workers in one of four jobs -- road drivers, local drivers, stevedores,
    and mechanics -- all of whom were exposed mainly to diesel aerosol
    (Zaebst et al., 1991). The overall arithmetic mean exposure was
    5.1 µg/m3 for 72 long-distance road drivers, 5.4 µg/m3 for 66
    local drivers, 26.6 µg/m3 for 80 mechanics, and 31.3 µg/m3 for 54
    stevedores exposed to diesel fork-lift trucks. The background
    concentrations in the same cities were 3.4 µg/m3 on 21 major
    highways and 1.4 µg/m3 in 23 residential areas. The exposures of the
    truck drivers were significantly higher than the background
    concentrations in residential areas but could not be distinguished
    from the background highway concentrations.

         Area air sampling in an American truck repair shop in 1987 gave
    the following results, depending on the sampling location: 0.225 and 
    0.32 µg/m3 submicrometre particles, 0.26 and 0.51 µg/m3 respirable
    particles, and 0.63 and 0.66 µg/m3 total particulate matter. The
    values for elemental carbon were 86-118 µg/m3 submicrometre
    particles and 159-214 µg/m3 total carbon. In personal samples from
    four mechanics, 79-193 µg/m3 submicrometre elemental carbon were
    found (US National Institute for Occupational Safety and Health,
    1990a).

         The concentrations of particles collected on fibreglass filters
    at 25 work-places in Germany where mainly fork-lift trucks and trucks
    were used were 0.02-0.8 mg/m3 in area samples and 0.13-1.2 mg/m3
    in personal samples (Lehmann et al., 1990). Area measurements in two
    German truck repair garages (five samples from each) showed
    concentrations of incinerable fine dust (according to German standard
    method TRGS 900) of 0.11-0.40 mg/m3 (Blome et al., 1990).

         Personal samples from truck drivers who loaded diesel-engined
    trucks onto car ferries in Sweden showed particle concentrations of
    0.1-1.0 mg/m3 (Ulfvarson et al., 1987). The drivers may also have
    been exposed to diesel exhaust from the ferries.

    B5.2.2  Bus garage and other bus workers

         In area samples taken in an American bus garage in 1989, the
    concentrations of particles were 0-0.1 mg/m3 submicrometre
    particles, 0-0.3 mg/m3 respirable particles, 0-0.22 mg/m3 total
    particulate matter, and 0.014-0.326 mg/m3 elemental carbon (US
    National Institute for Occupational Safety and Health, 1990b).

         The concentrations of smoke were measured in two diesel bus
    garages in London, United Kingdom, from high-volume samplers for
    successive periods through the day and night, which were linked to bus
    movements and the exposure of the maintenance staff. The contribution
    from the buses was assessed by subtracting the background values at
    control sites outside the garages. In the most polluted garage,
    0.3-1.2mg/m3 of smoke were measured during working periods (Waller
    et al., 1985).

         In Sweden, a particle concentration of 0.46 mg/m3 was found
    during personal sampling in periods of high activity (arrival and
    departure of buses) (Ulfvarson et al., 1987).

         Area air sampling of the concentrations of incinerable fine dust
    (according to German standard method TRGS 900) in two bus depots (six
    samples in each) and one bus garage (three samples) in Germany showed
    concentrations of 0.22-0.37 mg/m3 (Blome et al., 1990).

         The respirable dust levels in bus garages in Italy and the United
    States ranged from <0.01 to 0.73 mg/m3 (Gamble et al., 1987a). The
    total dust levels in bus garages in Denver, Colorado, United States,
    were 0.01-0.81 mg/m3 (Apol, 1983; Pryor, 1983).

    B5.2.3  Fork-lift truck operators

         Area samples taken in one American trucking terminal where
    diesel-powered fork-lift trucks were used in 1987 contained
    48-62 µg/m3 submicrometre particles, 34-57 µg/m3 respirable
    particles, and 62-97 µg/m3 total particles (US National Institute
    for Occupational Safety and Health, 1989a). The concentrations of

    elemental carbon in area samples from another American trucking
    terminal in 1988 were 40-44 µg/m3, and those in personal samples
    from fork-lift operators were 33-77 µg/m3 (US National Institute for
    Occupational Safety and Health, 1989b). Area samples taken in an
    American freight dock contained concentrations of <12 and 294 µg/m3
    respirable dust. The levels were significantly higher in areas where
    unfiltered rather than filtered diesel motors were used (US National
    Institute for Occupational Safety and Health, 1990c).

         A mean concentration of 31.3 µg/m3 of submicrometre elemental
    carbon was found in personal samples from stevedores operating
    diesel-powered fork-lift trucks. About 10 times less was found for
    operators of gasoline- or propane-powered trucks (Zaebst et al.,
    1991).

         The concentrations of particles in the breathing zone during use
    of diesel-powered fork-lift trucks for loading, unloading, and
    warehousing operations in American ammunition storage magazines were
    < 0.01 to1.3 mg/m3 (Ungers, 1984) and 0.5-5.0 mg/m3 (Ungers, 1985).

    B5.2.4  Railroad workers

         The mean levels of total particulate matter were 0.38 mg/m3
    (range, 0.1-0.8) in Finnish locomotive cabs and 1.99 mg/m3 (range,
    0.07-8.7) in roundhouses (Heino et al., 1978). In three German engine
    sheds, the ambient concentrations of incinerable fine dust were
    0.1-0.41mg/m3 (Blome et al., 1990).

         A summary of American literature showed time-weighted average
    levels of 0.05-0.07 mg/m3 (8-h) in tunnels, 0.16 mg/m3 (7.5-h)
    during freighting, 0.01 mg/m3 (5-h) in switchyards, and 0.27 mg/m3
    (8-h) in cabooses (Hobbs et al., 1977). The arithmetic mean
    concentrations of total respirable particles in personal samples from
    shop workers, engineers, brakemen, conductors, and farmers in the
    American railroad industry (after correction for the estimated
    contribution of cigarette smoke particulates) were 51-192 µg/m3
    (Woskie et al., 1988a).

    B5.2.5  Mine workers

         Diesel vehicles and machines have been used in nearly all
    American mines since the 1960s (Wichmann & Brüske-Hohlfeld, 1991). The
    mean concentration of diesel particulate matter in personal and area
    samples in mines where diesel equipment was used was 4.6 mg/m3
    (range, 0.2-14) (Holland, 1978). Total respirable dust levels of
    0.6-1.7 mg/m3 were found by personal sampling (Wheeler et al.,
    1981); average concentrations of 0.9-2.7 mg/m3 were found in
    full-shift personal samples, and the range for full-shift area samples
    was 0-16.1 mg/m3 (Reger et al., 1982). The 8-h time-weighted average
    level of respirable particulate matter in personal samples taken in an

    American molybdenum mine was 0.2-1.9 mg/m3 (Cornwell, 1982). Total
    respirable dust concentrations of 1.1-4.4 mg/m3 were reported in an
    American gold mine (Daniel, 1984). The US Mine Safety and Health
    Administration reported that the concentrations of diesel particulate
    matter in American mines other than coal mines were 0.3-1.5 mg/m3
    (Mauderly, 1992).

         Data from a salt mine in Canada showed a diesel particulate level
    of about 0.5 mg/m3 (Johnson & Carlson, 1986).

         In area samples from different work-places in eight German coal
    mines, the levels of airborne diesel particulate matter (incinerable
    fine dust) were 0.19-0.70 mg/m3 (Blome et al., 1990). In German
    potash-salt mines, depending on the district, 0.2-0.4 mg/m3 of
    diesel particles were measured (Bauer et al., 1990).

    B5.2.6  Fire fighters

         The average concentration of total particulate matter in personal
    samples from fire-fighters in Boston, New York City, and Los Angeles,
    United States, in 1985 was between < 0.1 and 0.48 mg/m3. On a typical
    day in Boston or New York City, exposure to diesel exhaust particles
    was estimated to be about 0.225 mg/m3 after correction for the
    contributions of background and smoking. Simulation of 'worst-case'
    exposures in Los Angeles fire stations gave a maximal concentration of
    0.748 mg/m3 total particles (Froines et al., 1987).

    B5.3  Biomonitoring

    B5.3.1  Urinary mutagenicity

         The occurrence of mutagens, measured by the Ames test, in the
    urine and faeces of eight car mechanics occupationally exposed to
    diesel exhaust was compared with that of a reference group of nine
    office workers. The samples were collected over 48 h; information was
    also gathered on dietary intake, smoking habits, and medication. No
    enhancement of the incidence or degree of faecal or urinary
    mutagenicity was seen (Willems et al., 1989).

         Post-shift urinary mutagenicity was measured by the Ames test in
    306 samples from 87 railroad workers with a range of exposures to
    diesel exhaust (Schenker et al., 1992). Exposure was measured
    throughout the work shift by constant-flow personal sampling pumps. In
    separate analyses of smokers and nonsmokers, no independent
    association was seen with exposure to diesel exhaust. The workers with
    the highest median exposure to respirable particles attributable to
    diesel exhaust (113 µg/m3) were railroad engineering shop workers.

         In studies of the effect of diesel exhaust on stevedores on
    'roll-on-roll-off' ships (Ulfvarson et al., 1987), no difference in
    mutagenicity to  Salmonella typhimurium TA98 or  Escherichia coli WP
    2  uvr A was seen in urine collected during periods of exposure and
    of no exposure. Similarly, no increase in urinary mutagenicity was
    found in samples from six volunteers before and after exposure to
    diesel exhaust collected from an experimental automobile run for 3.7 h
    at 60 km/h and 2580 rpm.

    B5.3.2  Other analyses

         Samples of total suspended particulate matter were collected in a
    repair shop for train engines, and the diesel exhaust concentrations
    were determined using 1-nitropyrene as a biomarker (see section B5.2).
    Urinary metabolites and the corresponding nitro-substituted
    derivatives were detected by immunoassay in the urine of three
    diesel-engine mechanics. They had a significantly increased excretion
    of metabolites in comparison with two office clerks (Scheepers et al.,
    1994b).

    B6.  KINETICS AND METABOLISM IN LABORATORY ANIMALS AND HUMANS

    B6.1  Deposition

         The mass median aerodynamic diameter of diesel exhaust particles
    in long-term studies in animals exposed by inhalation was 0.17-0.25 µm
    (Lewis et al., 1986; Lee et al., 1987; Wolff et al., 1987;
    Creutzenberg et al., 1990). The deposition efficiency of particles of
    this size range is lower than that of smaller particles, for which the
    deposition by diffusion increases with decreasing diameter. The
    deposition efficiency of particles with a diameter > 0.5 µm increases
    with the aerodynamic diameter, owing to sedimentation and impaction.
    The nose has hardly any filter effect for diesel particles because the
    fraction of particles < 0.4 µm deposited in the nose is very low
    (Heyder et al., 1986).

         After short-term inhalation by rats of 0.29-µm fused
    aluminosilicate particles, 8% was deposited in the bronchi and 13% in
    the lungs (Raabe et al., 1988). A total of 10-20% diesel exhaust
    particles was deposited in rats and guinea-pigs (Chan et al., 1981;
    Lee et al., 1983; Dutcher et al., 1984). In humans, the alveolar
    deposition efficiency of 0.2-µm particles was 10-20%, and no bronchial
    deposition was detected (Heyder et al., 1986).

         These experimental results agree well with those of a predictive
    model for deposition of inhaled diesel exhaust particles in laboratory
    animals and humans (Yu & Xu, 1987). In humans, the deposition
    efficiency was about 7% in the alveolar region and 4% in the
    tracheobronchial region. The model can also predict the fractions that
    will be deposited in different airway generations; maximal deposition
    occurs in the alveolar region (twentieth generation) of adults and of
    children over two years of age.

         The length and concentration of exposure had no significant
    effect on deposition in rats exposed to 0, 0.4, 3.5, or 7.1 mg/m3
    diesel exhaust particles for 6, 12, 18, or 24 months, when deposition
    of 0.1-µm gallium oxide particles was used as the indicator (Wolff et
    al., 1987).

    B6.2  Retention and clearance of particles

         The kinetics of inhaled substances are conveniently described by
    their pulmonary clearance and retention. These terms are often used
    interchangeably, but it should be kept in mind that the fraction
    cleared from the lung plus the fraction retained in the lung account
    for the total deposited amount (retention = deposition - clearance).
    Thus, in the following paragraphs, the term 'clearance' refers to a
    clearance rate ( r, per unit of time) and the term 'retention' to a
    biological half-time ( t ´), provided there is a constant rate of
    exponential clearance (ln2/ t ´).

         Clearance of diesel particles from the alveolar region is
    important in the long-term effects of particles on cells, as it is
    more than two orders of magnitude slower than mucociliary clearance,
    measured in several studies in rats administered labelled diesel or
    surrogate particles by inhalation. The pulmonary retention half-times
    in studies in which the lung burden of diesel soot was also reported
    are summarized in Table 33. The equivalent continuous exposure
    concentration is calculated by multiplying the actual concentration by
    the number of hours of weekly exposure divided by 168 h. This figure
    allows a comparison of exposure levels in studies with different
    exposure protocols. The assumption is that Habers' law ( c ×  t =
    constant) is applicable. This is probably valid until a certain burden
    of particulate is reached in the lung; however, when that level is
    exceeded (particle overload), the calculation is increasingly invalid.

         The retention half-time in these studies was 60-100 days in
    control rats with low lung burdens and up to 600 days in animals with
    high lung burdens. In a double logarithmic plot of retention
    half-times and lung burden, a linear relationship can be seen for lung
    burdens of diesel particles above 1mg per lung (Figure 1).

         The gaseous phase of diesel exhaust appears to have no effect on
    alveolar clearance in rats or hamsters (Heinrich et al., 1986a).

         The retardation of lung clearance is not specific to diesel
    exhaust but is known as 'lung overloading by particles'. This effect
    is characterized by (i) an overwhelming of normal alveolar
    macrophage-mediated clearance processes under certain conditions of
    exposure, (ii) resulting in lung burdens greater than predicted from

    deposition kinetics at low concentrations, (iii) with associated
    pathophysiological changes including altered macrophage function,
    inflammation, and pulmonary fibrosis, and (iv) an uncertain
    association with an increased incidence of lung tumours in rats
    (McClellan, 1990).

         The effect of overloading lung clearance mechanisms was
    investigated in NMRI and C57Bl mice, Wistar and Fischer 344 rats, and
    Syrian golden hamsters exposed to particulates, including test toner
    (polymer pigmented with carbon black), polyvinyl chloride, carbon
    black, diesel exhaust, and two crystalline forms of titanium dioxide
    (anatase and rutile), at concentrations of 0.8-64 mg/m3 for up to
    two years. In rats, lung overloading generally occurred after exposure
    to 0.5-1.5 mg per lung or about 1 mg of material per gram of lung
    tissue. A two- to 10-fold decrease in the clearance rate was seen
    after heavy particulate loading in all of these rodent species (Muhle
    et al., 1990).

         In comparisons of the effects of particles of different
    densities, the retained volume of particles appears to be a more
    useful parameter than the retained mass, as alveolar macrophages have
    an upper volumetric uptake limit. The particulate overload effect
    appears to be initiated in rats when the average particle volume per
    alveolar macrophage exceeds about 60m3, and alveolar
    macrophage-mediated particle clearance virtually ceases when the
    phagacytosed particulate volume exceeds an average of 600m3 (Morrow,
    1988).

         At a lung burden above threshold, signs of lung overloading in
    Fischer 344 rats persisted for 15 months after cessation of exposure
    (Bellmann et al., 1992). Retardation of alveolar clearance was also
    observed in hamsters and mice at a lung burden above about 1 mg per
    lung. For mice, no data were available with regard to lung burdens
    below 1 mg per lung, but retardation of alveolar clearance is probably
    already present at lower lung burdens. Lung overloading is thus seen
    in a variety of species with various materials. It is generally
    observed after the threshold lung burden of particles of low
    solubility and low acute toxicity is exceeded for a considerable time
    (Muhle et al., 1990). The effects of overloading should be considered
    in extrapolating the health effects seen at high concentrations in
    experiental animals to the low concentrations occurring in the
    environment or the work-place.

         The alveolar clearance rates of highly insoluble particles differ
    among species, with retention half-times ranging from 50 to 100 days
    in rodents to several hundred days in dogs and man (Figure 2). In
    human lung, the retention half-time of insoluble, nontoxic particles
    may be >500 days (Bailey et al., 1985a,b).


        Table 33.  Pulmonary retention of insoluble particles in rats after exposure to diesel exhaust, and lung burden of diesel soot
                                                                                                                                                

    Strain and sex   No./group     Duration of   Exposure                                         Pulmonary retention            Reference
                                   exposure                                                                                  
                                   (h/week)
                                                 Concentration (mg/m3)   Duration   Lung burden   Material          Half-time
                                                                         (weeks)    (mg/lung)     studied           (days)
                                                 Actual    Equivalent
                                                           continuous
                                                                                                                                                

    Fischer 344,     4             140           0         0             0          0             14C-Diesel        77           Chan et al.
    male                                         0.25      0.21          7          0.2           exhaust           90           (1984)
                                                 0.25      0.21          16         0.5                             92
                                                 6.00      5.00          1          0.7                             166
                                                 6.00      5.00          9          6.5                             562
                                                 6.00      5.00          16         11.8                            (> 1000)

    Wistar, female   5-7           90            0         0             13         0             59Fe2O3           61           Creutzenberg
                                                 0.84      0.45                     0.63          (0.3 µm)          94           et al. (1990);
                                                 2.5       1.34                     2.5                             119          Heinrich et
                                                 6.98      3.74                     8.2                             330          al. (1995)
                                                 0         0             52         0                               72
                                                 0.84      0.45                     2.8                             121
                                                 2.5       1.34                     10.7                            254
                                                 6.98      3.74                     35.7                            541
                                                 0         0             78         0                               96
                                                 0.84      0.45                     5.5a                            221
                                                 2.5       1.34                     12.2                            272
                                                 6.98      3.74                     58.8                            687
                                                                                                                                                
    Table 33 (contd)
                                                                                                                                                

    Strain and sex   No./group     Duration of   Exposure                                         Pulmonary retention            Reference
                                   exposure                                                                                  
                                   (h/week)
                                                 Concentration (mg/m3)   Duration   Lung burden   Material          Half-time
                                                                         (weeks)    (mg/lung)     studied           (days)
                                                 Actual    Equivalent
                                                           continuous
                                                                                                                                                

    Fischer 344,     8             35            0.15      0.03          18         0.035         Diesel            87           Griffis et al.
    male and                                     0.94      0.20                     0.22          exhaust           99           (1983)
    female                                       4.10      0.85                     1.89                            165

    Fischer 344,     4             140           6.00      5.00          1          1.7           14C-Diesel        61           Lee et al.
    male                                         6.00      5.00          3          4.4           exhaust           124          (1987)
                                                 6.00      5.00          6          10.4                            192

    Fischer 344,     15-16 (lung   35            0         0             104        0             134Cs-Fused       79           Wolff et al.
    male and         burden)                     0.35      0.07                     0.6           aluminosilicate   81           (1987)
    female           10            3.50          0.73                    11.5       particles     264
                     (retention)                 7.00      1.46                     20.5          (2 µm)            240

                                                                                                                                                

    a  Measured; value derived by interpolation from data at 52 and 104 weeks
        FIGURE 1

    FIGURE 2

         Retention of diesel particles in the lungs and lymph nodes has
    been reported in Fischer 344 and Wistar rats after long-term
    inhalation (Brightwell et al., 1986; Ishinishi et al., 1986; Mauderly
    et al., 1987; Strom et al., 1988; Creutzenberg et al., 1990; Strom et
    al., 1990). Some results with regard to lung burden are included in
    Table 33. Because clearance of lung burdens above about 1 mg per lung
    was impaired after long-term, continuous inhalation of a concentration
    equivalent to about 0.5mg/m3, no steady-state lung burden was found
    at serial sacrifice at up to two years. About 20% of the lung burden
    of diesel particles was found in lung-associated lymph nodes at that
    time.

         Mathematical models have been developed to calculate the
    retention of diesel particles in rats. Yu & Yoon (1990) and Yu et al.
    (1991) developed a dosimetry model to predict the retention of diesel
    particles and associated organic compounds in rats and humans of
    different ages. The model indicates that lung clearance would not be
    impaired in humans exposed to a concentration < 0.05 mg/m3 for
    24 h/day. The highest lung burden per unit of lung weight was
    calculated for five-year-old children. This model was used to assess
    the risk of inhaling diesel particles (Pepelko & Chen, 1993).

    B6.3  Retention and clearance of polycyclic aromatic hydrocarbons
          adsorbed onto diesel soot

         The clearance of particles from ciliated airways by the
    mucociliary escalator after bronchial deposition is virtually complete
    within 24 h. For soluble organic material like PAHs on the surface of
    carbon particles, diffusion-limited clearance through the airway
    epithelium to capillary blood was found to be slower in the bronchial
    region than in the alveoli (Gerde et al., 1991a, 1993a,b).

         PAHs in diesel soot adhere strongly to the surface of the
    particles. Several studies have shown that the association of PAHs
    with particles significantly slows the clearance of the PAHs from
    lungs in comparison with the clearance of inhaled PAHs unassociated
    with particulate matter.

         After inhalation by Fischer 344 rats of 3H-benzo[ a]pyrene
    adsorbed onto diesel soot (0.1% by mass) for 30 min, biphasic lung
    clearance was noted. In a first, fast phase, about 50% of the inhaled
    compound was cleared with a half-time of about 1 h, predominantly by
    mucociliary clearance. At the end of exposure, about 15% of the 3H
    label was found in blood, liver, and kidney, indicating rapid, direct
    absorption of this fraction of 3H-benzo[ a]pyrene and its metabolites
    into blood. The remaining 50% was retained in the lung, with a
    half-time of 18 days; this is one-third of the retention half-time of
    diesel particles. Thus, removal of benzo[ a]pyrene and its metabolites
    from the particle carrier may be the rate-limiting step of this
    clearance process (Sun et al., 1984). In contrast, more than 95% of
    pure 3H-benzo[a]pyrene was cleared from the lung within 12 h after
    inhalation (Sun et al., 1982).

         These results were confirmed and extended to another PAH,
    1-nitropyrene (Bond et al., 1986a,b), when the clearance of
    14C-labelled compound adsorbed to diesel soot was determined by the
    same method in Fischer 344 rats. After exposure to 490 ng/litre of
    pure labelled compound, about 90% was cleared in the first, fast
    clearance phase, with a retention half-time of 1 h; after exposure to
    650 ng/litre of 14C-1-nitropyrene adsorbed onto diesel particles,
    45% of the labelled compound was cleared from the lungs in the first
    phase, with a half-time of about 2 h. The remaining 55% was retained
    with a half-time of 36 days, which was twice the corresponding value
    for 3H-benzo[ a]pyrene in the study of Sun et al. (1984).

         On the basis of their data on the clearance of 3H-benzo[ a]pyrene
    and 14C-1-nitropyrene, Bond et al. (1986b) predicted equilibrium
    lung burdens in humans after long-term inhalation. Assuming inhalation
    of 3.5 g/m3 of particles containing 0.1% benzo[ a]pyrene and
    1-nitropyrene over an 8-h working day, they predicted equilibria of
    160 ng benzo[ a]pyrene and 31 ng 1-nitropyrene per lung. No data are
    available on changes in the retention of individual compounds after
    prolonged exposure to diesel exhaust.

         The clearance of a very low concentration (2.6 g/g particles) of
    3H-benzo[ a]pyrene adsorbed onto diesel particles was measured in
    Sprague-Dawley rats after intratracheal instillation (Bevan & Ruggio,
    1991). Only 25% was cleared in the early, fast phase; a long-term
    retention half-time of eight days was calculated for the remainder
    during a three-day follow-up, which was similar to the clearance in
    that period in the study of Sun et al. (1984).

    B6.4  Metabolism

         In the studies of Sun et al. (1984) and Bond et al. (1986a,b),
    the levels of 3H-benzo[ a]pyrene and 14C-1-nitropyrene and their
    metabolites in extracts of lung homogenates after inhalation of diesel
    exhaust particles were measured by high-performance liquid
    chromatography. Sun et al. detected oxidized metabolites 30 min after
    inhalation: 35% of the 3H-benzo[ a]pyrene had been metabolized in
    equal amounts to phenols (3-hydroxy- and 9-hydroxybenzo[ a]pyrene)
    and quinones (1,6 and 3,6 isomers). Twenty days later, only 13%
    phenols and 5% quinones were detected. In the study of Bond et al.
    (1986b), about 30% of the 3H-1-nitro-pyrene had been metabolized to
    acetylaminopyrenephenol (6 and 8 isomers) and about 40% remained
    unmetabolized 1 h after inhalation. Oxidative metabolism of
    14C-benzo[ a]pyrene, in solution or adsorbed onto diesel particles
    at a concentration of 500 pmol/106 cells, was also detected in cell
    cultures of pulmonary macrophages obtained from beagle dogs. The
    quantity of metabolites increased with incubation time up to 45 and

    125 pmol/106 cells in extracts and in the media, respectively, after
    48 h of incubation. After incubation for 6 h, the major metabolites
    found in extracts of cells were benzo[ a]pyrene-7,8-diol (0.5pmol/106
    cells) and benzo[a]pyrene-4,5-diol (0.2 pmol/106 cells); those in
    extracts of culture medium were benzo[ a]pyrene-7,8-diol (1.7 pmol/106
    cells) and benzo[ a]pyrene-9,10-diol (1.4 pmol/106 cells), in
    comparison with a total concentration of metabolites of 2.5 pmol/106
    cells in the extracts and 6 pmol/10 6 cells in the medium. Minor
    metabolites were benzo[ a]pyrene phenols (3-hydroxy- and
    9-hydroxy-benzo[ a]pyrene) and benzo[ a]pyrene quinones (6,12, 1,5,
    and 3,5 isomers). The quantities of metabolite were similar when
    macrophages were incubated with benzo[ a]pyrene in solution or
    adsorbed onto diesel particles (Bond et al., 1984).

         Using the 32P-postlabelling technique, Wong et al. (1986) found
    increased DNA adduct formation in the lungs of Fischer 344 rats after
    exposure for 31 months to diesel exhaust containing 7.1 mg/m3
    particles. After exposure of Fischer 344 rats and Syrian golden
    hamsters to dilutions of diesel engine exhaust for six months to two
    years, the level of haemoglobin adducts (2-hydroxyethylvaline and
    2-hydroxypropylvaline), corresponding to the metabolic conversion of
    5-10% of inhaled ethylene and propylene to its oxides, increased
    dose-dependently (Törnqvist et al., 1988). An adduct of
    benzo[ a]pyrenediol epoxide with deoxyguanosine was identified after
    inhalation of benzo[a]pyrene adsorbed onto carbon black (20 mg/g)
    (Wolff et al., 1989), demonstrating the production of epoxides of PAHs
    by oxidative metabolism.

         Methods for extracting diesel soot before testing for
    genotoxicity  in vitro have been discussed. As dichloromethane and
    dimethyl sulfoxide do not correspond to leaching conditions in the
    lung, it can be assumed that the bioavailability of genotoxic
    materials is lower in lungs than  in vitro. An investigation of
    dipalmitoyl phosphatidylcholine, the primary component of
    physiological surfactant, demonstrated that the genotoxicity
    associated with diesel particles inhaled into the lung can be
    activated by the solubilization and dispersion properties of pulmonary
    surfactants. This finding was confirmed in mammalian cells for sister
    chromatid induction (Keane et al., 1991) and for chromosomal
    aberrations and micronucleus formation (Gu et al., 1991, 1992).

    B7.  EFFECTS ON LABORATORY MAMMALS AND IN VITRO SYSTEMS

    B7.1  Single exposure

         Data on the acute toxicity of diesel exhaust are very limited;
    the two available studies suggest that diesel exhaust particles and
    their extracts have little acute toxicity.

         Male ICR mice were administered diesel exhaust particles by
    intratracheal instillation, and mortality due to lung oedema was
    assessed for up to 24 h. At a dose of 0.9 mg per animal, all animals
    died within 24 h. The LD50 value was 0.6 mg per animal, equivalent
    to about 20 mg/kg body weight. Diesel exhaust particles washed three
    times with methanol did not cause death at a dose of 1 mg per mouse,
    suggesting that the primary source of toxicity is the extractable
    organic compounds rather than the particles themselves. These
    compounds may produce superoxide radicals that damage endothelial
    cells (Sagai et al., 1993).

         The LD50 value for diesel exhaust extract in adult Syrian
    golden hamsters after intraperitoneal injection was 1280 mg/kg body
    weight (Pereira et al., 1982).

    B7.2  Short-term exposure

         Tracer-monitored DNA synthesis, indicating a proliferative
    response, was observed in whole lung tissue of groups of 3-10 Fischer
    344 rats exposed to diesel exhaust by inhalation at 6 mg/m3 for
    20h/day on seven days per week for periods of 1-14 days. The response
    was maximal after two days of exposure, showing a fourfold increase.
    The type II cell labelling index was significantly higher (fourfold)
    than that in controls after two and three days of exposure to diesel
    exhaust. Both values returned to control levels after a one-week
    recovery period. Experiments involving incorporation of 14C-palmitic
    acid revealed transient changes in lipid metabolism after one day of
    exposure, with a threefold increase in the amount of phosphatidyl-
    choline in lung tissue (Wright, 1986).

         Exposure of various species by inhalation to 6 mg/m3 resulted
    in slight alterations in lung function and histopathological changes.

         Groups of 10 male Chinese hamsters were exposed for 8 h/day on
    seven days per week for six months to particle concentrations of
    either 6 or 12 mg/m3 diesel exhaust.

         Measurements of pulmonary function indicated dose-related
    increases in lung weight and dose-related decreases in vital and
    diffusion capacity. Pathological evaluation revealed oedema,
    thickening of the alveolar lining, a possible increase in collagen,
    and the presence of numerous particle-laden macrophages, which almost
    filled the alveolar spaces (Vinegar et al., 1981).

         Groups of 41-51 male and female Hartley guinea-pigs were exposed
    for 20 h/day on seven days per week to exhaust from a light-duty
    diesel engine diluted to attain a concentration of 6 mg/m3
    particulate matter. Separate groups were exposed to raw exhaust and
    exhaust that had been irradiated with ultra-violet light for an
    average of about 1 h. Pulmonary function was measured, and
    electrocardiography was conducted on animals exposed for four weeks;

    those examined for histological changes were exposed for eight weeks.
    The only changes noted after four weeks were a 35% increase in
    pulmonary flow resistance and a small, but significant decrease in
    heart rate in animals exposed to irradiated exhaust. No changes in the
    electrocardiogram were noted. The pathological changes were limited to
    hyperplasia of alveolar lining cells and slight focal thickening of
    the interstitium in occasional alveoli with accumulation of
    particle-laden macrophages. Growth rates were unaffected (Wiester et
    al., 1980).

         As part of the same study, cats (Pepelko et al., 1981) and male
    rats (15 per group) (Pepelko, 1982) were exposed for four weeks to
    unirradiated exhaust under identical conditions. The pulmonary
    function of rats was unchanged; the only changes were significant
    increases in vital and total lung capacity, but these differences were
    quite small and of questionable significance. The only functional
    change noted in the cats was a decrease in maximal expiratory flow
    rate at 10% of vital capacity. The pathological changes were again
    limited to focal thickening of the alveolar walls near accumulations
    of particle-laden macrophages. In general, the functional and
    pathological changes were limited under these experimental conditions.

         Kaplan et al. (1982) exposed male Fischer 344 rats (168 per
    group), A/J mice (672 per group), and male Syrian golden hamsters (236
    per group) to 1.5 mg/m3 of diesel exhaust particles for 20 h/day on
    seven days per week for three months or to clean air. Animals were
    killed at the end of exposure and after an additional six-month
    recovery period. Mortality was not increased during the observation
    period, and body weight was not significantly altered in any species.
    At the end of exposure, the relative weight of the lung was
    significantly increased in rats. In mice, the pulmonary adenoma
    response was increased but not significantly. In all three species,
    grey to black discolouration of the lungs and lung-associated lymph
    nodes were observed. Microscopically, most of the particulate was
    found in macrophages, but some was free. After six months' exposure to
    clean air, the lungs of all species had considerably less pigment
    accumulated in foci.

         Sixteen male Fischer 344 rats were exposed to 6 mg/m3 diesel
    exhaust for periods ranging from 6 h to 63 days, with subsequent
    histological examination of the lungs. The numbers of diesel
    particulate-containing alveolar macrophages, type II alveolar cells,
    and inflammatory cells were found to be increased over those in eight
    controls. After a nine-week exposure, accumulation of macrophages was
    observed near the terminal bronchioles (White & Garg, 1981). In rats
    that inhaled 6 mg/m3 diesel exhaust for two weeks and were either

    killed immediately or allowed to recover for six weeks before
    sacrifice, there was no statistical difference in the number of
    macrophage aggregates (240 000 and 220 000 per lung); however, the
    aggregates in the recovery group had significantly higher average
    diameter and volume, demonstrating macrophage attachment after the end
    of exposure (Garg, 1983).

         The observations of Wiester et al. (1980) and Garg (1983)
    demonstrate the importance of alveolar macrophages in the clearance of
    particles of low solubility. These are engulfed by alveolar
    macrophages and eliminated via the mucociliary escalator and
    gastrointestinal tract, or, in the case of lung particle overloading,
    increasing fractions are sequestered in focal areas of the lung tissue
    and transferred to the lymph nodes (Muhle et al., 1990).

    B7.3  Long-term exposure and studies of carcinogenicity

    B7.3.1  Non-neoplastic effects

         No obvious signs of toxicity were observed in groups of 183 male
    and 183 female Fischer 344 rats exposed by inhalation to diesel
    exhaust at 0.35, 3.5, or 7.1 mg/m3 for 7 h/day on five days per week
    for up to 30 months (Mauderly et al., 1987). The body weights did not
    differ significantly from those of controls, and the exposures did not
    affect the life span of animals of either sex. Focal fibrotic and
    proliferative lung disease was observed in parallel with progressive
    accumulation of diesel soot in the lung. Henderson et al. (1988)
    reported the results of biochemical and cytological analyses of lung
    lavage fluid and lung tissue from rats and mice exposed in the same
    study, including determinations of lactic dehydrogenase, glutathione
    reductase, ß-glucuronidase, glutathione, and hydroxyproline, and the
    numbers of macrophages and neutrophils. No changes were seen in rats
    or mice exposed to 0.35 mg/m3 , but at higher concentrations dose-
    and duration-dependent increases in biochemical parameters and cell
    numbers in lavage fluid were observed, indicating a chronic
    inflammatory response. Lung weights were also increased in male and
    female mice exposed to the two higher levels for any length of time,
    in rats exposed to 7.0 mg/m3 for 12 months or longer, in female rats
    exposed to 3.5 mg/m3 for 18 months, and in males exposed to
    3.5 mg/m3 for 24 months. Tissue enzyme levels, measured in lung
    lavage fluid as an indicator of damage, were consistently raised in
    rats and mice at the two higher exposure levels and in lung tissue of
    rats and mice at various times (most pronounced at 7 mg/m3). These
    changes probably indicate an increase in lung tissue due to increased
    inflammatory cells and epithelial cell proliferation.

         In the same study, Mauderly et al. (1988) described focal
    proliferative, fibrotic, and emphysematous changes in the lungs of
    animals at the higher doses. Measurements  in vivo revealed a
    reduction in lung volume and compliance.

         In a follow-up study, Mauderly et al. (1990) exposed male Fischer
    344 rats that were either normal or had pre-existing pulmonary
    emphysema (elastase-treated) to diesel soot at 3.5 mg/m3 for 7 h/day
    on five days per week for 24 months, to determine whether
    emphysematous rats were more susceptible to diesel exhaust. No
    increase in susceptibility was seen. The emphysematous lungs had a
    lower accumulation of particles than the normal lungs, which prevented
    an overproportional response of the compromised animals.

         Male and female Fischer 344 rats exposed to diesel exhaust
    particles at 0.7, 2.2, or 6.6 mg/m3 for 16 h/day on five days per
    week for two years or to filtered diesel exhaust showed an
    exposure-related reduction in the rate of weight gain throughout the
    study at the highest dose and an increase in absolute lung weight at
    the two higher doses (Brightwell et al., 1986, 1989).

         Groups of 72 male and 72 female Fischer 344 rats exposed to
    diesel exhaust particles at 2 mg/m3 for 7 h/day on five days per
    week for 24 months showed alveolar type II cell hyperplasia,
    inhibition of long-term clearance of particles and pulmonary
    lipidosis. No immunological abnormalities, no induction of pulmonary
    or hepatic xenobiotic metabolizing enzymes, and no effects on
    mortality, body weight gain, organ:body weight ratios or clinical
    chemical parameters were seen (Lewis et al., 1986, 1989). In the same
    study, a dose-related depression in phagocytic activity was seen in
    alveolar macrophages (Castranova et al., 1985), and a trend to
    increased pulmonary arterial wall thickness was observed in comparison
    with controls (Vallyathan et al., 1986).

         In groups of 15 male monkeys, exposure to diesel exhaust (as
    described above) resulted in mild obstructive airway disease, but
    there was no evidence of restrictive lung disease. The effects were
    observed throughout the exposure period, after 6, 12, 18, and 24
    months (Lewis et al., 1989).

         Reduced body weight, significantly increased lung:body weight
    ratios, and type II cell proliferation were seen in female Fischer 344
    rats exposed to whole-fraction diesel exhaust at 5 mg/m3 for 8 h/day
    on seven days per week for 24 months (Iwai et al., 1986).

         Groups of 96 male and 96 female Syrian golden hamsters, female
    NMRI mice, and female Wistar rats were exposed to diesel engine
    emissions at 4.4 mg/m3 for 19 h/day on five days per week for
    120-140 weeks (Heinrich et al., 1986a; Stöber, 1986). The body weights
    of mice were significantly reduced, by about 12%. Lung weights were
    increased after two years of exposure, by two- to threefold in rats
    and mice and by 50-70% in hamsters. Different levels of histological
    response were seen in the three species. Hamsters had thickened
    alveolar septa and bronchioalveolar hyperplasia. Of the mice, 64% had

    bronchioalveolar hyperplasia (controls, 5%), 71% had multifocal
    alveolar lipoproteinosis, and 43% had multifocal interstitial fibrosis
    (controls, 4%). The nose, larynx, and trachea were not affected.
    Treated rats showed severe inflammatory changes in the lungs,
    thickened septa, foci of macrophages, and hyperplastic and metaplastic
    lesions, but no significant changes in respiratory rate, minute
    volume, compliance, or resistance were seen in 14 exposed rats.
    Significantly increased airway resistance and a significant decrease
    in dynamic lung compliance were observed in rats after two years of
    exposure. The hamsters had a significant increase in airway resistance
    and a nonsignificant reduction in lung compliance at termination of a
    one-year exposure; these changes persisted during the second year of
    exposure. In the same experiments, filtered diesel exhaust did not
    have the same effects as total exhaust in any of the three species
    (Heinrich et al., 1986a).

         Groups of 27 female Fischer 344 rats and about 50 male and 50
    female C57Bl/6N and ICR mice were exposed to particle concentrations
    of 2-4 mg/m3 for 4 h/day on four days per week for up to 24 months.
    After 12 months, the incidence of alveolar hyperplasia was increased
    in all animals. Exposed rats had more severe inflammation than
    controls, with inflammatory exudate into the nose, and bronchitis and
    pneumonia, but this effect was ascribed to the housing, which was not
    specific pathogen-free (Takemoto et al., 1986).

         Groups of nine male Hartley guinea-pigs were exposed at
    concentrations of 0.25, 0.75, or 1.5 mg/m3 for 20 h/day on 5.5 days
    per week for up to two years. An analysis of the alveolar capillary
    membrane by quantitative morphometry showed that it was thickened as a
    result of an increase in the absolute tissue volume of interstitium
    and type II cells. Exposure to 0.75 mg/m3 for six months caused
    fibrosis in regions of macrophage clusters and focal type II cell
    proliferation, as observed under the light microscope. No appreciable
    change in morphometric parameters was seen after exposure to the
    lowest dose, but the higher doses increased the thickness of alveolar
    septa and the numbers of various types of alveolar cells (Barnhart et
    al., 1981, 1982).

         As part of this investigation, Fischer 344 rats were exposed by
    inhalation to 0.25 or 1.5 mg/m3 diesel exhaust particles, and
    various biochemical and cytological studies were conducted up to 36
    weeks (Misiorowski et al., 1980). At the high dose, normalized lung
    weight, the rate of collagen synthesis, and the cellular and lipid
    content of lung tissue were significantly increased after six months.
    It was concluded that the reactivity of fibrogenic cells had been
    enhanced. In a long-term study of respiratory end-points in male
    guinea-pigs, including light and electron microscopy, lavage cytology,
    and lung tissue biochemistry, the lowest-observable-adverse-effect
    level (LOAEL) was 0.796 and the no-observable-adverse-effect level
    (NOAEL) was 0.258 mg/m3 (US Environmental Protection Agency, 1992b).

         Groups of 120 male Fischer 344 rats, 450 male A/J mice, and 120
    male Syrian hamsters were exposed to 0.25, 0.75, or 1.5 mg/m3 diesel
    exhaust particles for 20 h/day on seven days per week for 15 months.
    Mortality was not increased in any species. A dose-related deposition
    of pigment was observed histopathologically in alveolar macrophages
    and regional lymph nodes. Accumulation of alveolar macrophages was
    associated with an increase in connective tissue in alveolar walls,
    with mild fibrosis and proliferation of type II pneumocytes; these
    effects were termed pneumoconiosis. No treatment-related responses
    were found in organs other than the respiratory tract. The LOAEL was
    0.735 and the NOAEL, 0.242 mg/m3 (Kaplan et al., 1983).

         Groups of 120 male and 95 female Fischer 344/Jcl rats were
    exposed to light-duty diesel engine exhaust particulates at 0, 0.1,
    0.4, 1.1, or 2.3 mg/m3 or heavy-duty diesel engine exhaust
    particulates at 0, 0.5, 1.0, 1.8, or 3.7 mg/m3 for 16 h/day on six
    days per week for up to 30 months (Ishinishi et al., 1986, 1988;
    Suzuki et al., 1990). The body weights of females exposed to the
    highest dose of heavy-duty engine exhaust were decreased throughout
    the study by 15-20% in comparison with controls. No changes were
    observed histopathologically in the lungs at any concentration.
    Accumulation of particle-laden macrophages was observed at levels
    > 0.4 mg/m3. In areas of macrophage accumulation, bronchiolar
    epithelium replaced alveolar epithelium in the ducts. Proliferation of
    bronchiolar epithelium and type II cells and oedematous thickening and
    fibrosis of the alveolar septum were observed, which developed into
    small fibrotic lesions classified as hyperplastic. The incidences of
    hyperplastic lesions were 4, 4, 6, 12, and 87 per 124 rats exposed to
    light-duty engine exhaust and 1, 3, 7, 14, and 25 per 124 rats in
    those exposed to heavy-duty engine exhaust. The LOAELs were
    1.2 mg/m3 for light-duty and 1.0 mg/m3 for heavy-duty exhaust, and
    the NOAELs were 0.4 and 0.5 mg/m3, respectively (US Environmental
    Protection Agency, 1992b).

         In the same experimental series, Kato et al. (1992) studied
    non-neoplastic lesions in the respiratory organs of the rats every six
    months. The marked changes consisted of intake of diesel particles by
    type I epithelium, hypertrophy, and glandular proliferation of type II
    epithelium, with many extended microvilli and increased numbers and
    size of lamellar inclusion bodies, focal increase of collagen fibres
    in the interstitium, and infiltration of particle-laden macrophages,
    neutrophils, mast cells, and plasma cells into the interstitium of
    alveolar septa. The most marked changes in the airway were focal
    shortening of cilia and protrusions of nonciliated cells. These
    morphological changes appeared in all groups exposed to particle
    concentrations > 1 mg/m3 after six months of exposure and were more
    marked with increasing particle concentration and with time.

         Exposure of Syrian golden hamsters to 4 or 11 mg/m3 diesel
    exhaust for 7 h/day on five days per week for five months resulted in
    enlargement of liver sinusoids and loss of cristae in mitochondria
    (Meiss et al., 1981).

         The induction of pulmonary adenomas and carcinomas in rats after
    long-term inhalation of carbon black is proposed to be due partly to
    inflammatory cells and their mediators (Driscoll et al., 1995;
    Oberdörster et al., 1995; Driscoll et al., in press). After exposure
    of rats to carbon black at 1, 7, or 50 mg/m3 for 6 h/day on five
    days per week for 13 weeks, 353, 1923, and 7112 µg of carbon,
    respectively, were deposited per lung. After exposure to air or to
    1 mg/m3 of carbon black, no macrophage inflammatory protein-2
    (MIP-2) mRNA expression and no increase in inflammatory cells was
    observed; with 7 or 50mg/m3, both MIP-2 mRNA and the number of
    neutrophils in bronchoalveolar lavage fluid were increased, and the
     hprt frequencies were 3.2 and 4.2 times greater than in controls.

         Groups of 25 cats were exposed to either clean air or diesel
    exhaust for 8 h/day on seven days per week for two years at a
    concentration of 6 mg/m3 through week 62 and to 12 mg/m3 for weeks
    62-124. Some of the cats were then held in clean air for an additional
    six months. No response was seen after 61 weeks, but after 123 weeks
    signs of restrictive lung disease were observed. Total lung capacity,
    forced vital capacity, functional residual capacity, peak expiratory
    flow rates, and diffusion capacity were significantly decreased
    (Moorman et al., 1985). Morphological evaluation revealed
    exposure-related lesions in the centriacinar region of the lungs,
    bronchiolar epithelial metaplasia associated with the presence of
    ciliated and basal cells, and alveolar macrophages containing diesel
    particle-like inclusions. After six months' recovery in clean air, the
    metaplasia had disappeared but the fibrosis had progressed (Hyde et
    al., 1985). In exposed cats, the epithelium of the terminal and
    respiratory bronchioles consisted of ciliated, basal, and Clara cells,
    whereas controls had only Clara cells (Plopper et al., 1983). These
    observations indicate that the foci of the lesions were the
    epithelium, the interstitial compartments, and the centriacinar region
    of the lung. The metaplastic changes suggest that cancer may be
    induced by continuous exposure, while the other major effect is
    pulmonary fibrosis resulting in restrictive disease which is not
    ameliorated and in fact progresses after removal from exposure.
    Progression of effects is probably due to continued lung overloading
    with particulate matter, despite cessation of exposure.

         The dose-dependent effects on lung morphology, histopathology,
    biochemistry, and cytology after long-term exposure to diesel exhaust
    described above are reflected in the following findings seen in rats
    at a lung burden of about 0.5 mg/lung (Henderson et al., 1988; Muhle
    et al., 1990):

    --   The wet and dry weights of lungs increase by up to fourfold in 
         comparison with those of concurrent controls. This is due to
         severe  functional overloading in the lung, resulting in enhanced
         leukocyte  influx and tissue proliferation.

    --   Pulmonary inflammation is manifested by cytological (increased
         numbers of polymorphonuclear neutrophils) and biochemical (lactic
         dehydrogenase, collagen in lung lavage fluid) parameters.

    --   The retention of material is characterized by an increasing
         number of particle-laden macrophages and by black discolouration
         of lungs due to sequestration of accumulated particles.
         Additionally, increased transfer of material to the
         lung-associated lymph nodes is observed.

    --   Later findings during inhalation are proliferative alterations of
         the epithelial cells and onset of fibrosis.

    B7.3.2  Carcinogenicity

    B7.3.2.1  Inhalation

         Rats were exposed in several studies to diesel emissions
    containing > 2 mg/m3 of particles for more than two years,
    resulting in the induction of pulmonary tumours (Table 34). Many
    organic and inorganic gaseous components were also present in the test
    atmospheres, and the relatively high concentration of some pulmonary
    irritants like nitrogen dioxide, aldehydes, and sulfur dioxide may
    have contributed to the inflammatory response and to epithelial cell
    damage and subsequent cell proliferation. Exposure to gaseous PAHs,
    like naphthalene which is toxic to Clara cells, may influence the
    metabolic state of pulmonary cells. It is not clear however, whether
    the combined exposure to particles and gases in these experiments
    favoured tumour induction. It should be noted that Brightwell et al.
    (1986), Heinrich et al. (1986a), Iwai et al. (1986), Ishinishi et al.
    (1986), and Takaki et al. (1989) have reported that the gaseous phase,
    i.e. the proportion of exhaust without particles, does not induce
    pulmonary tumours.


        Table 34.  Carcinogenicity of diesel engine exhaust in rats
                                                                                                                                                

    Strain       No. and   Exposure concentration   Duration                Lung        Pulmonary tumours         Probability     Reference
                 sex       (mg/m3)a                 (observation            burden                                of error (%)b
                                                    period)                 (mg/lung)   No.   %      Mean
                           Actual    Continuous                                                      (f+m) (%)
                                                                                                                                                

    Wistar       40 m      8.3       1.5            6 h/d, 5 d/week, up     NR          1     17c                 NR              Karagianes
                 40 m      Clean air                to 20 months                        0                                         et al.
                                                    (20 months)                                                                   (1981)

    Fischer      72 f      6.6       3.1            16 h/d, 5 d/week, 24    NR          39    54     38.5         < 0.1           Brightwell
    344          71 m                               months (30 months)                  16    23                                  et al.
                 72 f      2.2       1.1                                                11    15     9.7          < 0.1           (1986, 1989)
                 72 m                                                                   3     4
                 71 f      0.7       0.4                                                      0      0.7
                 72 m                                                                   1     1
                           Filtered exhaust                                                   0
                                                                                              0
                 142 f     Clean air                                                    1     1      1.4
                 140 m                                                                  3     2

    Wistar       95 f      4.4       2.5            19 h/d, 5 d/week, 32    NR          17    18                  < 0.1           Heinrich et
                 92 f      Filtered exhaust         months (up to death)                      0                                   al. (1986a)
                 96 f      Clean air                                                          0
                                                                                                                                                

    Table 34 (contd)
                                                                                                                                                

    Strain       No. and   Exposure concentration   Duration                Lung        Pulmonary tumours         Probability     Reference
                 sex       (mg/m3)a                 (observation            burden                                of error (%)b
                                                    period)                 (mg/lung)   No.   %      Mean
                           Actual    Continuous                                                      (f+m) (%)
                                                                                                                                                

    Fischer
    344/Jcl
    Light-duty   60 f      2         1.1            16 h/d, 6 d/week, 30    NR          1     2      2.4          NR              Ishinishi
    engine       64 m                               months (30 months)                  2     3                                   et al.
                 59 f      1         0.6                                                2     3      4.1                          (1986,1988;
                 64 m                                                                   3     5                                   Takaki et al.
                 61 f      0.4       0.2                                                      0      0.8                          (1989)
                 64 m                                                                   1     2
                 59 f      0.1       0.06                                               2     3      1.6
                 64 m                                                                         0
                 59 f      Clean air                                                    2     3      3.3
                 64 m                                                                   2     3
    Heavy-duty   60 f      4         2.3                                                3     5      6.5          1.8
    engine       64 m                                                                   5     8
                 59 f      2         1.1                                                1     2      3.3
                 64 m                                                                   3     5
                 61 f      1         0.6                                                      0      0
                 64 m                                                                         0
                 59 f      0.4       0.2                                                1     2      0.8
                 64 m                                                                         0
                 59 f      Clean air                                                    1     2      0.8
                 64 m                                                                         0
                 16 m      Filtered exhaust                                                          0
                                                                                                                                                

    Table 34 (contd)
                                                                                                                                                

    Strain       No. and   Exposure concentration   Duration                Lung        Pulmonary tumours         Probability     Reference
                 sex       (mg/m3)a                 (observation            burden                                of error (%)b
                                                    period)                 (mg/lung)   No.   %      Mean
                           Actual    Continuous                                                      (f+m) (%)
                                                                                                                                                

    Fischer      19 f      4.9       1.6            8 h/d, 7 d/week, 24     NR          8     42                  < 1d            Iwai et al.
    344          16 f      Filtered exhaust         months (24 or 30                          0                                   (1986)
                 22 f      Clean air                months)                             1     5

    Fischer      15 f      2.0-4.0   0.2-0.4        4 h/d, 4 d/week, 24     NR                0                   NR              Takemoto
    344          15 f      Clean air                months (24 months)                        0                                   et al.
                                                                                                                                  (1986)

    Fischer      69 f      7         1.5            7 h/d, 5 d/week, 30     NR          13    19     16.1         < 0.1 (f)       Mauderly et
    344/Crl      74 m                               months (up to death)                10    14                  1.7 (m)         al. (1986)
                 68 f      3.5       0.7                                                2     3      4.6
                 63 m                                                                   4     6
                 68 f      0.35      0.07                                                     0      0.7
                 70 m                                                                   1     1
                 68 f      Clean air                                                    0     1.4
                 73 m                                                                   2     3

    Fischer      227 f+m   7         1.5            7 h/d, 5 d/week, 30     20.8                     12.8         < 0.1           Mauderly et
    344/N Crl    221 f+m   3.5       0.7            months (up to death)    11.5                     3.6          4.8             al. (1987)
                 223 f+m   0.35      0.07                                   0.6                      1.3
                 230 f+m   Clean air                (24 months)                                      0.9
                                                                                                                                                
    Table 34 (contd)
                                                                                                                                                

    Strain       No. and   Exposure concentration   Duration                Lung        Pulmonary tumours         Probability     Reference
                 sex       (mg/m3)a                 (observation            burden                                of error (%)b
                                                    period)                 (mg/lung)   No.   %      Mean
                           Actual    Continuous                                                      (f+m) (%)
                                                                                                                                                

    Wistar       100 f     7.0       3.8            18 h/d, 5 d/week, 24    63.9        22    22.0                < 0.1           Heinrich et
                 200 f     2.5       1.3            months (30 months)      23.7        11    5.5                 < 1             al. (1992,
                 198 f     0.8       0.5                                    6.3               0                                   1995);
                 217 f     Clean air                (24 months)             NR          1     0.5                                 Heinrich
                                                                                                                                  (1994)

    Fischer      212 f+m   6.3       3.0            16 h/d, 5 d/week, 24    85.4        38    17.9   17.9         NR              Nikula et al.
    344/N        210 f+m   2.4       1.1            months (24 months)      40.7        13    6.2    6.2                          (1994)
                 214 f+m   Clean air                (23 months)                         3     1.4    1.4

    Fischer      49 f      4.7       1.3            15 h/d, 3 d/week, 24,   NR          6     12.2                NR              Kawabata
    344/N        42 f      4.7       1.3            12, or 6 months                     8     19.0                                et al.
                 45 f      4.7       1.3                                                1     2.2                                 (1993)
                 48 f      Clean air                                        5           10.4

    Fischer      183 f+m   2.0       0.4            7 h/d, 5 d/week, 24     NR          8     4.4    4.4          NR              Lewis et al.
    344/N        180 f+m   Clean air                months                              6     3.3    3.3                          (1989)
                                                                                                                                                

    From Pott & Heinrich (1987) and supplemented; NR, not reported
    a  Mass median aerodynamic diameter, 0.17-0.25 µm; 'Continuous', equivalent continuous value
    b  Unilateral Fisher's exact test
    c  Only six animals investigated
    d  With Yates' correction
             No difference in lung tumour incidence was seen in groups of male
    and female rats exposed for 7 h/day on five days per week for 24
    months either to diesel exhaust particulates at 2 mg/m3 or to
    filtered air (Lewis et al., 1989).

         Exposure of six male Wistar rats to diesel exhaust soot at
    8.3 mg/m3 and coal dust for up to 20 months gave inconclusive
    results (Karagianes et al., 1981). The study had little statistical
    power owing to the small group size, and the insufficient duration of
    exposure may have prevented detection of tumours, although one mammary
    fibroadenoma and one lung bronchiolar adenoma were detected. One
    subcutal fibrosarcoma and one renal lymphoma were seen in controls.

         Eight bronchioalveolar adenomas and nine squamous-cell tumours
    were seen in 95 rats (18%) that had inhaled 4.4 mg/m3 diesel exhaust
    for 19 h/day on five days per week for 32 months. No tumours were
    observed in controls (Heinrich et al., 1986a). The dose-dependent
    carcinogenic potency of diesel exhaust was confirmed in a study in
    which rats were exposed to diesel exhaust at concentrations of 0.8,
    2.5, or 7.0 mg/m3 for 18 h/day on five days per week for 24 months
    (Heinrich et al., 1992; Heinrich, 1994; Heinrich et al., 1995).

         In male and female rats exposed to diesel exhaust at 6.6 mg/m3,
    a high incidence of lung tumours was observed in animals dying or
    sacrificed after 24 months (24/25 females, 12/27 males). The incidence
    of lung tumours (Table 34) was derived by pooling pulmonary adenomas;
    squamous-cell carcinomas; adenocarcinomas; mixed ademomas,
    adenocarcinomas, and squamous-cell carcinomas; and mesotheliomas.
    Non-neoplastic pulmonary lesions were not described (Brightwell et
    al., 1986, 1989).

         Groups of 24 female Fischer 344 rats were exposed to clean air,
    to whole-fraction (4.9 mg/m3 particulates) fuel, or to filtered
    diesel exhaust for 8 h/day on seven days per week for 24 months. After
    six months of exposure, the rats exposed to whole exhaust had lower
    body weights than those in the other two groups; after 18 months, both
    exposed groups showed body weight reduction. The relative lung weight
    was increased after 12 months' exposure to whole exhaust, and
    epithelial proliferation and significantly elevated numbers of lung
    tumours were seen in animals exposed to the whole fraction after six
    months. Increased incidences of leukaemia and mammary gland tumours
    were also reported in animals exposed to whole or filtered exhaust;
    however, because of the small numbers of animals, the significance of
    the results cannot be determined (Iwai et al., 1986).

         A total of 140 female and 140 male Fischer 344/N rats (in groups
    of three males and three females, five males and five females, or
    eight males and eight females, depending on the end-point) were
    exposed for 16 h/day on five days per week for up to 24 months,
    beginning at eight weeks of age, to particles of diesel exhaust or
    carbon black at 2.5 or 6.5 mg/m3 of air, or to clean air (Mauderly
    et al., 1994). Rats were killed after 3, 6, 12, 18, or 23 months of
    exposure for histopathological assessment and measurement of lung and
    lung-associated lymph node burdens of particles, lung weight,
    bronchoalveolar lavage indicators of inflammation, DNA adducts in
    whole lung and alveolar type II cells, and chromosomal anomalies in
    circulating lymphocytes. Diesel exhaust and carbon black had nearly
    identical effects, with a similar relation to exposure. There was a
    dose-related slowing of particle clearance after three months, with
    progressive accumulation of particles in lungs and lymph nodes.
    Inflammation, epithelial proliferation, and fibrosis were progressive
    and related to dose. Diesel exhaust slightly increased the number of
    lung DNA adducts, and the numbers of DNA adducts in type II cells were
    increased by both diesel exhaust and carbon black. No exposure-related
    chromosomal damage was found in circulating lymphocytes. The incidence
    of primary lung neoplasms was increased significantly and was related
    to the doses of both diesel exhaust and carbon black. The types of
    neoplasms were identical and included benign adenomas, malignant
    adenocarcinomas, squamous-cell carcinomas, and adenosquamous
    carcinomas; squamous cysts were also seen. The neoplasms had similar
    growth characteristics when transplanted; 50-67% of the transplanted
    squamous-cell carcinomas and 25-40% of the transplanted
    adenocarcinomas grew after injection into athymic mice (Table 35),
    whereas the squamous cysts did not.

         Fischer 344/Jcl rats were exposed to a range of concentrations of
    exhaust emission particulates from light-duty (up to 2 mg/m3) or
    heavy-duty diesel engines (up to 4 mg/m3) for 16 h/day on six days
    per week for up to 30 months. After 18 months, marked hyperplasia of
    type II cells was noted in the groups at higher exposures. After 30
    months, treatment-related pulmonary hyperplastic lesions were
    detected, and a dose-response relationship was seen for tumours in the
    groups exposed to heavy-duty but not light-duty diesel exhaust. The
    lung tumours were diagnosed as adenocarcinomas, squamous-cell
    carcinomas, or adenosquamous carcinomas. A significant correlation was
    observed between the incidence of anthracosis and areas with
    hyperplastic lesions; furthermore, anthracosis and hyperplastic
    lesions did not appear in lungs exposed to particle-free exhaust. The
    authors concluded that the occurrence of hyperplastic lesions was
    dependent on the presence of carbon particles (Ishinishi et al., 1986,
    1988; Takaki et al., 1989).

    Table 35.  Growth in athymic mice of lung tumours from rats exposed to
               diesel exhaust and carbon black
                                                                        

    Type of tumour    Carbon black                Diesel exhaust
                                                                        

                      Number       Growing        Number       Growing
                      implanted    No.    %       implanted    No.    %
                                                                        

    Adenocarcinom      8            2     25       10           4     40
    Squamous-cell      2            1     50        3           2     67
    carcinoma
    Squamous cyst     19            0      0        7           0      0
                                                                        

    From Mauderly et al. (1994)

         In a large-scale study of unfiltered diesel exhaust, 1097 Fischer
    344 rats were exposed for up to 30 months to 0.35, 3.5, or 7.0 mg/m3
    of diesel soot. A group of 365 rats exposed to clean air served as
    controls. Exposure to exhaust did not cause overt signs of toxicity,
    and no significant, exposure-related alterations in body weight were
    observed. Rats were killed at six-month intervals, examined for their
    lung burdens of diesel soot, and subjected to complete necropsy.
    Diesel soot accumulated at all doses but minimally in the low-dose
    group. An increased incidence of tumours was observed in the groups at
    the medium and high doses; most (81%) of the tumours were observed
    after two years of exposure. The inhaled soot concentration and the
    lung burden of diesel soot were highly significantly related to tumour
    incidence ( P < 0.001) (Mauderly et al., 1987). In a follow-up study
    with longer daily exposure (16 h/day), the dose-dependency of the
    tumour incidence was reproduced (Nikula et al., 1994). In this study,
    the carcinogenicity of diesel exhaust soot (8% extractable organic
    material; 2.5 mg/m3) was compared with that of carbon black (0.12%
    extractable organic material; 6.5 mg/m3 ) in 1152 male and female
    Fischer 344 rats exposed for 16 h/day on five days per week for up to
    24 months and examined histopathologically. The lung burden of
    particles was greater after exposure to diesel exhaust than to carbon
    black, but the neoplastic response in the lung was similar and usually
    occurred late in the study. When the lung burden of particles was used
    as the dose parameter,  the carcinogenicity of carbon black was
    greater than that of diesel exhaust soot.

         Groups of 50 four-week-old female Fischer 344 rats were exposed
    to diesel soot at a mean concentration of 4.73 mg/m3 for 15 h/day
    three times per week for 6, 12, or 24 months, and animals that died
    after 18-30 months were evaluated for tumours. After 30 months, all
    animals, including controls, were sacrificed and examined for lung

    tumours. The incidences of lung tumours were 10.4% in controls and
    2.2% after six months' exposure to diesel exhaust, 19.0% after 12
    months, and 12.2% after 24 months. No statistical analysis was
    presented, but it seems unlikely that there was a significant effect
    (Kawabata et al., 1993).

         Concurrent administration of known carcinogens with diesel
    exhaust has also been studied. Exposure of rats to diesel exhaust
    given an intraperitoneal injection of  N-nitrosodiisopropylamine or
     N-nitrosodiethylamine did not influence the rates of tumours induced
    by the nitrosamines (Takemoto et al., 1986). The effects of treatment
    of hamsters with  N-nitrosodiethylamine or benzo[ a]pyrene were not
    influenced by subsequent administration of diesel exhaust, whereas
    pretreatment of mice with benzo[ a]pyrene or dibenzanthracene followed
    by diesel exhaust gave inconsistent results. Exposure to diesel
    exhaust did not affect the total tumour rates in rats pretreated with
    6.25 or 12.5 g/kg body weight of  N-nitrosodipentylamine; however,
    the incidence of squamous-cell tumours was increased in both
    pretreated groups (Heinrich et al., 1986a).

         Hamsters were exposed for two years to unfiltered diesel exhaust
    at 0, 0.7, 2.2, or 6.6 mg/m3 or to filtered diesel exhaust, and
    groups of control and treated hamsters were administered
     N-nitrosodiethylamine subcutaneously before the beginning of
    exposure. The rate of pulmonary tumours was not increased by diesel
    exhaust, with or without pretreatment with the nitrosamine. Several
    hamsters that developed 'wet tail' infection after six months on test,
    resulting in significant mortality (46%), were treated with
    oxytetracycline, chloramphenicol, and dimetridazole (Brightwell et
    al., 1986, 1989).

         Overall, the only clear effect in these studies of
    co-carcinogenicity in rats was an increase in the percentage of
    malignant tumours in animals treated with diesel exhaust plus
    carcinogens in comparison with the carcinogen alone (Heinrich et al.,
    1986a).

         The results of studies in other species (mice, hamsters, and
    monkeys) treated by inhalation are given in Table 36. Heinrich et al.
    (1982) found significant increases in lung tumour incidence in mice,
    but this result could not be repeated in later studies (Heinrich et
    al., 1986a, 1992, 1995). Small but statistically significant increases
    in lung tumour incidence were reported in ICR and C57Bl mice (Takemoto
    et al., 1986) and in female Sencar mice (Pepelko & Peirano, 1983);
    however, because only small increases in non-malignant tumours were
    seen only in females in the study of Pepelko & Peirano (1983), the
    results must be considered to be weakly positive or equivocal.
    Hamsters showed no response to diesel exhaust.


        Table 36.  Long-term studies of diesel exhaust by inhalation in species other than rats
                                                                                                                                                

    Strain, species   No. and sex       Exposure                                          Results                                Reference
                                                                            
                                        Material        Particles   Duration
                                                        (mg/m3)
                                                                                                                                                

    NMRI mouse        2 × 96 f          (a) Exhaust     4           19 h/d, 5 d/week,     Lung tumours (adenomas and             Heinrich et al.
                                        (b) Filtered                27-28 months          adenocarcinomas) (a) 24/76 (32%),      (1986a)
                                        (c) Clean air                                     including 13 (17%) adenocarcinomas;
                                                                                          (b) 29/93 (31%), including 18 (19%)
                                                                                          adenocarcinomas; (c) 11/84 (13%),
                                                                                          including 2 (2.4%) adenocarcinomas
                                                                                          Incidence in historical controls,
                                                                                          < 32%

    ICR and           315 ICR,          Exhaust or      2.0-4.0     4 h/d, 4 d/week,      Lung tumours (adenomas and             Takemoto et al.
    C57Bl/6N          297 C57Bl/6N      clean air                   13-28 months          adenocarcinomas). ICR: 14/56;          (1986)
    mice, newborn     (sex                                                                7/60 in controls; C57Bl/6N:
                      unspecified)                                                        17/150; 1/51 in controls

    Sencar mouse      2 × 130 f         (a) Exhaust     12          8 h/d, 7 d/week,      Adenomas and carcinomas                Pepelko &
                      2 × 130 m         (b) Clean air               to 15 months of       (a) m: 5.9%; f: 16.3%                  Peirano (1983)
                                                                    age (from             (b) m: 3.8%; f: 7.2%
                                                                    conception to
                                                                    sacrifice)

    NMRI mouse        2 × 80 f          (a) Exhaust     7.5         18 h/d, 5 d/week,     Lung tumours                           Heinrich et al.
                                        (b) Clean air               13.5 months, 9.5      (a) 32.1%                              (1992, 1995)
                                                                    months' recovery      (b) 30%
                                                                                                                                                

    Table 36 (contd)
                                                                                                                                                

    Strain, species   No. and sex       Exposure                                          Results                                Reference
                                                                            
                                        Material        Particles   Duration
                                                        (mg/m3)
                                                                                                                                                

    NMRI and          3 × 120 f         (a) Exhaust     4.5         18 h/d, 5 d/week;     Lung tumours                           Heinrich et al.
    C57Bl/6N mice     3 × 120 f         (b) Filtered                NMRI, 23 months;      NMRI:  (a) 23.3%                       (1992, 1995)
                                        (c) Clean air               C57Bl, 24 months             (b) 46.7%
                                                                    + 6 months'                  (c) 30%
                                                                    recovery              C57Bl: (a) 8.5%
                                                                                                 (b) 3.6%
                                                                                                 (c) 5.1%

    Syrian golden     3 × 48 f          (a) Exhaust     3.9 ± 0.5   7-8 h/d, 5 d/week,    No effect on survival, no lung         Heinrich et al.
    hamster                             (b) Filtered                up to 30 months       tumours                                (1982)
                                        (c) Clean air

    Syrian golden     3 × 48 f          (a) Exhaust     4           19 h/d, 5 d/week,     Normal life span, no lung tumours      Heinrich et al.
    hamster           3 × 48 m          (b) Filtered                27-28 months                                                 (1986a)
                                        (c) Clean air

    Syrian golden     52 f + 52 m       (a) Exhaust     0.7, 2.2,   16 h/d, 5 d/week,     No significant increase in             Brightwell
    hamster           104 f + 104 m     (b) Filtered    6.6         24  months            frequency of tumours in                et al.
                      (controls)        (c) Clean air                                     respiratory tract                      (1986, 1989)

    Cat               2 × 25 m          (a) Exhaust     6 (1st      8 h/d, 7 d/week,      Lung function: decrease in closing     Pepelko &
                                                        yr), 12     24  months            volume after 1 year; reduction         Peraino (1983)
                                                        (2nd yr)                          in inspiratory, vital, and total
                                        (b) Clean air                                     lung capacity after 2 years;
                                                                                          diagnosis: pulmonary fibrosis of
                                                                                          the interstitial or
                                                                                          intra-alveolar type
                                                                                                                                                

    Table 36 (contd)
                                                                                                                                                

    Strain, species   No. and sex       Exposure                                          Results                                Reference
                                                                            
                                        Material        Particles   Duration
                                                        (mg/m3)
                                                                                                                                                

    Cynomolgus        4 × 15 m          (a) Coal dust   2.0         7 h/d, 5 d/week,      No difference in tumour incidence      Lewis et al.
    monkey                              (b) Exhaust     2.0         24 months             among groups                           (1986)
                                        (c) Coal dust   1.0
                                            + exhaust   1.0
                                        (d) Clean air
                                                                                                                                                
             It is clear that repeated inhalation of diesel exhaust at
    concentrations of more than about 2 mg/m3 (actual value),
    corresponding to a calculated equivalent continuous concentration of
    about 1 mg/m3, increases the incidence of pulmonary tumours. These
    tumours occur late in life after exposures of 24 months.

    B7.3.2.2  Other routes of exposure

         Studies have also been conducted by intratracheal instillation in
    rats and hamsters and by painting on mouse skin (Table 37). While the
    former are especially useful for hazard identification and for
    evaluating specific mechanistic aspects, they are less useful for the
    purposes of risk characterization and no attempt was made to identify
    critical studies. The results of the studies by intratracheal
    instillation, however, confirm that both diesel exhaust particles and
    carbon black induce lung tumours and demonstrate that the specific
    surface area of carbonaceous particles is correlated with the
    tumorigenic potency. Both of these findings support the suggestion
    that a nonspecific particle effect is of crucial importance for the
    induction of lung tumours by diesel exhaust. The dermal experiments,
    conducted by Nesnow et al. (1982a,b, 1983), have been used to estimate
    risk quantitatively, by the comparative potency method.

    B7.4  Dermal and ocular irritation; dermal sensitization

         No data were available.

    B7.5  Reproductive toxicity, embryotoxicity, and teratogenicity

    B7.5.1  Reproductive toxicity

         In a two-generation study of reproduction, 100 male and 100
    female CD-1 mice were exposed by inhalation to exhaust from a
    light-duty diesel engine for 8 h/day on seven days per week, at a
    concentration of 12 mg/m3. Most treatment-related effects were
    minimal. Overall fertility and survival rates were not significantly
    altered (Pepelko & Peirano, 1983).

         (C57Bl/6 × C3H)F mice (number not given) showed sperm
    abnormalities (reduced sperm count and weight of testis;
    teratospermia) after daily intraperitoneal injections of 50, 100, or
    200 mg/kg body weight of diesel exhaust particles for five days. The
    highest dose caused an eightfold increase in abnormalities over that
    in controls and a significant decrease in sperm number (Quinto & De
    Marinis, 1984).


        Table 37. Carcinogenicity of diesel exhaust after exposure other than by inhalation
                                                                                                                                                

    Strain,      No. and    Exposure                                               Schedule,         Results                           Reference
    species      sex                                                               duration
                            Route          Material                    Dose
                                                                       (mg)
                                                                                                                                                

    Fischer 344  (a) 31 f   Intratracheal  (a) Activated carbon        1/animal    10/week,          Survival rate, 71-83%             Kawabata
    rat          (b) 59 f   instillation   (b) Diesel particles        1/animal    30 months'        (lowest ingroup b). Malignant     et al.
                 (c) 53 f                  (c) None                                observation       lung tumours: (a) 7/23;           (1986)
                 (d) 27 f                  (d) Vehicle                                               (b) 20/42; (c) 0/44; (d) 1/23;
                                                                                                     P < 0.01. Benign and
                                                                                                     malignant tumours combined:
                                                                                                     (a) 11/23; (b) 31/42

    Osborne-     Groups     Lung           Organic material from                   Observation       (a) 1 bronchioalveolar adenoma;   Grimmer
    Mendel rat   of 35 f    implant        diesel exhaust:                         until             (b) 5 squamous-cell carcinomas;   et al.
                                           (a) Hydrophilic fraction    6.7         spontaneous       (c) 1 bronchioalveolar adenoma;   (1987)
                                           (b) Hydrophobic fraction    20.0        death             (d) 6 carcinomas;
                                           (c) Nonaromatic compounds   19.2                          (e) No tumours
                                               + PAHs (2 or                                          (f) 1 carcinoma
                                               3 rings)                                              (g) 7 carcinomas, 1 adenoma
                                           (d) PAHs ( > 4 rings)       0.2                           (h) No tumours
                                           (e) Polar PAHs              0.3                           (i) 1 adenoma
                                           (f) Nitro-PAHs              0.2
                                           (g) Reconstituted
                                               hydrophobic fraction    19.9
                                           (h) None
                                           (i) Vehicle
                                                                                                                                                

    Table 37 (contd)
                                                                                                                                                

    Strain,      No. and    Exposure                                               Schedule,         Results                           Reference
    species      sex                                                               duration
                            Route          Material                    Dose
                                                                       (mg)
                                                                                                                                                

    Wistar rat   (a) 40 f   Intratracheal  (a) Diesel soot (34 m2/g)   (a) 3       × 15              Primary lung tumours (%)          Pott &
                 (b) 58 f   instillation   (b) Diesel soot (70 m2/g)   (b) 3       × 10              (a) 65                            Roller
                 (c) 38 f                  (c) Diesel soot (70 m2/g)   (c) 3       × 20              (b) 60                            (1994);
                 (d) 37 f                  (d) Carbon black (270       (d) 3       × 15              (c) 66                            Pott et
                                               m2/g)                                                 (d) 65                            al.(1994)
                 (e) 37 f                  (e) Activated charcoal      (e) 3       × 10
                 (f) 39 f                      (860 m2/g)                  3       × 20              (e) 27
                 (g) 40 f                  (f) NaCl solution           (f) 3       × 10              (f) 0
                                           (g) NaCl solution           (g) 0.4     × 20 until        (g) 0
                                                                           ml      spontaneous
                                                                                   death or 131
                                                                                   weeks

    Wistar rat   Groups     Intratracheal  (a) Diesel soot (native)    (a) 1       × 15              Primary lung tumours (%):         Heinrich
                 of 48 f    instillation   (b) Diesel soot (toluene    (b) 2       × 15              (a) 17                            (1994)
                                               extract; 130 m2/g)          1       × 15
                                           (c) Carbon black            (c) 1       × 15              (b) 23
                                               (toluene extract;       (d) 1       × 15                   4
                                               270 m2/g; primary       (e) Vehicle control           (c) 21
                                               particle, 15 nm)
                                           (d) Carbon black (toluene                                 (d) 8
                                               extract; 270 m2/g;                                    (e) 0
                                               primary particle,
                                               15 nm)
                                           (e) NaCl solution
                                                                                                                                                

    Table 37 (contd)
                                                                                                                                                

    Strain,      No. and    Exposure                                               Schedule,         Results                           Reference
    species      sex                                                               duration
                            Route          Material                    Dose
                                                                       (mg)
                                                                                                                                                

    Syrian       Groups     Intratracheal  (a) Diesel particles        1.25,       1/week, 15        1 lung adenoma at high dose       Shefner
    golden       of 50 m;   instillation   (b) Diesel particles +      2.5,                          in groups (a) and (c) after 61    et al.
    hamster      various                       same amounts of         weeks                         weeks; no lung tumours in         (1982
                 controls                      ferric oxide                                          controls
                                           (c) Diesel particle
                                               extract
                                               + ferric oxide

    Syrian       (a) 3 ×    Intratracheal  (a) Exhaust extract         0.1, 0.5,   1/week, 15        Survival rates: (a) 95, 92,       Kunitake
    golden           62 m   instillation   (b) Vehicle                 1           weeks             71%; (b) 98%. No difference in    et al.
    hamster      (b) 59 m                  (c) 0.5 mg                                                tumour incidence between (a)      (1986)
                 (c) 62 m                      benzo[a]pyrene                                        and (b); 88% respiratory
                                                                                                     tumours in positive controls

    C57Bl        12 m,      Dermal         Acetone extract of          0.5 ml      3 times/week      16 mice dead by 10 weeks, 33      Kotin
    mouse        40 f,                     particles                               for life or       skin tumours by 13 months, 2      et al.
                 69                                                                22-23 months      skin papillomas, no tumours in    (1955)
                 controls                                                                            controls

    A mouse      50 m,      Dermal         Acetone extract of          0.5 ml      3 times/week      Males: 8 skin tumours by 16
                 25 f,                     particles                               for life or       1 papilloma, 3 squamous-cell
                 34                                                                22-23 months      carcinomas. Females: 20 skin
                 controls                                                                            tumours by 13 months,17 tumours
                                                                                                     at 13-17 months, no tumours in
                                                                                                     controls
                                                                                                                                                

    Table 37 (contd)
                                                                                                                                                

    Strain,      No. and    Exposure                                               Schedule,         Results                           Reference
    species      sex                                                               duration
                            Route          Material                    Dose
                                                                       (mg)
                                                                                                                                                

    Sencar       Groups     Dermal         Dichloromethane             0.1, 0.5,   1/week, 50-52     Skin carcinomas:                  Nesnow
    mouse        of 40 m                   extracts of particles       1.0, 2.0,   weeks             Engine A: 3% (m), 5% (f) at       et al.
                 + 40 f                    from engines (A, B, E)      or 4.0                        4 mg Engine B: 3% (m) at 0.5 mg   (1982a,b,
                                           Benzo[a]pyrene              12.6-                         Engine E: 3% (f) at 0.1 mg        1983)
                                                                       202 µg                        Positive control: 10-90%

    C57Bl/6N     Groups     Subcutaneous   Particles in olive oil      10, 25,     1/week, 5         First tumours palpated in:        Kunitake
    mouse        of         injection      containing 5% dimethyl      50, 100,    weeks; 18         week 47 at 25 mg/kg bw            et al.
                 15-30 f                   sulfoxide                   200, or     months'           week 30 at 50 mg/kg bw            (1986)
                                                                       500 mg/kg   observvation      week 27 at 100 mg/kg bw
                                                                       bw                            week 39 at 200 or 500 mg/kg bw
                                                                                                     Malignant fibrous histiocytomas
                                                                                                     in 5/22 mice at 500 mg/kg bw;
                                                                                                     0/38 in controls
                                                                                                                                                

    PAH, polycyclic aromatic hydrocarbon; TPA, 12-O-tetradecanoylphorbol 13-acetate
             Groups of 15 male cynomolgus monkeys were exposed by inhalation
    to 2 mg/m3 diesel particulates for 7 h/day on five days per week for
    two years. Sperm motility and velocity were similar to those of
    controls (Lewis et al., 1989).

         In a test for dominant lethal mutation, male Fischer 344 rats
    inhaled 2 mg/m3 diesel exhaust for 7 h/day on five days per week for
    six months and were subsequently mated with untreated females. Live
    and dead implants and preimplantation losses were analysed on days
    19-20 of gestation: no significant effects were observed (Lewis et
    al., 1989). In a similar test, 100 male and 54 female T stock mice
    were exposed to 6 mg/m3 diesel exhaust for 8 h/day on seven days per
    week for seven weeks. There were no significant dominant lethal
    effects in males or females. A reduced number of corpora lutea
    (reproductive function) was the only significant result (Pepelko &
    Peirano, 1983).

    B7.5.2  Embryotoxicity

         No increase in the frequency of sister chromatid exchange was
    seen in the livers of fetuses of Syrian golden hamsters that inhaled
    diesel exhaust particles at 12 mg/m3 for 8 h/day on days 5-13 of
    gestation. An intraperitoneal injection of 300 mg/kg body weight of
    diesel particulates on day 12 of gestation also had no significant
    effect on this end-point; however, intraperitoneal injection of 23%
    particulate mass extracted with methylene chloride on day 12 resulted
    in a dose-dependent increase in sister chromatid exchange frequency in
    fetal liver on day 13, and a doubling was seen at 320 mg/kg body
    weight. It was concluded that chemicals must be eluted from diesel
    particles in order for the genotoxic material to cross the placenta
    (Pereira et al., 1982).

    B7.5.3  Teratogenicity

         Sprague-Dawley rats and New Zealand white rabbits were exposed by
    inhalation to exhaust from a light-duty diesel engine, at a
    particulate matter concentration of 6 mg/m3, for 8 h/day on seven
    days per week.

         Twenty rats were exposed on days 5-16 of gestation and 20 rabbits
    on days 6-18. The numbers of viable fetuses per litter, dead fetuses
    per litter, resorptions per litter, implantation sites per litter,
    corpora lutea per litter, and average fetal weight did not differ
    significantly from those of controls (Pepelko & Peirano, 1983).

         Reproductive and developmental toxicity are considered unlikely
    to be critical end-points for diesel exhaust.

    B7.6  Mutagenicity and related end-points

    B7.6.1  In vitro

         The genetic effects of particles and particle extracts have been
    described (Lewtas & Williams, 1986; Morimoto et al., 1986; Henschler,
    1987). Diesel exhaust extracts rather than particles were used in most
    of these studies. Point mutations were observed  in vitro in bacteria
    and mammalian cells. Extracts caused chromosomal aberrations, DNA
    damage, sister chromatid exchange, and cell transformation (Table 38).
    The mutagenic potency of diesel engine exhaust depended in several
    studies on the characteristics of the diesel fuel and the type, age,
    and operating conditions of the engine (Henschler, 1987). As a rule,
    organic extracts of particles were mutagenic in the absence of an
    exogenous metabolic activating system (S9) (Ball et al., 1990);
    addition of S9 decreased (Lewtas, 1983) or suppressed (Morimoto et
    al., 1986) the mutagenic activity, perhaps by increasing protein
    binding of mutagenic material or by metabolic detoxification of
    directly acting mutagens, such as the nitroarenes, in the Ames test.
    It can therefore be concluded that PAHs and thioarenes, which must be
    metabolically activated, do not account for the mutagenic potency of
    diesel particles. Salmeen et al. (1985) and Beland et al. (1985)
    showed that nitroarenes are the main genotoxic agents in  Salmonella
     typhimurium T98 in the absence of S9. A number of other
    investigators have reported the presence of nitrated PAHs in extracts
    of diesel engine exhaust particles (Handa et al., 1983), including
    many three-, four-, and five-ring structures (Schuetzle et al., 1982).
    The concentration of 1-nitropyrene, one of the prevalent species, has
    been reported to be 15-25 mg/kg particulate matter, whereas
    benzo[ a]pyrene has been found at concentrations up to 50 mg/kg. The
    nitrated PAHs are important in the health effects of diesel exhaust
    since they are effective mutagens in microbial and human cell systems
    (Patton et al., 1986). Some nitrated PAHs are also carcinogenic in
    animals (Imaida et al., 1991). Ball & Young (1992) found that strain
    T102, which does not respond to the mutagenic action of nitroarenes,
    responds to a class of oxidizing compounds that can interact with DNA;
    these compounds are not nitroarenes or typical PAHs.

         The effects of diesel soot particles with and without adsorbed
    organic substances, diesel soot extract without particles, an isolated
    fraction of PAHs, and titanium dioxide and carbon black (Printex 90)
    as reference particles were investigated  in vitro on hamster lung
    epithelial cells (Mohr & Riebe-Imre, 1992). The substances were added
    to the cells at concentrations of 100-300 mg/litre. Diesel soot
    extract, but not the PAH moiety isolated from the extract, stimulated
    mixed-function oxygenases. The diesel soot extract was more cytotoxic
    than the PAH fraction for the epithelium of the hamster respiratory
    tract, but both mixtures induced the development of micronuclei. Under
    the study conditions, only the PAH fraction led to transformation of
    cells. The respiratory epithelium of hamsters was more sensitive than

    human lung epithelial cells investigated in parallel, perhaps because
    hamster cells are generally more easily transformed than human cells.
    Diesel soot particles, titanium dioxide, and Printex 90 had hardly any
    cytotoxic effect but triggered transformation. The authors concluded
    that a combination of effects of particles and PAHs are responsible
    for the degeneration of cells.

         Schiffmann & Henschler (1992) studied the effects of diesel soot
    extract and isolated fractions of PAHs, oxy-PAHs, and nitro-PAHs on
    hamster embryo fibroblasts  in vitro. The end-points investigated
    were induction of micronuclei, unscheduled DNA synthesis, and cell
    transformation. Diesel soot extract induced micronuclei at
    concentrations of 80-400 mg/litre, whereas the fractions of various
    PAHs were strongly cytotoxic. Diesel soot extract and the PAH fraction
    transformed the fibroblasts; the nitro- and oxy-PAH fractions had a
    weaker transforming effect, but their cytotoxic potency was clearly
    stronger. Kinetochore analysis showed a high percentage (37%) of
    chromosomes in micronuclei. No unscheduled DNA synthesis was induced.
    Micronuclei were also induced in human embryonic lung fibroblasts
    treated with diesel soot extract under comparable experimental
    conditions.

    B7.6.2  In vivo

         The genotoxicity of total diesel exhaust, diesel particles, and
    diesel soot extracts  in vivo has been investigated in somatic cells
    (Table 38). All of the substances induced genotoxic responses. Total
    exhaust induced only sister chromatid exchange in Syrian golden
    hamsters. The frequency of micronuclei was not increased significantly
    in mice, even at high doses (up to 640 mg/kg body weight). Organic
    extracts had clear genotoxic effects in various assays (Henschler,
    1987).

         No mutagenic activity was found in urine taken from rats exposed
    to diesel exhaust emission by inhalation at 2 mg/m3 for 7 h/day on
    five days per week for up to two years. A slight but nonsignificant
    increase in micronuclei was observed in the bone marrow of mice that
    had been exposed for six months; no increase in micronuclei was
    detected in rats over a 24-month period (Ong et al., 1985).

         In studies of heritable mutations in T stock mice, males were
    examined for point mutations, induction of dominant lethal mutations,
    translocations, and spermatogonial survival; females were examined for
    oocyte death and dominant lethal mutations. No significant effect was
    seen on germ cells. A reduced number of corpora lutea (reproductive
    function) was the only significant result (Pepelko & Peirano, 1983).

    B7.6.3  DNA adduct formation

         Extensive research has been done to determine whether the DNA
    adducts induced in lungs by diesel exhaust are related to later
    tumorigenesis.

         In rats that inhaled benzo[ a]pyrene absorbed to carbon black
    (20 mg/g), an adduct of the diol epoxide with deoxyguanosine was
    identified (Wolff et al., 1989), demonstrating that epoxides of PAHs
    are produced by oxidative metabolism. Haemoglobin and albumin adducts
    were also investigated.

         Methods have been developed to detect DNA adducts of
    1-nitropyrene or 1,6-dinitropyrene, which are markers of diesel engine
    exhaust. The adducts are analysed in rat tissue or peripheral blood
    lymphocytes by the 32P-postlabelling method and may be useful for
    the dosimetry of 1-nitropyrene or diesel exhaust particulates in
    occupational settings (El-Bayoumy et al., 1994). Studies on adduct
    formation in lung DNA induced by inhaled diesel exhaust have been
    conducted by Wong et al. (1986), Wolff et al. (1990), Bond et al.
    (1988, 1989, 1990a,b), and Gallagher et al. (1993, 1994), all
    involving the 32P-postlabelling technique. In general, these studies
    suggest that DNA adducts are good measures of the 'effective dose' of
    carcinogenic compounds.

         Wong et al. (1986) found increased DNA adduct formation in the
    lungs of Fischer 344 rats exposed to diesel exhaust particles at
    7.1 mg/m3 for 31 months. After Fischer 344 rats and Syrian golden
    hamsters were exposed to dilutions of diesel engine exhaust for six
    months to two years, the level of haemoglobin adducts (2-hydroxy-
    ethylvaline and 2-hydroxypropylvaline) increased dose-dependently,
    corresponding to metabolic conversion of 5-10% of inhaled ethylene and
    propylene to their oxides (Törnqvist et al., 1988).

         Bond et al. (1988) designed experiments to determine the location
    of DNA adducts in the respiratory tract of rats exposed to exhaust.
    Fischer 344 rats were exposed for 12 weeks to diesel exhaust soot at
    10 mg/m3; they were then sacrificed, and various regions of the
    respiratory tract were removed and analysed for DNA adducts. Adducts
    were found only in peripheral lung tissue and nasal tissue; the total
    levels were highest in the peripheral tissue (about 18 adducts per 109
    bases). Thus, the levels of total DNA adducts and exhaust-induced
    adducts are highest in the region of the rat respiratory tract where
    tumours are formed after exposure to a carcinogenic concentration of
    diesel exhaust (Mauderly et al., 1987).

         Bond et al. (1990c) exposed groups of Fischer 344 rats to soot at
    0.35, 3.5, 7.0, or 10 mg/m3 for 12 weeks and found that the levels
    of DNA adducts were similar, at about 14 adducts per 109 bases. This
    level is nearly twice as high as that found in sham-exposed rats.

    Thus, DNA adduct formation in lungs is independent of the
    concentration of exhaust at the levels tested. One explanation for
    these results (Bond et al., 1990a) is that the lung enzymes
    responsible for the metabolism of soot-associated chemicals to
    metabolites that bind to DNA were saturated at the concentrations
    tested. These data also suggest that other factors are important in
    the carcinogenicity of diesel exhaust, since the number of adducts was
    elevated at an exposure level (0.35 mg/m3) that did not increase
    lung tumour incidence. The induction of DNA adducts at low
    concentrations is likely to result in tumour initiation by the organic
    compounds present; at higher concentrations, the particles will induce
    cell death, and subsequent proliferation may well act as a promotional
    event. The combination may then result in detectable increases in
    tumour incidence.

         Bond et al. (1990a) also investigated the time course for DNA
    adduct formation and persistence. Fischer 344 rats were exposed to
    soot at 7 mg/m3 for up to 12 weeks and were sacrificed at 2, 4, 8,
    12, 14, and 16 weeks after the start of exposure. DNA adducts
    accumulated slowly in the lung, and the number was highest at the end
    of exposure, representing about 160% of the level seen in controls
    exposed to air only. The levels declined rapidly after termination of
    exposure and were not significantly different from those of controls
    four weeks later. Thus, steady-state levels of DNA adducts would be
    reached during long-term exposure to diesel exhaust.

         Bond et al. (1989, 1990c) and Wolff et al. (1990) investigated
    whether exposure to carbon particles without associated mutagenic
    organic chemicals also increases DNA adduct levels. Rats were exposed
    either to diesel exhaust or carbon black at 0, 3.5, or 10 mg/m3 for
    12 weeks. Solvent-extractable organic compounds made up about 30% of
    the diesel soot but only about 0.04% of the carbon black particles.
    Exposure to the highest levels of both carbon black and diesel exhaust
    increased the DNA adduct levels in lungs, although the levels in rats
    exposed to diesel exhaust were about 30% higher than those in rats
    exposed to carbon black. Heavy exposure to carbon particles can
    therefore increase DNA adduct formation, although the concentrations
    are lower than those of exhaust-related adducts.

         Bond et al. (1990b) investigated whether DNA adducts could be
    formed in specific cells of the rat lung, i.e. alveolar type II cells.
    After exposure of rats to 6.2 mg/m3 diesel exhaust or carbon black
    particles for 12 weeks (16 h/day on five days per week), type II cells
    were isolated and their DNA analysed for adducts. Both diesel exhaust
    and carbon black resulted in about a fourfold increase in total DNA
    adducts in type II cells. DNA adduct levels in peripheral lung tissue
    were increased by 60-80% in male and female Fischer 344 rats and
    female cynomolgus monkeys but not in female B6C3F1 mice or female
    Syrian hamsters after inhalation of diesel soot at 8.1 mg/m3 for 12
    weeks (Bond et al., 1989, 1990c).

         Gallagher et al. (1993, 1994) exposed female Wistar rats to
    diesel soot at 7.5 mg/m3 and to carbon black at 11.3 mg/m3 for 24
    months and measured DNA adducts in lungs. The mean adduct levels were
    similar in the two groups but were not significantly greater than
    those in controls. Time-course studies indicated that the levels in
    the lungs of rats exposed to diesel exhaust were lower at 24 months
    than after two or six months of exposure, presumably as a result of
    dilution due to increased cell proliferation. The level of a single
    DNA adduct, thought to be derived from a nitro-PAH in diesel exhaust
    and not observed in rats exposed to carbon black or titanium oxide,
    was elevated over that in controls after two, six, and 24 months. The
    DNA adduct levels in control rats increased over the duration of the
    24-month study as an effect of age.

         Gallagher et al. (1993) compared DNA adduct formation after
    exposure to diesel emissions  in vitro and  in vivo. After exposure
    of human lymphocytes to diesel extract, five major DNA adducts were
    detected, one of which was characterized as an adduct with
    benzo[a]pyrene. One was also detectable in reaction products of calf
    thymus DNA and diesel particle extract  in vitro and in skin and lung
    DNA from mice treated dermally with 50 mg diesel extract  in vivo. The
    differences in DNA patterns  in vitro and  in vivo (skin and lung)
    may reflect differences in metabolic pathways.

         Taken together, the studies of DNA adducts suggest that some
    organic chemicals in diesel exhaust can form DNA adducts in lung
    tissue and may play a role in the carcinogenic effects. As pointed out
    by Bond et al. (1989), however, DNA adducts alone cannot explain the
    carcinogenicity of diesel exhaust, and other factors, such as chronic
    inflammation and cell proliferation, are also important.

    B7.7  Special studies

    B7.7.1  Immunotoxicity

         During the clearance process, one possible pathway is
    translocation of diesel particulates into the lymphatic channels.
    After male guinea-pigs were exposed to diesel exhaust at a particle
    concentration of 1.5mg/m3 for four or eight weeks, the B- and T-cell
    counts in lymph nodes were not altered, and there were no significant
    changes in blood or spleen (Dziedzic, 1981).

         Fischer 344 rats were exposed to 2 mg/m3 diesel exhaust for
    7h/day on five days per week for 12 or 24 months and the immunological
    function of splenic B and T cells was measured by enumerating
    antibody-producing cells in the spleen or monitoring the proliferative
    response. No changes were observed (Mentnech et al., 1984).


        Table 38.  Genotoxicity of diesel exhaust, particles, and extracts
                                                                                                                                                

    Test organism             End-point                   Exhaust         Particles     Extract      Reference
                                                                                                                                                

    Mutagenicity
    Bacteria
    S. typhimurium            Point mutation (his)                        +             +            Huisingh et al. (1978)
    S. typhimurium            Point mutation (his)                        +             +            Loprieno et al. (1980)
    S. typhimurium            Point mutation (his)                                      +            Clark & Vigil (1980)
    S. typhimurium            Point mutation (his)                        +             +            Li & Royer (1982)
    S. typhimurium            Point mutation (his)                                      +            Salmeen et al. (1984)
    S. typhimurium            Point mutation (his)                                      +            Ong et al. (1985)
    S. typhimurium            Point mutation (his)                                      +            Bechtold et al. (1986)
    S. typhimurium            Point mutation (his;                                      +            Whong et al (1986)
                              SOS umu test)
    S. typhimurium            Point mutation (his)                                      +            Wallace et al. (1987)
    S. typhimurium            Point mutation (his)                        +a                         Wallace et al. (1990)
    S. typhimurium            Point mutation (his)                                      +            Lewis et al. (1989)
    S. typhimurium            Point mutation (his)                                      +            Rasmussen (1990)
    S. typhimurium            Point mutation (his)                        +             +            Keane et al. (1991)
    S. typhimurium, E. coli   Point mutation (trp)                                      +            Crebelli et al. (1991)
    E. coli                   Point mutation (trp)                                      +            Lewtas (1983)

    Mammalian cells
    L5178Y mouse              Point mutation (tk)                                       +            Lewtas (1983)
                              lymphoma cells
    L5178Y mouse              Point mutation (tk)                                       +            Mitchell et al. (1981)
                              lymphoma cells
                                                                                                                                                

    Table 38 (contd)
                                                                                                                                                

    Test organism              End-point                      Exhaust      Particles     Extract     Reference
                                                                                                                                                

    Mutagenicity (contd)
    Chinese hamster            Point mutation (hprt)                       +                         Chescheir et al. (1981)
    ovary cells
    Chinese hamster            Point mutation (hprt)                                     (+)         Chescheir et al. (1981)
    ovary cells
    Chinese hamster            Point mutation (hprt)                                     (+)         Casto et al. (1981)
    ovary cells
    Chinese hamster            Point mutation (hprt,                                     +           Morimoto et al. (1986)
    lung V79 cells             ATPase)
    Human xeroderma            Point mutation (hprt)                       +             +           McCormick et al. (1980)
    pigmentosum fibroblasts
    Human lymphoblast          Point mutation (tk)                                       +           Barfknecht et al. (1981)
    TK6 cells
    Balb/c3T3 mouse            Point mutation (ATPase)                     +             (+)         Curren et al. (1981)
    fibroblasts

    DNA damage
    Syrian hamster             DNA chain breaks                                          -           Casto et al. (1981)
    embryo cells               (alkaline elution)                                        
    Human xeroderma            DNA damage                                  +             +           McCormick et al. (1980)
    pigmentosum fibroblasts
    Rat primary                Unscheduled DNA                                           +           Lewtas (1983)
    hepatocytes                synthesis
    Chinese hamster lung       Unscheduled DNA                             +a            +           Gu et al. (1994)
    V79 cells                  synthesis
                                                                                                                                                

    Table 38 (contd)
                                                                                                                                                

    Test organism              End-point                      Exhaust      Particles     Extract     Reference
                                                                                                                                                

    Chromosomal effects
    Chinese hamster            Sister chromatid exchange                                 +           Lewtas (1983)
    ovary cells
    Chinese hamster            Chromosomal aberrations                                   +           Lewtas (1983)
    ovary cells
    Chinese hamster            Chromosomal aberrations                     +b                        Hasegawa et al. (1988)
    lung V79 cells                                                         -c
    Chinese hamster            Sister chromatid exchange                   +                         Hasegawa et al. (1988)
    lung V79 cells
    Chinese hamster            Sister chromatid exchange                   +             +           Keane et al. (1991)
    lung V79 cells
    Human lymphocytes          Chromosomal aberrations                                   +           Lewtas (1983)
    Chinese hamster            Sister chromatid exchange                                 +           Morimoto et al. (1986)
    lung V79 cells
    Human lymphocytes          Sister chromatid exchange                   (+)                       Tucker et al. (1986)
    Human lymhoblastoid        Sister chromatid exchange                                 +           Morimoto et al. (1986)
    cells
    Hamster embryo             Micronuclei                                               +           Schiffmann & Henschler
    fibroblasts                                                                                      (1992)
    Chinese hamster lung       Micronuclei                                 +a            +           Gu et al. (1992)
    V79 and ovary cells
                                                                                                                                                

    Table 38 (contd)
                                                                                                                                                

    Test organism                   End-point                       Exhaust     Particles    Extract    Reference
                                                                                                                                                

    Cell transformation
    Balb/c3T3 mouse fibroblasts     Cell transformation                         (+)                     Curren et al. (1981)
    Balb/c3T3 mouse fibroblasts     Cell transformation                         +b                      Hasegawa et al. (1988)
                                                                                             (+)c
    Hamster lung epithelial cells   Cell transformation                         +            +          Mohr & Riebre-Imre (1992)
    Hamster embryo fibroblasts      Cell transformation                                      +          Schiffmann & Henschler (1992)

    In vivoa
    Mouse                           Micronuclei, bone marrow        -           -            +          Pereira (1982)
    Mouse                           Micronuclei, bone marrow        -                                   Lewis et al. (1989)
    Mouse                           Micronuclei, bone marrow        -                                   Morimoto et al. (1986)
    Mouse                           Micronuclei, bone marrow        -           -            -          Pepelko & Peraino (1983)
    Mouse                           Micronuclei, bone marrow        (+)                                 Ong et al. (1985)
    Rat                             Micronuclei, bone marrow        -                                   Ishinishi et al. (1988)
    Rat                             Micronuclei, bone marrow        -                                   Ong et al. (1985)
    Chinese hamster                 Micronuclei, bone marrow        (+)         -            -          Pepelko & Peraino (1983)
    Mouse                           Sister chromatid exchange,      -           +            +          Pereira (1982)
                                    bone marrow
    Rat                             Sister chromatid exchange,      -                                   Ishinishi et al. (1988)
                                    bone marrow
    Rat                             Sister chromatid exchange,      -                                   Morimoto et al. (1986)
                                    bone marrow
    Rat                             Sister chromatid exchange,      -                                   Ong et al. (1985)
                                    peripheral blood leukocytes
                                                                                                                                                

    Table 38 (contd)
                                                                                                                                                

    Test organism         End-point                            Exhaust       Particles     Extract     Reference
                                                                                                                                                

    Rat                   Sister chromatid exchange,           -                                       Lewis et al. (1989)
                          lymphocytes
    Syrian golden         Sister chromatid exchange,           +                                       Pereira (1982)
    hamster               lung cells (inhalation)
    Syrian golden         Sister chromatid exchange,                         +                         Guerrero et al. (1981)
    hamster               lung cells (instillation)
    Syrian golden         Sister cromatid exchange,                          +             +           Pereira (1982)
    hamster               lung cells (instillation)
    Syrian golden         Sister chromatid exchange                                        +           Morimoto et al. (1986)
    hamster               liver cells (transplacental)
    Syrian golden         hprt mutation                                                    +           Morimoto et al. (1986)
    hamster               (transplacental)
    Mouse                 Mutation, S. typhimurium             -                                       Morimoto et al. (1986)
                          (host-mediated assay)
    Mouse                 Heritable effects on germ            -                                       Pepelko & Peirano (1983)
                          cells (various assays; m/f)
    Drosophila            Heritable effects (sex-linked                                                Schuler & Niemeier (1981)
    melanogaster          recessive lethal mutation)           -                                       
                                                                                                                                                

     his, histidine independence;  trp, tryptophane independence;  hprt, hypoxanthine-guanine-phosphoryltransferase (8-aza- or
    6-thioguanine resistance);  tk, thymidine kinase (bromodeoxyuridine or trifluorothymidine resistance); ATPase, Na+/K+-ATPase
    (ouabain resistance); +, positive; (+), weakly positive
    a  Dispersed in artificial surfactant
    b  Light-duty engine
    c  Heavy-duty engine
    d  Modified from Henschler (1987) and supplemented
             Fischer 344 rats and CD-1 mice were exposed to diesel exhaust
    particles at 0.35, 3.5, or 7 mg/m3 for 6, 12, 18, or 24 months to
    investigate whether the accumulation of diesel particulates in lymph
    nodes (indicated by black discolouration) influences the subsequent
    response to immunization by sheep red blood cells, evaluated by the
    presence of immunoglobulin (Ig) M antibody-forming cells. In rats, the
    total number of lymphoid and antibody-forming cells was significantly
    increased at the medium and high aerosol concentrations. In mice, the
    number of lymphoid cells was increased only at the high concentration.
    The authors concluded that diesel exhaust particles have only a
    minimal effect on the immune and antigen filtration functions in
    lung-associated lymph nodes because the relative numbers of
    antibody-forming cells and specific IgM, IgG, and IgA antibodies in
    rat sera were not significantly changed (Bice et al., 1985).

         Five intranasal inoculations of various doses of a suspension of
    diesel engine exhaust particles in ovalbumin were administered to BDF1
    mice at intervals of three weeks. Ovalbumin IgE antibody titres,
    assayed by passive cutaneous anaphylaxis, were enhanced by doses as
    low as 1 µg (Takafuji et al., 1987, 1989). Similarly, primary IgE
    responses were increased after intraperitoneal administration to mice
    of ovalbumin or cedar pollen allergen mixed with diesel exhaust
    particulates (Muranaka et al., 1986).

         Female CR/CD-1 mice were exposed by inhalation to 6-7 mg/m3
    diesel exhaust for 2-6 h (acute), 8 h/day for 2-16 days (subacute),
    or 8h/day for about 300 days (chronic). Immediately after the end 
    of exposure, animals were exposed to the infectious aerosols
     S. typhimurium, Streptococcus pyogenes, or A/PR8-34 influenza virus.
    Increased susceptibility to post-infection mortality was seen with the
    bacterial but not the viral pathogens. Nitrogen dioxide and acrolein
    vapour had specific effects (Campbell et al., 1981b).

         CD-1 mice were exposed to 2 mg/m3 diesel exhaust for one to six
    months. After three months of exposure and subsequent infection with
    influenza virus, an enhanced severity of response was seen, involving
    significant lung consolidation and focal macular collections of
    particle-laden macrophages (Hahon et al., 1985).

    B7.7.2  Behavioural effects

         Sprague-Dawley rats were exposed for 8 h/day on seven days per
    week for 16 weeks to diesel exhaust diluted to a particulate matter
    concentration of 6 mg/m3. Spontaneous locomotor activity, measured
    weekly on Wahman LC-34 running wheels, was significantly decreased
    during 8, 9, 11, and 12 weeks of exposure. In rats exposed for
    20 h/day on seven days per week for six weeks, forced activity on a
    motorized treadmill was measured during the last week. Exhaustion
    occurred in less than half the time that it occurred in control
    animals. Groups of 10 rats were exposed to diesel exhaust at 6 mg/m3

    for 20 h/day on days 1-7 post partum and were then held in clean air
    until 15 months of age. In bar pressing acquisition training carried
    out at five-day intervals for the next 42 days, the rate of learning
    was much slower than in control rats. The difference was highly
    statistically significant (Pepelko & Peirano, 1983).

         As these three studies were conducted at a high concentration,
    6 mg/m3, the practical consequences of the effects observed are not
    known. The results, while limited in scope, indicate, however, that
    behavioural and neurophysiological effects may be important toxic
    end-points which should be investigated further.

    B7.8  Factors that modify toxicity; toxicity of metabolites

         Little information is available, but chemicals and other factors
    may modify the toxicity of diesel emissions. For instance,
    chlorophyllin, a derivative of the green pigment chlorophyll, has been
    reported to inhibit or reduce the mutagenic activity of diesel
    particulate extracts in the Ames test (Ong et al., 1986).

    B7.9  Mechanisms of toxicity; mode of action

         It is not clear whether DNA-reactive or non-DNA-reactive
    mechanisms, or a combination of the two, are responsible for the
    carcinogenic action of inhaled diesel exhaust in laboratory animals.
    The results of several studies indicated that many PAHs (e.g.
    benzo[ a]pyrene) cause a carcinogenic response in rodents, but it is
    not known whether the same PAHs when associated with diesel exhaust
    also induce a carcinogenic response. One hypothesis for the mechanism
    of the tumorigenic response is that organic chemicals (e.g. PAHs,
    nitro-PAHs) desorb from soot particles, are metabolized to reactive
    metabolites, and interact with lung DNA to initiate carcinogenesis.
    This would be a predominantly DNA-reactive mechanism. The observation
    that organic chemicals associated with diesel soot can be metabolized
    by lung cells to metabolites that can form DNA adducts after long-term
    exposure to diesel exhaust supports this hypothesis. Studies with
    carbon black essentially devoid of PAHs but which also form DNA
    adducts do not, however, corroborate the PAH-DNA-reactive concept.

         Another hypothesis is that the lung tumours that arise in rats
    exposed to high levels of diesel exhaust are due to overloading of the
    normal lung particle clearance mechanisms, accumulation of soot
    particles, and cell damage followed by regenerative cell
    proliferation. Enhanced cell proliferation may increase the mutation
    frequency of key target genes. In this hypothesis, genotoxic chemicals
    may not be causative factors in tumorigenicity. This view is supported
    by the observations that lung cancer can be induced in rodents by
    inhalation of highly insoluble particles of low toxicity that are
    virtually devoid of organic chemicals (e.g. talc, carbon black, coal
    dust, titanium dioxide), although high concentrations of these
    particles (overload) are typically necessary.

         A third hypothesis, which draws upon the above two mechanisms, is
    that carcinogenesis is initiated by exposure to organic compounds
    associated with diesel soot and promoted by the chronic inflammation,
    cytotoxicity, and cell proliferation arising from the high
    concentrations of particles deposited and retained in the lungs.
    Again, however, the tumorigenic response observed with particles alone
    (carbon black, titanium dioxide) does not support this hypothesis.

         A key consideration is the relative contribution of different
    mechanisms at different levels of exposure. As discussed later in this
    section, high concentrations of particles, including diesel exhaust,
    leading to high lung burdens, compromise normal clearance mechanisms.
    Therefore, certain mechanisms may be invoked at high lung burdens that
    may not occur at lower lung burdens. Genotoxic aromatic compounds may
    take on increasing importance at lower concentrations.

         Figure 3 is a diagram outlining the possible mechanism of action
    of diesel exhaust, including the effects of overload conditions.

    B7.9.1  Carcinogenic effects

    B7.9.1.1  DNA-reactive mechanisms

         Diesel exhaust contains hundreds of chemicals, including PAHs,
    nitro-PAHs, alkyl-PAHs, oxy-PAHs, oxy-nitro-PAHs, and aldehydes
    (Scheepers & Bos, 1992b), many of which are known mutagens and
    carcinogens in laboratory animals. For example, inhaled
    benzo[ a]pyrene (in tar and pitch) is carcinogenic in rat lung
    (Heinrich et al., 1994). Nitro-PAHs are potent mutagens and, in some
    instances, carcinogenic. In pharmacokinetic studies cited in this
    monograph, clearance of PAHs and nitro-PAHs associated with the diesel
    particulate fraction is retarded or delayed. Other studies have shown
    clearly that the organic compounds associated with diesel soot are
    bioavailable to lung cells. Mechanisms for the delayed clearance of
    the PAHs have been discussed (Bond et al., 1986b; Gerde et al.,
    1991b). PAHs require metabolic activation to metabolites (e.g.
    epoxides) that can potentially bind to DNA. In the case of the PAHs
    associated with diesel soot, it is likely that alveolar lung cells
    (e.g. type II cells) are responsible for their metabolic activation,
    so that they can bind to lung cell DNA (Bond et al., 1983).

         Additional evidence for a DNA-reactive mechanism in the
    carcinogenicity of diesel exhaust is that exposure of laboratory
    animals, including rats and monkeys, to diesel exhaust results in the
    formation of DNA adducts in lung cells (see section B7.6). DNA adducts
    and subsequent DNA replication can result in mutations that play a key

    FIGURE 3

    role in initiating the carcinogenic response. These mutations could
    involve activation of oncogenes or inactivation of tumour suppressor
    genes. There may be some as yet unidentified chemicals in diesel
    exhaust with high mutagenic potential; furthermore, chemical
    interactions (additive or inhibitory) may occur among the components
    of diesel exhaust (Ball & Young, 1992).

    B7.9.1.2  Cytotoxicity with regenerative cell proliferation

         Another mechanism by which diesel exhaust may induce
    carcinogenesis involves cell killing, or cytotoxicity, by the
    particles or the reactive gases in exhaust, followed by regenerative
    proliferation of lung cells. Enhanced cell replication may result in
    an increased frequency of spontaneous mutations. Mutations could also
    arise from PAH-DNA adducts or from oxidative DNA damage caused by the
    reactive oxygen and nitrogen species in diesel soot, resulting in the
    inflammatory response. Driscoll et al. (1994) showed that the hprt
    gene mutation frequency in rat lung epithelial cells was significantly
    increased when the cells were incubated  in vitro with inflammatory
    cells obtained by bronchoalveolar lavage from rats exposed to quartz
    particles. This appears to be an important mechanism in lung tumour
    induction by particles that elicit a chronic inflammatory response in
    the lung. Sagai et al. (1993) showed that diesel exhaust particles
    could produce superoxide anions and hydroxyl radicals in the absence
    of biological activation (detection by cytochrome c and electron spin
    resonance). Reactive oxygen species may induce specific DNA adducts,
    such as 8-hydroxydeoxyguanosine. As mentioned above, mutations could
    involve activation of oncogenes or inactivation of tumour suppressor
    genes.

         Several data sets cited in this monograph support this
    hypothesis. For example, carbon black particles, which are virtually
    devoid of PAHs but are morphologically similar to diesel soot
    particles, induce lung cancer in rats. High particle burdens in rodent
    lungs result in an inflammatory response (Henderson et al., 1988).
    Release of cytokines and reactive oxygen and nitrogen species during
    the inflammatory response to particles may result in cell death. Some
    evidence exists (Wright, 1986) that diesel exhaust enhances lung cell
    replication.

    B7.9.1.3  Effects of particles

         The importance of pulmonary particle burden on lung tumour
    induction has been demonstrated clearly in long-term studies by
    inhalation in rats. As reviewed above, rats have similarly increased
    lung tumour incidences when exposed to diesel exhaust or carbon black
    particles by inhalation for 24 months. Similarly, carbon black
    particles practically devoid of PAHs induce pulmonary tumours after

    intratracheal instillation. The very large surface area of carbon
    black and of diesel exhaust particles after desorption of adsorbed
    organic compounds  in vivo may be involved mechanistically in a
    tumorigenic effect. Heinrich (1994) showed that the tumour response to
    different types of carbon black particles instilled intratracheally
    correlated well with their respective surface areas. Pott (1991) and
    Pott et al. (1993) suggested that particles with a large surface area
    have greater effects in humans than in rats because insoluble
    particles with no specific toxicity remain significantly longer in the
    terminal airways of humans than of rats (retention half-time, about 70
    days in rats and about 500 days in humans).

         The correlation between particle surface area and lung tumour
    incidence was examined (Oberdörster & Yu, 1990) by evaluating
    published studies of inhalation of diesel and other particles. Tumour
    induction in rats was best correlated with the surface area of the
    particles retained in the lung rather than with the particle mass,
    particle volume, or number of particles, regardless of the PAH
    content. It was suggested that particle surface area and surface
    properties play a decisive role ('critical surface area') and that
    absorbed PAHs are not responsible for the tumour response in rats
    exposed to diesel exhaust. In the human situation, however, it could
    not be excluded that organic compounds and gas-phase components are
    also involved, since the human particulate lung burden is much lower
    than those achieved in rats after long-term inhalation.

         Inhalation of diesel engine exhaust can result not only in
    pulmonary tumours but also in inflammation and fibrosis and in a delay
    in alveolar (not bronchial) pulmonary clearance. The mechanism by
    which tumours develop due to particle overload and its associated
    pathological and anatomical changes may be restricted to rats and may
    not occur under environmental conditions in humans, since the lung
    burdens of humans do not reach the levels that induce lung tumours in
    rats. This is of importance for quantitative risk assessment. Only
    occupational exposure to diesel exhaust may result in lung burdens
    near or at overload conditions, particularly if the lung is already
    compromised by exposure to other dusts. Bohning et al. (1982) reported
    retarded particle clearance in smokers; in these people, additional
    exposure to diesel exhaust may induce overload and associated toxic
    effects. It is not known, however, whether the mechanisms of
    particle-induced lung tumours are the same in rats and humans, and
    this information is necessary for extrapolating data from rats to
    humans.

         Not only diesel soot (0.8, 2.5, or 7.0 mg/m3) but also carbon
    black nearly completely devoid of organic compounds (Printex 90,
    7.5-12 mg/m3; particle size, 10 nm) and ultrafine titanium dioxide
    particles (7.5-15 mg/m3; particle size, 20 nm) caused lung tumours
    in female Wistar rats exposed by inhalation for 18 h/day on five days
    per week for 24 months. The tumour rate increased with increasing

    particle concentrations, independently of the type of particles
    inhaled. No lung tumours were observed in the rats exposed to the
    lowest concentration of diesel particles. The authors concluded that
    the carcinogenic component of diesel exhaust is in the inner part of
    the diesel soot particle, the carbon core, and is not the relatively
    small amount of carcinogenic PAHs (3.9 g benzo[ a]pyrene per gram of
    diesel soot) (Heinrich et al., 1992; Heinrich, 1994; Heinrich et al.,
    1995). These results were confirmed in another two-year study, in
    which male and female rats were exposed by inhalation to various
    concentrations of carbon black and diesel exhaust. Lung tumours were
    observed with both particle types (Nikula et al., 1994).

         Diesel soot does not appear to have a specific carcinogenic
    effect in rats; rather, there is a nonspecific effect of particles.

    B7.9.1.4  Effects of polycyclic aromatic hydrocarbons

         Exposure of rats by inhalation to 2.6 mg/m3 of an aerosol of
    tar-pitch condensate with no carbon core but containing 50 µg/m3
    benzo[ a]pyrene and other PAHs for 10 months caused lung tumours at a
    rate of 39%. The same amount of tar-pitch vapour condensed onto the
    surface of carbon black particles at 2 and 6 mg/m3 resulted in
    tumour rates that were roughly two times higher (89 and 72%). Since
    exposure to 6 mg/m3 carbon black almost devoid of extractable
    organic material caused a lung tumour rate of 18%, the tumour rate of
    72% seen after combined exposure to tar-pitch vapour and carbon black
    particles indicates a syncarcinogenic effect of PAHs and carbon black.
    A possible mechanism is an effect of deposition of PAHs (Heinrich et
    al., 1994). As the level of benzo[ a]pyrene in the coal-tar pitch was
    about three orders of magnitude greater than those in diesel soot,
    PAHs may play a negligible role in the tumorigenicity of diesel soot
    in rats. The PAH profile in diesel soot is, however, quite different
    from that in coal-tar pitch, as diesel soot contains highly mutagenic,
    carcinogenic nitro-PAHs and other poorly characterized mutagens that
    are not present in coal-tar pitch or on some of the carbon black
    particles used in experimental studies (Heinrich et al., 1994; Nikula
    et al., 1994).

         In a study of various extracts of diesel exhaust particles,
    30-40% of the total mutagenicity could be attributed to a group of six
    nitroarenes (Salmeen et al., 1984).

         A diesel exhaust particle extract was separated into a water- and
    a lipid-soluble fraction, and the latter was further separated into a
    PAH-free, a PAH-containing, and a polar fraction by column
    chromatography. These fractions were then tested in Osborne-Mendel
    rats by pulmonary implantation at doses corresponding to the
    composition of the original diesel exhaust. The water-soluble fraction
    did not induce tumours; the incidences induced by the lipid-soluble

    fractions were 0% with the PAH-free fraction, 25% with the
    PAH-containing fraction, and 0% with the polar fraction. The
    PAH-containing fraction, comprising only 1% by weight of the total
    extract, was shown to be responsible for the carcinogenic activity
    (Grimmer et al., 1991).

         Various dichloromethane extracts, each representing a complex
    mixture, were obtained from particulate emissions of four
    diesel-fuelled and one gasoline-fuelled automobiles (combustion), a
    coke oven battery (pyrolysis), and a roofing tar pot (evaporation),
    and their tumorigenic potency was compared in Sencar mice. It was
    concluded that the benzo[ a]pyrene content alone could not explain
    the tumorigenic activity of the mixtures (Nesnow et al., 1982a, 1983).

         The lung tumour rates in rats exposed to atmospheres containing
    PAHs depend not only on the PAH concentrations of the exhaust gas but
    also on parameters such as the composition of the carrier particle
    (mass ratio of carbon core to adsorbed layer of organics), the
    dissolution rate of particle-attached organic compounds, the retention
    half-time, and the cytotoxic effect of the carrier particle in the
    lung (Heinrich et al., 1991).

         Extracted diesel soot given intratracheally to rats induced lung
    tumours, but native, unchanged diesel soot resulted in higher tumour
    rates than extracted soot. Carbon black also caused tumours after
    intratracheal administration, and the rate increased with decreasing
    size of the primary particles (Heinrich, 1994).

         Carbon black particles almost completely devoid of organic
    compounds (< 0.046% extractable organic compounds and 0.6 ng/g
    benzo[ a]pyrene) caused tumours in the lungs of 17% of rats after
    exposure to a concentration of 6 mg/m3 for 18 h/day on five days per
    week for 10 months. No lung tumours occurred in controls exposed to
    clean air. Thus, the particle effect may be responsible for induction
    of lung tumours in rats exposed to diesel engine exhaust, and the
    effect may be related to the surface area of the carbon particle. A
    PAH depot effect could lead, however, to retarded dissolution of PAHs
    from the carrier, resulting in very efficient use of the extremely
    small amount of carcinogenic PAHs retained in the lung after exposure
    to diluted diesel engine exhaust, which contained benzo[ a]pyrene at
    about 10 ng/m3 (Heinrich et al., 1991). Since diesel particles
    induce DNA adducts at lower concentrations than carbon black (Bond et
    al., 1989, 1990c; Wolff et al., 1990), PAHs may have tumour initiating
    properties that are too weak to be detected at low doses; at higher
    doses, promotional effects of the particles, perhaps with additional
    initiation, may synergize to produce detectable carcinogenic effects.

    B7.9.2  Non-carcinogenic effects

         Diesel exhaust contains various respiratory irritants in the gas
    phase and in particulate matter. Both can induce inflammatory
    responses in the airways and alveolar regions of the lung. Airway
    inflammation involves damage to epithelial cells, including lipid
    peroxidation of cell membranes by oxidizing gaseous pollutants such as
    nitrogen dioxide. Indirect effects of particles, resulting from
    phagocytosis, can include the formation and release of various
    mediators, including oxidants, such as superoxide anions and hydroxyl
    radicals, and cytokines. These mediators may play a role in focal loss
    or shortening of cilia, type II cell hypertrophy, and hyperplasia. The
    latter changes can lead to the hyperplastic lesions seen in animals
    exposed to diesel exahaust (Ishinishi et al., 1986, 1988; Suzuki et
    al., 1990; Kato et al., 1992). Phagocytosis and subsequent clearance
    of particles by alveolar macrophages can be compromised by high
    particle burdens (Wolff et al., 1989; Creutzenberg et al., 1990),
    which may also increase the access of particles to the interstitium,
    leading to focal fibrosis (Henderson et al., 1988).

         Inflammation, altered lung clearance, and hyperplastic lesions
    can be considered early markers of exposure to diesel exhaust and are
    the basis of the non-neoplastic effects used to determine both the
    NOAEL and the 'benchmark concentration' described in section B10.1.3.

    B8.  EFFECTS ON HUMANS

    B8.1  General population

    B8.1.1  Acute exposure: olfactory, nasal, and ocular irritation

         Acute exposure to diesel exhaust has been associated with
    irritation of the eyes and nose. Signs and symptoms reported after
    acute exposure to diesel exhaust are described in section B8.2.1.

         The exhaust from e.g. poorly maintained engines or engines under
    load may be visible as a black, smoky cloud. WHO (1987) considered
    that sensory effects were parameters that could be used in setting
    occupational exposure limits. The characteristic odour of diesel
    exhaust provides a warning of its presence. At higher concentrations
    and under certain operating conditions, the odour can be offensive.
    The odour-causing agents are not known; however, compounds with a
    relative molecular mass > 80 are probably those mainly responsible,
    and aliphatic olefins (C1-C6, including acetylene), hydrocarbons
    with more than six carbons, and aliphatic aldehydes are probably not
    involved. Substances with low odour thresholds, such as nitrogen
    dioxide (0.3 ppm) and acrolein (0.5 ppm), appear to contribute to the
    odour of diesel exhaust to minor extents (5.3 and 0.25%, respectively)
    (Oelert & Florian, 1972).

         The ability to detect the odour and the irritating effects of
    diesel exhaust vary. For example, when six subjects smelled diesel
    exhaust diluted with air, the dilution factors needed to achieve the
    odour detection threshold varied from 140 to 475. The concentrations
    of different constituents of the exhaust at the odour threshold also
    varied: formaldehyde, 0.012-0.088 ppm; acrolein, 0.011-0.046 ppm; and
    nitrogen dioxide, 0.11-2.28 ppm (Linnell & Scott, 1962). When six
    subjects were exposed to three concentrations of diesel exhaust for
    10 min, the mean concentrations at the odour threshold were 1.3-4.2 ppm
    nitrogen dioxide, 0.2-1 ppm sulfur dioxide, < 0.1 ppm formaldehyde,
    and < 0.05 ppm acrolein. Three of the subjects exposed to the highest
    level discontinued exposure before 10 min, while none at the lowest
    level discontinued exposure although some experienced eye irritation
    (Battigelli, 1965).

         The health effects of inorganic gases present in diesel exhaust
    are not specific and are therefore not discussed in detail; however,
    it must be noted that these gases contribute to the environmental
    burden.

    B8.1.2  Air pollution

         The concentrations of some air pollutants, such as particulates
    and sulfur dioxide from the burning of coal for domestic heating and
    industrial purposes, are in many parts of the world much lower than
    they were several decades ago, and the associated adverse health

    effects have diminished accordingly. A number of epidemiological
    studies have demonstrated, however, that the relatively low remaining
    levels of air pollutants are associated with a range of health
    indices, including day-to-day changes in mortality (Schwartz, 1993),
    visits to hospital emergency departments (Schwartz et al., 1993), and
    changes in measures of lung function (Pope & Dockery, 1992). In most
    studies, the associations have been strongest with the fine
    particulate component of pollution, some of which may be derived from
    diesel exhausts.

         Vehicle traffic, both gasoline and diesel, also contributes to
    the nitrogen dioxide content of urban air, and there is evidence,
    primarily from experimental studies (Bauer et al., 1986; Koenig et
    al., 1988; Grant et al., 1993) that it can induce short-term
    decrements in lung function in asthmatic and non-asthmatic patients.
    Nitrogen dioxide and volatile hydrocarbons are involved in the
    formation of ozone and associated photochemical pollutants.

         Associations have also been demonstrated between long-term
    exposure to urban air pollutants and death rates from certain chronic
    conditions. Confounding factors often make interpretation difficult,
    but in an investigation conducted in the United States (Dockery et
    al., 1993), air pollution attributable to fine particles was
    positively associated with lung cancer and cardiopulmonary disease,
    after adjustment for smoking and other relevant risk factors, but not
    with other causes considered collectively.

         In neither short- nor long-term studies can the possible role of
    diesel particulates be specified, but diesel emissions contribute to
    urban particulates, especially in Europe and developing countries. The
    acidic and sulfate components of particulates, which are derived
    primarily from stationary fuel-burning sources, appeared to be
    implicated in a number of studies.

    B8.2  Occupational exposure

    B8.2.1  Effects on the respiratory system

    B8.2.1.1  Symptoms

         Most of the information about symptoms after acute exposure to
    diesel exhaust comes from anecdotal reports of occupationally exposed
    individuals. It was noted in a bus garage in London, United Kingdom,
    that the buses produce a 'lachrymatory mist' when they were started up
    from cold (Commins et al., 1956). In a review of 13 cases of acute
    overexposure to diesel exhaust in five underground coal mines in Utah
    and Colorado, United States, between 1974 and 1985, interviews in 1986
    with the miners revealed that 12 had experienced symptoms of mucous
    membrane irritation, headache, and light-headedness; eight had
    reported nausea, and four reported a sensation of unreality and
    'heartburn' (Kahn et al., 1988).

         In a study designed to investigate the acute effects of diesel
    exhaust on respiratory symptoms, a questionnaire was administered to
    232 male workers in four diesel bus garages, and personal samples of
    nitrogen dioxide and respirable particulate were obtained. After
    adjustment for age and smoking (current, ex-, never), workers with the
    highest exposure to respirable particulate (0.31 mg/m3) reported
    significantly more cough, itchy or burning eyes, headache, difficult
    or laboured breathing, a feeling of chest constriction, and wheeze
    than workers with lower exposures. Those exposed to nitrogen dioxide
    at >0.4 mg/m3 (> 0.3 ppm) also more frequently reported itchy or
    burning eyes, difficult or laboured breathing, a feeling of chest
    constriction, and wheezing. In comparison with a group of lead acid
    battery workers, the reporting of the following symptoms as
    'sometimes' or 'often' was significantly more prevalent among the
    workers in the diesel bus garage: eyes itch, burn, or water (49.5
    versus 23.5%), headache (24.2 versus 12.1%), difficult or laboured
    breathing (13.5 versus 3.6%), nausea (13.5 versus 4.5%), and wheeze
    (13.7 versus 4.5%) (Gamble et al., 1987a).

         In comparison with 11 office workers, 17 ferry stevedores had a
    greater prevalance of wheezing (24 versus 9%), chest tightness (29
    versus 9%), nasal complaints (47 versus 0%), chest pain (24 versus
    0%), and eye irritation (59 versus 27%); however, more of the
    stevedores smoked. The concentrations of nitrogen dioxide (0.3 mg/m3)
    and sulfur dioxide (0.7mg/m3) were both < 0.25 ppm; respirable
    particulate was not measured (Purdham et al., 1987). In a study in an
    iron ore mine where diesel engines were used underground, more workers
    who were smokers reported pressure over the chest or difficulty in
    getting air than the nonsmokers. No such difference was observed among
    smoking and nonsmoking surface workers. The underground workers who
    were smokers also had more episodes of a productive cough lasting
    three weeks during several winters (Jorgensen & Svensson, 1970).

    B8.2.1.2  Acute changes in pulmonary function

         The pulmonary function of 60 coal miners who worked in mines
    equipped with diesel engines was compared with that of 90 coal miners  
    not exposed to diesel exhaust. There were similar proportions of
    current smokers (45% in the exposed group and 43% in the unexposed
    group). Measurements made with personal dust samplers and passive
    nitrogen dioxide dosimeters showed average concentrations of
    0.04 mg/m3 (0.2 ppm) nitrogen dioxide, 0.4mg/m3 (0.3 ppm)
    formaldehyde, and 2.0 mg/m3 respirable dust in the exposed group;
    the unexposed group was exposed to 1.4 mg/m3 respirable dust. Forced
    vital capacity (FVC) and forced expiratory volume in 1 sec (FEV) were
    reduced during the work shift for both the diesel-exposed and
    unexposed groups. The reduction was slightly but not significantly
    greater in smokers than in ex- and nonsmokers. An additional analysis

    with adjustment for age, exposure to respirable dust, and years of
    underground mining still revealed no difference in shift-related
    pulmonary function between miners exposed and unexposed to diesel
    exhaust (Ames et al., 1982).

         No significant shift-related change in FVC or FEV was seen among
    workers on ferries transporting diesel trucks and other vehicles in
    comparison with controls (Purdham et al., 1987), and no significant
    change in spirometry measured over a shift was seen in 232 workers in
    four diesel bus garages (Gamble et al., 1987a). Ulfvarson et al.
    (1987), however, found significant decrements in FVC and FEV1 across
    a shift in 23 ferry workers who had had no exposure to diesel for 10
    days, with no difference between smokers and nonsmokers. The
    concentration range of total particles was 0.13-1 mg/m3, that of
    formaldehyde 0.04-0.6mg/m3 (< 0.03-0.5 ppm), and that of nitrogen
    dioxide 0.12-4.6mg/m3 (0.06-2.3 ppm). In a repetition of the study,
    a significant shift-related decrement was noted in FVC but not in
    FEV1. In a later study in which filters were put on the diesel
    trucks, a shift-related decrement in FVC was seen, but no subsequent
    shift-related changes in FEV were noted, with or without filters
    (Ulfvarson & Alexandersson, 19901).

         Three male railroad workers developed asthma after heavy exposure
    to locomotive emissions (Wade & Newman, 1993). The workers, none of
    whom were current smokers, had no previous history of asthma or of any
    respiratory disease, except one who had seasonal rhinitis. They were
    riding immediately behind the lead engines of the trains, where diesel
    exhaust was blown almost continually into the cab. In two, the acute
    onset of asthma occurred within the first hours of exposure.

    B8.2.1.3  Pulmonary effects

         In a study to investigate the effects of diesel exhaust on the
    cells found in bronchoalveolar lavage (BAL) fluid (Rudell et al.,
    1990), eight healthy, nonsmoking volunteers were exposed to diesel
    exhaust (for 60 min according to Rudell et al., 1989) at least three
    weeks after an initial BAL. The median steady-state concentrations
    measured in the exhaust were 4.6mg/m3 (3.7 ppm) nitric oxide,
    3.1 mg/m3 (1.6 ppm) nitrogen dioxide, 22.5mg/m3 (27 ppm) carbon
    monoxide, 0.5 mg/m3 formaldehyde, and 4.3 × 105 particles/cm3;
    according to Rudell et al. (1989), this particle concentration
    corresponds to a mass concentration of approximately 100 µg/m3. BAL
    was performed again 18 h after exposure. A significant reduction
    (P < 0.02) in the total number of mast cells was observed; the number
    of neutrophils was slightly but significantly (P < 0.05) increased in
    comparison with the values before exposure. The ratio of
    T-helper:suppressor cytotoxic cells was elevated ( P < 0.02), and the
    rate of phagocytosis of opsonized yeast cells by alveolar macrophages
     in vitro was reduced (P < 0.02). The number of lymphocytes remained
    unchanged.

    B8.2.2  Epidemiological studies (noncarcinogenic effects)

         Some but not all of the results from cross-sectional and
    longitudinal studies on workers with occupational exposure to low
    levels of diesel engine exhaust show decrements in lung function and
    an increased prevalence of respiratory symptoms.

    B8.2.2.1  Effects on the respiratory system

         A group of 550 underground workers and 273 surface workers in six
    coal mines where diesel-powered equipment was used were matched for
    smoking status, age, height, and years of underground mining to
    workers at other underground coal mines where diesel units were not
    used. The workers at the mines with diesel engines reported
    significantly more persistent cough (23.6%) than controls (16.5%) and
    more frequently had exacerbations of cough and phlegm (21.7 and 16.2%,
    respectively). Pulmonary function (FVC and FEV ) was decreased in both
    surface and underground workers in mines where diesel engines were
    used in comparison with surface and underground miners in other mines.
    There was no consistent relationship between years of underground
    mining and respiratory symptoms and pulmonary function; however, the
    mean time spent in underground mining was only 4.7 years. Full-shift
    area samples showed concentrations of respirable dust ranging from 0.4
    to 16.1 mg/m3 for various jobs; the mean values from personal
    samples were 0.93-2.73 mg/m3, and the nitrogen dioxide levels in
    personal samples were 0.16-0.26 mg/m3 (0.13-0.22 ppm) (Reger et al.,
    1982).

         Pulmonary function and respiratory symptoms were studied in 1976
    in 630 miners in six potash mines where diesel-powered vehicles had
    been introduced between 1950 and 1964. There was no consistent
    relationship with exposure to diesel exhaust, considered in several
    ways, including years of exposure, cumulative exposure to dust,
    nitrogen dioxide concentration, and prevalence of diesel use in the
    mine. The mean length of exposure to diesel engines was, however,
    relatively short: 5-14 years across mines. In addition, cumulative
    exposure was measured only to total dust rather than respirable dust.
    The mean values for total dust in the mines ranged from 9 to
    23 mg/m3 (personal sampling); the ratio of total dust to respirable
    dust ranged from 2 to 11 (area sampling), and the nitrogen dioxide
    concentration was 0.2-6.3mg/m3 (0.1-3.3 ppm) (personal sampling)
    (Attfield et al., 1982).

         A total of 259 miners in five salt mines where diesel equipment
    was used in some of the mines were studied with regard to the number
    of years worked underground and cumulative exposure to respirable
    particulate and nitrogen dioxide. These parameters were associated
    with phlegm after adjustment for age and smoking but not cough. The
    average exposure to respirable particulate in the five mines was

    0.2-0.7 mg/m3, and the average exposure to nitrogen dioxide was
    0.6-4.8mg/m3 (0.3-2.5 ppm). Pulmonary function (FEV 1, FVC, peak
    flow) was not related to the three parameters but was slightly reduced
    in comparison with that of other blue-collar workers. In another
    analysis in the same population, phlegm but not pulmonary function was
    associated with years of work underground in mines where diesel
    engines were used, after adjustment for age and smoking. There was a
    nonsignificant trend for an association between cough and shortness of
    breath with years of exposure (Gamble & Jones, 1983).

         Longitudinal changes in pulmonary function and chronic
    respiratory symptoms were studied between 1977 and 1982 in 280 miners
    exposed to diesel exhaust and 838 miners not exposed to diesel exhaust
    in American underground coal mines. The mean number of years of
    underground work ranged from 6.6 to 17.4 years across mines. Area
    samples taken in 1977 revealed 'low' levels of dust and diesel exhaust
    constituents, which were reported not to exceed 25% of current
    standards, but the actual values were not given. The exposed workers
    had smaller average decrements in FVC and FEV1 than workers not
    exposed to diesel exhaust after adjustment for age, smoking, and years
    of underground work. Additional analysis revealed no relationship with
    cumulative years of underground work, but the level of pulmonary
    function was not considered as a predictor of longitudinal change. The
    exposed miners also had a lower five-year incidence of cough, phlegm,
    and breathlessness than the miners not exposed to exhaust (Ames et
    al., 1984).

         Pulmonary function and respiratory symptoms were studied in 283
    male diesel bus workers in four garages in two cities. The number of
    years worked was a significant predictor of lower FVC and FEV1
    adjustment for age, height, race, and smoking status. In comparison
    after with another blue-collar population, the bus garage workers had
    a higher prevalence of cough and phlegm after adjustment for age and
    smoking. The prevalence of these symptoms was not related to the
    number of years worked (Gamble et al., 1987b).

    B8.2.2.2  Effects on the circulatory system

         No consistent excess of deaths due to cardiovascular disease has
    been identified in cohort studies of workers with potential exposure
    to diesel emissions. Motor vehicle examiners (Stern et al., 1981), who
    are probably exposed to exhaust from a variety of vehicles, had a
    slight, nonsignificant increase in deaths due to cardiovascular
    disease (standardized mortality ratio [SMR] = 105). No difference in
    mortality was seen among men working in potash mines with
    diesel-powered equipment in comparison with men in potash mines with
    no diesel engines (Waxweiler et al., 1973). In a pilot study of 129
    men employed in a bus company in 1951-59 for whom mortality was
    ascertained until 1978 (Edling & Axelson, 1984), a fourfold increase
    in cardiovascular deaths was seen in garage workers. The increase was

    calculated on the basis of at least 15 years since first exposure and
    10 or more years of exposure, but was based on only four deaths. This
    finding was not confirmed in a follow-up of 694 men through 1983
    (Edling et al., 1987). In a study of 8490 garage maintenance workers
    with at least one continuous year of service between 1967 and 1975, a
    deficit of deaths due to cardiovascular disease was seen (Rushton et
    al., 1983).

    B8.2.3  Epidemiological studies (carcinogenic effects)

         The relationship between cancers of the lung and bladder and
    occupational exposure to diesel exhaust has been evaluated in a number
    of epidemiological studies. The occupations examined included diesel
    truck drivers, bus garage workers, railroad workers, heavy equipment
    operators, and stevedores. The risk for lung cancer in relation to
    exposure to diesel exhaust, either self-reported or assumed from
    occupational categories, has also been evaluated in case-control
    studies. Only those studies that were considered relevant for an
    evaluation of the carcinogenic effects of diesel exposure are included
    in this assessment.

    B8.2.3.1  Lung cancer

         The epidemiological studies are summarized in Tables 39-41. Table
    39 is a summary of nine case-control and cohort studies of workers
    exposed to diesel exhaust, Table 40 summarizes nine population- and
    hospital-based studies, and Table 41 presents two studies based on
    information on exposure derived from registries (surveillance
    studies).

         Until recently, the epidemiological study of lung cancer and
    exposure to diesel exhaust was limited by failure to consider the
    latency and duration of exposure necessary for the development of lung
    cancer, both generally considered to be about 20 years (Schenker &
    Speizer, 1979). Studies of exposed workers have also been limited by a
    lack of worker- and industry-specific data on exposure. Such data were
    available only for the American railroad and trucking industries. For
    other occupational groups, such as bus garage workers and stevedores,
    indices of exposure were derived from vehicle use, fuel consumption,
    or years of exposure, or simply job title unsubstantiated by exposure
    assessment. In studies based in hospitals, the general population, or
    registries, information on usual work in a job with exposure to
    vehicle exhaust was derived from self- or surrogate reporting;
    self-reporting of exposure to diesel fumes has also been used. Studies
    with short durations of exposure and follow-up are not useful for
    determining whether diesel exhaust is a human carcinogen. Studies in
    which there is an imprecise definition of exposure to diesel exhaust
    (such as self-reported job title) do not allow detection of an effect,
    since the extent of exposure is unknown. In studies in which an
    elevated risk for lung cancer is noted, such imprecision can result in
    wide confidence intervals.

         The major potential confounding factor in occupational studies of
    lung cancer is tobacco smoking. Such a factor must be related not only
    to the outcome but also to the exposure. In order to assess whether
    tobacco use is a true confounder in a study of occupation, the smoking
    rates among individuals exposed and unexposed to diesel exhaust must
    be known; however, owing to the retrospective nature of many
    occupational studies, smoking histories are often not available. It is
    unlikely that the smoking habits of workers exposed and unexposed to
    diesel exhaust within the same occupational cohort are differentially
    related to the exposure, but when the effect of exposure is small, in
    the order of a relative risk of 1.5 (such has been found for diesel
    exhaust and lung cancer in most studies), differences in smoking that
    are unaccounted for could reduce the relative risk attributable to
    exposure to diesel exhaust. Interpretation of the results of studies
    lacking specific information about smoking is more uncertain when
    small relative risks are seen. Despite imprecise exposure histories,
    one strength of hospital- and registry-based studies of lung cancer is
    that smoking histories are often available.

    (a)  Occupationally based cohort and case-control studies

         Three studies have been conducted of railroad workers exposed to
    diesel exhaust. The transition from steam- to diesel-powered
    locomotives occurred in the United States during the 1950s. By 1959,
    95% of the locomotives in service were diesel powered. In a
    retrospective cohort study of American railroad workers, the mortality
    of 55 407 white male workers aged 40-64 in 1959 who had worked on the
    railroads for 10-20 years was ascertained through 1980, providing a
    22-year period of follow-up (Garshick et al., 1988). A yearly
    three-digit job code up to the time of death or retirement was
    available from the Railroad Retirement Board. The workers had held
    jobs in 39 selected categories, and an industrial hygiene survey was
    conducted to categorize the job codes into those associated with
    regular exposure to diesel exhaust (engineers, conductors, firemen,
    brakemen, and diesel locomotive repair shop workers) and those not
    exposed to diesel exhaust (clerks and signal maintainers). The
    concentration of respirable particulate, adjusted for cigarette smoke,
    was used as an indicator of exposure to diesel exhaust (Woskie et al.,
    1988a,b). A potentially confounding exposure in the railroad industry
    is asbestos, which was used to insulate the steam locomotives run
    previously. Thus, workers in steam locomotive repair shops would have
    been exposed to asbestos. Workers who had the longest potential
    duration of exposure, i.e. those aged 40-44 in jobs with exposure to
    diesel exhaust in 1959, had a relative risk of dying of lung cancer of
    1.45 (95% confidence interval [CI] = 1.11-1.89). Workers aged 45-49 in
    1959, who would have had slightly less exposure, had a relative risk
    of 1.33 (95% CI = 1.03-1.73). Exclusion of the workers with potential
    past exposure to asbestos did not appreciably change these relative
    risks.

         The US Environmental Protection Agency has supported an effort to
    obtain more detailed information on exposure in this study in order to
    estimate unit risk (Pepelko & Chen, 1993). Adequate exposure-response
    curves could not be drawn for years of exposure or for measurements of
    current exposure. Possible reasons include an imprecise assessment of
    past exposure, since only current measurements of exposure were
    available; changes in exposure over time, since improved ventilation
    and new diesel locomotives have been introduced, resulting in reduced
    exposure; and the lack of exposure histories for the workers who would
    have been exposed to diesel exhaust before 1959. In addition, the
    industrial hygiene survey was performed in four smaller railroads,
    where exposure may not reflect that throughout the industry; the
    presence of non-diesel particulate in the respirable samples collected
    could have led to exposure misclassification. Nevertheless, the
    industrial hygiene survey validated the classification of jobs into
    those associated and not associated with exposure.

         In a case-control study of American railroad workers by the same
    investigators (Garshick et al., 1987), deaths occurring between 1
    March 1981 and 28 February 1982 were recorded, and 1256 from lung
    cancer were matched by age to deaths from causes other than cancer or
    accidents. Smoking histories were obtained from next-of-kin. After
    adjustment for smoking and past exposure to asbestos, workers aged 64
    at the time of death and with 20 years of work in a job with exposure
    to diesel exhaust had an odds ratio of 1.41 (95% CI = 1.06-1.88).

         A cohort study addressed mortality between 1965 and 1977 among
    all 43 826 male railroad workers who had retired from the Canadian
    National Railway Company before 1965 but who were still alive and
    those who retired between 1965 and 1977. Occupation at the time of
    retirement was used to classify workers as unexposed, possibly
    exposed, or probably exposed to diesel fume. The relative risk of
    dying of lung cancer was 1.20 for workers with possible exposure
    (P = 0.013) and 1.35 for those probably exposed (P < 0.001). Similar
    relative risks were obtained after exclusion of workers involved in
    locomotive repair, who were most likely to have had past exposure to
    asbestos. These results are consistent with those of the studies of
    American railroad workers, although the degree to which the workers
    were actually exposed to diesel fumes is not known. Only retired
    workers were studied, excluding younger, active workers who would have
    had the most exposure. No information was given about the duration of
    exposure of the retired workers to diesel exhaust since many would
    have worked primarily before the transition to diesel locomotives
    (Howe et al., 1983).


        Table 39.  Occupation-based case-control and cohort studies of lung cancer in groups exposed to diesel engine exhaust
                                                                                                                                                

    Population                       Exposure assessment        Results                       Limitations                       Reference
                                                                                                                                                

    43 826 male pensioners of the    Exposure classified by     RR = 1.20 (P = 0.013)         Incomplete exposure               Howe et al.
    Canadian National Railway        experts on occupation at   and 1.35 (P = 0.001) for      assessment owing to               (1983)
                                     time of retirement:        possible and probable         lack of lifetime occupational
                                     unexposed, possibly        exposure, respectively.       history. Difficult to
                                     exposed, probably          Significant trend in RR       separate combustion products
                                     exposed                    with increasing likelihood    from diesel exhaust,
                                                                of exposure                   since the cohort would
                                                                                              have worked in both the
                                                                                              steam and diesel eras;
                                                                                              many would have had
                                                                                              relatively short exposure
                                                                                              to diesel. No data on
                                                                                              smoking; method of
                                                                                              categorizing exposure
                                                                                              not validated or described
                                                                                              in detail

    Incident deaths between          Personal exposure          Adjusted for smoking and      Possible misclassification        Garshick et al.
    1 March 1981 and 28              assessed by industrial     to asbestos, OR = 1.41        of exposure to diesel             (1987)
    February 1982 in US railroad     hygiene sampling in 39     (1.06-1.88) (< 64 years of    exhaust since job title might
    workers. 1256 cases of           job categories; yearly     age) for 20 years of work     not reflect exposure in each
    lung cancer and 2385 matched     job title used to          in a diesel-exposed job.      case; however, this should
    controls (up to two each, on     dichotomize exposure       For > 20 years of work in a   reduce the OR.
    age and date of death). Cases    into exposed and           diesel-exposed job relative
    and controls had worked for      unexposed between 1959     to 0-4 years, OR = 1.64
    the railroad for > 10 years.     and retirement             (1.18-2.29)
                                                                                                                                                

    Table 39 (contd)
                                                                                                                                                

    Population                       Exposure assessment        Results                       Limitations                       Reference
                                                                                                                                                

    55 407 white male railroad       Industrial hygiene data    RR = 1.45 (1.11-1.89) in      No data on smoking                Garshick et al.
    workers, aged 40-64 and          used to categorize jobs    workers aged 40-44 in                                           (1988)
    with 10-20 years of work in      as exposed or unexposed    1959; RR = 1.33 (1.03-
    1959. Mortality due to lung      to diesel exhaust; jobs    1.73) in workers aged 45-
    cancer (1694 deaths)             well categorized for       49. Unexplained by exposure
    determined retrospectively       most of the cohort         to asbestos. Workers with
    through 1980                                                highest potential
                                                                cumulative exposure to
                                                                diesel exhaust had highest
                                                                risk for cancer

    About 20 000 male London         Five job categories used   SMRs for all five job         Duration of work not              Waller (1981)
    Transport workers aged 45-64.    to define exposure,        categories < 100 for lung     considered; impossible to
    667 cases of lung cancer         validated by               cancer (healthy worker        follow individuals after
    ascertained among active         environmental              effect), but highest SMR in   retirement; no adjustment for
    workers > 25 years (1950-74)     concentrations of          highest exposure group, not   smoking
                                     benzo[a]pyrene and         significantly greater than
                                     particles measured         least exposed (underground
                                     in 1957 and 1979           train crew)

    8490 male London Transport       100 job titles grouped     SMR = 101 overall, but job    No validation of exposure         Rushton et al.
    maintenance workers (cohort      into 20 broad categories   category 'general hand'       by air sampling; short            (1983)
    mortality study). Mortality of                              showed increased SMR for      (average, 6 years) follow-up;
    workers employed for > 1 year                               lung cancer with no           no adjustment for smoking
    between 1 January 1967                                      apparent link to diesel
    and 12 March 1975                                           exhaust
                                                                                                                                                

    Table 39 (contd)
                                                                                                                                                

    Population                       Exposure assessment        Results                       Limitations                       Reference
                                                                                                                                                

    695 workers in bus garages       Intensity of exposure to   SMR = 122 (71-196); for       No adjustment for smoking         Gustavsson et
    who had worked for > 6 months    exhaust and asbestos       increasing exposure index,                                      al. (1990)
    in 1945-70. 20 deaths from       according to industrial    RR = 1.34 (1.09-1.64), 1.81
    lung cancer identified (nested   hygiene estimate;          (1.20-2.71), 2.43
    case-control study)              duration of exposure       (1.32-4.47)
                                     based on company records

    Cohort study of 107 563 US       Usual occupation (truck    SMR = 153 (23 cases)          No cases among non-smokers,       Walrath et al.
    veterans alive in 1954 and       driver)                    among people who had          but SMR calculated from           (1985)
    followed-up to 1 January 1970                               ever smoked; no statistics    smoking-specific rates
                                                                given

    994 male lung cancer deaths      Longest job held: diesel   OR = 1.55 (0.97-2.47) for     Extent of actual exposure         Steenland et
    in 1982-83 and 1085 controls     truck driver, gasoline     long-haul drivers with > 18   to diesel exhaust unknown         al. (1990, 1992)
    (excluding lung cancer,          truck driver, both         years of employment; OR =
    bladder cancer, accidents).      types, truck mechanic,     1.89 (1.04-3.42) for diesel
    Cases and controls from          stevedore. Exposure        truck drivers with > 35
    Teamsters Union who had filed    validated by industrial    years of employment
    claims (requires 20-year         hygiene survey             (adjusted for smoking)
    tenure)

    Retrospective cohort study of    Cohort assumed to be       Lung cancer: SMR = 168        No adjustment for smoking;        Gustafsson et
    stevedores in Sweden first       exposed on basis of job    (136-207) for incidence;      exposure assumptions not          al. (1986)
    employed before 1974 for a       duties, but no formal      SMR = 132 (105-166) for       fully justified; no
    continuous period of 6 months.   assessment made            mortality                     calculations done to explore
    6071 workers employed in 1961                                                             employment time as a surrogate
    with mortality determined                                                                 for exposure
    through 1980. 89 cases of
    lung cancer noted in Swedish
    Cancer Registry (71 deaths)
                                                                                                                                                

    Table 39 (contd)
                                                                                                                                                

    Population                       Exposure assessment        Results                       Limitations                       Reference
                                                                                                                                                

    Stevedores in Swedish            Estimated from use of      Adjusted for smoking          Crude adjustment for              Emmelin et al.
    population defined by            fuel in each port and      (yes/no) depending on         smoking; no industrial            (1993)
    Gustafsson et al. but limited    years of work since use    index of exposure used;       hygiene data to validate
    to workers in ports for which    of diesel equipment        exposure-response             estimates of exposure to
    data data on diesel fuel                                    relationship noted,           diesel exhaust
    consumption were available.                                 relative odds for medium
    Case-control study with 50                                  exposure ranging from 1.5
    cases of lung cancer detected                               to 2.7 and that for high
    in 1960-82 matched to 154                                   exposure ranging from 2.9
    controls on port and date of                                to 6.8. Lower limit of 90%
    birth                                                       confidence interval < 1,
                                                                except for one
                                                                high-exposure category
                                                                                                                                                

    RR, relative risk; OR, crude odds ratio; SMR, standardized mortality ratio; in parentheses, 95% confidence interval (CI), unless
    otherwise stated
             Three studies reported findings among bus garage workers. Waller
    (1981) studied lung cancer occurrence among London Transport staff on
    the basis of job category. Diesel buses were introduced in the 1930s,
    gradually replacing gasoline-powered vehicles and, in the 1950s,
    electric trolley-buses. Mortality from lung cancer was ascertained for
    active workers aged 45-64 between 1950 and 1974. A lower risk was seen
    for people in all job categories in comparison with the general
    population of the London area, indicating a 'healthy worker effect'.
    The highest SMR was seen for workers who had the most exposure to
    diesel exhaust (bus garage engineers), although the SMR was not
    significantly different in workers with no exposure. Although this
    study showed no relationship between job category and mortality from
    lung cancer, the study had significant limitations. No retired workers
    were followed up, so that workers who would have had significant
    exposure were excluded. Since the results were not presented on the
    basis of years of work (years of exposure to diesel fumes), workers
    with little exposure were grouped with those with long exposure,
    making it more difficult to detect an effect of diesel exhaust.

         Rushton et al. (1983) ascertained the mortality of 8490 London
    Transport bus garage workers who had been employed for at least one
    year between 1967 and 1975. One hundred job titles were grouped into
    20 broad categories. The overall SMR for lung cancer was not elevated,
    and there was no relationship with any job with exposure to diesel
    exhaust; however, the SMRs were not presented on the basis of years of
    work in a job with exposure, and the duration of follow-up was short,
    with a mean of 5.9 years.

         Using a nested case-control design, Gustavsson et al. (1990)
    studied a cohort of 695 men who had worked in five bus garages in
    Stockholm, Sweden, in 1945-70. Diesel-powered buses were first
    introduced in Stockholm in the 1930s, and after 1945 all of the
    internal combustion engines in buses were diesel-powered. No
    measurements of exposure were available, so exposure to diesel exhaust
    was estimated by industrial hygienists on the basis of garage
    ventilation, work practices, and the number of buses in the garages
    and was graded on a five-point scale, each increase corresponding to a
    50% increase in intensity. An index of cumulative exposure to diesel
    exhaust was calculated for each worker by multiplying the exposure
    level for each work period by the duration in years. Past exposure to
    asbestos was estimated on the basis of historical measurements of
    asbestos fibres obtained during brake-repair operations. Twenty cases
    of lung cancer were identified and matched by age to six controls. The
    relative risk for lung cancer increased with increasing index of
    exposure to diesel exhaust: That at the lowest exposure level was 1.34
    (95% CI = 1.09-1.74), that at an intermediate level was 1.81
    (95% CI = 1.20-2.71), and that at the highest level was 2.43 (95%
    CI = 1.32-4.47). No such increase was observed with a similar index of
    exposure to asbestos. No information was available on smoking.

         Diesel-powered trucks were introduced in the United States in the
    1950s and 1960s. Trucking companies had completed the transition to
    diesel trucks by 1960, while independent drivers and non-trucking
    companies would have completed the transition later. Steenland et al.
    (1990, 1992) studied 994 deaths from lung cancer occurring in 1982 and
    1983 among male members of the Teamsters Union who had filed claims
    for pension benefits (requiring at least 20 years of membership). Work
    histories and smoking histories were obtained from next-of-kin; job
    categories were also available from the Union records. The odds ratio
    for lung cancer among long-haul drivers, who drove mostly diesel
    trucks, with 18 or more years of employment after 1959 was 1.55 (95%
    CI = 0.97-2.47) after adjustment for age, smoking, and exposure to
    asbestos. Teamsters with 35 years or more of employment whose main job
    was a diesel truck driver had a relative risk for lung cancer of 1.89
    (95% CI = 1.04-3.42). For drivers of gasoline trucks, the relative
    risk was only 1.34 and did not achieve conventional levels of
    significance (95% CI = 0.81-2.22).

         In an industrial hygiene survey (Zaebst et al., 1991; Steenland
    et al., 1992) that accompanied this study, measurements of elemental
    carbon were used to estimate exposure to diesel exhaust. The truck
    drivers appeared to have received much of their current exposure to
    diesel exhaust from background levels on the road rather than directly
    from their engines. No historical measurements of exposure were
    available.

         Gustafsson et al. (1986) studied the mortality of 6071 stevedores
    in Sweden who had been employed for at least six months between 1961
    and 1974 through 1980. The stevedores had been exposed to both
    gasoline and diesel exhausts; diesel-powered trucks were first used in
    Swedish ports in the late 1950s, with a rapid increase in use in the
    1960s. The SMR for lung cancer was 132 (95% CI = 105-166). The
    year-specific lung cancer rates in the stevedores between 1961 and
    1980 were greater than those in the Swedish male population.

         In order to refine the assessment of exposure among the
    stevedores and to adjust for smoking, a case-control study was
    conducted in this cohort (Emmelin et al., 1993). Cases and controls
    were selected from among male stevedores who had been employed for at
    least six months in 1950-74; the cases were those occurring in
    1960-82. On the basis of 50 cases and 154 controls, with adjustment
    for smoking (yes/no), the odds ratio for lung cancer was seen to
    increase with increases in three indices of exposure: years since
    diesel equipment was used in the port, estimates of cumulative fuel
    consumption, and years that fuel use was above a minimal level in the
    port. The increases were consistent with an exposure-response
    relationship. Only one of the point estimates of the odds ratios for
    lung cancer reached statistical significance.


        Table 40.  Population- and hospital-based studies of lung cancer in groups exposed to diesel engine exhaust
                                                                                                                                                

    Population                        Exposure assessment          Results                         Limitations                       Reference
                                                                                                                                                

    7518 (3539 men, 3979              Occupation determined        OR = 1.52 for truck drivers     Exposure estimate based on        Williams
    women) incident invasive          at interview                 (P > 0.05)                      self-reporting; not validated     et al.
    cancers from the Third                                                                         47% non-response; controls        (1977)
    National Cancer Survey. Lung                                                                   consisted of other cancers,
    cancer cases: 432 in men,                                                                      probably diluting risk
    128 in women. Combined other                                                                   estimate; few cause-specific
    cancer sites used as                                                                           cancers and individual
    controls                                                                                       occupations

    Cohort study of 107 563 US        Usual occupation (truck      SMR = 153 (23 cases)            No cases in nonsmokers,           Walrath
    veterans alive in 1954 and        driver)                      among people who had            but SMR calculated using          et al.
    followed-up to 1 January 1970                                  ever smoked; no statistics      smoking-specific rates            (1985)

    589 cases of lung cancer          Job title from next-of-kin   Adjustment for smoking;         No validation of exposure         Damber &
    reported to the Swedish           for occupations held for     (0.6-2.2) for professional      groupings, so actual exposure     Larsson
    Cancer registry 1972-77 with      > 1 year                     drivers (> 20 years of          to diesel exhaust uncertain       (1987)
    death before 1979; 562                                         employment) with dead
    matched dead controls (sex,                                    controls; OR = 1.1 (0.6-2.2)
    age, year of death,                                            with living controls
    municipality) drawn from
    National Registry of Causes
    of Death; matched living
    controls (sex, year of birth,
    municipality) drawn from
    National Population Registry
                                                                                                                                                

    Table 40 (contd)
                                                                                                                                                

    Population                        Exposure assessment          Results                         Limitations                       Reference
                                                                                                                                                

    Montreal, Canada,                 Interview with subject;      Adjustment for smoking;         Job title used to indicate        Siemiatycki
    hospital-based cases and          job title translated into    OR = 1.2 for squamous-cell      exposure to diesel exhaust;       et al.
    controls; 857 cases of lung       likely exposure to           lung carcinoma (90% CI,         not validated; exposure           (1988)
    cancer and 1523 controls (other   combustion products by       1.0-1.5) for any exposure       ill-defined
    cancers than lung)                industrial hygienist         but not elevated for other
                                                                   cell types. No relationship
                                                                   with years of exposure or
                                                                   intensity

    461 981 male volunteers           Self-reported occupation     Adjustment for smoking;         Exposure information              Boffetta
    enrolled in the American          and exposure to diesel       > 16 years' exposure to         based on self-reporting;          et al.
    Cancer Society prospective        exhaust                      diesel exhaust, OR = 1.21       not validated                     (1988)
    mortality study in 1982; aged                                  (0.94-1.56); for 1-15 years,
    40-79 at enrolment; 2-year                                     OR = 1.05 (0.80-1.39).
    follow-up; 378 622 men with                                    Lung cancer mortality also
    known exposure to diesel                                       elevated in miners and heavy
    exhaust                                                        equipment operators

    1260 cases of lung cancer         Job title obtained from      For motor vehicle drivers       Crude classification of           Benhamou
    and 2084 controls matched         occupational history         (adjusted for smoking),         exposure                          et al.
    for age, sex, hospital            obtained at interview;       OR = 1.42 (1.07-1.89); for                                        (1988)
    admission in France in 1976-80    exposure based on any        all transport equipment
                                      work in a job                operators, OR = 1.35
                                                                   (1.05-1.75). Increase in risk
                                                                   with duration of employment
                                                                                                                                                

    Table 40 (contd)
                                                                                                                                                

    Population                        Exposure assessment          Results                         Limitations                       Reference
                                                                                                                                                

    Pooled data from three            Job title from interviews    Smoking-adjusted OR =           No validation of exposure         Hayes
    case-control studies with         with subjects or             1.5 (1.1-1.9) for truck         classification; unclear if        et al.
    2291 male lung cancer cases       next-of-kin; years of work   drivers with > 10 years of      most exposure was to              (1989)
    and 2570 controls. Analysis       in motor exhaust-related     work and job history obtained   diesel exhaust or if job
    limited to 1444 cases and         occupation calculated        by direct interview. For all    titles selected as 'exposed'
    1893 controls who provided                                     motor exhaust-related           were actually exposed to
    information at interview                                       occupations and  > 10 years     diesel exhaust
                                                                   of work, OR = 1.5 (1.2-1.9)

    Hospital-based study of 2584      Occupational titles and      Smoking-adjusted OR for         Exposure histories not            Boffetta
    cases in 18 hospitals in six      self-reporting of exposure   occupations with probable       validated                         et al.
    US cities; 5099 controls                                       exposure to diesel exhaust                                        (1990)
                                                                   = 0.95 (0.78-1.16). For
                                                                   self-reported exposure,
                                                                   OR = 1.21 (0.78-2.02)

    Detroit, USA, area; 3792          Interview to obtain full     For drivers of heavy trucks     Exposure histories not            Swanson
    cases and 1966 controls           occupational history;        with > 20 years of work,        validated                         et al.
    identified through cancer         analysis by job title        OR = 2.5 (1.4-4.4), with                                          (1993)
    surveillance system                                            adjustment for smoking;
                                                                   exposure-response
                                                                   relationship with years of
                                                                   work. For drivers of light
                                                                   trucks with > 20 years of
                                                                   work, OR = 2.1 (0.9-4.6)
                                                                                                                                                

    OR, odds ratio; SMR, standardized mortality ratio; in parentheses, 95% confidence interval (CI), unless otherwise stated
        (b)  Population- and hospital-based studies

         In a study of 107 563 American veterans who were alive in 1954
    and were followed to the end of 1969, the SMR among smokers who
    reported their occupation as truck or tractor driver was 153, based on
    23 cases. Most of these individuals would have had little exposure to
    diesel exhaust, however, since diesel trucks were introduced in the
    United States only in the 1950s and 1960s (Walrath et al., 1985).

         In the Third National Cancer Survey in the United States,
    self-reporting of work as a truck driver and adjustment for smoking
    resulted in an odds ratio of 1.52, which was not statistically
    significant (Williams et al., 1977). In a study of 589 cases of lung
    cancer reported to the Swedish Cancer Registry between 1972 and 1977
    among people who had died before 1979, the odds ratio for > 20 years
    of work as a professional driver was 1.2 (95% CI = 0.6-2.2) after
    adjustment for smoking (Damber & Larsson, 1987). In a hospital-based
    study of 857 cases of lung cancer and controls, occupational histories
    obtained by interview were interpreted by an industrial hygienist.
    After adjustment for smoking, the odds ratio for squamous-cell
    carcinoma of the lung was 1.2 (90% CI = 1.0-1.5) in association with
    any exposure to diesel exhaust (Siemiatycki et al., 1988). The
    mortality of 378 622 men between 1982 and 1984 was analysed in the
    American Cancer Society prospective study of cancer on the basis of
    self-reporting of exposure to diesel exhaust. After adjustment for
    age, smoking, and other occupational exposures, including asbestos,
    coal-tar and pitch, and gasoline exhaust, the relative risk for lung
    cancer for men with 16 years of exposure was 1.21 (95% CI =
    0.94-1.56). For truck drivers who reported exposure to diesel exhaust,
    the relative risk was 1.22 (95% CI = 0.77-1.95) (Boffetta et al.,
    1988).

         In a study in France, 1260 male cases of lung cancer and 2084
    controls matched on age and hospital were collected between 1976 and
    1980; job titles and smoking histories were obtained by interview, and
    odds ratios were calculated on the basis of any work in a particular
    occupation. After adjustment for smoking, the odds ratios were 1.35
    (95% CI = 1.05-1.75) for transport equipment operators and 1.42 (95%
    CI = 1.07-1.89) for motor vehicle drivers. There was no increase in
    risk with increasing years of work in a job, but details of the
    analysis were not presented (Benhamou et al., 1988).

         The data from three case-control studies carried out by the
    American National Cancer Institute between 1976 and 1983 were pooled
    in order to study lung cancer in people with occupations associated
    with motor vehicles. The analysis was limited to 1444 male patients

    with lung cancer and 1893 controls who provided information on smoking
    and occupation at an interview. The odds ratio for lung cancer for 10
    or more years of employment in any occupation related to exhaust from
    either diesel or non-diesel engines was 1.5 (95% CI = 1.2-1.9), after
    adjustment for age and smoking. In truck drivers, the odds ratio was
    1.5 (95% CI = 1.1-1.9) (Hayes et al., 1989).

         In a case-control study of patients with lung cancer in 18
    hospitals in six American cities, 2584 cases seen between 1969 and the
    late 1980s were matched to at least one control on the basis of age,
    sex, hospital, and year of interview. A smoking history and usual
    occupation were obtained by interview. After 1985, reports of
    self-reported exposure to diesel exhaust were obtained for 477 lung
    cancer patients and 946 controls. Exposure was graded as 'probable' if
    the individual had usually worked on railroads (the specific job was
    not used) or in a variety of jobs related to motor vehicles. For these
    individuals, the odds ratio for lung cancer was 1.31 (95% CI =
    1.09-1.57); for those reporting usual occupation as a truck driver,
    the odds ratio was 1.31 (95% CI = 1.03-1.67). After adjustment for
    cigarette use, age, race, and date of interview, however, the odds
    ratios for lung cancer were reduced to 0.95 (95% CI = 0.78-1.16) for
    individuals with probable exposure and 0.88 (95% CI = 0.67-1.15) for
    truck drivers. For self-reported exposure to diesel exhaust, the odds
    ratio was 1.21, after adjustment for smoking, age, and other potential
    confounding variables including exposure to asbestos, education, and
    race, but did not achieve statistical significance (95% CI =
    0.73-2.02) (Boffetta et al., 1990).

         In a study of 3792 cases of lung cancer in men in Detroit, United
    States, newly diagnosed in 1984-87, an occupational history and
    information on smoking were collected by interview with the subject or
    a surrogate. Controls were men with colon or rectal cancer. After
    adjustment for smoking, the odds ratio for lung cancer was 2.5 (95% CI
    = 1.4-4.4) among white men who had driven heavy trucks for 20 years
    and 2.1 (95% CI = 0.9-4.6) for drivers of light trucks for 20 years.
    There was a significant increase in the odds ratio for lung cancer
    with increasing years of work for drivers of both types of truck
    (Swanson et al., 1993).

    (c)  Registry-based, surveillance studies

         Table 41 summarizes two studies in which information on exposure
    was based on occupational titles obtained from censuses. No
    information was available on smoking or on length of employment. Both
    studies reported an elevated risk for lung cancer. Ahlberg et al.
    (1981) found a relative risk of 1.33 (95% CI = 1.13-1.56) for
    professional drivers, and Hansen (1993) found an SMR for all
    respiratory cancer in truck drivers of 160 (95% CI = 128-198).


        Table 41. Surveillance studies of lung cancer in groups exposed to diesel engine exhaust
                                                                                                                                                

    Population                        Exposure assessment        Results                         Limitations                  Reference
                                                                                                                                                

    Swedish census-based cancer       Listing of occupation in   RR = 1.33 (1.13-1.56)           No validation of exposure;   Ahlberg et al.
    incidence registry; 154 cases     census                                                     no adjustment for smoking    (1981)
    among professional drivers

    Retrospective cohort study of     Job title in census        84 cancers in truck drivers;    No adjustment for smoking;   Hansen (1993)
    14 225 truck drivers in 1970;                                SMR =160 (128-198);             information on specific
    mortality determined until                                   expected number based           exposure or length of
    1980. Referent group of persons                              on rates in referent            employment
    with no exposure to combustion                               population
    products on basis of census
    job title
                                                                                                                                                

    RR, relative risk; SMR, standardized mortality ratio; in parentheses, 95% confidence interval

    Table 42.  Case-control studies of urinary bladder cancer in individuals with possible exposure to diesel exhaust
                                                                                                                                                

    Population                         Exposure assessment          Results                                                    Reference
                                                                                                                                                

    Incident cases in 480 men          Any exposure to diesel and   Based on 15 cases and  controls with discordant            Howe et al.
    and 152 women in three             traffic fumes obtained at    exposure histories, OR for exposure to diesel and          (1980)
    Canadian provinces, April          interview                    traffic fumes = 2.8 (0.8-11.8); not adjusted for
    1974-June 1976; each case                                       smoking
    matched by age and sex to
    a neighbourhood control

    Incident cases in 303 white        Occupation and industry      For any employment in trucking, RR = 2.2 (1.1-4.4)         Silverman et al.
    men in metropolitan Detroit,       obtained at interview        based on 28 cases and 13 controls. For any work in         (1983)
    USA, December 1977-November                                     railroad and railway express services, RR = 1.9
    1978; 2986 controls                                             (0.9-3.8) based on 22 cases and 12 controls, not
    selected by random-digit                                        adjusted for smoking. For any employment as a truck
    dialling but of same age                                        driver (42 cases and 18 controls), adjusted for
    distribution as cases                                           smoking, RR = 2.1 (1.4-4.4). For > 10 years as truck
                                                                    driver, smoking-adjusted RR = 5.5 (1.8-17.3), and for
                                                                    truck drivers who ever drove vehicles with diesel
                                                                    engines, smoking-adjusted RR = 11.9 (2.3-61.1).
                                                                    Elevated risk for truck drivers with employment after
                                                                    rather than before 1950, when diesel truck use became
                                                                    more prevalent
                                                                                                                                                

    Table 42 (contd)
                                                                                                                                                

    Population                         Exposure assessment           Results                                                     Reference
                                                                                                                                                

    White residents of New             Occupation and exposure       For any employment as a truck driver (35 cases, 53          Hoar & Hoover
    Hampshire and Vermont,             to diesel fuel or engines     controls), OR = 1.5 (0.9-2.6), not adjusted for smoking.    (1985)
    USA, who died of bladder           by next-of-kin                Those employed 1930-49 had the highest OR (2.6;
    cancer 1975-79 (on death                                         1.3-5.1) after adjustment for coffee drinking and
    certificate) matched to one                                      smoking. For 26 cases and 39 controls who reported
    control with cause of death                                      exposure to diesel fuel or engines, OR = 1.5 (0.8-2.8),
    other than suicide, matched                                      not adjusted for smoking. Significant trend for
    on state, sex, race, and age,                                    increased relative odds up to 39 years of exposure
    and a second control with                                        to diesel exhaust
    same criteria but also matched
    on county. Information obtained
    for 325 cases and 673 controls

    Cases in 512 men in Turin,         Occupation obtained at        In truck drivers, based on 16 cases and 16 controls,        Vineis &
    Italy, matched by age to 596       interview                     OR = 1.2 (0.6-2.5), not adjusted for smoking                Magnani (1985)
    hospital controls without cancer

    Cases in 194 men in 18 hospitals   Usual occupation obtained;    Only 16 cases in occupations with potential exposure        Wynder et al.
    in six US cities, January          likelihood of exposure to     to diesel exhaust. ORs all < 1 for warehousemen, bus        (1985)
    1981-May 1983. Controls            diesel exhaust determined     and truck drivers, and heavy equipment operators. In
    were persons hospitalized at       on basis of estimates of      railroad workers, based on two cases and one control,
    the same time as the case but      percentage of  workers with   OR = 2. For any exposure, with adjustment for smoking,
    for diseases not related to        potential exposure in that    OR = 0.87 (0.47-1.58). For high exposure, OR = 1.68
    cigarette smoking. Each case       occupation: high exposure,    (0.49-5.73), not adjusted for smoking
    matched to three controls          > 20% of workers; moderate
    (total, 582) by age, race, year    exposure, 10-19% of workers
    of interview, and hospital of
    admission
                                                                                                                                                

    Table 42 (contd)
                                                                                                                                                

    Population                         Exposure assessment           Results                                                     Reference
                                                                                                                                                

    1909 new cases in white men        Occupation obtained at        After adjustment for smoking, RR for men usually            Silverman
    from 10 metropolitan areas in      interview                     working as a truck driver = 1.5 (1.1-2.0), based on 99      et al. (1986)
    the USA, 1977-78. Up to two                                      cases and 123 controls. For those ever employed as
    controls (total, 3569) selected                                  truck driver, increased risk with increasing years of
    by random-digit dialling and                                     employment; however, truck drivers first employed
    matched for age and                                              < 40 years since diagnosis had no increased risk. OR
    geographic area with case                                        for bus driver as usual occupation = 1.5 (0.6-3.9),
                                                                     based on 9 cases and 13 controls, adjusted for smoking

    Cases in 99 men in La Plata,       Occupational history          Based on 20 cases ever employed as a truck or railway       Iscovich et al.
    Argentina, matched on age to       obtained at interview         driver, RR = 4.31 (P < 0.05); reduced whenr                 (1987)
    99 hospital controls without                                     adjusted for smoking, but no details provided
    cancer and 99 neighbourhood
    controls

    Cases in 371 men and women         Occupational history          After adjustment for age and sex, RR = 1.55 (1.06-2.28)     Jensen et al.
    in the Danish Cancer Registry.     at interview                  any employment as a land transport worker; based            (1987)
    Controls (771) selected randomly                                 on 51 cases and 73 controls. After adjustment for
    from general population                                          smoking, age, and sex, years of work as a land transport
                                                                     worker, and years of work as a bus, taxi, or truck
                                                                     driver significantly associated with increased risk

    Deaths in 731 men in Ohio,         Yearly listing of             Work as a truck driver > 20 years, OR = 12 (P < 0.01),      Steenland
    USA, 1960-82, identified on        occupation obtained from      six cases and one control; for > 20 years as a railroad     et al. (1987)
    death certificates and             city commercial directories   worker, OR = 2.21 (P < 0.01), based on 22 cases, not
    matched to six controls without                                  adjusted for smoking
    tumours of the urinary tract or
    pneumonia on age, year of
    death, and race
                                                                                                                                                

    Table 42 (contd)
                                                                                                                                                

    Population                         Exposure assessment           Results                                                     Reference
                                                                                                                                                

    Cases in 826 men and women         Occupational history and      For any exposure of men to exhausts, OR adjusted for        Risch et al.
    diagnosed 1979-82 in               exposure to 'exhausts'        cigarette smoking = 1.16 (0.91-1.48). For men ever          (1988)
    Edmonton, Calgary, Toronto,        obtained at interview         exposed 8-28 years before diagnosis, OR = 1.21
    and Kingston, Canada,                                            (0.93-1.58). For men in jobs with any contact with
    matched by age, sex, and                                         diesel or traffic fumes 8-28 years before diagnosis,
    area of residence to 792                                         adjusted for smoking, OR = 1.69 (1.24-2.31);
    randomly selected population                                     significant trend with increasing years of exposure
    controls

    Cases in 486 men and women         Interview with subject;       After adjustment for smoking, OR = 1.0 (90% CI,             Siemiatycki
    in Montreal, Canada, hospitals     job title translated into     0.8-1.2) for any exposure to diesel exhaust, based          et al. (1988)
    and 2196 controls with             likely exposure to            on 82 cases with exposure
    diseases other than lung           combustion products by
    or kidney cancer                   industrial hygienists

    Cases in 136 men in 18             Usual occupation and          For any exposure, OR = 1.24 (0.77-2.00), based on           Iyer et al.
    hospitals in six US cities.        self-reporting of exposure    41 cases, with adjustment for smoking                       (1990)
    Two controls without               to diesel exhaust obtained
    tobacco-related diseases matched   at interview; grouped into
    to each case on age (< 2 years),   probable and possible
    race, hospital, and year of        exposure
    interview
                                                                                                                                                

    Table 42 (contd)
                                                                                                                                                

    Population                        Exposure assessment            Results                                                   Reference
                                                                                                                                                

    256 cases and 287 controls in     Job titles for all jobs held   For any exposure to diesel exhaust, RR = 1.7              Steineck et al.
    population-based study of all     obtained by postal             (0.9-3.3), based on 25 cases and 19 controls, adjusted    (1990)
    urothelial cancer in Stockholm,   questionnaire; industrial      for year of birth and smoking. For subjects considered
    Sweden, 1985-87, among            hygeienists reviewed titles    to have moderate or high exposure, RR = 1.1 (0.3-4.3),
    men born 1911-45                  and categorized subjects as    but based on only four cases and five controls
                                      exposed to various
                                      substances

    Cases in 658 men in seven         Interview with subject; job    After adjustment for smoking, OR for work in road         Cordier et al.
    French hospitals, 1984-87,        title and duties translated    transport = 1.02 (0.62-1.69), based on 36 cases and       (1993)
    matched for age, race, and        into likely exposure to        35 controls. For stevedores, OR = 1.31 (0.87-1.98);
    place of residence to 658         diesel exhaust by industrial   for material-handling equipment operator, OR = 7.67
    controls selected randomly        hygienists                     (0.96-61.4); and for transport equipment operators,
    among patients admitted                                          OR = 0.88 (0.62-1.26). For diesel fumes, OR = 0.99
    for diagnoses other than                                         (0.32-3.03), based on seven cases and two controls
    cancer

    Cases in 153 men in one           Occupational history           For any work as a road transport worker,                  Notani et al.
    hospital in Bombay, India,        obtained at interview          smoking-adjusted OR = 1.36 (0.5-3.7), based on eight      (1993)
    1986-90, matched by age to                                       cases and nine controls
    212 controls with oral or
    pharyngeal cancer or benign
    oral disease
                                                                                                                                                

    OR, odds ratio; RR, relative risk; in parentheses, 95% confidence interval (CI), unless otherwise stated
        B8.2.3.2  Urinary bladder cancer

         Fifteen case-control studies of urinary bladder cancer in
    relation to presumed exposure to diesel exhaust are summarized in
    Table 42. The limitation of all of these studies is that exposure to
    diesel exhaust was not clearly characterized. Howe et al. (1980) used
    any self-reported exposure to fumes that included diesel fumes, while
    Hoar & Hoover (1985) used reports by next-of-kin of work with any
    diesel fuel or engines or work as a truck driver. Risch et al. (1988)
    obtained a history of exposure to exhaust by interview. Wynder et al.
    (1985) estimated the likelihood of exposure to diesel exhaust during a
    worker's lifetime on the basis of usual occupation, obtained at
    interview, and attempted to divide exposure into high and moderate. In
    a later study, self-reported exposure to diesel exhaust was recorded
    (Iyer et al., 1990). Silverman et al. (1983), Vineis & Magnani (1985),
    Silverman et al. (1986), Jensen et al. (1987), Iscovich et al. (1987),
    and Notani et al. (1993) based their analyses of the risk for bladder
    cancer on employment in an occupational category with a high
    likelihood of exposure to diesel exhaust, such as truck driver,
    railroad worker, transport worker, bus driver, and other related
    occupations. In the studies of Siemiatycki et al. (1988) and Steineck
    et al. (1990), occupational histories were reviewed by industrial
    hygienists to link occupational titles to possible exposure. Cordier
    et al. (1993) also used expert review of occupational histories to
    estimate exposure to diesel exhaust on the basis of job title, but
    additionally examined specific job titles. Furthermore, exposure was
    usually assumed to have occurred if the person had ever been employed
    in an occupation or had ever been exposed to diesel exhaust or other
    fumes. Silverman et al. (1983, 1986) and Jensen et al. (1987) used
    self-reported years of work as a truck driver. Self-reporting of
    exposure poses a potential problem of differential recall among cases
    and controls, as cases may recall exposure to diesel fumes or work in
    a diesel-exposed job more readily than controls. Steenland et al.
    (1987) used city directories listing occupation to obtain independent
    occupational histories.

         As the available occupational histories were crude, it was also
    difficult to assess latency or duration of exposure. Silverman et al.
    (1983) found that the relative risk (adjusted for smoking) was 5.5
    (95% CI = 1.8-17.3) for truck drivers with 10 or more years of work
    experience and 11.9 (95% CI = 2.3-61.1) for those with a history of
    driving vehicles with diesel engines. Truck drivers employed after
    1950, when diesel truck use became prevalent, had the highest risk. In
    a later, larger study (Silverman et al., 1986), men who usually worked
    as a truck driver or deliveryman had a relative risk of 1.5 (95% CI =
    1.1-2.0), adjusted for smoking. The risk increased with increasing
    years of work; however, truck drivers who had started work fewer than
    40 years before diagnosis, who would have driven primarily diesel
    engines, did not have an increased risk for bladder cancer. Hoar &
    Hoover (1985) also noted increased relative odds with increasing years
    of employment in jobs with reported exposure to diesel exhaust.

         An additional limitation of the 15 case-control studies is that
    results indicating an effect of presumed exposure to diesel exhaust or
    a specific occupation are based on relatively few cases and controls,
    even when conventional levels of statistical significance were
    achieved. The study of Silverman et al. (1986), in which exposure was
    defined as working as a truck driver or deliveryman, had the largest
    number of exposed individuals (99 cases and 123 controls). Four
    additional cohort studies not listed in the table (Howe et al., 1983;
    Rushton et al., 1983; Schenker et al., 1984; Boffetta et al., 1988),
    which were designed mainly to examine lung cancer and exposure to
    diesel exhaust, also had too few cases of bladder cancer on which to
    base meaningful conclusions. One cohort study of Danish bus drivers
    (Netterstrom, 1988) showed an elevated SMR of 153, which was of
    borderline statistical significance (95% CI = 91-217), but this result
    was based on 13 cases.

         Misclassification of exposure to diesel exhaust would hide any
    increase in risk for bladder cancer, since many unexposed individuals
    would be included with those classified as exposed. Diesel exhaust is
    responsible for most of the respirable particles in mixed vehicle
    exhausts, and exposure to respirable particles with adsorbed PAHs is
    presumed to result ultimately in an increased risk for bladder cancer
    due to metabolic transformation and urinary excretion of carcinogens.
    The risk for bladder cancer is also increased in cigarette smokers and
    in workers exposed to dyes containing aromatic amines (Ruder et al.,
    1990). Coffee has been suggested to be a risk factor for bladder
    cancer but has not been clearly implicated (Howe et al., 1980). The
    classification of exposure to diesel exhaust is sufficiently crude in
    the papers summarized in Table 42, however, that even when smoking was
    considered, the effect of other, unmeasured confounding exposures
    cannot be excluded. The actual contribution of diesel exhaust to an
    increased risk of bladder cancer is uncertain, even when an effect of
    presumed exposure was noted (Silverman et al., 1983; Hoar & Hoover,
    1985; Silverman et al., 1986; Iscovich et al., 1987; Steenland et al.,
    1987). Although the literature suggests that truck drivers have an
    increased risk for bladder cancer, the specific exposure responsible
    has not been determined.

    B9.  EFFECTS ON OTHER ORGANISMS IN THE LABORATORY AND THE FIELD

         Virtually no data are available relating specifically to the
    effects of diesel fuel exhaust emissions.

         The growth and photosynthesis of liquid cultures of the green
    alga  Chlamydomonas reinhardtii were examined after incubation with
    iso-octane extracts of diesel particulate exhaust containing 51
    compounds (mainly polynuclear aromatic ketones and pure PAHs)
    identified by gas chromatographic-mass spectrometric analysis. At
    concentrations up to 0.125% by volume, dose-dependent growth
    retardation of the algae was observed. Higher concentrations (no

    details given) caused death. The algae adapted to sublethal
    concentrations of the extract over a period of several days;
    thereafter, toxic extracts at concentrations up to 2.5% by volume
    affected neither growth nor photosynthesis. It was noted that the
    amounts of certain components decreased during incubation, suggesting
    uptake into the cells (Liebe & Fock, 1992).

    B10.  EVALUATION OF HUMAN HEALTH RISKS AND EFFECTS ON THE ENVIRONMENT

         Hundreds of chemical compounds are released during the combustion
    of diesel fuel. The characteristics and amount of exhaust depend on
    the type of diesel engine, its operating conditions and adjustment,
    and the composition of the fuel. The following evaluation focuses
    mainly on the risks to human health associated with exposure to diesel
    particulate matter, since the large amount of soot particles emitted
    is the main characteristic of diesel exhaust.

         In general, the risk assessment paradigm proposed by the National
    Academy of Sciences of the United States (US National Reserach
    Council, 1983) will be followed. The four steps in this process are:
    hazard identification, dose-response assessment, exposure assessment,
    and risk characterization. The outcome of these individual steps is
    critical for taking risk management decisions, including setting
    exposure standards that have consequences for public health, and for
    social, economic, and political issues.

         The epidemiological studies in humans and the studies in
    experimental animals considered useful for human health risk
    assessment are discussed in this section. Both carcinogenic and other
    effects are considered, but a threshold is assumed to exist only for
    non-cancer effects. An IPCS method for deriving health-based guidance
    exposure limits provides further details (IPCS, 1993).

    B10.1  Exposure of the general population

         The exposure of the general population to diesel exhaust varies
    with proximity to diesel-powered vehicles. For example, higher levels
    of diesel exhaust constituents are present in busy streets and parking
    areas than in rural areas. The concentrations of the main gaseous
    components resulting from diesel exhaust can be estimated from
    specific emission factors, the contribution of diesel vehicles to
    total traffic, and atmospheric dispersion models. As assumptions are
    necessary to attribute the emissions of the gaseous components to
    diesel sources, the validity of such estimates is considered to be
    limited.

         On a global basis, the contribution of diesel exhaust to the
    total emissions of particulate matter varies widely, depending on the
    percentage of diesel vehicles in the total volume of traffic,
    maintenance of individual engines, fuel quality, and emission control
    techniques. Such factors also lead to wide local variations in the
    concentrations to which humans are exposed.

         Measurements and calculations of ambient concentrations of diesel
    particulates near roads provide a daily average range of 8-42 µg/m3.
    The estimated annual average concentrations in Germany are 5-10 µg/m3
    in urban areas and 1.5 µg/m3 in rural areas, and those in the United
    States are 1-2 µg/m3 in urban areas and 0.6-1 µg/m3 in rural
    areas. As the proportion of diesel vehicles is smaller in the United
    States than in Europe, the concentrations in 1990, estimated from
    national averages, were 2.0 µg/m3 in an urban area and 1.1 µg/m3
    in a rural area, with predictions of 1.2 g/m3 for urban areas and
    0.6 µg/m3 for rural areas in 1995.

         Diesel particulates form part of the fine particle range
    (< 2.5 µm diameter). The average concentration of this fraction of
    suspended particulate matter as a whole, measured in six cities in the
    United States during 1979-85, was 11-30 µg/m3 or, in terms of
    inhalable particles (<10-15 µm diameter; equivalent to particulate
    matter with a diameter < 10 µm [PM10]), 18-46 µg/m3 (Dockery et
    al., 1993). In these locations, diesel particulates probably
    represented less than 10% of the suspended particulates measured as
    PM10; considerably higher proportions were found in London and other
    large cities in the United Kingdom on the basis of data on emissions
    (Quality of Urban Air Review Group, 1993), but few data are available
    that are related directly to ambient concentrations of the diesel
    component. The one experimental study that has been reported (Horvath
    et al., 1985), carried out in Vienna, Austria, using chemically
    labelled fuel, yielded a value for the background concentration of
    diesel particulates within the city of 11 µg/m3.

    B10.2  Occupational exposure

         Occupational exposure to airborne diesel particulate matter has
    been well determined for only two occupational groups, railroad
    workers (Howe et al., 1983; Garshick et al., 1987, 1988) and workers
    in the trucking industry (stevedores, local and road truck drivers,
    and mechanics) (Emmelin et al., 1993). The levels of exposure during
    work shifts are about 51-192 µg/m3 for railroad workers exposed to
    diesel locomotive exhaust, roughly 25 µg/m3 for truck drivers, and
    156 µg/m3 for stevedores (assuming that elemental carbon represented
    20% of respirable particulate). These levels would be expected to vary
    with the amount of ventilation available. Exposure to diesel exhaust
    can also vary over a worker's lifetime, even within the same job. In

    the railroad industry, exposure to diesel exhaust was greater in the
    years after the introduction of diesel equipment. Reports of working
    conditions in the 1950s and 1960s in diesel repair shops cited 'smoky'
    atmospheres that were probably related to use of diesel equipment and
    poor ventilation (Woskie et al., 1988b). No direct information is
    available that allows reconstruction of accurate levels of past
    exposure.

         Reports of concentrations of diesel exhaust in other industries
    are not accurate because of the presence of other dusts, although
    attempts have been made to measure the contribution of diesel exhaust
    (Daniel, 1984; Lehmann et al., 1990). Even for American railroad
    workers (Woskie et al., 1988a), the contribution of other dusts could
    not be determined, although cigarette smoke particulate was taken into
    consideration. Consequently, the concentrations of diesel exhaust
    reported in some jobs on railroads where other dusts (such as sand)
    were present might have been lower.

    B10.3  Non-neoplastic effects

    B10.3.1  Hazard identification

    B10.3.1.1  Humans

         Diesel exhaust contains gases that irritate the nose, throat, and
    eyes. It also contains particles which, together with the gases, can
    cause airway irritation. Diesel exhaust may be recognized by its
    characteristic odour. After acute exposure, it may induce mucous
    membrane irritation, headache, and light-headedness. When volunteers
    were exposed to diesel exhaust with an estimated particle
    concentration of 100 µg/m3 for 1 h, several indicators of an
    inflammatory response were observed in bronchioalveolar lavage fluid
    (see section B8.2.1). In a well-conducted study on diesel-bus garage
    workers, those with the highest exposures to respirable particulate
    (> 0.31 mg/m3) reported significantly more cough, itchy or burning
    eyes, headache, and wheeze, after adjustment for age and smoking, than
    people with lower exposure. The contribution of diesel exhaust to the
    respirable particulate is not known. Exposure to nitrogen dioxide at
    > 0.3 ppm also resulted in more eye irritation and wheeze. Studies of
    exposure indicate, however, variability in the ability of individuals
    to detect the odour and the irritating effects of diesel exhaust.
    Several cases of persistent asthma and asthma attacks have been
    reported after acute exposure.

         No consistent change in pulmonary function has been reported over
    a work shift, although in one study of ferry workers, a decline in FVC
    and FEV1 was noted. Specific measurements of diesel exhaust are
    lacking in these studies, although total dust and respirable dust were
    sometimes measured. Cases of persistent asthma have been reported in
    railroad workers acutely exposed to apparently high levels of whole
    exhaust, but the mechanism is unknown.

         Studies of the possible chronic effects of exposure on
    respiratory parameters are limited to occupational cohorts who are
    still at work and thus have relatively short durations of exposure.
    Such cohorts are likely to be healthier than older cohorts, which
    include retirees who had longer exposure. Although excess respiratory
    symptoms and reduced pulmonary function have been reported in some
    studies, it is not clear whether these are long-term effects of
    exposure.

    B10.3.1.2  Experimental animals

         Long-term studies in laboratory animals are described in section
    B7.3. Many end-points were measured in the respiratory system of
    several species of rodent after two or more years of exposure and at
    several intermediate times. Adverse effects were seen after a
    sufficient particle load had accumulated in the lung; and the
    appearance of the effects is determined by both the concentration and
    the duration of exposure. The results of these studies illustrate the
    response of the respiratory tract in terms of biochemistry,
    histopathology, cytology, pulmonary function, inflammation, and
    clearance of particles. Early events in the pathogenesis of this
    response include phagocytosis of inhaled particles. As the lung burden
    increases, particle-filled macrophages are observed in the alveoli. At
    lung burdens associated with particle overload, migration of these
    macrophages out of the lung appears to be inhibited, and an early
    event in the development of lung damage is the accumulation and
    aggregation of clusters of particle-filled macrophages. Degeneration
    of macrophages is observed, and the numbers of macrophages and
    neutrophils in the alveolar and interstitial spaces increase.

         Damage to the alveolar epithelium in areas surrounding these
    macrophage clusters includes cell damage and proliferation.
    Biochemical changes are seen in lavage fluid and lung tissue. The
    alterations to particle-filled macrophages result in decreased
    function and impaired particle clearance, the latter leading to a more
    rapid increase in the lung burden of particles. Late effects include
    inflammatory and proliferative responses, leading to fibrotic
    responses. The sequence of these events has not been distinguished
    clearly in experimental work; they are likely to be interrelated and
    to occur, to some extent, concurrently.

         The possible mechanism of action of diesel exhaust (discussed in
    section B7.9) includes a role of the gas phase, as opposed to the
    particle phase, in causing lung damage. The non-carcinogenic toxicity
    of diesel exhaust is considered to be due to the particle content,
    since no effects are seen in rodents exposed to diesel exhaust that
    has been filtered to remove particulate matter. Another issue in the

    dosimetry of diesel exhaust is the role of the insoluble carbon core
    in relation to that of the adsorbed organic material. As discussed in
    section B7.9, studies that show similarities in non-carcinogenic
    responses to exposure to diesel exhaust and to carbon particles
    without an organic component indicate a role of long-term retention of
    the insoluble carbon core.

    B10.3.2  Dose-response assessment

    B10.3.2.1  Epidemiological studies

         Variability in the reporting of odour and irritation indicates
    that a subsegment of the population is sensitive to diesel exhaust;
    however, the proportion of the population with this characteristic
    cannot be estimated. Chronic effects have not been seen in
    occupationally exposed people, but excess respiratory symptoms have
    been reported after acute exposure. It is not possible to determine
    the dose of diesel exhaust (on the basis of exposure to gas or
    particles) that produces these symptoms.

    B10.3.2.2  Studies in experimental animals

         In order to select the pivotal study for establishing the NOAEL
    or uncertainty factor, the most sensitive, relevant studies must be
    identified. All of the long-term studies focus on effects on the
    respiratory tract, which are clearly those most relevant. Studies of
    relatively low concentrations are listed in Table 43, which gives the
    target and actual exposure concentrations, the equivalent continuous
    exposure of the animal, the equivalent continuous human exposure, and
    the effect level. The equivalent continuous exposure of animals is
    calculated by multiplying the actual exposure concentration by the
    (number of hours of daily exposure/24) and by the (number of days per
    week of exposure/7). This value is necessary for comparing exposure
    concentrations among studies with different protocols, i.e. different
    numbers of hours of exposure per day and/or days per week.

         The equivalent continuous human exposure is calculated from the
    particle deposition-clearance model of Yu & Yoon (1990), discussed in
    section B6.2. This model is based on studies of rats since only those
    studies provide data on the exposure-response relationship for
    inhibition of particle clearance.

         The model of Yu & Yoon (1990) is a sophisticated approach to
    defining the relationship between the inhaled concentration and the
    dose in lung tissue. It combines detailed models of rat and human lung
    structure, aerodynamic models of the air flow in the airways, particle
    deposition dynamics, and information on clearance of particles from
    rat and human lung. The model is specific to diesel particles because
    the characteristics of the particles used in the model are derived

    from studies of diesel particles and because the information on lung
    clearance is based on studies of diesel exhaust in rats. The model has
    been adapted for diesel exhaust particles by allowing evaluation of
    the insoluble carbon core and of the tightly and loosely bound organic
    compounds separately. Values for deposition and clearance are combined
    with input from experimental studies to calculate the burden of
    particles in the lungs of the animals at the end of the study. The
    continuous concentration that would result in the same lung burden in
    humans is then calculated, as the mass of particles per unit of
    alveolar surface area or unit of lung weight. This lung burden is
    assumed to result in a similar effect in humans and in rats. With
    these considerations, the equivalent human concentrations are
    predicted on the basis of the assumption that the retained particle
    mass per unit alveolar surface area is the appropriate dose measure
    for extrapolation between species. The effect level used in the risk
    assessment is based on an evaluation of the adversity of the effect.

         For the purposes of risk characterization, the NOAEL or, if that
    is not available, the LOAEL for the critical effect is related to
    exposure. The critical effect is the first adverse effect that appears
    when the critical concentration is reached at the target site. A
    decision about whether an effect is critical is a matter of expert
    judgement. An adverse effect is defined as a change in morphology,
    physiology, growth, development, or the life span of an organism which
    results in impairment of functional capacity or of capacity to
    compensate for additional stress or an increase in susceptibility to
    the harmful effects of other environmental influences (International
    Union for Pure and Applied Chemistry, 1993).

         The NOAEL is defined as the greatest concentration or amount of a
    substance, found by experiment or observation, that causes no
    alteration in morphology, functional capacity, growth, development, or
    the life span of target organisms that are distinguishable from those
    observed in normal (control) organisms of the same species and strain
    under the same defined conditions of exposure (International Union for
    Pure and Applied Chemistry, 1993).

         The LOAEL is defined as the lowest concentration or amount of a
    substance, found by experiment or observation, that causes an adverse
    alteration in morphology, functional capacity, growth, development, or
    the life span of a target organism distinguishable from those of
    normal (control) organisms of the same species and strain under
    defined conditions of exposure (International Union for Pure and
    Applied Chemistry, 1993).


        Table 43.  Exposure of experimental animals, equivalent continuous human exposure, and effect levels of long-term inhalation in studies
               of rats described in section B7.3
                                                                                                                                                

    Length of       End-points       Target         Actual         Equivalent           Equivalent            Effect       Study
    exposure        evaluateda       exposure       exposure       continuous           continuous            levelb
                                     (mg/m3)        (mg/m3)        animal exposure      human exposure
                                                                   (mg/m3)              (mg/m3)
                                                                                                                                                

    7 h/d,          HP, LL,          0.35           0.353          0.0735               0.042                 N            Mauderly
    5 d/week,       LTB, TC,         3.5            3.47           0.723                0.360                 A            et al. (1987)
    30 months       LB               7              7.08           1.47                 0.582                 A

    Light-duty,     HP               0.1            0.11           0.063                0.038                 N            Ishinishi
    16 h/d,                          0.4            0.41           0.23                 0.139                 N            et al. (1986,
    6 d/week,                        1              1.18c          0.67                 0.359                 A            1988)
    30 months                        2              2.32           1.3                  0.571                 A

    Heavy-duty,     HP               0.4            0.46           0.26                 0.155                 N            Ishinishi
    16 h/d,                          1              0.96           0.55                 0.303                 A            et al. (1986,
    6 d/week,                        2              1.84           1.05                 0.493                 A            1988)
    30 months                        4              3.72           2.13                 0.911                 A

    110 h/week,     HP               0.25           0.258          0.17                 NA                    N            Barnhart
    6 months                         0.75           0.796          0.52                 NA                    A            et al. (1981)
                                     1.5            1.53           1.0                  NA                    A

    7 h/d,          HP, SB           2              1.95           0.57                 0.336                 A            Lewis et al.
    5 d/week,                                                                                                              (1989)
    24 months
                                                                                                                                                

    Table 43 (contd)
                                                                                                                                                

    Length of       End-points       Target         Actual         Equivalent           Equivalent            Effect       Study
    exposure        evaluateda       exposure       exposure       continuous           continuous            levelb
                                     (mg/m3)        (mg/m3)        animal exposure      human exposure
                                                                   (mg/m3)              (mg/m3)
                                                                                                                                                

    19 h/d,         HP, LL,          4              4.24           2.40                 1.02                  A            Heinrich et
    5 d/week,       TC, LB                                                                                                 al. (1986a)
    32 months

    18 h/d,         HP, LL,          0.8            0.84           0.45                 0.23                  A            Heinrich et
    5 d/week,       TC, LB           2.5            2.50           1.34                 0.59                  A            al. (1995)
    30 months                        7.5            6.98           3.74                 1.56                  A
                                                                                                                                                

    NA, not available (guinea-pigs were used)
    a  HP, histopathological examination; LL, lung lavage fluid; LTB, lung tissue biochemistry; TC, tracer clearance; SB, serum biochemistry;
       LB, lung burden
    b  N, no effect; A, adverse effect
    c  According to Suzuki et al. (1990), 1.08 mg/m3
             The selected NOAEL is divided by uncertainty factors to account
    for data gaps. For these data, uncertainty exists in two areas:
    extrapolation from animals to humans and accounting for sensitive
    subpopulations. The factor for accounting for sensitive subpopulations
    is a default value of 10, as it is considered that there are no data
    suggesting a different factor. An additional default value of 10 is
    usually applied for interspecies extrapolation, which is nominally
    considered to consist of 100.4 (2.5) for toxicodynamics and 100.6
    (4.0) for toxicokinetics (IPCS, 1994).

         The LOAELs and NOAELs are selected on the basis of the following
    reasoning. The LOAEL is the lowest level at which adverse effects are
    seen in studies of adequate quality. The NOAEL, the highest level at
    which no effect is seen in the available studies, must be lower than
    the selected LOAEL. The LOAELs and NOAELs are identified by a
    comparison with the equivalent continuous human exposure level. The
    LOAELs in these studies are strikingly similar, with values of 0.360,
    0.359, 0.303, 0.336, and 0.23 mg/m3 in studies from the Inhalation
    Toxicology Research Institute (United States; Mauderly et al., 1987),
    the Health Effects Research Program on light- and heavy-duty diesel
    engine exhausts (Japan; Ishinishi et al., 1986, 1988), the National
    Institute for Occupational Safety and Health (United States; Lewis et
    al., 1989), and the Fraunhofer Institute for Toxicology and Aerosol
    Research (Germany; Heinrich et al., 1986a, 1995), respectively (see
    Table 43). The consistency of these levels of effect, in view of the
    diversity of end-points represented, gives a high level of confidence
    in the end-points.

         The study at the Inhalation Toxicology Research Institute was
    selected as the most representative because it covers the greatest
    variety of end-points, including histopathology, bronchiolar lavage
    biochemistry and cytology, lung tissue biochemistry, and particle
    clearance; LOAELs were observed for all of these end-points. The
    studies of the Health Effects Research Program showed a similar LOAEL
    but only for histological changes in the lung. The value for
    equivalent continuous human exposure is not available from the study
    carried out by the General Motors Corporation (Barnhart et al., 1981)
    because the main results were for guinea-pigs; although rats were
    included in the study, detailed results were not presented. The
    appropriate NOAEL in these studies is not selected on the basis of the
    LOAEL. As close dosing intervals were used in the studies of of the
    Health Effects Research Program, the NOAEL was selected on the basis
    of studies of light-duty diesel engines, as 0.139 mg/m3.

    B10.3.3  Exposure assessment

         Exposure to diesel exhaust is discussed in sections B10.1 and  
    B10.2.

    B10.3.4  Risk characterization

    B10.3.4.1  Humans

         A quantitative assessment of the risk for humans of
    non-carcinogenic effects of exposure to diesel exhaust cannot be made
    on the basis of studies in humans. A substantial volume of literature
    shows an association between acute exposure to fine particulates and
    morbidity and mortality in the general population. It is somewhat
    uncertain whether there is a direct causal link and, if so, whether it
    is related to a specific component of the suspended particulates.
    Attention has been focused on fractions of suspended particulate
    matter, currently measured as PM10 (diameter < 10 µm) or PM2.5
    (diameter < 2.5 µm), comprised mainly of diesel particulates in the
    submicrometre range, which represent an increasing proportion in some
    countries. There is also evidence that exposure to particulates is
    involved in respiratory symptoms and in mortality from certain chronic
    respiratory conditions.

    B10.3.4.2  Experimental animals

         Three approaches were used in assessing the risk for
    non-carcinogenic effects of exposure to diesel exhaust, on the basis of
    exposure (dose)-response assessment in experimental studies (see
    section B10.3.2.2), and for deriving guidance values for exposure.
    Approaches 1 and 2 are based on the NOAEL from studies in rats. The
    difference between the two is that approach 1 involves a dosimetric
    extrapolation model (Yu & Yoon, 1990) for converting the actual
    concentration to which animals were exposed to an equivalent
    continuous human exposure, thereby reducing the uncertainty in
    interspecies extrapolation.

         In approach 2, no dosimetric conversion is performed but the
    usual uncertainty factors for interspecies and intraspecies
    extrapolation are applied. In contrast to these NOAEL-defined
    approaches, approach 3 is based on the principle of the 'benchmark
    dose', in which a concentration of exposure to diesel exhaust is
    derived from dose-response relationships observed in rats.

         The model of Yu & Yoon (1990) takes into consideration the
    specific characteristics of particle deposition and retention in rat
    and human lung. Application of the model to studies of inhalation in
    experimental animals is shown schematically in Figure 4. The deposited
    and retained doses in the lung are calculated from the concentrations
    in rats and expressed per unit lung weight or per unit alveolar
    surface area. Under the assumption that the retained dose per lung
    will lead to the same effect in rodent and human lung, a human
    equivalent concentration can be calculated (as shown in Table 43), to
    obtain an equivalent continuous human exposure. The equivalent

    continuous exposure of animals shown in Table 43 was derived either by
    calculating the continuous exposure at which the lung burden attained
    in a study by inhalation (if measured) would have been reached, or by
    calculating the lung burden in rats at the end of exposure on the
    basis of the actual exposure parameters (also shown in Table 43) and
    then calculating as above. The correlation between continuous exposure
    in animals and humans in the model of Yu & Yoon (1990), assuming that
    overload induces prolonged lung clearance, is shown in Figure 5. Use
    of this approach, rather than simple adjustment of discontinuous to
    continuous exposure by the  c x  t constant used in other assessments,
    should reduce the uncertainty associated with the transformation.

          Approach 1. The NOAEL of 0.41 mg/m3 from the study in rats
    exposed by inhalation to light-duty engine exhaust (Ishinishi et al.,
    1986, 1988; Table 43) is converted to an equivalent continuous
    exposure of 0.23 mg/m3 in rats and then to an equivalent continuous
    exposure of 0.139 mg/m3 , assumed to be the NOAEL in humans.
    Application of a sophisticated dosimetric model decreases the
    uncertainty in interspecies extrapolation from 10 to 100.4 (IPCS,
    1994). Application of the usual uncertainty factor of 10 for
    intraspecies differences results in a total uncertainty factor of
    10 × 100.4 = 25.

         The guidance value derived from this approach is

               0.139 mg/m3  = 5.6 µg/m3.
              --------------- 
                   25

    No additional correction factor is required, resulting in a guidance
    value (for the general population) of 5.6 µg/m3.

         The inherent uncertainties in the dosimetric model are due to the
    assumptions that must be made. A principal assumption is necessary to
    estimate inhibition of clearance in humans, since data on this aspect
    do not exist. It is assumed that clearance in humans is inhibited at
    the same lung burden (mass per alveolar surface area) as in rats. The
    other principal assumption is that the correct dose measure for lung
    damage is mass of particle core per alveolar surface area. Since the
    damage is localized to specific areas, another measure may be more
    appropriate.

    FIGURE 4

          Approach 2. In this approach, the equivalent continuous animal
    exposure based on the NOAEL of 0.41 in rats (light-duty diesel
    exhaust; Ishinishi et al., 1986, 1988; Table 47) is used to derive the
    guideline value. Because the dosimetric model is not used, the default
    value of 10 is applied for interspecies uncertainty and an uncertainty
    factor of 10 is added for intraspecies differences. The guidance value
    obtained with this approach is

                        0.23 mg/m3
                      ------------- = 2.3 µg/m3
                           100

          Approach 3. An alternative to using the NOAEL is to derive a
    'benchmark dose', as described by Crump (1984). In studies by
    inhalation, the concentration rather than the dose is considered to be
    more precise, and the 'benchmark concentration' is used. Like the
    benchmark dose, this term covers the entire exposure-response
    relationship in a given study rather than relying on only one data
    point representing the NOAEL or LOAEL. This approach reduces the
    uncertainty inherent in defining an NOAEL or LOAEL by considering the
    upper 95% confidence interval of the full exposure-response curve from
    a study in experimental animals, in which the lower 95% confidence
    limit of a concentration corresponds to a 1, 5, or 10% increase in
    response, defined as the percentage of animals responding in a
    specific group. This exposure concentration is then the benchmark
    concentration of the study. It is converted to the human benchmark
    concentration on the basis of differences in pulmonary dosimetry
    between the two species, which additionally reduces the use of
    uncertainty factors. Because the full range of experimental results
    from a specific study is used, the benchmark concentration approach
    reduces statistical uncertainty.

         In principle, this approach involves three steps. The first is
    selection of the appropriate experimental study and end-point for
    establishing an exposure-response curve. The second is calculation of
    the benchmark concentration for the animals from a mathematical
    description of the exposure-response curve and determination of the
    95% confidence interval. Thirdly, the human benchmark concentration is
    calculated; for exposure by inhalation, the exposure of the animals is
    extrapolated dosimetrically to the human situation, with application
    of uncertainty factors the size of which is determined as discussed
    above. These factors generally consist of one for interspecies
    extrapolation and one for sensitive subpopulations. The first two
    steps of the benchmark concentration approach are shown schematically
    in Figure 6.

    FIGURE 5

    FIGURE 6

         Two sets of data were used to derive the benchmark concentration
    for diesel exhaust. The first was that of Ishinishi et al. (1988) from
    a study in which rats were exposed to exhaust from a light-duty diesel
    engine for two years at four concentrations. Hyperplastic lesions,
    shown histopathologically to be a sensitive indicator of lung damage,
    were used to establish a well-described exposure-response
    relationship. The second data set was that of Creutzenberg et al.
    (1990) for female Wistar rats exposed for 96 h per week for a total of
    78 weeks to three concentrations of diesel exhaust. The most sensitive
    end-points for exposure-related non-neoplastic changes were lung
    clearance of particles and the occurrence of polymorphonuclear
    neutrophils (PMN) in lung lavage fluid as indicators of inflammation.
    Since these measurements are laborious, however, they were performed
    only in a subset of the exposed animals, so that only six to eight
    rats were studied per group.

         The responses in the study of Ishinishi et al. (1988) were
    expressed in terms of individual animals affected per total number of
    exposed animals in each group, whereas the data of Creutzenberg et al.
    (1990) were reported as group mean values plus or minus the standard
    deviation, which is less useful for the benchmark concentration
    approach. Individual responses in the latter study were, however,
    provided by Bellmann (personal communication), so that the data from
    this study could also be expressed in terms of percentage of animals
    with impaired lung clearance. Impairment of lung function was
    considered to be significant when the calculated pulmonary retention
    half-time of the administered particles was at least 3.5 times greater
    than their normal average half-time in the lungs of control animals.

         Evaluation of the end-point chronic inflammation (PMN in lung
    lavage fluid) proved to be more difficult, since all of the rats at
    the lowest exposure level (0.8 mg/m3) in the study of Creutzenberg
    et al. (1990) had increased PMN levels (100% response). Lung lavage
    was performed at 22 and 24 months of exposure, and since the responses
    at these two times were similar, the data were combined to increase
    the numbers of animals per group to 11-14. The individual responses
    (percentage of PMN among the total number of cells in the lavage
    fluid) were used to derive the curve and the 95% confidence interval.
    An excess of PMN of up to 3% of the total number of cells over the
    background level was used to define the benchmark concentration. The
    highest value observed in the control group was 2.75% PMN after
    22 months of exposure.

         The three data sets and the results of the probability function,
    using a Weibull model to calculate the benchmark concentration, are
    given in Tables 44, 45, and 46 and are illustrated in Figures 7 and 8.

    Table 44.  Benchmark concentration for rats after long-term exposure
               to diesel exhaust by inhalation in study by Ishinishi et al.
               (1988); end-point: hyperplastic lesions of the lung
                                                                        

    Concentration (mg/m3)     Rats that responded
    (equivalent continuous                                              
    exposure of rats)         Actal/observed    Predicted    Percent
                              numbers           number
                                                                        

    0                         4/125               4.6           3.2
    0.063                     4/125               4.6          3.25
    0.23                      6/125               4.7           4.8
    0.67                      12/123             12.1          9.76
    1.3                       87/124             87.0          70.2
                                                                        


    Probability function [p(conc)] - A0]/(1 - A0) = [1 - exp(-A1 *
    Conc)K]

                   A0 = 3.7196 × 10-2
                   A1 = 3.740
                   K  = 4.3504

    Lower confidence limit of the benchmark concentration (BC) for a given
    risk:

                                              

                             Risk     Rat BC
                                      (mg/m3)
                                              

                             0.1      0.634
                             0.05     0.511
                             0.01     0.313
                                              

    The respective benchmark concentrations from the data sets for rats
    are 0.634, 0.119, and 0.090 mg/m3, corresponding to a 10% response
    (the lower 95% confidence limit on the exposure concentration for
    hyperplastic lung injury and impaired lung clearance) or a 3% excess
    of PMN in lavage fluid (the lower 95% confidence limit on the exposure
    concentration for chronic alveolar inflammation).

         The resulting human benchmark concentrations for the three
    end-points were calculated from the model of Yu & Yoon (1990),
    applying uncertainty factors of 10 to account for sensitive
    subpopulations (human intraspecies differences) and 100.4 (2.5) for
    potentially different toxicodynamics (see above), to be: 14 µg/m3
    for hyperplastic lung lesions, 3 µg/m3 for impaired lung clearance,
    and 2 µg/m3 for chronic alveolar inflammation. Although the
    mathematical description of the exposure-response relationships in
    rats is very good for all three sets of data, they should be viewed as
    hypothetical since they are based on dosimetric conversions from rats
    to humans and on the application of uncertainty factors.

    Table 45.  Benchmark concentration for  rats after long-term exposure
               to diesel exhaust by inhalation in study by Creutzenberg
               et al. (1990); end-point: impaired lung clearance
                                                                        

    Concentration (mg/m3)     Rats that responded
    (equivalent continuous                                              
    exposure of rats)         Actal/observed   Predicted    Percent
    numbers                   number
                                                                        

    0                         0/6              0                0
    0.45                      2/8              1.7             25
    1.34                      3/6              3.0             50
    3.74                      5/6              5.2           83.3
                                                                        

    Probability function
              [p(conc)] - A0]/(1 - A0) = [1 - exp(-A1 * Conc - D0)A2]

                   A0 = 0.0
                   A1 = 0.522
                   A2 = 1.00
                   D0 = 0.0  (threshold)

    Lower confidence limit of the benchmark concentration (BC) for a given
    risk:

                                                

                                Risk     Rat BC
                                         (mg/m3)
                                                

                                0.1      0.119
                                0.05     0.058
                                0.01     0.011
                                                

         The non-cancer guidance values and the benchmark concentrations
    derived from these approaches are summarized in Table 47.

         More sensitive end-points with respect to the adverse effects of
    diesel exhaust on the lower respiratory tract are impaired clearance
    in the deep lung and chronic alveolar inflammation, rather than
    hyperplastic lung lesions. Chronic alveolar inflammation was also a
    significant finding in the long-term study of rats exposed to diesel
    exhaust by Henderson et al. (1988), who found a significantly
    increased percentage of PMN in lavage fluid from the group with the
    lowest exposure (0.35 mg/m3) after two years. For all three types of
    end-points, however, there may be a threshold below which no change is
    to be expected. The threshold concept is not included in the
    mathematical model, and it would be difficult to do so at this point
    in the absence of specific experimental data and data on variation
    between human subjects.

         These estimates are for the effects of long-term exposures, and
    the concentrations are expressed as annual means over a lifetime. They
    could apply to long-term effects on health, such as the association
    between mortality from certain chronic conditions and exposure to
    particulates in air pollution seen in a study in six cities in the
    United States (Dockery et al., 1993). Their relevance to human
    experience must also be considered in the context of particulate air
    pollutants in general, however, and they have no direct bearing on the
    acute effects of exposure to particulates, for which guidance values
    are required in terms of 24-h mean concentrations.

    B10.4  Neoplastic effects

    B10.4.1  Hazard identification

    B10.4.1.1  Lung cancer: occupational exposure

         In the 1970s and 1980s, the recognition that diesel exhaust
    contains respirable particles and that known carcinogenic substances
    are adsorbed on the surface of these particles led to the hypothesis
    that inhalation of diesel exhaust could result in lung cancer in
    humans. As lung cancer develops slowly, over many years, studies of
    individuals with long, well-defined exposure and follow-up (> 20
    years) were considered to be the most informative. Four studies of
    occupationally exposed individuals meet these criteria (Garshick et
    al., 1987, 1988; Gustavsson et al., 1990; Emmelin et al., 1993; see
    Table 43).

    Table 46.  Benchmark concentration for rats after long-term exposure to
               diesel exhaust by inhalation in study by Creutzenberg et al.
               (1990); end-point: chronic alveolar inflammation
                                                                        

    Concentration (mg/m3)     PMN in lung lavage fluid (%)    No. rats
    (equivalent  continuous                                   responding
    exposure of rats          Actual/observed   Predicted     per no.
                                                              exposed
                                                                        

    0                         0.7               0             0/11 
    0.45                      11.1              11.7          13/13
    1.34                      29.9              27.6          14/14 
    3.74                      50.8              51.9          14/14
                                                                        


    Dose-response function: F(dose) = Q(0) × (Dose - (Dose - D0)Q(2)
    Dose = log(1 + d) (d = rat exposure concentration)

                   Q(0) =  0.6.53
                   Q(1) =  77.010
                   Q(2) =  1.056
                   D0   =  0.0 (threshold)

    Lower confidence limit of the benchmark concentration (BC) for a given
    excess polymorphonuclear neutrophil (PMN) response (% of total lavaged
    cells):

                                                    
 
                           PMN in excess    Rat BC
                           of control (%)   (mg/m3)
                                                    

                           3                0.090
                           2                0.059
                           1                0.029
                                                    

         Transition from steam to diesel locomotives occurred in the
    American railroad industry during the 1950s. In a case-control study
    of American railroad workers (Garshick et al., 1987), exposure was
    estimated on the basis of yearly job (exposed or unexposed to diesel
    exhaust) from 1959 to death or retirement; deaths were identified for
    12 months in 1981-82. In an additional study of these workers
    (Garshick et al., 1988), a retrospective cohort was defined on the
    basis of work in a job with exposure to diesel exhaust in 1959, and

    FIGURE 7

    FIGURE 8

    deaths were recorded for 1959-80. Although the previous exposure
    levels of the railroad workers were not available, the classification
    of workers into groups on the basis of exposure was validated by an
    industrial hygienist and a review of work practices. Air sampling was
    conducted to establish current exposure levels in various jobs
    (Hammond et al., 1988; Woskie et al., 1988a,b).

         Gustavsson et al. (1990) studied lung cancer among bus garage
    workers in Stockholm, Sweden, where all buses with internal combustion
    engines have been diesel-fuelled since 1945. A cohort was established
    of people who had worked in bus garages for at least six months
    between 1945 and 1970, and mortality was determined for 1952-86. A
    nested case-control study was performed within this cohort. Although
    actual exposure was not measured, relative exposure to diesel exhaust
    was graded on the basis of a scale established by an industrial
    hygienist in a review of work practices, bus engine operation, and
    shop ventilation.

    Table 47.  Summary of non-cancer guidance values and benchmark
               concentrations
                                                                        

    Approach                                           Guidance value
                                                       or benchmark
                                                       concentration
                                                       (µg/m3)
                                                                        

    NOAEL with dosimetric conversion from rats to humans    5.6a

    NOAEL without dosimetric conversion from rats to        2.3
    humans

    Benchmark concentration with dosimetric conversion
    from rats to humans
    Chronic alveolar inflammation                           2a
    Impaired lung clearance                                 3a
    Hyperplastic lesions                                    14a

    Benchmark concentration without dosimetric
    conversion from rats to humans

    Chronic alveolar inflammation                           0.9
    Impaired lung clearance                                 1.2
    Hyperplastic lesions                                    6.3
                                                                        

    NOAEL, no-observed-adverse-effect level
    Normalized for lung surface area in rats and humans; after
    normalization for lung weight, the benchmark concentration increases
    by a factor of 4.

         Emmelin et al. (1993) studied Swedish stevedores in 15 ports
    where diesel equipment had been introduced between 1957 and 1963; they
    also performed a nested case-control study on cases of lung cancer
    identified between 1960 and 1982. Exposure to diesel exhaust was
    estimated on the basis of three indices derived from estimated diesel
    fuel consumption and years of work after the introduction of diesel
    equipment in each port.

         All four studies showed an increased risk for lung cancer with
    exposure to diesel exhaust. The relative risks reported ranged from
    about 1.4 for railroad workers with the longest duration of exposure
    in both studies of such populations to 1.3-2.4 for bus garage workers,
    depending on the exposure category. The stevedores also had an
    increased relative odds ratio for lung cancer with increasing exposure
    with respect to all three indices of exposure examined, with an
    increase of three- to sixfold in the highest exposure categories and
    1.5-2.7-fold in the lowest exposure category. The point estimates
    were, however, imprecise and had wide confidence intervals: for all
    but one exposure category, the lower limit of the 90% confidence
    intervals presented by the authors included 1.0.

         The effects of cigarette smoking could be adjusted for in the
    case-control studies of railroad workers (Garshick et al., 1987) and
    stevedores (Emmelin et al., 1993). The smoking histories of the
    railroad workers were obtained from next-of-kin. Various regression
    models (including both dose and duration of smoking) were examined in
    detail in order to adjust adequately for the effects of smoking. No
    matter how it was accounted for, the risk for lung cancer based on
    work in a job with exposure to diesel exhaust was similar to the
    unadjusted odds ratio. The study of stevedores (Emmelin et al., 1993)
    was smaller than that of railroad workers, and in order to reduce the
    number of strata for analysis workers were stratified only as smoker
    or nonsmoker. In the two studies in which information on smoking was
    not available (Garshick et al., 1988; Gustavsson et al., 1990), the
    analysis was based on a comparison with workers in the same
    occupational cohort, making confounding by smoking less likely.

         The results of most of the other studies summarized in Tables
    39-41 support those of the four studies discussed above. The relative
    risks were generally in the range 1.2-1.9 but did not always achieve
    statistical significance. Furthermore, exposure to diesel exhaust was
    less precisely defined in these studies; it was usually based on
    self-reporting, the report of a surrogate, or a census report of
    occupation (Table 41). Self-reporting or the report of a surrogate
    would reflect actual exposure to diesel exhaust or work in a job in
    which exposure was likely; however, defining exposure in this less
    precise fashion make it harder to detect an effect of the exposure and
    leads to lower relative risks and wider confidence intervals.

         In three other studies, job titles obtained from an employer were
    used to define exposure. In the study of Howe et al. (1983), the job
    held at the time of retirement from the railroads was used to
    categorize exposure, and mortality was assessed for 1965-77.
    Significantly elevated risks for lung cancer were noted among workers
    who had probably been exposed to diesel exhaust (relative risk = 1.35)
    and among those who had possibly been exposed (1.20); however, since
    only retired workers were studied, actual exposure to diesel exhaust
    would have been short for many workers. Rushton et al. (1983) and
    Waller (1981) used job titles within the London Transport organization
    to indicate potential exposure to diesel exhaust. Neither study showed
    an elevated risk for lung cancer, but the study of Rushton and
    coworkers was characterized by a short duration of exposure and
    follow-up, whereas in the study of Waller, deaths occurring after
    employment were not included.

         The relative risks for lung cancer as a result of exposure to
    diesel exhaust are generally low, and risks of this magnitude are more
    susceptible to chance and to the effects of unmeasured confounding
    factors and imprecision in adjusting for known confounding factors. As
    discussed above, the elevated risk for lung cancer observed in the
    four most informative studies is unlikely to be due to confounding by
    cigarette smoking and is probably due to exposure to diesel exhaust.
    Other studies, although limited primarily by the exposure
    ascertainment, support this assessment.

    B10.4.1.2  Urinary bladder cancer: occupational exposure

         The hypothesis that exposure to diesel exhaust results in cancer
    of the urinary bladder is based on the fact that individuals who smoke
    cigarettes, a source of PAHs (as is diesel particulate), excrete
    mutagenic substances in their urine, and cigarette smoking is
    associated with an increased risk for bladder cancer. In a recent
    study of railroad workers with current exposure to diesel exhaust
    (median concentration of respirable particulates over a work shift,
    54-113 µg/m3; adjusted for smoking), mutagenicity in urine at the
    end of a shift was not associated with exposure (Schenker et al.,
    1992).

         The evidence that bladder cancer is associated with exposure to
    diesel exhaust is based on an association between work in industries
    or occupations with potential exposure to diesel exhaust. The relative
    risks reported (some of which are adjusted for smoking) are generally
    in the range 1.2-2 (Hoar & Hoover, 1985; Silverman et al., 1986),
    although higher risks have been reported. Exposure was assessed mainly
    by self-reporting and reporting of occupation at interview. The main
    occupation in which there was considered to be potential exposure to
    diesel exhaust was truck driving. Measurements of exposure of truck
    drivers by Zaebst et al. (1991) indicate that current exposure is
    likely to be low, but there is no knowledge about past exposure
    levels.

         Thus, although work in occupations related to motor vehicles has
    been associated with bladder cancer, the extent of exposure of these
    workers to diesel exhaust is unknown.

    B10.4.2  Dose-response assessment

    B10.4.2.1  Lung cancer

         Job category and work in an occupation with exposure to diesel
    exhaust have been used as indicators of exposure. When years of
    exposure and derived indices of exposure have been used, an
    exposure-response relationship has generally been found. Measurements
    of total and respirable dust in exposed workers do not reflect actual
    exposure to diesel exhaust particles, however, since there are other
    sources of dust in an occupational environment. The exposure of only
    two occupational groups, American railroad workers (Woskie et al.,
    1988b) and truck drivers (Zaebst et al., 1991), to diesel exhaust
    particles has been quantified (see section B5.2). Current exposure was
    measured for both these groups, but past exposure levels could not be
    determined (Woskie et al., 1988b), so that studies of workers do not
    permit determination of a dose-response relationship for inhaled
    diesel particles.

    B10.4.2.2  Urinary bladder cancer

         Years of work in an occupation related to motor vehicles (mainly
    truck driving) was examined in some, but not all, studies, but no
    measurements were available to estimate exposure to diesel exhaust in
    these settings.

    B10.4.3  Exposure assessment

         The ranges of exposures observed for workers are presented in
    sections B5.2 and B10.2.

    B10.4.4  Risk characterization

    B10.4.4.1  Human lung cancer

         Thus, historical measurements of exposure to diesel exhaust are
    unreliable and exist only for current workers in two industries. A
    quantitative risk assessment cannot be conducted on the basis of
    epidemiological data in which job title was used as a surrogate of
    exposure. Attempts to obtain estimates of risk (presented in section
    B10.4.1) in the retrospective cohort study of American railroad
    workers (Garshick et al., 1988) were not successful (see below for
    discussion). Consequently, there are no human data suitable for
    estimating unit risk.

    B10.4.4.2  Human urinary bladder cancer

         The risk for urinary bladder cancer cannot be assessed from the
    available epidemiological data.

    B10.4.4.3  Risk characterization based on studies in experimental
               animals

    (a)  Introduction

         Most human tumours arise in the bronchial region, and although
    the frequency of adenocarcinomas has increased in recent years
    (Martini, 1993), squamous bronchogenic carcinomas still represent the
    majority (about 60%) of lung tumours in humans (Auerbach & Garfinkel,
    1991; Devesa et al., 1991; Campobosso et al., 1993). The tumours
    observed in rats are located almost exclusively in the bronchoalveolar
    regions. This difference raises the question of whether the rat is the
    appropriate species for extrapolating to humans. Although it may be
    the wrong model, however, studies of hamsters and monkeys give
    negative results, so the rat is the most appropriate model from a
    conservative viewpoint and is therefore used for quantitative risk
    extrapolation.

         The objective is to assess quantitatively the potential risk for
    lung cancer in humans posed by exposure to diesel exhaust emissions in
    the ambient air. Ideally, human risk due to exposure to an
    environmental pollutant should be predicted on the basis of human
    experience. Although several epidemiological studies are available on
    bus and railroad workers, these results alone are not adequate to
    assess the potential risk of cancer in humans because of the lack of
    reliable information on the conditions of exposure of these workers.
    The challenge is to assess the risk of exposure to diesel exhaust on
    the basis of all of the available information, for both experimental
    animals and humans. In contrast to the sparse human data, there is a
    rich pool of information on diesel-induced lung tumours in two strains
    of rat. One approach to integrating this information is to make a
    quantitative risk assessment on the basis of information from
    bioassays and on the relevant biological mechanism and then to
    evaluate these animal-based results against available human
    experience. This approach was followed in this monograph.

         Although, the available experimental data on the possible impact
    of extractable organic matter and PAHs on the lung tumour response
    associated with exposure to diesel soot are not strong (Heinrich et
    al., 1991; Heinrich, 1994; Heinrich et al., 1995), a mathematical
    approach to assessing the risk of exposure to diesel exhaust should
    take into account the effects of both particles (carbon core) and
    extractable organic matter, because: (i) organic compounds include a
    variety of PAHs and nitroaromatic compounds, many of which are known

    to be carcinogenic; (ii) the results of recent studies on inert
    particles and carbon black in rats (Mauderly et al., 1991; Heinrich,
    1994; Heinrich et al., 1994) strongly support the hypothesis that the
    carbon cores of diesel particles are the component primarily
    responsible for the induction of lung cancer; (iii) PAHs alone are
    unlikely to be responsible for the tumour response; (iv) the
    disproportionately high tumour incidence in animals with heavy
    exposure coincides with a disproportionate increase in the cumulative
    lung burden of diesel particles. Although a qualitative description of
    the biological mechanism of diesel exhaust-induced tumours is
    plausible, however, the lack of quantitative information on the
    dynamics of tumour initiation, promotion, and progression vitiates the
    construction of a biologically based dose-response model.

         The crudest dose-response model is obtained by fitting observed
    dose-response data to a mathematical function that is a monotonic,
    non-decreasing function of exposure, using the inhaled concentration
    and observed tumour incidence, without considering information on
    pharmacokinetics and the mechanism of carcinogenesis. This approach
    may result in uncertain estimates of risk at low doses, because, while
    the model may adequately fit the dose-response data for high doses at
    which a tumour response is observed, it may greatly underestimate or
    overestimate the risk at low doses.

         Most previous risk estimates of diesel exhaust-induced cancer
    risk were developed before the results of many of the bioassays became
    available (Pepelko & Chen, 1993). Nevertheless, most of the estimates
    are similar (within an order of magnitude), except for those
    calculated on the basis of human data, which are higher than those
    calculated from other databases. Because of great uncertainties
    associated with the available human data, it is considered more
    appropriate to calculate risk on the basis of animal data and then use
    human data to evaluate their validity.

         The approach adopted in this assessment is the linearized
    multistage model (Anderson et al., 1983), in which burden per lung
    surface area (milligrams per square centimetre) is used as an
    equivalent dose in animals and humans. The use of lung (alveolar)
    surface area is justified by the fact that the lung tumours observed
    in rats are derived exclusively from epithelial cells in the alveolar
    region of the lung. Normalization on the basis of lung weight can also
    be done by using a factor of 4, for increasing the dose or decreasing
    the risk, for the difference in the ratio of rat:human lung surface
    area and lung weight. The risk prediction calculated by this approach
    is presumably conservative (i.e. overpredicts the risk) because the
    dose-response function is assumed to be non-threshold and linear at
    low doses.

         An alternative model that incorporates a biological mechanism is
    constructed using statistically estimated parameters. The model
    assumes that carcinogenic agents present in the organic fraction act
    directly on the target cells, primarily via initiation. It is further
    assumed that the majority of the particles are ingested by
    macrophages. Particle-laden macrophages are then induced to secrete a
    variety of mediators (e.g. reactive oxygen species and cytokines),
    which diffuse to the target cells and induce initiation,
    proliferation, and conversion of initiated cells to malignant cells.
    This alternative model can be used only as a tool to evaluate the
    impact of various biological assumptions and not for risk prediction,
    because the parameters of the model are not measured in the
    laboratory.

    (b)  Approaches to quantification of human risk from exposure to
         diesel exhaust

         Several issues must be addressed in estimating the risk of
    exposure to diesel exhaust, including the critical target site, the
    fraction of exhaust responsible for tumour induction, the availability
    of dosimetric methods for accurately extrapolating dose from that for
    experimental animals chronically exposed to high concentrations of
    exhaust to that of humans exposed at ambient concentrations, and the
    most suitable low-dose extrapolation model.

         The critical target organ was considered to be the lung. Although
    Iwai et al. (1986) reported induction of both malignant lymphomas and
    lung tumours in rats exposed to diesel exhaust, the lung was the only
    target site in other experimental studies of this species. Although
    potentially carcinogenic agents present in diesel exhaust may be
    absorbed from the lungs, enter the bloodstream, and be transported
    systemically, no data are available to evaluate this possibility.
    Organic compounds adsorbed on particles may also reach systemic
    targets via the gastrointestinal tract, as particles deposited in the
    conducting airways are cleared rapidly to the oropharynx and
    swallowed. A considerable volume of particles is also likely to be
    ingested as a result of grooming when the whole animal is exposed
    (Wolff et. al., 1982); however, because the half-times for the elution
    of organic compounds from particles are considerably longer than their
    passage through the gastrointestinal tract, the fraction absorbed is
    expected to be small. In any case, there is little evidence for any
    systemic effects of diesel exhaust.

         The site of action in the lungs is assumed to be the epithelial
    lining of the alveoli and small airways. According to Mauderly et al.
    (1987), inflammation and tumours appear to arise from this tissue.
    Although interstitial events (e.g. fibrosis) have been suggested to be
    associated with induction of lung tumours by particles (Kuschner,
    1968), no data are available to support this view with respect to
    diesel exhaust.

         Accurate extrapolation of dose from studies of animals exposed to
    high concentrations of exhaust to humans exposed to ambient
    concentrations requires a variety of adjustments for species
    differences in deposition efficiency and respiratory ventilation
    rates. As the normal retention half-times in the alveolar region are
    several times longer in humans than rats (Chan et al., 1981; Bohning
    et al., 1982), the lung burden of humans may be underestimated if this
    difference is not taken into account. The high exposure concentrations
    used in some experimental studies, however, result in greatly slowed
    or even completely inhibited clearance (Griffis et al., 1983). In
    order to extrapolate dose accurately from experimental studies to
    humans, the detailed dosimetric model developed by Yu et al. (1991)
    was used, which accounts for species differences in respiratory
    ventilation rates, deposition efficiency, normal particle clearance
    rates, transport of particles to lung-associated lymph nodes, and lung
    surface area. It also accounts for inhibition of particle clearance
    due to lung overload. In this model, dose is estimated in terms of
    particle concentration per unit of lung surface area.

         Two approaches were used to derive unit risk estimates: the
    linearized, multistage, low-dose extrapolation model and a model based
    on the biological mechanism discussed above. In the linearized
    multistage model, the lung burden of carbon core per lung surface area
    is used as the dosimetric parameter. A particle-based assessment was
    considered to be reasonable for two principal reasons: (i) exposure to
    the vapour phase alone did not result in detectable tumour induction
    in rats (Brightwell et al., 1986; Iwai et al., 1986; Heinrich et al.,
    1986a); and (ii) exposure to carbon black, which is similar in
    composition to the carbon core of whole diesel exhaust but contains
    only negligible amounts of organic compounds, was about as effective
    in inducing lung cancer as whole diesel exhaust (Mauderly et al.,
    1991; Nikula et al., 1994; Heinrich et al., 1995). Although use of the
    carbon core as a dosimetric parameter implies that it is primarily the
    insoluble carbon core of diesel particles that is responsible for the
    carcinogenic effects of diesel exhaust, it can also be viewed simply
    as a marker of exposure to diesel exhaust, with no biological
    implication. Similarly, organic compounds could be used as markers of
    exposure for the dose-response calculation. When this is done, the
    resulting unit risk estimates (not presented here) are very similar
    (within 25%) to those calculated with the carbon core as the surrogate
    of exposure.

         The second approach is based on the assumption that, even though
    the concentration of carcinogenic compounds on diesel particles is
    low, they nevertheless can act in concert with the particles to induce
    carcinogenesis. An alternative low-dose extrapolation model was
    therefore developed that allows for the possibility that various PAHs
    and nitroaromatic compounds may induce organ-specific adducts that
    contribute to cell initiation. This model is described in Appendix A1.

         Both approaches incorporate the detailed dosimetric model of Yu
    et al. (1991) to estimate dose at the airway and alveolar surfaces.
    Risk is based on the assumption that an equivalent concentration
    (dose) per unit of alveolar surface area results in equivalent risks
    in humans and rats.

    (c)  Evidence to support a carcinogenic effect of the carbon core

         The study of Heinrich et al. (1994) on carbon black and coal-tar
    pitch provides information useful for evaluating the assumptions used
    to construct the model that reflects current thinking about the
    mechanism of action of diesel exhaust in inducing lung tumours. One
    relevant question about diesel exhaust is whether lung tumours are
    induced by the carbon core alone or in combination with organic
    compounds.

         In this study, female Wistar rats were exposed to carbon black or
    tar pitch, or both for 10 months (43 weeks) or 20 months (86 weeks)
    and followed until the end of the study, which lasted for
    two-and-a-half years. No lung tumours were observed in the control
    group. For each exposed group, the probability of dying from lung
    cancer was calculated by a Kaplan-Meier survival analysis. Since these
    data are used solely for interpretation and not for risk prediction,
    it is statistically more appropriate to use only data on mortality,
    excluding those tumours found at terminal sacrifice. Figure 9 shows
    the probability of dying from lung cancer for the rats that were
    exposed either to 6 mg/m3 of carbon black, or to coal-tar pitch
    containing 50 µg/m3 of benzo[a]pyrene, or to a combination of the
    two for 10 months, and then followed until the end of the study. In
    Figure 10, each plot represents the time at which a tumour occurred.
    To provide a better visual presentation, the probabilities at these
    points are connected by straight lines. (Note: It is statistically
    more appropriate to present these probabilities by a step function
    rather than by interpolation when the Kaplan-Meier procedure is used.)

         Two conclusions can be drawn from Figure 10: (1) Tumours appeared
    much earlier in animals exposed to coal-tar pitch alone or in
    combination with carbon black than in animals exposed to carbon black
    alone. For instance, the first tumour occurred 665 days after the
    start of exposure to carbon black, 406 days after exposure to coal-tar
    pitch, and only 310 days after exposure to the combination. (2) Carbon
    black and coal-tar pitch together caused a higher rate of mortality
    from tumours than the sum of mortality caused by carbon black and
    coal-tar pitch separately. The probability of dying from a lung tumour
    before terminal sacrifice at 912 days was 0.28 for carbon black, 0.52
    for coal-tar pitch, and 0.99 for the combination. These two
    observations suggest that carbon black and coal-tar pitch act on
    different stages of tumorigenesis. They are consistent with the
    hypothesis that carbon black can increase the proliferation rates of
    normal and cells at different stages of differentiation initiated
    spontaneously or by coal-tar pitch.

         When the rate of mortality from tumours among rats exposed to
    diesel exhaust is compared with that of groups exposed to carbon
    black, coal-tar pitch, or the combination (Figure 9), the rate among
    the animals exposed to diesel exhaust is comaprable to that of rats
    exposd to carbon black for 10 to 20 months. This implies that carbon
    black alone can induce a tumour response similar to that induced by
    diesel exhaust. Figure 9 also shows that doubling the duration of
    exposure to carbon black from 10 to 20 months does not increase the
    probability of dying from a lung tumour before terminal sacrifice;
    however, doubling the duration of exposure to coal-tar pitch
    significantly increases the probability of dying from a lung tumour.
    Figure 11 shows that the shapes of the curves for age-dependent
    mortality from cancer for animals exposed to diesel exhaust and
    coal-tar pitch for 20 months are similar. It should be kept in mind
    that the PAH content of the atmosphere containing both carbon black
    and coal-tar pitch was about 1000 times higher than that of diesel
    engine exhaust. These observations may have profound implications for
    elucidating the mechanisms of induction of lung tumours by diesel
    exhaust, but more refined analyses are necessary before any conjecture
    can be made.

    (d)  Data available for calculating risk

         Seven bioassays have shown lung tumour responses in rats
    (Brightwell et al., 1986; Heinrich et al., 1986b; Ishinishi et al.,
    1986; Iwai et al., 1986; Mauderly et al., 1987; Nikula et al., 1994;
    Heinrich et al., 1995). Four studies were chosen for use in
    calculating unit risk (Table 48) because they involved multiple
    exposure groups. Data on time to event (death with or without a
    tumour) are available in the studies of both Mauderly et al. (1987)
    and Heinrich et al. (1995) and were used in all of the risk
    calculations.

    (e)  Calculation of unit risks

         The unit risk from exposure to an air pollutant is defined as the
    95% upper bound of the increased lifetime cancer risk for an
    individual continuously exposed to a concentration of 1 µg/m3 in
    ambient air. Unit risk is a convenient tool for calculating lifetime
    risk due to exposure to a pollutant when low-dose linearity is
    assumed. Under this assumption, the risk due to exposure to  d µg/m3
    of pollutant can be calculated from  u ×  d if the pollutant has a
    unit risk of  u/µg per m3. This calculation is valid, however, only
    when the level of risk is low. The results of unit risk calculations
    are summarized in Table 49.

    FIGURE 9

    FIGURE 10

    FIGURE 11


        Table 48.  Incidence of lung tumours in rats exposed to diesel engine exhaust
                                                                                                                                                

    Strain       Exposure                            Dose metric                                        Lung tumour        Reference
    (sex)                                                                                               incidence
                 Schedule         Concentration      Weekly exposure      Lung particle burden
                                  (mg/m3)            (mg/m3 × h)          (mg/cm2  lung surface)a
                                                                                                                                                

    Fischer      7 h/d,           0                       0               0                                 2/230          Mauderly et al.
    344          5 d/week         0.35                   12               6.4 × 10-5                        3/223          (1987)
    (f+m)                         3.50                  122               2.8 × 10-3                        8/222
                                  7.08                  248               6.0 × 10-3                       29/227

    Wistar       18 h/d,          0                       0               0                                 1/217          Heinrich et al.
    (f)          5 d/week,        0.84                   76               5.8 × 10-4                        0/198          (1995)
                 2 years,         2.50                  225               5.2 × 10-3                       11/200
                 < 6 months       6.98                  628               1.5 × 10-2                       22/100
                 follow-up

    Fischer      16 h/d,          0                       0               0                                 1/123          Ishinishi et al.
    344          6 d/weekb        0.46                   44               2.5 × 10-4                        1/123          (1986)
    (f+m)                         0.96                   92               2.0 × 10-3                        0/125
                                  1.84                  177               4.2 × 10-3                        4/123
                                  3.72                  357               8.8 × 10-3                        8/124

    Fischer      16 h/d,          0                       0               0                                 4/250          Brightwell et al.
    344          5 d/week         0.7                    56               3.5 × 10-4                        1/112          (1986)
    (f+m)                         2.2                   176               4.2 × 10-3                       14/112
                                  6.6                   528               1.3 × 10-2                       55/111
                                                                                                                                                

    a  Calculated from the model of Yu et al. (1991)
    b  Data on heavy-duty diesel engine exhaust used because those for light-duty engines showed no statistically significant difference
             Particle-based model: The linearized multistage model, which is
    used as a conservative approach to estimating risk, has the
    mathematical form P = 1 exp(-Z), where Z is either  q0 +  q1 ×  d + ...
    +  qm ×  dm of a polynomial concentration  d, or ( Q0 +  Q1 ×  d + ...
    +  Qm ×  dm) ×  tk, a polynomial of concentration  d multiplied by a time
    factor,  tk, when data on time to event are used. In this case, the
    lifetime risk is calculated from actuarial life tables using the
    survival probability of control Fischer 344 rats within the National
    Toxicology Program, provided by Portier et al. (1986). The range of
    extrapolation is about three orders of magnitude in the present
    assessment. Denote P0 the lifetime cancer risk at concentration 0.
    Because the extra risk (P-P0)/(1-P0) is dominated by the linear term
     q1 ×  d at low concentrations, the 95% upper bound of q1 is used
    to represent unit risk.

         In extrapolating risk from animals to humans, a dose metric that
    induces the same tumour incidence rate in animals and humans must be
    assumed (i.e. dose equivalence). To calculate equivalent doses, a
    mathematical model is used to adjust for the dosimetric parameters
    that determine the lung burden of particulate matter in rats and
    humans and to correct to dose per unit lung surface area. This dose is
    considered to be equally potent in inducing lung tumours in animals
    and in humans. The model accounts for differences between animals and
    humans in regional deposition efficiency, particle clearance rates at
    low doses and at doses that result in impaired clearance, and lung
    surface area. In these calculations, the mass fraction of organic
    compounds adsorbed on particles is assumed to be 20%; one-half of the
    mass is assumed to be composed of slowly eluted organic compounds
    (half-life, 30 days) and the other half of rapidly eluted organic
    compounds (half-life, 1.3 h). The remainder is considered to be
    inorganic carbon. At higher concentrations, particle clearance slows
    and may even stop, and the lung particle burdens increase continually
    during exposure. The organic constituents, however, are eluted from
    the particles fairly quickly and reach a steady state even during
    continued exposure to high concentrations. The lung burdens of organic
    compounds are therefore less affected by inhibition of clearance by
    overloading.

         In the linearized multistage model, determination of the dose of
    the carbon core is problematic because the lung burdens after low and
    high exposures differ drastically over time. This difference suggests
    that use of a lung burden at a fixed time (e.g. one year after the
    start of exposure) to represent dose may not be appropriate. In this
    assessment, the average lung burden is used as the dose at the target
    organ. The average lung burden is calculated by dividing the area
    under the curve of lung burden over time by the corresponding period
    of the experiment for which the curve was calculated. Figure 12 shows
    a comparison of the lung burden predicted from the model and that
    observed in the laboratory (Muhle et al., 1994) at specific times. It
    suggests that the model used to calculate lung burden is adequate.

    Table 49.  Estimates of unit risk per microgram of particles per
               cubic metre of diesel exhaust
                                                                        

    Upper 95% confidence limit of cancer       Study
    risk due to exposure to 1 µg/m3 of
    diesel particulate matter
                                                                        

    3.4 × 10-5                                 Mauderly et al. (1987)
    1.6 × 10-5a                                Ishinishi et al. (1988)
    7.1 × 10-5                                 Brightwell et al. (1986)
    3.4 × 10-5                                 Heinrich et al. (1995)
    3.4 × 10-5                                 Geometric mean of
                                               four studies
                                                                        

    If milligrams per lung weight are used instead of milligrams per lung
    surface as the equivalent dose, the risk estimates are reduced by a
    factor of 4.

    a  Heavy-duty diesel engine

          Biologically based model: A second approach to estimating risk
    was used because it was considered more desirable to base risk on a
    biologically based dose-response model. Although the data are at
    present insufficient to replace the linearized multistage model, which
    is considered more conservative, the implications of hypothetical
    mechanisms of cancer induction by diesel particles can nevertheless be
    investigated. The biological issues considered include the effects on
    the carcinogenic process of particle-adsorbed organic compounds and of
    a variety of mediators secreted by particle-laden macrophages. A
    stochastic model was developed, which:

    --   accounts for the possible effects of both the carbon particles
         and their associated organic compounds:

    --   allows evaluation of the contribution to tumour induction of   
         both carbon particles and organic compounds;

    --   allows for changing parameters with increasing lung burden;   
         and

    --   assumes that cell proliferation and tumour induction are
         stochastic. (For instance, it is not appropriate to assume that
         all cells divide at the same age.)

    Unlike the linearized multistage model, this approach does not require
    a constant dose metric but allows for varying lung burden over time.

    FIGURE 12

         The model, which is described in Appendix B10.3, allows for
    initiation by both the carbon and the organic fraction and for the
    proliferative effects of the carbon fraction. Although these
    mechanisms remain to be proven, it is assumed that carcinogenic agents
    present in the organic fraction act directly on the target cells,
    primarily by initiation. It is further assumed that most of the
    particles are ingested by macrophages. Particle-laden macrophages are
    then induced to secrete a variety of mediators (e.g. reactive oxygen
    species and cytokines), which diffuse to the target cells, inducing
    initiation, proliferation, and conversion of initiated cells to
    malignant cells.

         The results reported by Mauderly et al. (1987) for tumour
    induction were used to estimate the model parameters. These are based
    on the development of malignant tumours rather than all tumours as in
    the first method. This was necessary in order to ensure that the data
    used to estimate the parameters represented the same biological
    mechanism. Lung burdens were calculated from the same dosimetry model
    used in the linearized multistage model.

    (f)  Results of unit risk calculations

          Particle-based model: Unit risks were derived from the linearized
    multistage approach for the tumour incidences seen in four bioassays
    (Brightwell et al., 1986; Ishinishi et al., 1986; Mauderly et al.,
    1987; Heinrich et al., 1995) and the corresponding equivalent doses
    (Table 48). The resulting unit risk estimates, listed in Table 49,
    range from 1.6 to 7.1 × 10-5/µg particles per m3 with a geometric
    mean of 3.4 × 10-5/µg per m3.

         In these calculations, the relationship between air concentration
    (micrograms per cubic metre) and lung burden (milligrams) in humans is
    used to determine the lung burden resulting from lifetime exposure to
    1 g/m3 of diesel exhaust particulate matter. The particle burden in
    terms of mass per unit lung surface area is then multiplied by the
    slope derived from the bioassay data. For instance, when the data of
    Mauderly et al. (1987) are used, the slope of the curve for
    carcinogenicity in rats (i.e. the upper 95% confidence limit of the
    linear coefficient in the multistage model), expressed in terms of
    equivalent dose (micrograms of carbon particulate matter per square
    centimetre) is 1.7 × 10-2/µg per cm2. According to the dosimetry
    model, an air concentration of 1 µg/m3 of particulate matter
    corresponds to a mass of 1230 µg of carbon particles per human lung.
    Because the lung epithelial surface area, including the alveolar
    region and conducting airways, is assumed to be 627 000 cm2, a unit
    risk of 3.4 × 10-5/µg per m3 is derived from 1.7 × 10-2/µg per cm2
    × 1230 µg/627 000 cm2.

          Biologically based model: The unit risk estimated by the
    alternative model, based on the data on malignant tumours from the
    study of Mauderly et al. (1987), is equal to 1.65 × 10-5/µg particles
    per m3. This is lower than the estimate of 3.4 × 10-5/µg per m3
    derived from the same study with the linearized multistage approach.
    Application of the linearized multistage model only to malignant
    tumours seen in that study, rather than all lung tumours, resulted in
    a unit risk estimate of 1.74 × 10-5/µg particles per m3 (Table 50).
    Thus, the unit risk estimates obtained by the two approaches are
    virtually identical. The estimated risks may differ somewhat with
    increasing doses, because the slopes are not identical at all exposure
    levels. It should be noted, however, that the unit risk predicted by
    the alternative model is derived under the assumption that particles
    continue to exert an effect on cell initiation or proliferation (or
    both) at low doses. There is considerable uncertainty about the
    effects of particles at low doses. It has been claimed that particles
    do not induce initiation or proliferation at low doses (Vostal, 1986).
    At present, the evidence is inadequate to support or refute this
    claim. Moreover, even if macrophage overload is required, because of
    uneven distribution of particles, some macrophages may become
    overloaded even at low concentrations. Because of this uncertainty,
    the biologically based model, like the linearized multistage model,
    does not depart from linearity at low doses; however, if initiation
    and proliferation do not occur at low doses, the risk may be much
    smaller.

         In an estimate of unit risk for rats, based on the results of
    seven studies by inhalation and using linear interpolation and the
    linearized multistage model, the risks were 7 × 10-5/µg of diesel
    particles per m3 and 10 × 10-5/µg carbon core particles per m3
    (Csicsaky et al., 1993; Pott et al., 1993; Roller & Pott, 1994).

    (g)  Results and implications of the biologically based model

         On the assumption that particles continue to affect cell
    initiation or proliferation at low doses, the risks calculated with
    this model are comparable to those obtained with the linearized
    multistage model. Excess risks due to various exposures are shown in
    Tables 50 and 51. Those in Table 50 are the risks predicted for humans
    exposed continuously (24 h/day) from the two models: it is interesting
    to note that the results are similar. Table 51 shows the excess risks
    due to exposure to 2.6 µg/m3 of diesel particulate emission for
    16 h/day on seven days per week and to 15 µg/m3 for 8 h/day on five
    days per week. The first concentration was reported by the US
    Environmental Protection Agency Office of Mobile Sources to be the
    annual mean exposure of the American population to diesel particulate
    matter in 1986; the second concentration was reported to be that to
    which workers are exposed on urban freeways (Carey, 1987). The
    retention half-time for insoluble particles after exposure to

        Table 50.  Comparison of excess risk for humans due to continuous exposure to
               various concentrations of diesel exhaust emissions using two
               different models
                                                                                                

    Exposure concentration    Biologically based (alternative)       Linearized (µg/m3)
                                                                     multistage
                              Maximum likelihood   Upper 95%         model
                              estimate             bound estimate
                                                                                                

    0.01                      7.68 × 10-8          1.35 × 10-7       1.71 × 10-7
    0.1                       8.12 × 10-7          1.71 × 10-6       1.72 × 10-7
    1.0 (unit risk)           8.16 × 10-6          1.65 × 10-5       1.74 × 10-5
    100                       5.58 × 10-4          9.63 × 10-4       1.74 × 10-4
    1000                      2.60 × 10-2          4.22 × 10-2       3.33 × 10-2
                                                                                                

    a  Slope - 9.04 per mg/cm2 of lung surface, using carbon core as dosimetric.
    Only malignant tumours are used in the calculations.
    
    2.6 µg/m3 is shown to increase from 296 days for members of the
    general population with normal respiratory function to 519 days for
    those with a smoking history of 20 pack-years (Bohning et al., 1982).
    This information was used to reduce the alveolar clearance rate for
    the dosimetric calculations to that used for other risk calculations.
    Interactive effects of smoking and diesel exhaust are not considered
    in the risk calculations owing to lack of data. The studies of carbon
    black and coal-tar pitch (Heinrich, 1994; Heinrich et al., 1994)
    indicate that smokers have a higher risk for lung cancer than
    nonsmokers when they are exposed to diesel exhaust.

         The excess lifetime risks shown in Tables 50 and 51 are
    standardized by the actuarial life-table approach, using the survival
    probability of control animals in the US National Toxicology Program
    provided by Portier et al. (1986). This approach gives a weighted
    average of the probability of cancer occurrence over an entire
    lifetime, weighted by survival probability.

    Table 51.  Excess lifetime risk for humans due to exposure to diesel
               exhaust emissions under various exposure scenarios
                                                                        

    Exposure pattern              Biologically based model     Linearized
                                                               multistage
                                  Maximum       Upper 95%      modela
                                  likelihood    bound
                                  estimate      estimate
                                                                        

    General population (normal    1.41 × 10-5   2.44 × 10-5
    respiratory function;
    nonsmoker): 2.6 µg/m3,
    16 h/day, seven days
    per week

    General population            2.32 × 10-5   3.61 × 10-5    5.38 × 10-5
    (20-pack-year smoker)b:
    2.6 µg/m3, 16 h/day,
    seven days per week

    Occupationally exposed:       3.12 × 10-5   5.17 × 10-5    6.18 × 10-5
    15 µg/m3 8 h/day, five
    days per week
                                                                        

    a  Calculated using carbon core as dosimetric; only malignant
       tumours are used.
    b  In this calculation, smoking affects only lung clearance rate;
       biological interaction between smoking and exposure to diesel
       exhaust is not considered. The retention half-times for
       insoluble particles increased from 296 days for persons with
       normal respiratory function to 519 days for 20-pack-year smokers
       (Bohning et al., 1982).

    Some implications of the alternative model are:

    (1)  At a low exposure concentration, the decrease in diesel-induced 
         initiation results in a greater reduction of risk; that is, the
         number of initiated cells (cancer risk) is reduced more
         efficiently with a low than with a high exposure. If only organic
         compounds induce initiation when the concentration is low, they
         play a more important role than particles in inducing tumours at
         low concentrations, whereas the roles are reversed when the
         concentration is high.

    (2)  Although cells initiated by diesel exhaust play an important role
         in cancer induction, either organic compounds or the carbon core
         alone could induce initiation by increasing their respective
         proportions. Thus, although initiated cells are important for
         tumour induction, they may be induced by any agent that initiates
         tumours (e.g. smoking).

    (3)  A small change in the rate of proliferation induced by diesel
         exhaust could disproportionately change cancer risk. As this
         parameter is assumed to be a function of the dose of carbon core,
         lung overloading has a significant effect on cancer incidence. In
         the absence of better information, it is assumed in this
         assessment that the carbon core continues to have a proliferating
         effect at low doses.

         These observations suggest that, although the effect of particle
    overload on cell proliferation is important, initiation by the carbon
    core or organic compounds or both is also essential. Although this
    conclusion is only tentative, because the model parameters are
    estimated statistically on the basis of bioassays conducted at high
    concentrations, it does suggest the importance of studying the role of
    the carbon core and organic compounds in initiation and promotion at
    low and high exposure concentrations. Does the relative initiation
    potential of organic compounds and the carbon core differ with
    concentration? These observations also suggest that a subcohort of
    workers who were smokers and were exposed to high concentrations of
    diesel exhaust for a long time would have a greater risk of dying from
    lung cancer.

    (h)  Comparison of risk estimates derived from experimental studies
         and human experience

         The bioassay-based risk estimates, which range from 1.6 × 10-5 to
    7.1 × 10-5, with a geometric mean of 3.4 x 10-5, can be compared with
    human experience on the basis of three data sets: those of an
    epidemiological study conducted on London Transport employees (Waller,
    1981) and a subsequent analysis (Harris, 1983) and those on American
    railroad workers (Garshick et al., 1987, 1988). Although these data
    cannot be used to calculate unit risk, mainly because of a lack of
    reliable information on exposure, they can be used to evaluate the
    validity of unit risk estimates based on the results of experimental
    studies.

         Attempts have already been made to estimate the potential cancer
    risk due to exposure to diesel exhaust on the basis of epidemiological
    data. From the results of the study of London Transport workers,
    Harris (1983) estimated that the increase in the relative risk for
    lung cancer associated with exposure to 1 µg/m3-year of diesel exhaust
    was 1.2 × 10-4, with an upper bound of the 95% confidence limit of
    4.8 × 10-4. (Generally, when data from an epidemiological study with

    negative results are used to estimate cancer risk, the upper bound is
    used to calculate unit risk.) The resulting unit risk estimate is 2 ×
    10-3, which is about 60 times higher than the mean unit risk estimate
    derived from bioassays, 3.4 × 10-5, and about 30 times higher than the
    upper end of the range of unit risk estimates from bioassays, 7.1 ×
    10-5. Therefore, the risk estimate based on bioassays is not
    inconsistent with that for humans, since 2 × 10-3 is the estimated
    upper bound in an epidemiological study with negative results.

         McClellan et al. (1989) reported risk estimates based on the
    study of Garshick et al. (1987), in which lung cancers in railroad
    workers were evaluated. Assuming exposure to concentrations of 500 and
    125 µg/m3, the upper bounds of the 95% confidence interval for
    lifetime cancer risk were estimated to be 6 × 10-4 and 2 × 10-3,
    respectively. The lower of the two unit risk estimates is only about
    one order of magnitude higher than the risk estimates based on
    bioassays.

         An epidemiological study potentially more suitable for
    quantitative risk assessment was reported by Garshick et al. (1988),
    which was based on a large number of subjects; a small but significant
    increase in the rate of mortality from lung cancer was seen in some
    subcohorts. The US Environmental Protection Agency has supported an
    effort to derive a unit risk estimate from this study, and data on
    exposure were estimated by Woskie et al. (1988a,b) and Hammond et al.
    (1988). These data were analysed in a variety of ways, using relative
    risk and absolute risk dose-response models and by classifying
    individuals into various exposure categories, including job, duration
    of employment, age, and exposure markers. Even though at least 50
    analyses were carried out, an adequate dose-response relationship
    could not be obtained; these data were therefore not used to estimate
    unit risk.

         The lack of a dose-response relationship is not totally
    unexpected, given the low increase in mortality rate in the study of
    Garshick et al. (1988) and the uncertain estimates of exposure. The
    data on exposure were derived from air samples collected during a
    limited period (1981-83) on four small railroads operating in a
    limited geographical area (northern United States), in which
    concentrations of respirable particulate matter were measured rather
    than diesel exhaust per se; the measured concentrations were then
    adjusted to derive markers of exposure. The data were used to estimate
    exposure to diesel exhaust of railroad workers throughout the United
    States 30 or more years earlier. Diesel equipment and working
    conditions have, however, changed since the 1940s when diesel engines
    first began to be used in large numbers. Woskie et al. (1988b)
    summarized anecdotal reports of smoky working conditions in diesel
    repair shops during the 1950s and 1960s and reported that the limited
    data available on levels of nitrogen oxide during those periods

    indicated high levels of diesel exhaust. By the time samples were
    collected, however, the smoky conditions would have been largely
    mitigated by improved ventilation and the advent of less smoky diesel
    engines. The study of Garshick et al. (1988) gives the relative risks
    for dying from lung cancer in exposed as opposed to unexposed railroad
    workers classified into five subcohorts by age in 1959. The risks
    ranged from 0.96 (95% CI, 0.74-1.33) to 1.45 (1.11-1.89). The highest
    relative risk, 1.45, which was observed in workers who were 40-45
    years old in 1959, was used here to evaluate the validity of unit risk
    based on biossays results. Assuming that this subcohort of workers was
    exposed to diesel exhaust for 8 h/day on five days per week from age
    35 to age 65, the background mortality rate from lung cancer in this
    subcohort would be 0.038 (0.63 × 0.06), as the unexposed workers in
    the same age group had a relative risk of 0.63 and the corresponding
    lifetime mortality rate from lung cancer in the general American white
    male population is about 0.06. If the lung cancer risk due to
    1 µg/m3 is assumed to be 3.4 × 10-5, the concentration of diesel
    exhaust in the working environment was at least 400 µg/m3, calculated
    as follows: The risk due to 1 µg/cm2 of particles is 0.017 (which
    results in a unit risk of 3.4 × 10-5). For a lower bound of the
    relative risk of 1.1, a lung burden,  d, that satisfies the
    relationship 0.1 × 0.038 = 0.017 d would be needed; that is,  d =
    0.22 µg/cm2, which is equivalent to an air concentration of
    0.4 mg/m3 by the dosimetry model of Yu et al. (1991). If the highest
    unit risk estimate derived from Brightwell et al. (1986), 7.1 ×
    10-5, is used, the required minimal air concentration would be about
    0.2 mg/m3. These calculations imply that a diesel exhaust
    concentration of at least 0.2 mg/m3 was necessary to observe a
    statistically significant increase in the mortality rate from lung
    cancer in this study. This concentration appears to be reasonable in
    the light of the working conditions described by Woskie et al.
    (1988a).

         Although the risk estimates derived from bioassays are lower than
    those derived from human data and may possibly over-predict risk,
    since the non-linearity of the human response is not considered in
    this model, they are not inconsistent, for the following reasons:

    (1)  The risk estimates based on epidemiological studies are not
         derived from all of the available data but on only a subset with
         the highest response. The estimates are therefore higher than
         those that would be derived from the whole data set.

    (2)  When a single data point (i.e. an overall relative risk and an
         averaged exposure concentration) is used in the calculations, the
         resulting slope for potency will be overestimated if the
         dose-response relationship is not linear over all exposure
         concentrations. Assume, for example, that the response follows
         the simple multistage model  P(d) = 1 - exp[-( q0 +  qd +  q2 d2)].
         The relative risk,  R, at concentration  d is  R(d) =  P(d)/P0.

         Using this mathematical expression, it is easy to demonstrate
         that the slope factor calculated from [ P(d) - 1]  P0/ d is
         greater at high doses (including the averaged concentration used
         in the risk calculation) than at low doses where the
         dose-response function is dominated by  q1. Epidemiologists
         have long had a similar (but not identical) concern about the use
         of averaged data, since ecological associations are not
         necessarily consistent with those measured at the individual
         level (see Cohen, 1994; Greenland & Robins, 1994; Piantadosi,
         1994).

    (3)  Some occupational groups were exposed to considerably higher
         concentrations of diesel exhaust in the past than presently. For
         example, the average particle concentration in a Finnish
         roundhouse was reported to be 2 mg/m3 (Heino et al., 1978). The
         lung burdens would thus be greater than those predicted on the
         basis of current exposures, and the unit risk estimates would be
         higher

    (i)  Summary and conclusions

         A dosimetric model that accounts for differences between
    experimental animals and humans in lung deposition efficiency, lung
    particle clearance rates, lung surface area, ventilation, and the
    rates of elution of organic chemicals from the particle surface was
    used to calculate equivalent human doses as particle concentration per
    unit lung surface area. After dosimetric adjustment, four risk
    estimates were derived by a linearized multistage model from three
    bioassays with Fischer 344 rats and one with female Wistar rats, which
    ranged from 1.6 to 7.1 × 10-5/µg particles per m3, with a geometric
    mean of 3.4 × 10-5. This quantitative assessment of the carcinogenic
    risk due to exposure to diesel engine emissions is reasonable,
    because:

    --   The estimates are based on several well-designed, well-executed,
         long-term bioassays.

    --   Epidemiological studies indicate that humans are susceptible to
         the induction of lung cancer after inhalation of diesel exhaust.

    --   Dosimetry modelling, especially to account for inhibition of
         particle clearance at high doses, has allowed accurate
         extrapolation of doses from animals to humans.

    --   The doses are based on actual concentrations of particulate
         matter per unit of lung surface area.

    --   Use of an alternative model that attempts to account for the
         possible biological effects of the organic and carbon core
         fractions did not appreciably change the unit risk estimate.

    --   The risk estimates based on the results of bioassyas are not
         inconsistent with the available human experience.

         Nevertheless, a number of uncertainties remain, the most
    significant of which are:

    --   In any interspecies extrapolation there may be an inherent
         difference in sensitivity to the agent being assessed.

    --   The assumption of equivalent sensitivity across species is based 
         on concentration per unit of lung surface area, and use of other
         assumptions of dose equivalence may lead to different estimates
         of risk; however, the estimates should not vary by more than one
         order of magnitude.

    --   Although linearized low-dose extrapolation methods are used, it
         is still uncertain whether inflammatory cells secrete mediators
         that induce cancer in lung epithelial cells when the particle
         burden is smaller than that necessary to inhibit clearance. Even
         if macrophages are activated at low particle burdens, it is
         uncertain whether the responses of epithelial cells are linear at
         very low concentrations; however, non-linear carcinogenic
         responses have been observed in bioassays.

    --   Although the unit risk estimate is corroborated in the
         alternative model, uncertainty about the response at low doses
         remains because the estimates in the alternative model are based
         on the assumption that particles continue to induce cell
         initiation and/or proliferation at low doses. The actual risk
         would be much lower if this assumption does not hold.

         The risk at low doses derived from the linearized multistage
    model may thus be overestimated if particles no longer induce cell
    initiation and/or proliferation. The model was selected for
    calculating risk because a conservative model was needed in order to
    ensure protection of public health and because adequate data were not
    available to use fully the alternative model constructed for this
    assessment. Thus, the risk derived from the bioassays should be viewed
    as hypothetical.

         These unit risk estimates should not be used to evaluate the
    carcinogenic risk of other types of particulate matter present in
    ambient air, which may have different solubilities, surface areas, and
    free radical contents, which factors greatly affect carcinogenic
    potency.

         It could be argued that since the types and location of tumours
    seen in rats after exposure to particles are different from those
    found in humans, the experimental data are not relevant for humans.
    The tumours diagnosed in rats after long-term exposure to carbon black
    particles and diesel exhaust include benign adenomas and malignant
    adenocarcinomas, squamous-cell carcinomas, adenosquamous carcinomas,
    and squamous cysts (Mauderly et al., 1994; see section B7.3.2). After
    injection into athymic mice, cells from 50-67% of squamous-cell
    carcinomas and 25-40% of adenocarcinomas were found to grow (Table
    35), whereas those from squamous cysts did not (Mauderly et al.,
    1994).

         Squamous cysts have been defined by other authors as benign
    cystic keratinizing squamous-cell tumours (Kittel et al., 1993;
    Dungworth et al., 1994; Heinrich et al., 1995) and found not to grow
    after transplantation (Heinrich et al., 1995). As pointed out by Mohr
    (1992), there is, however, controversy about the correct terminology
    of this type of lesion. In 1995, an international group of
    pathologists re-evaluated lung tissue sections from rats exposed to
    particles by inhalation and identified four distinct lesions:
    squamous-cell metaplasia, pulmonary keratinizing cyst, cystic
    keratinizing epithelioma (considered to be benign), and squamous-cell
    carcinoma (G. Oberdörster, personal communication).

         Regardless of the correct classification of lesions, the
    important fact is that malignant tumours are induced in rats after
    long-term inhalation of deisel exhaust and carbon black. The tumour
    response of human beings may not, however, be the same, either
    qualitatively or quantitatively. In fact, the response of rats to
    long-term exposure to high concentrations of particles differs from
    that of other species, including mice and hamsters. In the absence of
    sufficient information about specific mechanisms unique to rats,
    however, there is no justification for excluding data on this species
    from extrapolations to the human situation (US Environmental
    Protection Agency, 1986).

         With respect to the mechanisms of induction of lung tumours by
    particles in rats, it has been shown that particles devoid of
    polyaromatic compounds can increase mutation rates in pulmonary
    epithelial cells, which is an important step in tumour development via
    cell transformation. In a recent study in which rats were exposed to
    carbon black for 13 weeks, a significant influx of inflammatory cells
    was seen at concentrations of 7 and 50 mg/m3 but not at 1 mg/m3
    (Oberdörster et al., 1995), and a significant, three- to fourfold
    increase in the frequency of  hprt mutations was seen in alveolar
    epithelial cells (Driscoll et al., 1995 and in press; see section
    B7.6). Thus, burdens of particles of low toxicity per se, similar to
    those reached in long-term studies in rats, can be mutagenic, possibly
    through the involvement of DNA damaging oxidants derived from
    inflammatory cells.

    Appendix B10.1  Construction of a biologically based (alternative)
                    model

    1.  Preliminary considerations

         In order to evaluate the effects of various biological
    assumptions on the assessments of the risk of exposure to diesel
    exhaust, a mathematical dose-response model must be constructed that
    takes into account the proposed biological mechanisms. As a
    significant issue in assessing the risk of diesel exhaust is the
    effect of lung overloading on tumour induction, the model should have
    the following properties:

    --   It should depend on dose metrics for both organic compounds and
         the carbon core, and it should account for the contributions of
         each to tumour induction and formation both separately and
         jointly.

    --   It should allow for changes in the model parameters with time due
         to increasing lung burden during exposure.

    --   It should view cell proliferation and tumour induction and
         formation stochastically: it is not realistic to assume
         deterministic clonal growth. For instance, it should not be
         assumed that all cells divide at the same age.

         It is therefore assumed that a normal cell can be initiated by
    both organic compounds and the carbon core. The initiation rate is
    denoted by µ1, which is a function of the background rate and that
    induced by diesel exhaust (as specified below). Because an initiated
    cell eventually either dies or enters the cell cycle (including cells
    in quiescence, G0), it is reasonable to assume that the lifetime of
    an initiated cell follows a certain probability distribution. In this
    model, a cell in G0 phase is equivalent to one with a certain
    probability of a very long lifetime (i.e. in the right-hand tail of
    the distribution of the cell's lifetime). At the end of its lifetime,
    it either dies (death) with probability  b, divides into two daughter
    cells (birth) with probability  a, or divides into one initiated cell
    and one malignant cell (second transition) with probability µ2;
    alpha + ß + µ2 = 1. Instead of assuming that a single malignant cell
    is equivalent to a tumour, as in the model proposed by Moolgavkar &
    Venzon (1979) and Moolgavkar & Knudson (1981), it is assumed here that
    a malignant cell has a certain probability of becoming a tumour; this
    probability is assumed to be dose-dependent, thus allowing for an
    evaluation of the effect of dose on tumour progression. It should be
    noted that the proposed model does not exclude the possibility that
    there may be more than one step in the 'initiation' of a normal cell.
    The rate of initiation used in the model should be viewed as a net
    rate that represents several genetic alterations and repairs.

    2.  Mathematical model and relationship of parameters to lung burden

         A dose-response function  P(t:d,D) is constructed for the
    probability of cancer by time (age)  t, which depends on both organic
    compounds,  d, and particles (carbon core),  D, and incorporates the
    biologically based concept discussed above. Because the model
    parameters that are not observed directly in the laboratory can be
    estimated statistically only from the results of bioassays at high
    concentrations, the model should not be considered a real model of
    diesel-induced carcinogenesis; uncertainty about extrapolation to low
    doses remains.

         A model with the features presented above was originally proposed
    by Chen & Farland (1991) and was extended into one with time-variable
    parameters by Tan & Chen (1992). This model was used as the basis for
    constructing a biologically based dose-response model. A brief
    mathematical description is presented in Appendix B10.3.

         The data on time to event from Mauderly et al. (1987) were used
    to estimate model parameters. These data are useful because they
    contain information on natural mortality and serial sacrifice of
    animals with and without (malignant) tumours, which is valuable for
    estimating tumour latency. In order to use this information to
    calculate maximum likelihood estimates of parameters, an 'E-M'
    algorithm was derived. In the E-M algorithm, each iteration involves
    an 'expectation' step (E) and a 'maximization' step (M) (see Appendix
    B10.2).

    (a)  Model parameters and notations

         The following parameters are incorporated into the dose-response
    model, which includes the rate of initiation (µ1), the rate of
    proliferation (gammaalpha), the rate of conversion (gammaµ2), and
    the probability of progression ( q). The rate of death of the
    initiated cells is implicitly defined by gamma(1-µ2-alpha). The
    parameters are all dose dependent.

     D:      dose of carbon core in milligrams per square centimetre of
            lung epithelial surface; varies over time;

     d:      dose of organic compounds in milligrams per square centimetre
            of lung epithelial surface;

     µ1:     dose-related initiation rate (per cell per day); depends on
            µ0 (background rate),  d, and  D by µ1 = µ0 (1 +  ad +  bD), where
             a and  b are parameters to be estimated statistically;

     µ2:     probability that a malignant cell will be produced by the end
            of the lifetime of an initiated cell;

     alpha:   probability that an initiated cell will divide into two
            daughter cells by the end of its lifetime;

     q:      probability that a single malignant cell will develop into a
            malignant tumour;

    gamma:  1/gamma is the mean lifetime of an initiated cell in days; the
            lifetime ends when the cell goes into mitosis or dies. If it
            is assumed that the probability that a cell will go into
            mitosis is about the same as the probability that it will die,
            the mean cell lifetime can be conveniently interpreted as time
            to mitosis (i.e. cell turnover time); thus, a shorter cell
            lifetime implies more frequent cell division. Time to mitosis
            is a random variable, not a fixed constant as in the
            assumption of the model of Greenfield et al. (1984), which was
            used by Cohen & Ellwein (1988) to analyse the results of
            bioassays to dectect urinary bladder cancer.

     N:      number of (normal) target cells.

    (b)  Practical considerations

         The E-M algorithm developed in Appendix B10.2 is an elegant
    procedure that can be used by statistical theory alone to test
    hypotheses about whether a particular parameter is influenced by
    organic compounds and the carbon core, individually or together. For
    instance, it could be postulated that the parameter the reciprocal of
    which represents mean cell lifetime is given by gamma( d,Di) = gamma0 +
    gamma11 d + gamma12 Di, and then proceed to test the null hypothesis
    that gamma11 = 0, i.e. no effect of organic compounds on the cell
    lifetime. This temptation must, however, be resisted, because too many
    parameters would have to be be estimated. Therefore, the biologically
    plausible assumption that parameters  q and gamma depend only on the
    lung burden of the carbon core,  D,
    is used.

         The study of Mauderly et al. (1987) lasted about 940 days. In
    order to construct a dose-response model including time-dependent lung
    burden, the time interval (0-940) is divided into five sub-intervals,
    each spanning six months, except for the last, which spans 730-940
    days. The deposition-retention model developed by Yu et al. (1991),
    corresponding to a concentration of diesel exhaust emissions in
    ambient air of milligrams per cubic metre, is used to calculate
    ( d,Di), where  i = 1, 2, ..., 5; the dose of organic compounds,
     d, does not change with time because it reaches a steady state soon
    after exposure begins; and  Di is the lung burden of carbon core
    during the  ith sub-interval.

         The assumptions for the dose-parameters relationship are:

    --   The rate of initiation associated with a lung burden { d,Di,  i
         = 1, 2, ..., 5} is given by µi( d,Di) = µ0(1 +  a *  d +  b *  Di) for
          i = 1, 2, ..., 5. This is the only parameter that is assumed to
         depend on both  d and  D.

    --   The probability of tumour formation from a malignant cell is
         assumed to be dependent on lung burden,  D, from  q( Di) =  q0
          q0 Di,  i = 1, 2, ..., 5. In order to simplify calculation, the
         possibility that  q0 is also dependent on organic compounds,  d, is
         not considered.

    --   The cell lifetime, gamma, is assumed to be related nonlinearly to
         lung burden,  D, from gamma( Di) = gamma0 + gamma1 Log(1 +  Di),
          i = 1, 2, ..., 5.

         In order to reduce the number of parameters that are to be
    estimated from the data of Mauderly et al. (1987), some of the
    background parameters for the dose-response model (µ0,  q0, and
    gamma0) are estimated from the historical control rates for Fischer
    344 rats in the US National Toxicology Program (reconstructed from
    Portier et al., 1986). The dose-related parameters are then estimated
    with the E-M algorithm, which is described in Appendix B10.2. The
    parameters estimated for the model resulting from the tumour response
    data of Mauderly et al. (1987) and the corresponding dosimetric
    parameters (Table 52) are given in Table 53.

        Table 52.  Dosimetric parameters (milligrams per square centimetre of lung surface) used
               in modelling
                                                                                                

    Exposure   d             D1            D2            D3            D4            D5
    concn
    (mg/m3)
                                                                                                

    0.35       2.5 × 10-6    6.2 × 10-5    8.8 × 10-5    9.0 × 10-5    9.0 × 10-5    9.0 × 10-5

    3.50       3.6 × 10-5    7.5 × 10-4    2.4 × 10-3    3.9 × 10-3    5.3 × 10-3    6.3 × 10-3

    7.08       7.3 × 10-5    2.0 × 10-3    5.5 × 10-3    8.6 × 10-3    1.1 × 10-2    1.4 × 10-2
                                                                                                

     d, organic compounds;  Di, i = 1, 2, ..., 5, are average lung burdens of carbon core over
    five time intervals. Values calculated from the retention model of Yu et al. (1991)
    
    Table 53.  Maximum likelihood estimates for model parameters
                                                               

    Parameter                     Estimate
                                                               

    µ                             1.033 × 10-7
    a                             1.103 × 104
    b                             3.214 × 102
    µ2                            7.907 × 10-7
    q0                            1.035 × 10-1
    q1                            5.332 × 10-2
    gamma0                        1.662 × 10-2
    gamma1                        2.647 × 10-2
    alpha                         5.443 × 10-1
    N (given)                     8.80 × 107
                                                               

    For definitions of parameters, see text. Background parameters µ0, q0,
    and gamma0 are estimated separately from historical control data from
    the US National Toxicology Program. The number of target cells, N, is
    assumed to be 10 times the number of type II cells in mice, which is
    given by Kauffman (1974). It is not essential that N be given
    accurately because Nµ0 appears as a single term in the model; the
    estimated µ0 will compensate for the underestimation or overestimation
    of N.

    Appendix B10.2  E-M algorithm

         The E-M algorithm, derived below, was used to calculate maximum
    likelihood estimates of the parameters for the alternative model. The
    data used were taken from Mauderly et al. (1987) and include the time
    when an animal died, naturally or at sacrifice, with or without
    (malignant) tumours. The theory of the algorithm is given by Dempster
    et al. (1977).

         Assume that the distinct times at which animals died are  t1
    < t2 < ...< tm. The observations can be classified as follows:

     a1x (i):    observed number of natural deaths without tumour at time
                ti in treatment group  x (The four groups are  x =
               1, 2, 3, 4.)

     a2x (i):    observed number of natural deaths with tumour at time
                ti in treatment group  x

     b1x (i):    series sacrificed at time  ti without tumour in
               treatment group  x

     b2x (i):    series sacrificed at time  ti with tumours in
               treatment group  x.

         Let  Td represent the time an animal died and  T the time a
    tumour developed.

    alphax (i) =  Pr Td =  ti |  Td >  ti,  T >  ti,,  x} (conditional probability
                of death without tumour)

    ßx( i |  u) =  Pr Td =  ti |  Td >  ti,  T Epsolin ( tu-1,  tu],  x} (related
                to deaths with tumours)

    Define

    EQUATION 1

    and

     Sx (t) =  Pr T >  t |  x} = exp[- t0integralt  h(x)dx].

         The function  Sx (t) is the probability of being tumour-free by
    time  t.

         The exact forms of the hazard function,  h(x), and  Sx (t) are
    given in Appendix B10.3.

    Let

     a2x( i |  u) =   number of natural deaths at  ti, with a tumour
                  developing during ( tu-1,  tu], in treatment group
                   x, u <  i,

     b2x( i |  u) =   number of animals sacrificed at  ti, with a tumour
                  developing during ( tu-1,  tu], in treatment group
                   x, u <  i.

    Then

    EQUATION 2

    Let

    EQUATION 3

    and

    EQUATION 4

    where

    EQUATION 5

         Given  a2x( x), { a2x( i |  u),  u = 1, ...,  i} is an ( i-1)-dimension
    multinomial with parameter { a2x( i),  Px( i |  u),  u = 1, ...,  i}.

    Thus,

          E[ a2x( i |  u) |  a2x( i)] =  a2x( i) Px( i |  u).

         Similarly, { b2x( i |  u),  u = 1, ...,  i}, is an ( i - 1)-dimension
    multinomial with parameters { b2x( i),  Qx( i |  u),  u = 1, ...,  i}, and

          E[ b2x( i |  u) |  b2x( i)] =  b2x( i)  Qx( i |  u).

         It can be shown that the likelihood function is proportional to

    EQUATION 6

    where

    EQUATION 7

    Let

    EQUATION 8

    Let

    EQUATION 9

    be a vector of parameters in function  S;

    alphax =  [alphax(1), alphax(2), ..., alphax( m)], and
    ßx( u)  =  [ßx(1 |  u), ßx(2 |  u, ..., ßx( m |  u)]

    be vectors of parameters related to conditional probabilities of death
    with and without tumours. These parameters and those in thetax were
    estimated by the E-M algorithm described below.

    The M step:

         Given initial values  a2x( i |  u) and  b2x( i |  u), estimate

    EQUATION 10

    The E step:

    Given the estimated values of alphax( i), ßx( i), and thetax from the M
    step, compute  Px( i |  u) and  Qx( i |  u) and obtain estimates of
     a2x( i |  u) and  b2x( i |  u) by

    EQUATION 11

         With the estimated values of  a2x( i |  u) and  b2x( i |  u) available
    from the E step, go back to the M step and repeat the process until
    the estimates are stable.

    Appendix B10.3  A tumour growth model

         A tumour growth model with piece-wise constant parameters taken
    from Tan & Chen (1992) is an extension of a stochastic model developed
    by Chen & Farland (1991). The biological justification of this model
    is similar to that of the two-stage model proposed by Greenfield et
    al. (1984), which was used by Cohen & Ellwein (1988) to analyse
    urinary bladder tumour occurrence. The two models differ, however,
    with respect to their mathematical formulations; the one adopted in
    this report is a stochastic model, whereas the other is a
    deterministic model and does not allow for estimation of parameters
    because it does not have complete mathematical expression.

         Although its most general form is not used here because of lack
    of data, the stochastic model of Chen & Farland (1991) has two
    desirable features: (i) it allows for any cell growth distribution
    (e.g. Gompertz), rather than only exponential distribution as in other
    models; and (ii) it incorporates the birth and death of tumour cells,
    rather than assuming that a tumour is born once a single tumour cell
    occurs, as did Moolgavkar & Venzon (1979) and Moolgavkar & Knudson
    (1981). Therefore, if information on cell lifetime distribution and
    the progression of tumour development is available, a reasonably
    realistic model can be constructed.

         For completeness, a brief description of the model is presented
    here. The following notations are required:

     N(t):         number of normal (target) cells at time  t,
    µ1:           rate of initiation, and
    contour       the probability density function for the lifetime of an
    integral (t):  an initiated cell.

         At the end of its lifetime, an initiated cell either divides
    (mitosis) or dies (programmed or nonprogrammed death). If it enters
    mitosis, it either divides either into two initiated cells with
    probability a or into one initiated cell and one malignant cell with
    probability µ2. At the end of a cell's lifetime, the probability of
    it dying is ß = 1 - alpha - µ2. A similar set-up (to allow for any
    cell lifetime distribution) can be made for a malignant cell; however,
    we confined ourselves to a simpler version, assuming that the lifetime
    of a malignant cell follows an exponential distribution. Thus, we
    assume that a malignant cell follows a simple birth-death process; it
    can either divide into two malignant cells at a rate alpham or die
    at a rate ßm.

         When the parameters are constant over time (age), the hazard
    function is given by

            h(t) = µ1µ2 0integralt  N(s)m( t - s)  ds

    where

    EQUATION 13

    where  y1 <  y2 are two real roots of alpha y2 - (alpha + ß + µ2 q) y +
    ß = 0; alpha + ß + µ2

    = 1;  q = 1 - ßm/alpham;  A(t) = / 0integralt  a(x)d x, where  a(t) =
    contour integral( t)/[1 -  F(t)] is the hazard function of the cell
    lifetime and F(t) is the cumulative function of contour integral (t).

    Two cases of special interest are  a(t) = gamma when an exponential
    distribution is assumed and  a(t) = exp(-gamma t), when the Gompertz
    distribution is assumed.

         When an exponential distribution (i.e.  a(t) = gamma or  A(t) =
    gamma t) and  q = 1 are assumed, the model is equivalent to the
    model of Moolgavkar & Venzon (1979) and Moolgavkar & Knudson (1981). A
    special case that may be more appropriate than the exponential
    distribution is that when the Gompertz distribution is assumed (i.e.
     A(t) = [1 - exp(-gamma t)]/gamma).

         For the model with time-dependent parameters, assume that the
    study begins at time t0. Divide the time scale ( t0,  t] into
     k sub-intervals  Lj = ( tj-1,  tj],  j = 1, 2, ... k-1 and  Lk = ( tk-1,  tk]
    where  tk =  t. (Note that these sub-intervals may not be the same as
    those defined by deaths or sacrifice previously.) The parameters that
    vary over sub-intervals ( ti-1,  ti],  i = 1, 2, ...,  k are µ1j, alphaj,
    ßj, µ2j,  N, and those parameters related to contour integral( t). The
    hazard functionis given by

    EQUATION 14

    where

    EQUATION 15

    and

    EQUATION 16

    where  y1j <  y2j are two real roots of a alpha y2 - (alphaj + ßj +
    µ2j qj) y + ßj = 0; alphaj + ßj + µ2j = 1;  qj = 1 - ßmj/alphamj,  j = 1,
    2, ...,  k.

         When an exponential distribution (i.e.  Aj( t) = gammaj( t) and
     qj = 1) is assumed, the model is equivalent to the model of
    Moolgavkar & Venzon (1979) and Moolgavkar & Knudson (1981), with
    piece-wise constant parameters. A special case that may be more
    appropriate than the exponential distribution is that when the
    Gompertz distribution is assumed (i.e.  Aj( t) = {1 - exp[-gammaj t)]}
    /gammaj).

         In the alternative model for exposure to diesel exhaust, in which
    the total time is divided into five (i.e.  k = 5) sub-intervals, Tan
    & Chen (1992) showed that, under the assumption of exponential cell
    lifetime distribution, the tumour-free distribution function,  Sx( t),
    can be written:

    EQUATION 17

    where  sj =  tj if  j <  k and  sj =  t if  j = k, and

    EQUATION 18

    where,

     WI            = [alpha + ß + µ2 q)2 - 4alphaß]´
     ZI            = alpha - ß - µ2 q, and
    DeltaI( s,t)   = gammai( t - s) if both  s and  t are in the same closed
                    sub-interval [ ti-1,  ti] and

    EQUATION 19

    if  sepsilon  Li,  tepsilon Lj with  tj <  tj.

    B11.  RECOMMENDATIONS

    B11.1  Recommendations for the protection of human health

         Diesel exhaust contributes to the total effect of combustion
    products on the environment. The data reviewed in this monograph
    support the conclusion that inhalation of diesel exhaust is of concern
    with respect to both neoplastic and non-neoplastic diseases. The
    particulate phase appears to have the greatest effect on health, and
    both the particle core and the associated organic materials have
    biological activity, although the gas-phase components cannot be
    disregarded. The following actions are recommended for the protection
    of human health.

    --   Diesel exhaust emissions should be controlled as part of the
         overall control of atmospheric pollution, particularly in urban
         environments.

    --   Emissions should be controlled strictly by regulatory inspections
         and prompt remedial action.

    --   Urgent efforts should be made to reduce emissions, specifically
         of particulates, by changing exhaust train techniques, engine
         design, and fuel composition.

    --   In the occupational environment, good work practices should be
         encouraged, and adequate ventilation must be provided to prevent
         excessive exposure.

    B11.2  Recommendations for the protection of the environment

         Too few data are available on which to base specific suggestions
    with regard to diesel exhaust, except as part of the general control
    of emissions.

    B11.3  Recommendations for further research

         The following recommendations are intended to help reduce the
    uncertainty associated with assessing the risks of exposure to diesel
    exhaust for human health.

    --   Research is required on methods for determining concentrations of
         diesel particulates in the presence of fine particles from other
         sources, in order to improve assessments of the exposures of
         occupational groups and the general population indoors and
         outdoors. The quality and quantity of emissions from engines that
         are not properly maintained or tuned should be investigated as
         part of these studies.

    --   The effects of diesel exhaust emissions on lung clearance should
         be studied in animals and humans at concentrations including
         those likely to be encountered by humans.

    --   The mechanisms involved in the etiology of tumours induced in
         rats exposed to particulate must be investigated by all available
         techniques.

    --   The relative roles of the gas and particulate phases of diesel
         exhaust in causing adverse effects after short- and long-term
         exposure should be investigated.

    --   Epidemiological investigations into the effects of diesel exhaust
         should be continued in order to assess issues of dose and latency
         in human populations with respect to lung cancer; and longer-term
         studies should be conducted in populations exposed to diesel
         exhaust with respect to noncarcinogenic effects.

    --   Further steps should be undertaken to improve the combustion and
         exhaust emission characteristics of diesel engines.

    B12.  PREVIOUS EVALUATIONS BY INTERNATIONAL BODIES

         The carcinogenic risks for human beings were evaluated by a
    working group convened by the International Agency for Research on
    Cancer in 1988 (International Agency for Research on Cancer, 1989b).
    The conclusions were:

         'There is  sufficient evidence for the carcinogenicity in
    experimental animals of whole diesel engine exhaust.

         'There is  inadequate evidence for the carcinogenicity in
    experimental animals of gas-phase diesel engine exhaust (with
    particles removed).

         'There is  sufficient evidence for the carcinogenicity in
    experimental animals of extracts of diesel engine exhaust particles.

         'There is  limited evidence for the carcinogenicity in humans of
    diesel engine exhaust.

         'There is  limited evidencefor the carcinogenicity in humans of
    engine exhausts (unspecified as from diesel or gasoline engines).

    'Overall evaluation

    'Diesel engine exhaust  is probably carcinogenic to humans (Group2A).'

    REFERENCES

    Acher AJ, Boderie P, & Yaron B (1989) Soil pollution by petroleum
    products: I. Multiphase migration of kerosene components in soil
    columns. J Contam Hydrol, 4: 333-345.

    Ackman RG & Noble D (1973) Steam distillation: A simple technique for
    recovery of petroleum hydrocarbons from tainted fish. J Fish Resour
    Board Can, 30: 711-714.

    Agency for Toxic Substances and Disease Registry (1995) Toxicological
    profile for fuel oils. Atlanta, GA, 204 pp.

    Ahlberg J, Ahlbom A, Lipping H, Norell S, & Österblom L (1981) [Cancer
    among professional drivers.] Laekartidn, 78: 1545-1546 (in Danish).

    Albrechcinski TM, Michalovic JG, Wattle BJ, & Wilkinson EP (1985) A
    laboratory investigation of the fate of diesel emissions in the
    atmosphere. Task 2 (Final Report No. CRC-APRAC-CAPA-13-76-05;
    PB 86-215837). Buffalo, NY, Calspan Corp., 163 pp.

    Amann CA & Siegla DC (1982) Diesel particulates -- What they are and
    why. Aerosol Sci Technol, 1: 73-101.

    American Automobile Manufacturers Association (1993) World motor
    vehicle data. Detroit, pp 9, 50, 51, 335.

    American Petroleum Institute (1979) Teratology study in rats, diesel
    fuel (Publication No. 27-32174). Washington DC, 23 pp.

    American Petroleum Institute (1980a) Acute toxicity tests:
    API 78-3, No. 2 home heating oil (10% cat.) (Publication
    No. 27-32773). Washington DC, 52 pp.

    American Petroleum Institute (1980b) Acute toxicity tests:
    API 78-2, No. 2 home heating oil (30% cat.) (Publication
    No. 27-32771). Washington DC, 48 pp.

    American Petroleum Institute (1980c) Acute toxicity tests:
    API 78-4, No. 2 home heating oil (50% cat.) (Publication
    No. 27-32068). Washington DC, 52 pp.

    American Petroleum Institute (1980d) Acute toxicity tests. API jet
    fuel A (Publication no. 27-32815). Washington DC, 48 pp.

    American Petroleum Institute (1980e) Mutagenicity evaluation of jet
    fuel A in the mouse dominant lethal assay. Final report (Publication
    No. 28-31345). Washington DC, 36 pp.

    American Petroleum Institute (1981) Mutagenicity evaluation of diesel
    fuel in the mouse dominant lethal assay, final report (Publication
    No. 33-30493). Washington DC, 9 pp.

    American Petroleum Institute (1982a) Acute toxicity studies:
    Hydrodesulfurized middle distillate, API sample 81-09 (Publication
    No. 30-32347). Washington DC, 24 pp.

    American Petroleum Institute (1982b) Acute toxicity studies:
    Hydrodesulfurized middle distillate, API sample 81-10 (Publication
    No. 30-32348). Washington DC, 26 pp.

    American Petroleum Institute (1982c) Acute toxicity studies
    hydrodesulfurized kerosine. Sample 81-07 (Publication No. 30-31986).
    Washington DC, 112 pp.

    American Petroleum Institute (1983a) LC50 acute inhalation toxicity
    evaluation of a petroleum derived hydrocarbon in rats:
    Hydrodesulfurized middle distillate, API sample 81-09 (Publication
    No. 30-32856). Washington DC, 47 pp.

    American Petroleum Institute (1983b) 28-Day dermal toxicity study in
    the rabbit: Hydrodesulfurized middle distillate, API sample 81-10
    (Publication No. 30-32298). Washington DC, 13 pp.

    American Petroleum Institute (1983c) LC50 acute inhalation toxicity
    evaluation of a petroleum derived hydrocarbon in rats (Publication
    no. 30-32855). Washington DC, 32 pp.

    American Petroleum Institute (1984) Dermal sensitization study in
    guinea pig, closed patch technique: Hydrodesulfurized middle
    distillate, sample 81-10 (Publication No. 31-31414). Washington DC,
    12 pp.

    American Petroleum Institute (1985a) 28-Day dermal toxicity study in
    the rabbit: API 83-08, light catalytically cracked distillate
    (CAS 64741-59-9) (Publication No. 32-32753). Washington DC, 27 pp.

    American Petroleum Institute (1985b) 28-Day dermal toxicity study in
    the rabbit: API 83-11, straight run middle distillate (CAS 64741-59-9)
    (Publication No. 32-32747). Washington DC, 29 pp.

    American Petroleum Institute (1985c) Acute oral toxicity study in
    rats, acute dermal toxicity study in rabbits, primary dermal
    irritation study in rabbits, primary eye irritation study in rabbits,
    dermal sensitization study in guinea pigs: API 83-07, light
    catalytically cracked distillate (CAS 64741-59-9) (Publication
    No. 33-30162). Washington DC, 41 pp.

    American Petroleum Institute (1985d) Acute oral toxicity study in
    rats, acute dermal toxicity study in rabbits, primary dermal
    irritation study in rabbits, primary eye irritation study in rabbits,
    dermal sensitization study in guinea pigs, API 83-11, straight run
    middle distillate (CAS 64741-44-2) (Publication No. 32-32857).
    Washington DC, 34 pp.

    American Petroleum Institute (1985e) Acute oral toxicity study in
    rats, acute dermal toxicity study in rabbits, primary dermal
    irritation study in rabbits, primary eye irritation study in rabbits,
    dermal sensitization study in guinea pigs. API 83-09, straight run
    kerosine (CAS No. 8008-20-6) (Publication No. 32-32858). Washington
    DC, 39 pp.

    American Petroleum Institute (1985f) The evaluation of the
    carcinogenicity of certain petroleum fractions (Publication
    No. 32-30964). Washington DC, 148 pp.

    American Petroleum Institute (1986a) Acute inhalation toxicity
    evaluation in rats: API 83-07, light catalytically cracked distillate
    (CAS 64741-59-9) (Publication No. 33-30549). Washington DC, 15 pp.

    American Petroleum Institute (1986b) Acute inhalation toxicity
    evaluation in rats: API 83-08, light catalytically cracked distillate
    (CAS 64741-59-9) (Publication No. 33-30444). Washington DC, 16 pp.

    American Petroleum Institute (1986c) Four week subchronic inhalation
    toxicity study in rats -- final report: API 81-07, hydrodesulfurized
    kerosene (petroleum) (CAS 64742-81-0); API 81-09, hydrodesulfurized
    middle distillate (petroleum) (CAS 64742-80-9); API 81-10,
    hydrodesulfurized middle distillate (petroleum) (CAS 64742-80-9)
    (Publication No. 33-32724). Washington DC, 31 pp.

    American Petroleum Institute (1987a) Acute inhalation toxicity
    evaluation of a petroleum derived hydrocarbon in rats: API 83-11,
    straight run middle distillate (CAS 64741-59-9) (Publication
    No. 34-30635). Washington DC, 53 pp.

    American Petroleum Institute (1987b) Acute inhalation toxicity
    evaluation of a petroleum derived hydrocarbon in rats -- API 83-09.
    Final report (Publication No. 34-30634). Washington DC, 10 pp.

    American Petroleum Institute (1987c) Mutagenicity of API 81-10,
    hydrode-sulfurized middle distillate (CAS 64742-80-9) (coded as
    API 86-10) in a mouse lymphoma mutation assay. Final report
    (Publication No. 34-32643). Washington DC, 21 pp.

    American Petroleum Institute (1989a) Twenty-four month dermal
    carcinogenesis/chronic toxicity screening bioassay of refinery streams
    in C3H/ Hej mice. API 190r. Final report: lifetime carcinogenicity
    evaluation (Publication No. 36-33220). Washington DC, 26 pp.

    American Petroleum Institute (1989b) Lifetime dermal
    carcinogenesis/chronic toxicity screening bioassay of refinery streams
    in C3H/HE mice. API 135r. Final report: lifetime carcinogenicity
    evaluation (Publication No. 36-31364). Washington DC, 41 pp.

    American Society for Testing and Materials (1988) ASTM Designation:
    D 975-81. Standard specification for diesel fuel oils. Philadelphia,
    Committee on Standards, pp 351-361.

    American Society for Testing and Materials (1992) ASTM Designation:
    D 975-92a. Standard specification for diesel fuel oils. Philadelphia,
    Committee on Standards, pp 310-311.

    Ames RG, Attfield MD, Hankinson JL, Hearl FJ, & Reger RB (1982) Acute
    respiratory effects of exposure to diesel emissions in coal miners.
    Am Rev Respir Dis, 125: 39-42.

    Ames RG, Hall DS, & Reger RB (1984) Chronic respiratory effects of
    exposure to diesel emissions in coal mines. Arch Environ Health,
    39: 389-394.

    Anderson JW (1977a) Effects of petroleum hydrocarbons on the growth of
    marine organisms. In: McIntyre AD & Whittle KJ, eds, Petroleum
    hydrocarbons in the marine environment (Rapports et proces-verbaux des
    réunions, Vol. 171). Copenhagen, International Council for the
    Exploration of the Sea, pp 157-165.

    Anderson JW (1977b) Responses to sublethal levels of petroleum
    hydrocarbons: Are they sensitive indicators and do they correlate with
    tissue contamination? In: Wolfe DA, ed., Fate and effects of petroleum
    hydrocarbons in marine ecosystems and organisms. Oxford, Pergamon
    Press, pp 95-114.

    Anderson JW (1979) An assessment of knowledge concerning the fate and
    effects of petroleum hydrocarbons in the marine environment. In:
    Vernberg WB, Calabrese A, Thurberg FP, & Vernberg FJ, eds, Marine
    pollution: Functional responses. New York, Academic Press, pp 3-21.

    Anderson JW, Neff JM, Cox BA, Tatem HE, & Hightower GM (1974)
    Characteristics of dispersions and water-soluble extracts of crude and
    refined oils and their toxicity to estuarine crustaceans and fish. Mar
    Biol, 27: 75-88.

    Anderson JW, Dixit DB, Ward GS, & Foster RS (1976) Effects of
    petroleum hydrocarbons on the rate of heart beat and hatching success
    of estuarine fish embryos. In: Vernberg FJ & Vernberg WB, eds,
    Pollution and physiology of marine organisms, Vol. 2. New York,
    Academic Press, pp 241-258.

    Anderson E, Albert R, Anderson L, Bayliss D, Bayard S, Chen C, Chu M,
    Gibb H, Haberman B, Hiremath C, McGaughy R, Singh D, & Thorslund T
    (1983) Quantitative approaches in use to assess cancer risk.
    Risk Anal, 3: 277-295.

    Apol A (1983) Health hazard evaluation. Regional transportation
    district; Denver (Report No. HETA 82-137-1264; PB84-209386).
    Cincinnati, OH, National Institute for Occupational Safety and Health,
    pp 1-23.

    Atlas RM & Bartha R (1973) Fate and effects of polluting petroleum in
    the marine environment. Res Rev, 49: 49-84.

    Atlas RM, Boehm PD, & Calder JA (1981) Chemical and biological
    weathering of oil, from the Amoco Cadiz oil spillage, within the
    littoral zone. Estuarine Coastal Mar Sci, 12: 589-608.

    Attfield MD, Trabant GD, & Wheeler RW (1982) Exposure to diesel fumes
    and dust at six potash mines. Ann Occup Hyg, 26: 817-831.

    Auerbach O & Garfinkel L (1991) The changing pattern of lung
    carcinoma. Cancer, 68: 1973-1977.

    Bailey MR, Fry FA, & James AC (1985a) Long-term retention of particles
    in the human respiratory tract. J Aerosol Sci, 16: 295-305.

    Bailey MR, Hodgson A, & Smith H (1985b) Respiratory tract retention of
    relatively insoluble particles in rodents. J Aerosol Sci, 16: 279-293.

    Baker JM (1970) The effect of fuel oils on plants. Environ Pollut,
    1: 27-44.

    Ball JC & Young WC (1992) Evidence for a new class of mutagens in
    diesel particulate extracts. Environ Sci Technol, 26: 2181-2186.

    Ball JC, Greene B, Young WC, Richert JFO, & Salmeen IT (1990)
    S9-Activated Ames assays of diesel-particle extracts. Detecting
    indirect-acting mutagens in samples that are direct-acting. Environ
    Sci Technol, 24: 890-894.

    Barbella R, Bertoli C, Ciajolo A, & D'Anna A (1988) Soot and unburnt
    liquid hydrocarbon emissions from diesel engines. Combust Sci Technol,
    59: 183-198.

    Barfknecht TR, Andon BM, Thilly WG, & Hites RA (1981) Soot and
    mutation in bacteria and human cells. In: Cooke M & Dennis AJ, eds,
    Chemical analysis and biological fate: Polynuclear aromatic
    hydrocarbons. Columbus, OH, Battelle Press, pp 231-242.

    Barnett CJ & Kontogiannis JE (1975) The effect of crude oil fractions
    on the survival of a tidepool copepod,  Tigriopus californicus.
    Environ Pollut, 8: 45-54.

    Barnhart MI, Chen S-T, Salley SO, & Puro H (1981) Ultrastructure and
    morphometry of the alveolar lung of guinea pigs chronically exposed to
    diesel engine exhaust: six months' experience. J Appl Toxicol,
    1: 88-103.

    Barnhart MI, Salley SO, Chen S-T, & Puro H (1982) Morphometric
    ultrastructural analysis of alveolar lungs of guinea pigs chronically
    exposed by inhalation to diesel exhaust (DE). In: Lewtas J, ed.,
    Toxicological effects of emissions from diesel engines. New York,
    Elsevier, pp 183-200.

    Barrientos A, Ortuno MT, Morales JM, Martinez Tello F, & Rodicio JL
    (1977) Acute renal failure after use of diesel fuel as shampoo. Arch
    Int Med, 137: 1217.

    Battigelli MC (1965) Effects of diesel exhaust. Arch Environ Health,
    10: 165-167.

    Bauer MA, Utell MJ, Morrow PE, Speers DM, & Gibb FR (1986) Inhalation
    of 0.30 ppm nitrogen dioxide potentiates exercise-induced bronchospasm
    in asthmatics. Am Rev Respir Dis, 134: 1203-1208.

    Bauer HD, Dahmann D, Kollmeier FW, & Lindecke B (1990) [Diesel engine
    emissions in potash-salt mines.] Kompass, 100: 516-524 (in German).

    Bechtold WE, Henderson TR, & Brooks AL (1986) Isolation,
    identification and bacterial mutagenicity of 2-nitro-9-fluorenone from
    diesel-exhaust particle extracts. Mutat Res, 173: 105-109.

    Beck LS, Hepler DI, & Hansen KL (1984) The acute toxicology of
    selected petroleum hydrocarbons. In: MacFarland HN, Holdsworth CE,
    MacGregor JA, Call RW, & Kane ML, eds, Proceedings of the symposium:
    The toxicology of petroleum hydrocarbons, May 1984. Washington DC,
    American Petroleum Institute, pp 1-12.

    Behn U, Kaschani DT, Lützke K Rentel S & Burk HD (1985) ['PAH-sample
    system' for industrial plants. Trial and first results.] Erdöl Kohle
    Erdgas Petrochem, 38: 131 (in German).

    Behymer T & Hites R (1984) Similarity of some organic compounds in
    spark-ignition and diesel engine particulate extracts. Environ Sci
    Technol, 18: 203-206

    Beland FA, Heflich RH, Howard PC, & Fu PP (1985) The in vitrometabolic
    activation of nitro polycyclic aromatic hydrocarbons. In: Harvey RG,
    ed., Polycyclic hydrocarbons and carcinogenesis. Washington DC,
    American Chemical Society, pp 371-396.

    Beliles RP & Mecler FJ (1982) Inhalation teratology of jet fuel A,
    fuel oil and petroleum naphtha in rats. In: MacFarland HN, Holdsworth
    CE, MacGregor JA, Call RW, & Kane ML, eds, Proceedings of the
    symposium: The toxicology of petroleum hydrocarbons, May 1982.
    Washington DC, American Petroleum Institute, pp 234-238.

    Bellmann B, Muhle H, Creutzenberg O, & Mermelstein R (1992)
    Irreversible pulmonary changes induced in rat lung by dust overload.
    Environ Health Perspectives, 97: 189-191.

    Benhamou S, Benhamou E, & Flamant R (1988) Occupational risk factors
    of lung cancer in a French case-control study. Br J Ind Med,
    45: 231-233.

    Bennett D, Girling AE, & Bounds A (1990) Ecotoxicology of oil
    products: Preparation and characterisation of aqueous test media.
    Chemosphere, 21: 659-669.

    Bevan DR & Ruggio DM (1991) Bioavailability  in vivo of benzo[ a]pyrene
    adsorbed to diesel particulate. Toxicol Ind Health, 7: 125-139.

    Bice DE, Mauderly JL, Jones RK, & McClellan RO (1985) Effects of
    inhaled diesel exhaust on immune responses after lung immunization.
    Fundam Appl Toxicol, 5: 1075-1086.

    Biles R, McKee RH, Lewis SC, Scala RA, & DePass LR (1988) Dermal
    carcinogenic activity of petroleum-derived middle distillate fuels.
    Toxicology, 53: 301-314.

    Blackburn GR, Deitch RA, Schreiner CA, Mehlman MA, & Mackerer CR
    (1984) Estimation of the dermal carcinogenic activity of petroleum
    fractions using a modified Ames assay. Cell Biol Toxicol, 1: 67-80.

    Block RN, Allworth N, & Bishop M (1991) Assessment of diesel
    contamination in soil. In: Calabrese EJ & Kostecki PT, eds,
    Hydrocarbon contaminated soils, Vol. I. Remediation techniques,
    environmental fate, risk assessment, analytical methodologies,
    regulatory considerations. Chelsea, MI, Lewis Publishers, pp 135-148.

    Blome H, Heidermanns G, & Timmer J (1990) [Assessment of workplace
    atmospheres in case of diesel motor vehicles.] Staub Reinhalt Luft,
    50: 93-97 (in German).

    Blumer M, Souza G, & Sass J (1970) Hydrocarbon pollution of edible
    shellfish by an oil spill. Mar Biol, 5: 195-202.

    Boehm PD & Quinn JG (1974) The solubility behavior of No. 2 fuel oil
    in sea water. Mar Pollut Bull, 5: 101-105.

    Boffetta P, Stellman SD, & Garfinkel L (1988) Diesel exhaust exposure
    and mortality among males in the American Cancer Society prospective
    study. Am J Ind Med, 14: 403-415.

    Boffetta P, Harris RE, & Wynder EL (1990) Case-control study on
    occupational exposure to diesel exhaust and lung cancer risk. Am J Ind
    Med, 17: 577-591.

    Bohning D, Atkins H, & Cohn S (1982) Long-term particle clearance in
    man: Normal and impaired. Ann Occup Hyg, 26: 259-271.

    Bokn T (1987) Effects of diesel oil and subsequent recovery of
    commercial benthic algae. Hydrobiology, 151-152: 277-284.

    Bond JA, Mitchell CE, & Li AP (1983) Metabolism and macromolecular
    covalent binding of benzo[ a]pyrene in cultured Fischer-344 rat lung
    type II epithelial cells. Biochem Pharmacol, 32: 3771-3776.

    Bond JA, Butler MM, Medinsky MA, Muggenburg BA, & McClellan RO (1984)
    Dog pulmonary macrophage metabolism of free and particle-associated
    [14C]benzo[a]pyrene. J Toxicol Environ Health, 14: 181-189.

    Bond JA, Sun JD, Medinsky MA, Jones RK, & Yeh HC (1986a) Deposition,
    metabolism, and excretion of 1-[14C]nitropyrene and 1-[14C]nitropyrene
    coated on diesel exhaust particles as influenced by exposure
    concentration. Toxicol Appl Pharmacol, 85: 102-117.

    Bond JA, Sun JD, Mitchell CE, Dutcher JS, Wolff RK, & McClellan RO
    (1986b) Biological fate of inhaled organic compounds associated with
    particulate matter. In: Lee SD, Schneider D, Grant JD, & Verkerk PJ,
    eds, Aerosols. Chelsea, MI, Lewis Publishers, pp 579-592.

    Bond JA, Wolff RK, Harkema JR, Mauderly JL, Henderson RF, Griffith WC,
    & McClellan RO (1988) Distribution of DNA adducts in the respiratory
    tract of rats exposed to diesel exhaust. Toxicol Appl Pharmacol,
    96: 336-346.

    Bond JA, Harkema JR, Henderson RF, Mauderly JL, McClellan RO, & Wolff
    RK (1989) Molecular dosimetry of inhaled diesel exhaust. In: Mohr U,
    Bates DV, Dungworth DL, Lee PN, McLellan RO, & Roa FJC, eds,
    Assessment of inhalation hazards: Integration and extrapolation using
    diverse data. Berlin, Springer, pp 315-324.

    Bond JA, Johnson NF, Snipes MB, & Mauderly JL (1990a) DNA adduct
    formation in rat alveolar type II cells: Cells potentially at risk for
    inhaled diesel exhaust. Environ Mol Mutag, 16: 64-69.

    Bond JA, Harkema JR, Henderson RF, Mauderly JL, McClellan RO, & Wolff
    RK (1990b) The role of DNA adducts in diesel exhaust-induced pulmonary
    carcinogenesis. Prog Clin Biol Res, 340C: 259-269.

    Bond JA, Mauderly JL, & Wolff RK (1990c) Concentration- and
    time-dependent formation of DNA adducts in lungs of rats exposed to
    diesel exhaust. Toxicology, 60: 127-135.

    Booth M & Reglitzky AA (1991) Diesel fuel quality in an
    environmentally conscious world. In: The diesel engine -- energy stake
    and environmental constraints. Cologne, TÜV Rheinland-OPET,
    pp 289-299.

    Boudet F, Fabre M, Boe M, Delon M, Ruiz J, & Lareng L (1983) Toxic
    pneumopathy after voluntary ingestion of one-and-a-half litres of
    diesel fuel. Toxicol Eur Res, 5: 247-249.

    Braddock JN & Perry NK Jr (1986) Gaseous and particulate emissions
    from gasoline- and diesel-powered heavy duty trucks (Paper
    No. 860617). In: International Congress and Exposition, Detroit, MI,
    24-28 February 1986. Warrendale, PA, Society of Automotive Engineers,
    39 pp.

    Brightwell J, Fouillet X, Cassano-Zoppi A-L, Gatz R, & Duchosal F
    (1986) Neoplastic and functional changes in rodents after chronic
    inhalation of engine exhaust emissions. In: Ishinishi N, Koizumi A,
    McClellan RO, & Stöber W, eds, Carcinogenic and mutagenic effects of
    diesel engine exhaust. Amsterdam, Elsevier, pp 471-485.

    Brightwell J, Fouillet X, Cassano-Zoppi A-L, Bernstein D, Crawley F,
    Duchosal F, Gatz R, Perczel S, & Pfeifer, H (1989) Tumours of the
    respiratory tract in rats and hamsters following chronic inhalation of
    engine exhaust emissions. J Appl Toxicol, 9: 23-31.

    Broman D, Näf C, Lundbergh I, & Zebühr Y (1990) An in situ study on
    the distribution, biotransformation and flux of polycyclic aromatic
    hydrocarbons (PAHs) in an aquatic food chain (seston  Mytilus edulis L.
     Somateria mollissima L.) from the Baltic: An ecotoxicological
    perspective. Environ Toxicol Chem, 9: 429-442.

    Bruner RH (1984) Pathologic findings in laboratory animals exposed to
    hydrocarbon fuels of military interest. In: Mehlman MA, Hemstreet GP,
    Thorpe JT, & Weaver NK, eds, Renal effects of petroleum hydrocarbons
    (Advances in modern environmental toxicology, Vol. 7). Princeton, NJ,
    Princeton Scientific Publishers, pp 133-140.

    Bruner RH, Kinkead ER, O'Neill TP, Flemming CD, Mattie DR, Russell CA,
    & Wall HG (1993) The toxicologic and oncogenic potential of JP-4 jet
    fuel vapors in rats and mice: 12-month intermittent inhalation
    exposures. Fundam Appl Toxicol, 20: 97-110.

    Bury RB (1972) The effects of diesel fuel on a stream fauna. Calif
    Fish Game, 58: 291-295.

    Byrne CJ & Calder JA (1977) Effect of the water-soluble fractions of
    crude, refined and waste oils on the embryonic and larval stages of
    the Quahog clam Mercenaria sp. Mar Biol, 40: 225-231.

    Campbell J, Scholl J, Hibbler F, Bagley S, Leddy D, Abata D, & Johnson
    J (1981a) The effect of fuel injection and timing on the physical,
    chemical, and biological character of particulate emissions from a
    direct injection diesel. In: Diesel combustion and emissions,
    Part III, SP-495. International Off-Highway Meeting and Exposition,
    Milwaukee, Wisconsin, 14-17 September 1981. Warrendale, PA, Society of
    Automotive Engineers, pp 71-108.

    Campbell KI, George EI, & Washington IS Jr (1981b) Enhanced
    susceptibility to infection in mice after exposure to dilute exhaust
    from light duty diesel engines. Environ Int, 5: 377-382.

    Campobosso O, Andrion A, Ribotta M & Ronco G (1993) The value of the
    1981 WHO histological classification in inter-observer reproducibility
    and changing pattern of lung cancer. Int J Cancer, 53:205-208

    Carey PM (1987) Air toxics emissions from motor vehicles (Technical
    report) (EPA-AA-TSS-PA-86-5). Ann Arbor, MI, US Environmental
    Protection Agency, 129 pp.

    Carpenter CP, Geary DL Jr, Myers RC, Nachreiner DJ, Sullivan LJ, &
    King JM (1976) Petroleum hydrocarbon toxicity studies. XI. Animal and
    human response to vapors of deodorized kerosene. Toxicol Appl
    Pharmacol, 36: 443-456.

    Carr RS & Reish DJ (1977) The effect of petroleum hydrocarbons on the
    survival and life history of polychaetous annelids. In: Wolfe D, ed.,
    Fate and effects of petroleum hydrocarbons in marine ecosystems and
    organisms. Oxford, Pergamon Press, pp 168-173.

    Carstens T & Sendstad E (1979) Oil spill on the shore of an
    ice-covered fjord in Spitzbergen. In: Proceedings of the 79th
    international conference on port and ocean engineering under Arctic
    conditions, Trondheim, 13-18 August. Trondheim, Norwegian Institute of
    Technology, pp 1227-1242.

    Casto BC, Hatch GG, Huang SL, Lewtas J, Nesnow S, & Waters MD (1981)
    Mutagenic and carcinogenic potency of extracts of diesel and related
    environmental emissions:  In vitro mutagenesis and oncogenic
    transformation. Environ Int, 5: 403-409.

    Castranova V, Bowman L, Reasor MJ, Lewis T, Tucker J, & Miles PR
    (1985) The response of rat alveolar macrophages to chronic inhalation
    of coal dust and/or diesel exhaust. Environ Res, 36: 405-419.

    Chan K-Y & Chiu SY (1985) The effects of diesel oil and oil
    dispersants on growth, photosynthesis, and respiration of  Chlorella
     salina. Arch Environ Contam Toxicol, 14: 325-331.

    Chan TL, Lee PS, & Hering WE (1981) Deposition and clearance of
    inhaled diesel exhaust particles in the respiratory tract of Fischer
    rats. J Appl Toxicol, 1: 77-82.

    Chan TL, Lee PS, & Hering WE (1984) Pulmonary retention of inhaled
    diesel particles after prolonged exposures to diesel exhaust. Fundam
    Appl Toxicol, 4: 624-631.

    Chen B & Farland W (1991) Incorporating cell proliferation in
    quantitative cancer risk assessment: Approach, issues, and
    uncertainties. In: Butterworth B, Slaga T, Farland W, & McClain M,
    eds, Chemically induced cell proliferation: Implications for risk
    assessment. New York, Wiley-Liss, pp 481-499.

    Chescheir GM, Garrett NE, Shelburne JD, Huisingh JL, & Waters MD
    (1981) Mutagenic effects of environmental particulates in the
    CHO\HGPRT system In: Waters MD, Sandhu SS, Huisingh JL, Claxton L, &
    Nesnow S, eds, Short-term bioassays in the analysis of complex
    environmental mixtures II, Vol. 2. New York, Plenum Press, pp 337-350.

    Chia F-S (1973) Killing of marine larvae by diesel oil. Mar Pollut
    Bull, 4: 29-30.

    Clark CR & Vigil CL (1980) Influence of rat lung and liver homogenates
    on the mutagenicity of diesel exhaust particulate extracts. Toxicol
    Appl Pharmacol, 56: 110-115.

    Clark CR, Walter MK, Ferguson PW, & Katchen, M (1988) Comparative
    dermal carcinogenesis of shale and petroleum-derived distillates.
    Toxicol Ind Health, 4: 11-22.

    Cohen BL (1994) Invited commentary: In defense of ecologic studies for
    testing a linear no threshold theory. Am J Epidemiol, 139: 765-768.

    Cohen S & Ellwein L (1988) Cell growth dynamics in long-term bladder
    carcinogenesis. Toxicol Lett, 43: 151-173.

    Coleman WE, Munch JW, Streicher RP, Ringhand HP, & Kopfler FC (1984)
    The identification and measurement of components in gasoline,
    kerosene, and No. 2 fuel oil that partition into the aqueous phase
    after mixing. Arch Environ Contam Toxicol, 13: 171-178.

    Coley TR (1989) Diesel fuel additives influencing flow and storage
    properties. In: Owen K, ed., Gasoline and diesel fuel additives. New
    York, John Wiley & Sons, pp 105-132.

    Commins BT, Waller RE, & Lawther PJ (1956) Smoke in a London diesel
    bus garage. An interim report. Br Med J, ii: 753-754.

    Conaway CC, Schreiner CA, & Cragg ST (1984) Mutagenicity evaluation of
    petroleum hydrocarbons. In: MacFarland HN, Holdsworth CE, MacGregor
    JA, Call RW, & Lane ML, eds, Applied toxicology of petroleum
    hydrocarbons (Advances in modern environmental toxicology, Vol 6).
    Princeton, NJ, Princeton Scientific Publishers, pp 89-107.

    CONCAWE (1985) Health aspects of petroleum fuels. Potential hazards
    and precautions for individual classes of fuels (Report No 85/51). The
    Hague, 54 pp.

    CONCAWE (1986) The relationship between automotive diesel fuel
    characteristics and engine performance (Report No. 86/65). The Hague,
    30 pp.

    CONCAWE (1987) Diesel fuel quality and its relationship with emissions
    from diesel engines (Report No. 10/87). The Hague, 21 pp.

    CONCAWE (1990a) Motor vehicle emission regulations and fuel
    specifications -- 1990 update (Report No. 2/90). Brussels, Health
    Management Group, 92 pp.

    CONCAWE (1990b) The sulphur content of diesel fuel and its
    relationship with particulate emissions from diesel engines (Report
    No. 90/54). Brussels, Health Management Group, 42 pp.

    CONCAWE (1991) Middle distillates -- A review of the results of a
    CONCAWE programme of short-term biological studies (Report No. 91/51).
    Brussels, Health Management Group, pp 1-13.

    CONCAWE (1992) The chemical composition of diesel particulate
    emissions (Report No. 92/51). Brussels, Health Management Group,
    24 pp.

    CONCAWE (1993) Ecotoxicological testing of petroleum products: test
    methodology (Report No. 92/56). Brussels, Petroleum Products Ecology
    Group, 18 pp.

    CONCAWE (1995) Kerosines/jet fuels (Product No. 94/106). Brussels,
    Petroleum Products and Health Managements Groups, 13 pp.

    CONCAWE (in press) Gas oils (diesel fuels/heating oils) Brussels,
    Petroleum Products and Health Management Groups, 45 pp.

    Cooper JR & Mattie DR (1993) Abstracts of the 32nd Annual Meeting of
    the Society of Toxicology. Toxicologist, 13: 78.

    Cordier S, Clavel J, Limaset JC, Boccon-Gibod L, Le Moual N, Mandereau
    L, & Hermon D (1993) Occupational risks of bladder cancer in France: A
    multicentre case-control study. Int J Epidemiol, 22: 403-411.

    Corner EDS, Harris RP, Kilvington CC, & O'Hara SCM (1976) Petroleum
    compounds in the marine food web: Short-term experiments on the fate
    of naphthalene in  Calanus. J Mar Biol, 56: 121-133.

    Cornwell RJ (1982) Health hazard evaluation determination report.
    Climax Molybdenum Company, Climax, CO (PB84-14850 1; Report
    No MHETA-81-108-9004). Morgantown, WV, National Institute for
    Occupational Safety and Health, 23 pp.

    Cowan MJ & Jenkins, LJ (1980) Navy toxicity study of shale and
    petroleum JP-5 aviation fuel and diesel fuel marine. In: Griest WH,
    Guerin MR, & Coffin DL, eds, Health effects investigation of oil shale
    development. Ann Arbor, MI, Ann Arbor Science Publishers, pp 129-139.

    Cox BA (1974) Responses of the marine crustaceans  Mysidopsis almyra
    Bowman,  Penaeus aztecus Ives, and  Penaeus setiferus (L.) to
    petroleum hydrocarbons. Diss Abstr Int, B35: 3739.

    Cox BA, Anderson JW, & Parker JC (1975) An experimental oil spill: The
    distribution of aromatic hydrocarbons in the water, sediment, and
    animal tissues within a shrimp pond. In: Proceedings of the Conference
    on Prevention and Control of Oil Pollution, 25-27 March 1975, San
    Francisco, California. Washington DC, American Petroleum Institute,
    pp 607-612.

    Crebelli R, Fuselli S, Conti G, Conti L, & Carere A (1991)
    Mutagenicity spectra in bacterial strains of airborne and engine
    exhaust particulate extracts. Mutat Res, 261: 237-248.

    Creutzenberg O, Bellmann B, Heinrich U, Fuhst R, Koch W, & Muhle H
    (1990) Clearance and retention of inhaled diesel exhaust particles,
    carbon black, and titanium dioxide in rats at lung overload
    conditions. J Aerosol Sci, 21: 455-458.

    Crisp AJ, Bhalla AK, & Hoffbrand BI (1979) Acute tubular necrosis
    after exposure to diesel oil. Br Med J, ii: 177.

    Croxall JP (1977) The effects of oil on seabirds. In: McIntyre AD &
    Whittle KJ, eds, Petroleum hydrocarbons in the marine environment
    (Rapports et proces-verbaux des réunions, Vol. 171). Copenhagen,
    International Council for Exploration of the Sea, pp 191-195.

    Crump KS (1984) A new method for determining allowable daily intakes.
    Fundam Appl Toxicol, 4: 854-871.

    Csicsaky M, Roller M, & Pott F (1993) [Quantitative risk assessment of
    selected carcinogenic substances in the work area (Special publication
    No. S/31)]. Dortmund, Federal Institute for Labour Protection,
    pp 18-29 (in German).

    Cuddihy RG, Griffith WC, & McClellan RO (1984) Health risks from
    light-duty diesel vehicles. Low risk of lung cancer from exposure to
    diesel exhaust is projected from laboratory and epidemiologic data.
    Environ Sci Technol, 18: 14A-21A.

    Curren RD, Kouri RE, Kim CM, & Schechtman LM (1981) Mutagenic and
    carcinogenic potency of extracts from diesel related environmental
    emissions: Simultaneous morphological transformation and mutagenesis
    in BALB/c3T3 cells. Environ Int, 5: 411-415.

    Custance SR, Sullivan MJ, McCaw PA, & Kopf AC (1993) Environmental
    fate of the chemical mixtures: Crude oil, JP-5, mineral spirits, and
    diesel fuel. In: Kostecki PT, & Calabrese EJ, eds, Hydrocarbon
    contaminated soils and groundwater, Vol. 3. Boca Raton, FL, Lewis
    Publishers, pp 205-212.

    Dalbey W, Lock S, Garfinkel S, Jenkins R, Holmberg R, & Guerin M
    (1982) Inhalation exposures of rats to aerosolized diesel fuel. In:
    MacFarland HN, Holdsworth CE, MacGregor JA, Call RW, & Kane ML, eds,
    Proceedings of the symposium on toxicology of petroleum hydrocarbons.
    Washington DC, American Petroleum Institute, pp 13-25.

    Dalbey W, Henry M, Holmberg R, Moneyhun J, Schmoyer R, & Lock S (1987)
    Role of exposure parameters in toxicity of aerosolized diesel fuel in
    the rat. J Appl Toxicol, 7: 265-275.

    Damber LA & Larsson LG (1987) Occupation and male lung cancer: A
    case-control study in northern Sweden. Br J Ind Med, 44: 446-453.

    Dambo WB (1993) Tolerance of the periwinkles Pachymelania aurita
    (Muller) and  Tympanotonus fuscatus (Linne) to refined oils. Environ
    Pollut, 79: 293-296.

    Daniel JH Jr (1984) Diesels in underground mining: A review and an
    evaluation of an air quality monitoring methodology (Document
    No. Rl-8884; US NTIS PB84-214444). Washington DC, US Department of the
    Interior, Bureau of Mines.

    Das PKMK & Konar SK (1988) Acute toxicity of petroleum products, crude
    oil and oil refinery effluent on plankton, benthic invertebrates and 
    fish. Environ Ecol, 6: 885-891.

    Das M & Misra MP (1988) Acne and folliculitis due to diesel oil.
    Contact Derm, 18: 120-121.

    Davis SJ & Gibbs CF (1975) The effect of weathering on a crude oil
    residue exposed at sea. Water Res, 9: 275-285.

    Davison W, Franklin CE, McKenzie JC, & Dougan MCR (1992) The effects
    of acute exposure to the water soluble fraction of diesel fuel oil on
    survival and metabolic rate of an Antarctic fish ( Pagothenia
     borchgrevinki). Comp Biochem Physiol C, 102: 185-188.

    Davison W, Franklin CE, McKenzie JC, & Carey PW (1993) The effects of
    chronic exposure to the water soluble fraction of fuel oil on an
    Antarctic fish Pagothenia borchgrevinki. Comp Biochem Physiol,
    C104: 67-70.

    Deichmann WB, Kitzmiller KV, Witherup S, & Johansmann R (1944)
    Kerosene intoxication. Ann Intern Med, 21: 803-823.

    Dempster A, Laird N, & Rubin D (1977) Maximum likelihood from
    incomplete data via the EM algorithm. R Stat Soc, B39: 1-38.

    Dennington VN, George JJ, & Wyborn CHE (1975) The effects of oils on
    growth of freshwater phytoplankton. Environ Pollut, 8: 233-237.

    Devesa SS, Shaw GL, & Blot WJ (1991) Changing patterns of lung cancer
    incidence by histological type. Cancer Epidemiol Biomarkers Prev,
    1: 29-34

    Dineen D (1991) Remediation options for diesel-contaminated soil. In:
    Calabrese EJ & Kostecki PT, eds, Hydrocarbon contaminated soils,
    Vol. 1. Remediation techniques, environmental fate, risk assessment,
    analytical methodologies, regulatory considerations. Chelsea, MI,
    Lewis Publishers Inc., pp 181-191.

    Dixit D & Anderson JW (1977) Distribution of naphthalenes within
    exposed  Fundulus similus and correlations with stress behavior. In:
    Proceedings of 1977 oil spill conference, New Orleans, March 1977.
    Washington DC, American Petroleum Institute.

    Dockery DW, Arden Pope C III, Xu X, Spengler JD, Ware JH, Fay ME,
    Ferris BG Jr, & Speizer FE (1993) An association between air pollution
    and mortality in six US cities. New Engl J Med, 329: 1753-1759.

    Dolan DF, Kittelson DB, & Pui, DYH (1980) Diesel exhaust particle size
    distribution measurement techniques (Paper No. 800187). Warrendale,
    PA, Society of Automotive Engineers, pp 149-164.

    Dorie LD, Bagley ST, Leddy DG, & Johnson JH (1987) Characterization of
    mutagenic subfractions of diesel exhaust modified by ceramic
    particulate traps. Environ Sci Technol, 21: 757-765.

    Driscoll KE, Carter JM, Howard BW, & Hassenbein DG (1994) Mutagenesis
    in rat lung epithelial cells after  in vivo silica exposure or
     ex vivo exposure to inflammatory cells. Am J Resp Crit Care Med,
    149: A553.

    Driscoll KE, Carter JM, Oberdörster G, Howard BW, & Hassenbein DG
    (1995) Inflammation, cytokine expression and mutagenesis in rat lung
    after carbon black inhalation. Toxicologist, 15: 46.

    Driscoll KE, Carter JM, Howard BW, Hassenbein D, Pepelko W, Baggs RB,
    & Oberdörster G (in press) Pulmonary inflammatory, chemokine and
    mutagenic responses in rats after subchronic inhalation of carbon
    black. Toxicol Appl Pharmacol.

    Du C-J, Kittelson DB, & Zweidinger RB (1984) Measurements of
    polycyclic aromatic compounds in the cylinder of an operating diesel
    engine (SAE Technical Paper No. 840364). In: International congress
    and exposition, Detroit, Michigan, 27 February-2 March 1984.
    Warrendale, PA, Society of Automotive Engineers, Engineering Resource
    for Advancing Mobility, 12 pp.

    Dungworth DL, Mohr U, Heinrich U, Ernst H, & Kittel B (1994)
    Pathological effects of inhaled particles in rat lungs: Associations
    between inflammatory and neoplastic processes. In: Mohr U, Dungworth
    DL, Mauderly JL, & Oberdörster, eds, Toxic and carcinogenic effects of
    solid particles in the respiratory tract. Washington DC, International
    Life Sciences Institute Press, pp 75-98.

    Dunlap LE & Beckmann DD (1988) Soluble hydrocarbons analysis from
    kerosene/diesel type hydrocarbons. In: Proceedings of the conference
    on petroleum hydrocarbons and organic chemicals in ground water:
    Prevention, detection and restoration, 9-11 November 1988, The Westin
    Galleria, Houston, Texas. Dublin, OH, National Water Well Association,
    pp 37-45

    Dutcher JS, Sun JD, Lopez JA, Wolf I, Wolff RK, & McClellan RO (1984)
    Generation and characterization of radiolabeled diesel exhaust. Am Ind
    Hyg Assoc J, 45: 491-498.

    Dziedzic D (1981) Differential counts of B and T lymphocytes in the
    lymph nodes, circulating blood and spleen after inhalation of high
    concentrations of diesel exhaust. J Appl Toxicol, 1: 111-115.

    Edgerton SA, Coutant RW, & Henley MV (1987) Hydrocarbon fuel spill
    dispersion on water: a literature review. Chemosphere, 16: 1475-1487.

    Edling C & Axelson O (1984) Risk factors of coronary heart disease
    among personnel in a bus company. Int Arch Occup Environ Health,
    54: 181-183.

    Edling C, Anjou C-G, Axelson O, & Kling H (1987) Mortality among
    personnel exposed to diesel exhaust. Int Arch Occup Environ Health,
    59: 559-565.

    Egebäck KE & Bertilsson BM (1983) Chemical and biological
    characterization of exhaust emissions from vehicles fueled with
    gasoline, alcohol, LPG and diesel (Report No. SNV PM 1635). Solna,
    National Swedish Environment Protection Board, 188 pp.

    El-Bayoumy K, Johnson BE, Roy AK, Upadhyaya P, & Partian SJ (1994)
    Biomonitoring of nitropolynuclear aromatic hydrocarbons via protein
    and DNA adducts (Research report No. 64). Cambridge, MA, Health
    Effects Institute, pp 1-39.

    Elbers G & Muratyan S (1991) [Problems of traffic-related outdoor air
    measurement of particles (diesel soot) (VDI Report No. 888)]. In:
    [Carcinogenic substances in the environment: Origin, measurement,
    risk, minimization.

    Meeting of the Commission for Clean Air of the Society of German
    Engineers (VDI) and the German Standards Institute (DIN), Mannheim,
    23-25 April]. Düsseldorf, Society of German Engineers Press,
    pp 143-170 (in German).

    Elbers G & Richter J (1994) [Measurement of traffic-induced soot
    immissions.] Staub Reinhalt Luft, 54: 19-24 (in German).

    Emmelin A, Nyström L, & Wall S (1993) Diesel exhaust exposure and
    smoking: A case-referent study of lung cancer among Swedish dock
    workers. Epidemiology, 4: 237-244.

    Engelhardt FR, Wong MP, & Duey ME (1981) Hydromineral balance and gill
    morphology in rainbow trout Salmo gairdneri, acclimated to fresh and
    sea water, as affected by petroleum exposure. Aquat Toxicol,
    1: 175-186.

    Environment Canada (1990) Canadian emissions inventory of common air
    contaminants -- 1985 (Environmental Protection Series No. EPS/5/AP/3),
    Ottawa.

    Environment Canada (1992) Summary of Canadian new vehicle emissions
    standards, June 1992. Unpublished results.

    Eppley ZA (1992) Assessing indirect effects of oil in the presence of
    natural variation: The problem of reproductive failure in South Polar
    skuas during the  Bahia Paraiso oil spill. Mar Pollut Bull,
    25: 307-312.

    Ernst VV, Neff JM, & Anderson JW (1977) The effects of the
    water-soluble fractions of No. 2 fuel oil on the early development of
    the estuarine fish,  Fundulus grandis Baird and Girard. Environ
    Pollut, 14: 25-35.

    European Commission (1993) Guideline 93/12/EEC of the Council from 23
    March 1993 on sulfur content of certain liquid fuels (EEC Directive,
    L74/81), Luxembourg, 3 pp.

    European Committee for Standardization (1993) [Automotive fuels,
    diesel. Requirements and methods of test. European Standard EN 590.]
    Brussels, 5 pp (in German).

    Fabri J, Dabelstein W, & Reglitzky A (1990) Motor fuels. In: Elvers B,
    Hawkins S, & Schulz G, eds, Ullmann's encyclopedia of industrial
    chemistry, 5th Ed, Vol A16. Weinheim, VCH, pp 719-753.

    Farrington JW, Davis AC, Frew NM, & Rabin KS (1982) No. 2 fuel oil
    compounds in  Mytilus edulis. Retention and release after an oil
    spill. Mar Biol, 66: 15-26.

    Federal Environment Agency (1991) [Air and noise levels caused by
    traffic -- emissions, immissions and effects (Report No. 40/91)].
    Berlin, 194 pp (in German).

    Federal Environment Agency (1992) [Transport in Germany-Energy
    consumption and emissions of air pollutants due to transport in German
    Democratic Republic, Berlin (East) and the Federal Republic of Germany
    in 1988 and in Germany in 2005 (UBA-FB 50473/85; Report No. 5/92)].
    Berlin (in German).

    Federal Ministry for Transport (1992) [Traffic in statistical
    numbers.] Berlin, German Institute for Economic Research,
    pp 132, 133, 280, 281 (in German).

    Federal Office for Motor Vehicles (1993) [Emission and noise values of
    passenger cars, light vans and station wagons with EEC licence], 3rd
    Ed. Flensburg, 260 pp (in German).

    Federal Office for Motor Vehicles (1994) [Emission values of vehicles
    with EEC licence], 4th Ed. Flensburg, 192 pp (in German).

    Flowers TH, Pulford ID, & Duncan HJ (1984) Studies on the breakdown of
    oil in soil. Environ Pollut Ser, B8: 71.

    Fortnagel M (1992) [Possibilities of the reduction of harmful
    substances of diesel engines from the view of the automotive
    industry.] In: [Effects of diesel engine exhausts on health]. Munich,
    Society for Research on Radiation and the Environment, Research Centre
    for Environment and Health, pp 53-56 (in German).

    Fossato VU & Canzonier WJ (1976) Hydrocarbon uptake and loss by the
    mussel  Mytilus edulis. Mar Biol, 36: 243-250.

    Frankenberger WT Jr, Emerson KD, & Turner DW (1989) In situ
    bioremediation of an underground diesel fuel spill: A case history.
    Environ Manage, 13: 325-332.

    Fraser Williams (Scientific Systems) Ltd (1985) Oil and hazardous
    materials: technical assistance data system (OHM/TADS;
    Version 5.00/9.5-December, 1985) Poynton, Cheshire.

    Frehland E, Weiss A, & Meduna V (1992) [Microbial communities in
    diesel contaminated soil]. In: Behrens D & Wiesner J, eds, [Microbial
    cleaning of soil].

    Frankfurt/Main, German Society for Chemical Apparatus, Chemical
    Technology and Biotechnology, pp 347-349.

    Froines JR, Hinds WC, Duffy RM, Lafuente EJ, & Liu WCV (1987) Exposure
    of firefighters to diesel emissions in fire stations. Am Ind Hyg
    Assoc J, 48: 202-207.

    Fromme H (1990) [The importance of diesel engine exhaust gases for
    ecological health protection and hygienic environmental protection.]
    Öff Gesundheitswesen, 52: 625-629 (in German).

    Gaddo P, Settis M, & Giacomelli L (1984) Artifact formation during
    diesel particulate collection. Orbassano, Fiat Engineering, Ecological
    Service, pp 437-449.

    Gairing M, Naber D, Schäfer A, Lange W, Reglitzky A & Le'Jeune A
    (1994) [The influence of fuel property on the exhaust emissions of
    modern diesel motors from Mercedes-Benz]. Motortech Z, 55: 3-11
    (in German, with English abstract)

    Galin T, McDowell C, & Yaron B. (1990) The effect of volatilization on
    the mass flow of a non-aqueous pollutant liquid mixture in an inert
    porous medium: Experiments with kerosene. J. Soil Sci, 41: 631-641.

    Gallagher J, George M, Kohan M, Thompson C, Shank T, & Lewtas J (1993)
    Detection and comparison of DNA adducts after  in vitro and  in vivo
    diesel emission exposures. Environ Health Perspectives, 99: 225-228.

    Gallagher J, Heinrich U, George M, Hendee L, Philips DH, & Lewtas J
    (1994) Formation of DNA adducts in rat lung following chronic
    inhalation of diesel emissions, carbon black and titanium dioxide
    particles. Carcinogenesis, 15: 1291-1299.

    Gamble JF & Jones WG (1983) Respiratory effects of diesel exhaust in
    salt miners. Am Rev Respir Dis, 128: 389-394.

    Gamble J, Jones W, & Minshall S (1987a) Epidemiological-environmental
    study of diesel bus garage workers: Acute effects of NO2 and
    respirable particulate on the respiratory system. Environ Res,
    42: 201-214.

    Gamble J, Jones W, & Minshall S (1987b) Epidemiological-environmental
    study of diesel bus garage workers: Chronic effects of diesel exhaust
    on the respiratory system. Environ Res, 44: 6-17.

    Garg BD (1983) Histologic quantitation of macrophage aggregates in the
    lungs of diesel exhaust exposed rats. Acta Stereol, 2
    (Suppl. 1): 235-238.

    Garshick E, Schenker MB, Muñoz A, Segal M, Smith TJ, Woskie SR,
    Hammond SK, & Speizer FE (1987) A case-control study of lung cancer
    and diesel exhaust exposure in railroad workers. Am Rev Respir Dis,
    135: 1242-1248.

    Garshick E, Schenker MB, Muñoz A, Segal M, Smith TJ, Woskie SR,
    Hammond SK, & Speizer FE (1988) A retrospective cohort study of lung
    cancer and diesel exhaust exposure in railroad workers. Am Rev Respir
    Dis, 137: 820-825.

    Gaworski CL, MacEwen JD, Vernot EH, Bruner RH, & Cowan MJ Jr (1984)
    Comparison of the subchronic inhalation toxicity of petroleum and oil
    shale JP-5 jet fuels. In: MacFarland HN, Holdsworth CE, MacGregor JA,
    Call RW, Kane ML, eds, Applied toxicology of petroleum hydrocarbons
    (Advances in modern environmental toxicology, Vol. 6). Princeton, NJ,
    Princeton Scientific Publishers, pp 33-47.

    Gearing PJ, Gearing JN, Pruell RJ, Wade TL, & Quinn JG (1980)
    Partitioning of No. 2 fuel oil in controlled estuarine ecosystems.
    Sediments and suspended particulate matter. Environ Sci Technol,
    14: 1129-1136.

    Gerde P, Medinsky A, & Bond JA (1991a) The retention of polycyclic
    aromatic hydrocarbons in the bronchial airways and in the alveolar
    region. A theoretical comparison. Toxicol Appl Pharmacol,
    107: 239-252.

    Gerde P, Medinsky MA, & Bond JA (1991b) Contemporary issues in
    toxicology. Particle-associated polycyclic aromatic hydrocarbons. A
    reappraisal of their possible role in pulmonary carcinogenesis.
    Toxicol Appl Pharmacol, 108: 1-13.

    Gerde P, Muggenburg BA, Sabourin PJ, Harkema JR, Hotchkiss JA, Hoover
    MD, & Henderson RF (1993a) Disposition of polycyclic aromatic
    hydrocarbons in the respiratory tract of the beagle dog. II. The
    conducting airways. Toxicol Appl Pharmacol, 121: 319-327.

    Gerde P, Muggenburg BA, & Henderson RF (1993b) Disposition of
    polycyclic aromatic hydrocarbons in the respiratory tract of the
    beagle dog. III. Mechanisms of the dosimetry. Toxicol Appl Pharmacol,
    121: 328-334.

    German Scientific Association for Petroleum, Natural Gas, and Coal
    (1991) [Middle distillates. Analytical investigations and mutagenicity
    studies. Research report No. 412-1.] Hamburg, 65 pp (in German).

    German Institute for Scientific Research (1993) [Transport in
    figures]. Berlin, p 133

    Gibson TL, Ricci AI, & Williams RL (1981) Measurement of polynuclear
    aromatic hydrocarbons, their derivatives, and their reactivity in
    diesel automobile exhaust. In: Cooke M & Dennis AJ, eds, Polynuclear
    aromatic hydrocarbons, fifth international symposium on chemical
    analysis and biological fate. Columbus, OH, Battelle Press,
    pp 707-717.

    Grant LD, Raub JA, Graham JA, Kotchmar DJ & Tilton BE (1993) Health
    effects of motor vehicle-related criteria air pollutants. In:
    Proceedings of the IPCS/Australia international workshop on human
    health and environmental effects of motor vehicle fuels and their
    exhaust emissions, Sydney, 6-10 April 1992. Canberra, Commonwealth
    Department of Health, Housing, Local Government and Community
    Services, pp 101-173.

    Greenfield R, Ellwein L & Cohen S (1984) A general probabilistic model
    of carcinogenic analysis: Experimental bladder cancer. Carcinogenesis,
    5: 437-445.

    Greenland S & Robins J (1994) Invited commentary: Ecologic
    studies -- biases, misconceptions, and counterexamples. Am J Epidemiol,
    139: 747-760.

    Griffis LC, Wolff RK, Henderson RF, Griffith WC, Mokler BV, &
    McClellan RO (1983) Clearance of diesel soot particles from rat lung
    after a subchronic diesel exhaust exposure. Fundam Appl Toxicol,
    3: 99-103.

    Grimmer G, Hildebrandt A, & Böhnke H (1973) Investigations on the
    carcinogenic burden by air pollution in man. II. Sampling and
    analytics of polycyclic aromatic hydrocarbons in automobile exhaust
    gas. 1. Optimization of the collecting arrangement. Zbl Bakt Hyg I Abt
    Orig B, 158: 22-34.

    Grimmer G, Naujack K-W, & Schneider D (1979) Changes in PAH-profiles
    in different areas of a city during the year. In: Björseth A & Dennis
    A, eds, Polynuclear aromatic hydrocarbons: Chemistry and biological
    effects. Fourth international symposium on polynuclear aromatic
    hydrocarbons. Columbus, OH, Battelle Press, pp 107-125.

    Grimmer G, Naujack K-W, & Schneider D (1982) Profile analysis of
    polycyclic aromatic hydrocarbons by glass capillary gas chromatography
    in atmospheric suspended particulate matter in the nanogram range
    collecting 10 m3 of air. Fresenius Z Anal Chem, 311: 475-484.

    Grimmer G, Brune H, Deutsch-Wenzel R, Dettbarn G, Jacob J, Naujack
    K-W, Mohr U, & Ernst H (1987) Contribution of polycyclic aromatic
    hydrocarbons and nitro-derivatives to the carcinogenic impact of
    diesel engine exhaust condensate evaluated by implantation into the
    lungs of rats. Cancer Lett, 37: 173-180.

    Grimmer G, Jacob J, Dettbarn G, & Naujack K-W (1988) Effect of the
    pH-value of diesel exhaust on the amount of filter-collected
    nitro-PAH. In: Cooke M & Dennis AJ, eds, Polynuclear aromatic
    hydrocarbons. Tenth international symposium: A decade of progress.
    Columbus, OH, Battelle Press, pp 341-351.

    Grimmer G, Brune H, Dettborn G, Jacob J, Misfeld J, Mohr U, Naujack
    K-W, Timm J, & Wenzel-Hartung R (1991) Relevance of polycyclic
    aromatic hydrocarbons as environmental carcinogens. Fresenius J Anal
    Chem, 339: 792- 795.

    Grundmann R & Rehm HJ (1991) Biodegradation of diesel-fuel. Use of
    free and immobilized mixed cultures in soils. Erdöl Kohle Erdgas
    Petrochem 44: 149-150.

    Gu Z-W, Zhong B-Z, Whong W-Z, Wallace WE, & Ong T (1991) Mutagenicity
    and clastogenicity studies of diesel emission particles in cultured
    mammalian cells. Environ Mol Mutag, 17 (Suppl. 19): 28.

    Gu Z-W, Zhong B-Z, Nath B, Whong W-Z, Wallace WE, & Ong T-M (1992)
    Micronucleus induction and phagocytosis in mammalian cells treated
    with diesel emission particles. Mutat Res, 279: 55-60.

    Gu Z-W, Zhong B-Z, Keane MJ, Whong W-Z, Wallace WE, & Ong T (1994)
    Induction of unscheduled DNA synthesis in V79 cells by diesel emission
    particles dispersed in simulated pulmonary surfactant. Ann Occup Hyg,
    38 (Suppl 1): 345-349.

    Guerrero RR, Rounds DE, & Orthoefer J (1981) Genotoxicity of Syrian
    hamster lung cells treated  in vivo with diesel exhaust particulates.
    Environ Int, 5: 445-454.

    Guillemin MP, Herrera H, Huynh CK, Droz P-O, & Vu Duc T (1992)
    Occupational exposure of truck drivers to dust and polynuclear
    aromatic hydrocarbons: A pilot study in Geneva, Switzerland. Int Arch
    Occup Environ Health, 63: 439-447.

    Gusein-Zade KM (1974) [The results of dermatological examinations of
    engine drivers working in oil fields of Apsheron.] Vestn Dermatol
    Venereol, 10: 47-49 (in Russian).

    Gustafsson L, Wall S, Larsson L-G, & Skog B (1986) Mortality and
    cancer incidence among Swedish dock workers -- A retrospective cohort
    study. Scand J Work Environ Health, 12: 22-26.

    Gustavsson P Plato N Lidström E-B, & Hogstedt C (1990) Lung cancer and
    exposure to diesel exhaust among bus garage workers. Scand J Work
    Environ Health, 16: 348-354.

    Hahon N, Booth JA, Green F, & Lewis TR (1985) Influenza virus
    infection in mice after exposure to coal dust and diesel engine
    emissions. Environ Res, 37: 44-60.

    Hammerle RH, Ketcher DA, Horrocks RW, Lepperhoff G, Hüthwohl G, &
    Lüers B (1994a) Emissions from current diesel vehicles (SAE Technical
    Paper Series No. 942943). Warrendale, PA, Society of Automotive
    Engineers.

    Hammerle R, Schuetzle D, & Adams W (1994b) A perspective on the
    potential development of 'environmentally acceptable' light-duty
    diesel vehicles. Environ Health Perspectives, 101: 1-20.

    Hammond S, Smith T, Woskie S, Leaderer B, & Bettinger N (1988) Markers
    of exposure to diesel exhaust and cigarette smoke in railroad workers.
    Am Ind Hyg Assoc J, 49: 516-522.

    Hampton CV, Pierson WR, Schuetzle D, & Harvey TM (1983) Hydrocarbon
    gases emitted from vehicles on the road. 2. Determination of emission
    rates from diesel and spark-ignition vehicles. Environ Sci Technol,
    17: 699-708.

    Handa T, Yamanuchi T, Ohnishi M, Hisamatsu Y, & Ishii T (1983)
    Detection and average content levels of carcinogenic and mutagenic
    compounds from the particulates on diesel and gasoline engine
    mufflers. Environ Int, 9: 335-341.

    Hansen ES (1993) A follow-up study on the mortality of truck drivers.
    Am J Ind Med, 23: 811-821.

    Hare CT (1986) The effects of diesel fuel properties on particulate
    emissions. In: Lee SD, Schneider T, Grant LD, & Verkerk PJ, eds,
    Aerosols: Research, risk assessment and control strategies. Chelsea,
    MI, Lewis Publishers, pp 501-514.

    Harris J (1983) Diesel emissions and lung cancer. Risk Anal,
    3: 83-100.

    Harrison W, Winnik MA, Kwong PTY, & MacKay D (1975) Crude oil spills.
    Disappearence of aromatic and aliphatic components from small
    sea-surface slicks. Environ Sci Technol, 9: 231-234.

    Hartung R & Hunt GS (1966) Toxicity of some oils to waterfowl. J Wildl
    Manage, 30: 564-570.

    Hartung A, Kraft J, Schulze J, Kiess H, & Lies K-H (1984)
    Identification of nitrated polycyclic aromatic hydrocarbons in diesel
    particulate extracts and their potential formation as artifacts during
    particulate collection. Chromatographia, 19: 269-273.

    Hasegawa MM, Nishi Y, Tsuda H, Inui N, & Morimoto K (1988) Effects of
    diesel exhaust particles on chromosome aberration, sister chromatid
    exchange and morphological transformation in cultured mammalian cells.
    Cancer Lett, 42: 61-66.

    Hawthorne SB & Miller DJ (1986) Extraction and recovery of organic
    pollutants from environmental solids and Tenax-GC using supercritical
    CO2. J Chromatogr Sci, 24: 258-264.

    Hayes RB, Thomas T, Silverman DT, Vineis P, Blot WJ, Mason TJ, Pickle
    LW, Correa P, Fontham ETH, & Schoenberg JB (1989) Lung cancer in motor
    exhaust-related occupations. Am J Ind Med, 16: 685-695.

    Hedtke SF & Puglisi FA (1982) Short-term toxicity of five oils to four
    freshwater species. Arch Environ Contam Toxicol, 11: 425-430.

    Heino M, Ketola R, Mäkelä P, Mäkinen R, Niemelä R, Starck J, &
    Partanen T (1978) Work conditions and health of locomotive engineers.
    1. Noise, vibration, thermal climate, diesel exhaust constituents,
    ergonomics. Scand J Work Environ Health, 4: 3-14.

    Heinrich U (1994) Carcinogenic effects of solid particles. In: Mohr U,
    Dungworth DL, Mauderly JL, & Oberdörster G, eds, Toxic and
    carcinogenic effects of solid particles in the respiratory tract.
    Washington DC, International Life Science Institute Press, pp 57-73.

    Heinrich U, Peters L, Funcke W, Pott F, Mohr U, & Stöber W (1982)
    Investigation of toxic and carcinogenic effects of diesel exhaust in
    long-term inhalation exposure of rodents. In: Lewtas S, ed.,
    Toxicological effects of emissions from diesel engines. Amsterdam,
    Elsevier, pp 225-242.

    Heinrich U, Muhle H, Takenaka S, Ernst H, Fuhst R, Mohr U, Pott F, &
    Stöber W (1986a) Chronic effects on the respiratory tract of hamsters,
    mice and rats after long-term inhalation of high concentrations of
    filtered and unfiltered diesel engine emissions. J Appl Toxicol,
    6: 383-395.

    Heinrich U Pott F & Rittinghausen S (1986b) Comparison of chronic
    inhalation effects in rodents after long-term exposure to either coal
    oven flue gas mixed with pyrolized pitch or diesel engine exhaust. In:
    Ishinishi N, Koizumi A, McClellan RO, & Stöber W, eds, Carcinogenic
    and mutagenic effects of diesel engine exhaust. Amsterdam, Elsevier,
    pp 441-457.

    Heinrich U, Pott F, & Roller M (1991) [Polycyclic aromatic
    hydrocarbons -- Animal results and epidemiological findings relevant to
    risk assessment]. In: [Carcinogenic substances in the environment:
    Origin, measurement, risk, minimization. Meeting of the Commission for
    Clean Air of the Society of German Engineers (VDI) and the German
    Standards Institute (DIN), Mannheim, 23-25 April]. Düsseldorf, Society
    of German Engineers Press, pp 71-92 (in German).

    Heinrich U, Fuhst R, & Mohr U (1992) [Experimental inhalation studies
    with laboratory animals on the question of the tumour-inducing effect
    of diesel engine exhausts and two test dusts.] In: Effects of diesel
    engine exhausts on health. Munich, GSF Research Centre for Environment
    and Health Inc., pp 21-30 (in German).

    Heinrich U, Peters L, Creutzenberg O, Dasenbrock C, & Hoymann HG
    (1994) Inhalation exposure of rats to tar/pitch condensation aerosol
    or carbon black alone or in combination with irritant gases. In: Mohr
    U, Dungworth DL, Mauderly JL, & Oberdörster G, eds, Toxic and
    carcinogenic effects of solid particles in the respiratory tract.
    Washington DC, International Life Sciences Institute Press,
    pp 433-441.

    Heinrich U, Fuhst R, Rittinghausen S, Creutzenberg O, Bellmann B, Koch
    W, & Levsen K (1995) Chronic inhalation exposure of Wistar rats, and
    two different strains of mice to diesel engine exhaust, carbon black
    and titanium dioxide. Inhal Toxicol, 7: 533-556.

    Hellstrom T & Doving KB (1983) Perception of diesel oil by cod
    ( Gadus morhua L.). Aquat Toxicol, 4: 303-315.

    Henderson TR, Li AP, Royer RE, & Clark CR (1981) Increased
    cytotoxicity and mutagenicity of diesel fuel after reaction with
    NO2. Environ Mutag, 3: 211-220.

    Henderson RF, Pickrell JA, Jones RK, Sun JD, Benson JM, Mauderly JL, &
    McClellan RO (1988) Response of rodents to inhaled diluted diesel
    exhaust: Biochemical and cytological changes in bronchoalveolar lavage
    fluid and in lung tissue. Fundam Appl Toxicol, 11: 546-567.

    Henschler D ed. (1987) [Diesel engine emissions]. In: [Maximal working
    place concentration. Chemical compounds in the work area hazardous to
    health]. Weinheim, VCH, 25 pp (in German).

    Herbes SE & Whitley TA (1983) Characterization and toxicity of
    water-soluble photooxidants produced during irradiation of coal
    liquids by sunlight. Environ Pollut, 6: 221-240.

    Heyder J, Gebhart J, Rudolf G, Schiller CF, & Stahlhofen W (1986)
    Deposition of particles in the human respiratory tract in the size
    range 0.005-15 µm. J Aerosol Sci, 17: 811-825.

    Hildemann LM, Markowski GR, & Cass GR (1991) Chemical composition of
    emissions from urban sources of fine organic aerosol. Environ Sci
    Technol, 25: 744-759.

    Hoar SK & Hoover R (1985) Truck driving and bladder cancer mortality
    in rural New England. J Natl Cancer Inst, 74: 771-774.

    Hobbs JR, Walter RA, Hard T, & Devoe D (1977) Train generated air
    contaminants in the train crew's working environment (FRA/ORD-77/08:
    US NTIS PB 265-355). Springfield, VA, National Technical Information
    Service, pp 1-46.

    Holland WD (1978) Determination of breathing zone concentrations of
    contaminants from emissions from diesel powered vehicles in
    underground mines (BuMines OFR 24-80; US NTIS PB80-150766).
    Springfield, VA, National Technical Information Service, pp 1-127.

    Holt S (1987) The effects of crude and diesel oil spills on plant
    communities at Mesters Vig, northeast Greenland. Arct Alp Res,
    19: 490-497. Horvath et al. (1985)

    Horvath H, Kreiner I, & Norek C (1987) The role of the diesel aerosol
    in the air pollution of Vienna. J Aerosol Sci, 18: 817-819.

    Horvath H, Kreiner I, Norek C, Preining O, & Georgi B (1988) Diesel
    emissions in Vienna. Atmos Environ, 22: 1255-1269.

    Howe GR, Burch JD, Miller AB, Cook GM, Estève J, Morrison B, Gordon P,
    Chambers LW, Fodor G, & Winsor GM (1980) Tobacco use, occupation,
    coffee, various nutrients, and bladder cancer. J Natl Cancer Inst,
    64: 701-713.

    Howe GR, Fraser D, Lindsay J, Presnal B, & Zhang Yu S (1983) Cancer
    mortality (1965-77) in relation to diesel fume and coal exposure in a
    cohort of retired railway workers. J Natl Canc Inst, 6: 1015-1019.

    Huisingh J, Bradow R, Jungers R, Claxton L, Zweidinger R, Tejada S,
    Bumgarner J, Duffield F, Waters MD, Simmon VF, Hare C, Rodriguez C, &
    Snow L (1978) Application of bioassay to the characterization of
    diesel particle emissions. In: Waters MD, Nesnow S, Huisingh JL,
    Sandhu SS, & Claxton L, eds, Application of short-term bioassays in
    the fractionation and analysis of complex environmental mixtures. New
    York, Plenum Press, pp 383-418.

    Hyde DM, Plopper CG, Weir AJ, Murnane RD, Warren DL, Last JA, &
    Pepelko WE (1985) Peribronchiolar fibrosis in lungs of cats
    chronically exposed to diesel exhaust. Lab Invest, 52: 195-206.

    Iida N, Suzuki Y, Sato GT, & Sawada T (1986) Effects of intake oxygen
    concentration on the characteristics of particulate emissions from a
    DI diesel engine (Paper No. 861233). Warrendale, PA, Society of
    Automotive Engineers.

    Imaida K, Hirose M, Tay L, Lee M-S, Wang Y, & King CM (1991)
    Comparative carcinogenicities of 1-, 2-, and 4-nitropyrene and
    structurally related compounds in the female CD rat. Cancer Res,
    51: 2902-2907.

    International Agency for Research on Cancer (1989a) Diesel fuels. In:
    IARC monographs on the evaluation of the carcinogenic risk of
    chemicals to humans, Vol. 45, Occupational exposures in petroleum
    refining; crude oil and major petroleum fuels. Lyon, pp 219-237.

    International Agency for Research on Cancer (1989b) Diesel and
    gasoline engine exhausts. In: IARC monographs on the evaluation of the
    carcinogenic risk of chemicals to humans, Vol. 46, Diesel and gasoline
    engine exhaust and some nitroarenes. Lyon, pp 41-185.

    International Standards Organization (1987) Petroleum products -- Fuels
    (class F) -- Specifications of marine fuels (No. ISO 8217-1987/E).
    Geneva, 7 pp.

    International Union for Pure and Applied Chemistry (1993) Glossary for
    chemists of terms used in toxicology (IUPAC Recommendations 1993).
    Pure Appl Chem, 65: 2003-2122.

    IPCS (1986) Environmental health criteria, 52: Toluene. Geneva, WHO.

    IPCS (1993) Environmental health criteria, 150: Benzene. Geneva, WHO,
    156 pp.

    IPCS (1994) Environmental health criteria, 170: Assessing human health
    risks of chemicals: Derivation of guidance values for health-based
    exposure limits. Geneva, WHO, 73 pp.

    Iscovich J, Castelletto R, Estève J, Muñoz N, Colanzi R, Coronel A,
    Deamezola I, Tassi V, & Arslan A (1987) Tobacco smoking, occupational
    exposure and bladder cancer in Argentina. Int J Cancer, 40: 734-740.

    Ishinishi N, Kuwabara N, Nagase S, Suzuki T, Ishiwata S, & Kohno T
    (1986) Long-term inhalation studies on effects of exhaust from heavy
    and light duty diesel engines on F344 rats. In: Ishinishi N, Koizumi
    A, McClellan RO, & Stöber W, eds, Carcinogenic and mutagenic effects
    of diesel engine exhaust. Amsterdam, Elsevier, pp 329-348.

    Ishinishi N, Kuwabara N, Takaki Y, Nagase S, Suzuki T, Nakajima T,
    Maejima K, Kato A, & Nakamura M (1988) Long-term inhalation
    experiments on diesel exhaust. In: Diesel exhaust and health risk.
    Results of the HERP studies. Tsukuba, Health Effects Research
    Programme, pp 11-84.

    Israel GW, Zierock KH, & Mollenhauer K (1982) [Distribution and
    chemical composition of particle emissions from different diesel
    motors] (VDI Report No. 429). Düsseldorf, Association of German
    Engineers Press, pp 279-286 (in German).

    Iwai K, Udagawa T, Yamagishi M, & Yamada H (1986) Long-term inhalation
    studies of diesel exhaust on F344 SFP rats. Incidence of lung cancer
    and lymphoma. In: Ishinishi N, Koizumi A, McClellan RO, & Stöber W,
    eds, Carcinogenic and mutagenic effects of diesel engine exhaust.
    Amsterdam, Elsevier, pp 349-360.

    Iyer V, Harris RE, & Wynder EL (1990) Diesel exhaust exposure and
    bladder cancer risk. Eur J Epidemiol, 6: 49-54.

    Jaenicke R (1986) Physical characterization of aerosols. In: Lee SD,
    Schneider T, Grant LD, & Verkerk PJ, eds, Aerosols: Research, risk
    assessment and control strategies. Chelsea, MI, Lewis Publishers,
    pp 97-106.

    Jensen TE & Hites RA (1983) Aromatic diesel emissions as a function of
    engine conditions. Anal Chem, 55: 594-599.

    Jensen OM, Wahrendorf J, Knudsen JB, & Sorensen BL (1987) The
    Copenhagen case-referent study on bladder cancer. Risks among drivers,
    painters and certain other occupations. Scand J Work Environ Health,
    13: 129-134.

    Jinno K, Hoshino T, Hondo T, Saito M, & Senda M (1986) Identification
    of polycyclic aromatic hydrocarbons in extracts of diesel particulate
    matter by supercritical fluid chromatography coupled with an
    ultraviolet multichannel detector. Anal Chem, 58: 2696-2699.

    Johnson JH & Carlson DH (1986) The application of advanced measurement
    and control technology to diesel-powered vehicles in an underground
    salt mine. Montreal, Canadian Institute of Mining and Metallurgy,
    pp 206-237.

    Jorgensen H & Svensson Å (1970) Studies on pulmonary function and
    respiratory tract symptoms of workers in an iron ore mine where diesel
    trucks are used underground. J Occup Med, 12: 348-354.

    Joumard R & Perrin M-L (1988) Measurement of particle and gaseous
    pollution of the atmosphere due to buses. Sci Total Environ,
    76: 55-62.

    Juhnke I & Lüdemann D (1978) [Results of a study of 200 chemical
    compounds for acute fish toxicity with the golden orfe test.] Z Wasser
    Abwasser Forsch, 11: 161-164 (in German).

    Kahn G, Orris P, & Weeks J (1988) Acute over exposure to diesel
    exhaust: Report of 13 cases. Am J Ind Med, 13: 405-406.

    Kainz RJ & White LE (1982) Consequences associated with the inhalation
    of uncombusted diesel vapor. In: MacFarland HN, Holdsworth CE,
    MacGregor JA, Call RW, & Kane ML, eds, Proceedings of the symposium:
    The toxicology of petroleum hydrocarbons, May 1982. Washington DC,
    American Petroleum Institute, pp 320-327.

    Kainz RJ & White LE (1984) Depressant effects associated with the
    inhalation of uncombusted diesel vapor. In: MacFarland HN, Holdsworth
    CE, MacGregor JA, Call RW, & Kane ML, eds, Applied toxicology of
    petroleum hydrocarbons (Advances in modern environmental toxicology,
    Vol. 6). Princeton, NJ, Princeton Scientific Publishers, pp 233-243.

    Kaplan HL, MacKenzie WF, Springer KJ, Schreck RM, & Vostal JJ (1982) A
    subchronic study of the effects of exposure of three species of
    rodents to diesel exhaust. In: Lewtas J, ed., Toxicological effects of
    emissions from diesel engines. Amsterdam, Elsevier, pp 161-182.

    Kaplan HL, Springer KJ, & MacKenzie WF (1983) Studies of potential
    health effects of long-term exposure to diesel exhaust emissions.
    Final report (SwRI Project No. 01-0750-103; SFRE Project No. 1239).
    San Antonio, TX, Southwest Research Institute, 59 pp.

    Karagianes MT, Palmer RF, & Busch RH (1981) Effects of inhaled diesel
    emissions and coal dust in rats. Am Ind Hyg Assoc J, 42: 382-391.

    Karl DM (1992) The grounding of the Bahia Paraiso: Microbial ecology
    of the 1989 Antarctic oil spill. Microbiol Ecol, 24: 77-89.

    Kato A, Kyono H, & Kuwabara N (1992) [Electron-microscopic
    observations on rat lungs after long term inhalation of diesel
    emissions -- Non-neoplastic lesions]. Jpn J Chest Dis, 30: 238-247
    (in Japanese, with English abstract).

    Kauffman S (1974) Kinetics of alveolar epithelial hyperplasia in lungs
    of mice exposed to urethane. I. Quantitative analysis of cell
    populations. Lab Invest, 30: 170-175.

    Kawabata Y, Iwai K, Udagawa T, Tukagoshi K, & Higuchi K (1986) Effects
    of diesel soot on unscheduled DNA synthesis of tracheal epithelium and
    lung tumor formation. In: Ishinishi N, Koizumi A, McClellan RO, &
    Stöber W, eds, Carcinogenic and mutagenic effects of diesel engine
    exhaust. Amsterdam, Elsevier, pp 213-222.

    Kawabata Y, Udagawa T, Higuchi K, Yamada H, & Iwai K (1993) One year
    exposure to diesel engine exhaust causes lung tumors. In: Mohr U,
    Dungworth DL, Mauderly JL & Oberdörster G, eds, Toxic and carcinogenic
    effects of solid articles in the respiratory tract. Washington DC,
    International Life Sciences Institute Press, pp 429-431.

    Keane MJ, Xing S-G, Harrison JC, Ong T, & Wallace WE (1991)
    Genotoxicity of diesel-exhaust particles dispersed in simulated
    pulmonary surfactant. Mutat Res, 260: 233-238.

    Kemp PF, Swartz RC, & Lamberson JO (1986) Response of the
    phoxocephalid amphipod,  Rhepoxynius abronius, to a small oil spill
    in Yaquina Bay, Oregon. Estuaries, 9: 340-347.

    Kennicutt MC II, (1990) Oil spillage in Antarctica: Initial report of
    the National Science Foundation-sponsored quick response team on the
    grounding of the  Bahia Paraiso. Environ Sci Technol, 24: 620-624.

    Kennicutt MC II & Sweet ST (1992) Hydrocarbon contamination on the
    Antarctic peninsula: III. The  Bahia Paraiso -- Two years after the
    spill. Mar Pollut Bull, 25: 303-306.

    Kennicutt MC II, Sweet ST, Fraser WR, Stockton WL, & Culver M (1991a)
    Grounding of the Bahia Paraiso at Arthur Harbor, Antarctica. 1.
    Distribution and fate of oil spill related hydrocarbons. Environ Sci
    Technol, 25: 509-518.

    Kennicutt MC II, Sweet ST, Fraser WR, Culver M, & Stockton WL (1991b)
    The fate of diesel fuel spilled by the Bahia Paraiso in Arthur
    Harbour, Antarctica. In: Proceedings 1991 international oil spill
    conference. (Prevention, behavior, control, cleanup), 4-7 March 1991,
    San Diego, California. Washington DC, American Petroleum Institute,
    pp 493-500.

    Kennicutt MC II, McDonald TJ, Denoux GJ, & McDonald SJ (1992a)
    Hydrocarbon contamination on the Antarctic peninsula. I. Arthur
    Harbor -- Subtidal sediments. Mar Pollut Bull, 24: 499-506.

    Kennicutt MC II, McDonald TJ, Denoux GJ, & McDonald SJ (1992b)
    Hydrocarbon contamination on the Antarctic peninsula. II. Arthur
    Harbor -- Inter- and subtidal limpets (Nacella concinna). Mar Pollut
    Bull, 24: 506-511.

    Kittel B, Ernst H, Dungworth DL, Rittinghausen S, Nolte T, Kamino K,
    Stuart B, Lake SG, Cardesa A, Moraweitz G, & Mohr U (1993)
    Morphological comparison between benign keratinizing cystic squamous
    cell tumours of the lung and squamous lesions of the skin in rats.
    Exp Toxicol Pathol, 45: 257-267

    Klingenberg H, Schürmann D, & Lies KH (1991) [Diesel motor
    exhaust -- Formation and determination]. In: [Carcinogenic substances in
    the environment: Origin, measurement, risk, minimization. Meeting of
    the Commission for Clean Air of the Society of German Engineers (VDI)
    and the German Standards Institute (DIN), Mannheim, 23-25 April].
    Düsseldorf, Society of German Engineers Press, pp 119-131 (in German)

    Knave B, Olson BA, Elofson S, Gamberale F, Isaksson A, Mindus P,
    Persson HE, Struwe G, Wennberg A, & Westerholm P (1978) Long term
    exposure to jet fuel. II. A cross-sectional epidemiologic
    investigation on occupationally exposed industrial workers with
    special reference to the nervous system. Scand J Work Environ Health,
    4: 19-45.

    Koenig JQ, Covert DS, Marshall SSG, Van Belle G, & Pierson WE (1988)
    The pulmonary effects of ozone and nitrogen dioxide alone and combined
    in healthy and asthmatic adolescent subjects. Toxicol Ind Health,
    4: 521-532

    Köhler M & Eichhoff HJ (1967) [Quick method for determination of
    aromatic hydrocarbons in air dust.] Z Anal Chem, 232: 401-409
    (in German).

    Kotin P, Falk HL, & Thomas M (1955) Aromatic hydrocarbons. III.
    Presence in the particulate phase of diesel-engine exhausts and the
    carcinogenicity of exhaust extracts. Arch Ind Health, 11: 113-120.

    Kraft J & Lies K-H (1981) Polycyclic aromatic hydrocarbons in the
    exhaust of gasoline and diesel vehicles (Technical paper No. 810082).
    Warrendale, PA, Society of Automotive Engineers, pp 1-11.

    Kühnhold WW (1977) The effect of mineral oils on the development of
    eggs and larvae of marine species. A review and comparison of
    experimental data in regard to possible damage at sea. In: McIntyre AD
    & Whittle KJ, eds, Petroleum hydrocarbons in the marine environment
    (Rapports et proces-verbaux des réunions, Vol. 171). Copenhagen,
    International Council for Exploration of the Sea, pp 175-183.

    Kunitake E, Shimamura K, Katayama H, Takemoto K, Yamamoto A, Hisanaga
    A, Ohyama S, & Ishinishi N (1986) Studies concerning carcinogenesis of
    diesel particulate extracts following intratracheal instillation,
    subcutaneous injection, or skin application. In: Ishinishi N, Koizumi
    A, McClellan RO, & Stöber W, eds, Carcinogenic and mutagenic effects
    of diesel engine exhaust. Amsterdam, Elsevier, pp 235-252.

    Kuschner M (1968) The causes of cancer. Am Rev Respir Dis,
    98: 573-590.

    Larson RA, Hunt LL, & Blankenship DW (1977) Formation of toxic
    products from a No. 2 fuel oil by photooxidation. Environ Sci Technol,
    11: 492-496.

    Larson RA, Bott TL, Hunt LL, & Rogenmuser K (1979) Photooxidation
    products of a fuel oil and their antimicrobial activity. Environ Sci
    Technol, 13: 965-969.

    Lawler GC, Loong W-A, & Laseter JL (1978) Accumulation of aromatic
    hydrocarbons in tissues of petroleum-exposed mallard ducks
     (Anas platyrhynchos). Environ Sci Technol, 12: 51-54.

    Lee RF (1977) Accumulation and turnover of petroleum in marine
    organisms. In: Wolfe DA, ed., Fate and effects of petroleum
    hydrocarbons in marine ecosystems and organisms. Oxford, Pergamon
    Press, pp 60-70.

    Lee RF (1981) Mixed function oxygenases (MFO) in marine invertebrates.
    Mar Biol Lett, 2: 87-105.

    Lee FS-C & Schuetzle D (1983) Sampling, extraction and analysis of
    polycyclic aromatic hydrocarbons from internal combustion engines. In:
    Björseth A, ed., Handbook of polycyclic aromatic hydrocarbons. New
    York, Marcel Dekker, pp 27-94.

    Lee FS-C, Pierson WR, & Ezike J (1980) The problem of PAH degradation
    during filter collection of airborne particulates -- An evaluation of
    several commonly used filter media. In: Björseth A & Dennis AJ, eds,
    Polynuclear aromatic hydrocarbons, 4th International Symposium:
    Chemistry and biological effects. Dearborn, MI, Engineering and
    Research Staff Research, Ford Motor Co., pp 543-563.

    Lee PS, Chan TL, & Hering WE (1983) Long-term clearance of inhaled
    diesel exhaust particles in rodents. J Toxicol Environ Health,
    12: 801-813.

    Lee PS, Gorski RA, Hering WE, & Chan TL (1987) Lung clearance of
    inhaled particles after exposure to carbon black generated from a
    resuspension system. Environ Res, 43: 364-373.

    Lee LS, Hagwall M, Delfino JJ, & Rao PSC (1992) Partitioning of
    polycyclic aromatic hycrocarbons from diesel fuel into water. Environ
    Sci Technol, 26: 2104-2110.

    Lehmann E, Rentel K-H, Allescher W, & Hohmann R (1990) [Measurement of
    diesel exhaust at the workplace.] Zentralbl Arbeitsmed Arbeitsschutz
    Prophyl Ergon, 40: 2-11 (in German).

    Levsen K (1988) The analysis of diesel particulate. Fresenius Z Anal
    Chem, 331: 467-478.

    Levsen K (1992) [Production and analysis of total diesel exhaust for
    efficiency research.] In: [Effects of diesel engine exhausts on
    health]. Munich, GSF Research Centre for Environment and Health Inc.,
    pp 16-20 (in German).

    Levsen K, Puttins U, Schilhabel J, & Priess B (1988) Artefacts during
    the sampling of nitro PAHs of diesel exhaust. Fresenius Z Anal Chem,
    330: 527-528.

    Lewis SC (1983) Crude petroleum and selected fractions. Prog Exp Tumor
    Res, 26: 68-84.

    Lewis TR, Green FHY, Moorman WJ, Burg JAR, & Lynch DW (1986) A chronic
    inhalation toxicity study of diesel engine emissions and coal dust,
    alone and combined. In: Ishinishi N, Koizumi A, McClellan RO, & Stöber
    W, eds, Carcinogenic and mutagenic effects of diesel engine exhaust.
    Amsterdam, Elsevier, pp 361-380.

    Lewis TR, Green FHY, Moorman WJ, Burg JR, & Lynch DW (1989) A chronic
    inhalation toxicity study of diesel engine emissions and coal dust,
    alone and combined. J Am Coll Toxicol, 8: 345-375.

    Lewtas J (1983) Evaluation of the mutagenicity and carcinogenicity of
    motor vehicle emissions in short-term bioassays. Environ Health
    Perspectives, 47: 141- 152.

    Lewtas J & Williams K (1986) A retrospective view of the value of
    short-term genetic bioassays in predicting the chronic effects of
    diesel soot. In: Ishinishi N, Koizumi A, McClellan R, & Stöber W, eds,
    Carcinogenic and mutagenic effects of diesel engine exhaust.
    Amsterdam, Elsevier, pp 119-140.

    Li AP & Royer RE (1982) Diesel-exhaust-particle extract enhancement of
    chemical-induced mutagenesis in cultured Chinese hamster ovary cells:
    Possible interaction of diesel exhaust with environmental carcinogens.
    Mutat Res, 103: 349-355.

    Liebe B & Fock HP (1992) Growth and adaptation of the green algae
    Chlamydomonas reinhardtii on diesel exhaust particle extracts. J Gen
    Microbiol, 138: 973-978.

    Lies K-H, Hartung, A Postulka, A, Gring H, & Schulze J (1986)
    Composition of diesel exhaust with particular reference to particle
    bound organics including formation of artifacts. In: Ishinishi N,
    Koizumi A, McClellan RO, & Stöber W, eds, Carcinogenic and mutagenic
    effects of diesel engine exhaust. Amsterdam, Elsevier, pp 65-82.

    Linnell RH & Scott WE (1962) Diesel exhaust composition and odor
    studies. J Am Poll Control Assoc, 12: 510-515.

    Lock S, Dalbey W, Schmoyer R, & Griesemer R (1984) Inhalation
    toxicology of diesel fuel obscurant aerosol in Sprague-Dawley rats:
    Final report, phase 3, subchronic exposures (TN AD-A150 100). Oak
    Ridge, TN, Oak Ridge National Laboratory.

    Lockhart WL, Danell RW, & Murray DAJ (1987) Acute toxicity bioassays
    with petroleum products: Influence of exposure conditions. In:
    Vandermeulen JH & Hrudey SE, eds, Oil in freshwater: Chemistry,
    biology, countermeasure technology. New York, Pergamon, pp 335-409.

    Loprieno N, DeLorenzo F, Cornetti GM, & Biaggini G (1980) Mutagenicity
    studies on diesel particles and particulate extracts. Health effects
    of diesel engine emissions (EPA NO 600/9-80-057a). In: Proceedings of
    an international symposium, Vol 1. Washington DC, US Environmental
    Protection Agency, pp 276-308.

    Lowe DM & Pipe RK (1986) Hydrocarbon exposure in mussels: A
    quantitative study of the responses in the reproductive and nutrient
    storage cell systems. Aquat Toxicol, 8: 265-272.

    Lowenthal DH, Zielinska B, Chow JC, Watson JG, Gautan M, Ferguson DH,
    Neuroth GR, & Stevens KD (1994) Characterization of heavy-duty diesel
    vehicle emissions. Atmos Environ, 28: 731-743.

    Lysyj I & Russell EC (1974) Dissolution of petroleum-derived products
    in water. Water Res, 8: 863-868.

    Mackay D & Shiu WY (1976) Aqueous solubilities of weathered northern
    crude oils. Bull Environ Contam Toxicol, 15: 101-109.

    Mackay D, Shiu WY, & Ma KC (1992) Illustrated handbook of
    physical-chemical properties and environmental fate for organic
    chemicals. Vol. 2. Polynuclear aromatic hydrocarbons, polychlorinated
    dioxins and dibenzofurans. Ann Arbor, MI, Lewis Publishers Inc.,
    p 253.

    Mackie PR, McGill AS, & Hardy R (1972) Diesel oil contamination of
    brown trout ( Salmo trutta L.). Environ Pollut, 3: 9-16.

    Mahoney BM & Haskin HH (1980) The effects of petroleum hydrocarbons on
    the growth of phytoplankton recognised as food forms for the eastern
    oyster,  Crassostrea virginica Gmelin. Environ Pollut, A22: 123-132.

    Marcus JM & Scott GI (1989) Effects of two diesel fuel mixtures on
    fecal coliform bacteria densities. Bull Environ Contam Toxicol,
    42: 395-401.

    Marklund S, Andersson R, Tysklind M, Rappe C, Egebäck K-E, Björkman E,
    & Grigoriadis V (1990) Emissions of PCDDs and PCDFs in gasoline and
    diesel fueled cars. Chemosphere, 20: 553-561.

    Martini N (1993) Operable lung cancers. CA Cancer J Clin, 43: 201-204.

    Mattie DR, Alden CL, Newell TK, Gaworski CL, & Flemming CD (1991) A
    90-day continuous vapor inhalation toxicity study of JP-8 jet fuel
    followed by 20 or 21 months of recovery in Fischer 344 rats and
    C57BL/6 mice. Toxicol Pathol, 19: 77-87

    Mauderly JL (1992) Diesel exhaust. In: Lippmann M, ed., Environmental
    toxicants. Human exposures and their health effects. New York, Van
    Nostrand Reinhold, pp 119-162.

    Mauderly JL, Jones RK, McClellan RO, Henderson RF, & Griffith WC
    (1986) Carcinogenicity of diesel exhaust inhaled chronically by rats.
    In: Ishinishi N, Koizumi A, McClellan RO, & Stöber W, eds,
    Carcinogenic and mutagenic effects of diesel engine exhaust.
    Amsterdam, Elsevier, pp 397-409.

    Mauderly JL, Jones RK, Griffith WC, Henderson RF, & McClellan RO
    (1987) Diesel exhaust is a pulmonary carcinogen in rats exposed
    chronically by inhalation. Fundam Appl Toxicol, 9: 208-221.

    Mauderly JL, Gillett NA, Henderson RF, Jones RK, & McClellan RO (1988)
    Relationships of lung structural and functional changes to
    accumulation of diesel exhaust particles. Ann Occup Hyg, 32: 659-669.

    Mauderly JL, Bice DE, Cheng YS, Gillett NA, Griffith WC, Henderson RF,
    Pickrell JA, & Wolff RK (1990) Influence of preexisting pulmonary
    emphysema on susceptibility of rats to inhaled diesel exhaust. Am Rev
    Respir Dis, 141: 1333-1341.

    Mauderly JL, Snipes MB, Barr EB, Bechtold WE, Bellinsky SA, Henderson
    RF, Mitchell CE, Nikula KJ, & Thomassen DG (1991) Influence of
    particle-associated organic compounds on carcinogenicity of diesel
    exhaust. In: Eighth Health Effects Institute Annual Conference,
    21-24 April 1991, Colorado Springs, CO. Cambridge, MA, Health Effects
    Institute, 13 pp.

    Mauderly JL, Snipes MB, Barr EB, Belinsky SA, Bond JA, Brooks AL,
    Change IY, Cheng YS, Gillett NA, Griffith WC, Henderson RF, Mitchell
    CE, Nikula KJ, & Thomassen DG (1994) Pulmonary toxicity of inhaled
    diesel exhaust and carbon black in chronically exposed rats. (HEI
    Research Report No. 68). Cambridge, MA, Health Effects Institute,
    97 pp.

    McClellan RO (1986) Health effects of diesel exhaust: A case study in
    risk assessment. Am Ind Hyg Assoc J, 47: 1-13.

    McClellan RO (1990) Particle overload in the lung: Approaches to
    improving our knowledge. J Aerosol Med, 3: 197-207.

    McClellan R, Cuddihy R, Griffith W, & Mauderly J (1989) Integrating
    diverse data sets to assess the risks of airborne pollutants. In: Mohr
    U, ed., Assessment of inhalation hazards. New York, Springer, pp 3-22.

    McCormick JJ, Zator RM, DaGue BB, & Maher VM (1980) Studies on the
    effects of diesel particulate on normal and xeroderma pigmentosum
    cells (EPA-NO 600/9-80-057a). In: Pepelko WE, Danner RM, & Clark NA,
    eds, Health effects of diesel engine emissions. Cincinnati, OH, US
    Environmental Protection Agency, pp 413-415.

    McKee RH & Plutnick RT (1989) Carcinogenic potential of gasoline and
    diesel engine oils. Fundam Appl Toxicol, 13: 545-553.

    McKee RH, Amoruso MA, & Freeman JJ (1994) Evaluation of the genetic
    toxicity of middle distillate fuels. Environ Mol Mutag, 23: 234-238.

    McLaren P (1985) Behaviour of diesel fuel on a high energy beach. Mar
    Pollut Bull, 16: 191-196.

    Meiss R, Robenek H, Schubert M, Themann H, & Heinrich U (1981)
    Ultrastructural alterations in the livers of golden hamsters following
    experimental chronic inhalation of diluted diesel exhaust emission.
    Int Arch Occup Environ Health, 48: 147-157.

    Menne RJ, Rechs M, Horrocks R & Harbolla B (1994) [Current and future
    trends of passenger car diesel engines]. Motortech Z, 55: 323
    (in German, with English abstract).

    Mentnech MS, Lewis DM, Olenchock SA, Mull JC, & Koller WA (1984)
    Effects of coal dust and diesel exhaust on immune competence in rats.
    J Toxicol Environ Health, 13: 31-41.

    Mercedes-Benz AG (1994a) [Exhaust emissions: Threshold limits,
    regulations and measurement of exhaust emissions; calculation of the
    fuel use from the exhaust test; passenger cars (May 1994)
    Mercedes-Benz]. Stuttgart, 19 pp (in German).

    Mercedes-Benz AG (1994b) Limit values for diesel exhaust components.
    Data sheet. Stuttgart, 7 pp.

    Mészáros A & Mészáros E (1989) Sulfate formation on elemental carbon
    particulate. J Aerosol Sci Technol, 10: 337-342.

    Metz N (1989) Diesel particulate emission in Germany (West). Exp
    Pathol, 37: 15-22.

    Meyers PA & Quinn JG (1973) Association of hydrocarbons and mineral
    particles in saline solution. Nature, 244: 23-24.

    Misiorowski RL, Strom KA, Vostal JJ, & Chvapil M (1980) Lung
    biochemistry of rats chronically exposed to diesel particulates. In:
    Pepelko WE, Danner RM, & Clarke NA, eds, Health effects of diesel
    engine emissions, Vol 1 (EPA-600/9-80-057a). Cincinnatti, OH, US
    Environmental Protection Agency, pp 465-480.

    Mitchell AD, Evans EL, Jotz MM, Riccio ES, Mortelmans KE, & Simmon VF
    (1981) Mutagenic and carcinogenic potency of extracts of diesel and
    related environmental emissions:  in vitro mutagenesis and DNA
    damage. Environ Int, 5: 393-401.

    Mohr, U, ed. (1992) International classification of rodent tumours.
    Part 1. The rat (IARC Scientific Publications No. 122). Lyon,
    International Agency for Research on Cancer.

    Mohr U & Riebe-Imre M (1992) [Tests on the  in vitro transformation
    of epithelial cells of the respiratory tract by automotive exhaust
    gases]. In: [Effects of diesel engine exhausts on health.] Munich, GSF
    Research Centre for Environment and Health Inc., pp 31-38 (in German).

    Moolgavkar S & Knudson A (1981) Mutation and cancer: A model for human
    carcinogenesis. J Natl Cancer Inst, 66: 1037-1052.

    Moolgavkar S & Venzon D (1979) Two-event models for carcinogenesis:
    Incidence curves for childhood and adult tumors. Math Biosci,
    47: 55-77.

    Moorman WJ, Clark JC, Pepelko WE, & Mattox J (1985) Pulmonary function
    responses in cats following long-term exposure to diesel exhaust.
    J Appl Toxicol, 5: 301-305.

    Morgan P & Watkinson RJ (1988) Hydrocarbon degradation in soils and
    methods for soil biotreatment. Crit Rev Biotechnol, 8: 305-333.

    Morimoto K, Kitamura M, Kondo H, & Koizumi A (1986) Genotoxicity of
    diesel exhaust emissions in a battery of  in-vitro short-term and
     in-vivo bioassays. In: Ishinishi N, Koizumi A, McClellan RO, &
    Stöber W, eds, Carcinogenic and mutagenic effects of diesel engine
    exhaust. Amsterdam, Elsevier, pp 85-101.

    Morrow JE (1973) Oil-induced mortalities in juvenile coho and sock-eye
    salmon. J Mar Res, 31: 135-141.

    Morrow PE (1988) Possible mechanisms to explain dust overloading of
    the lungs. Fundam Appl Toxicol, 10: 369-384.

    Muhlbaier Dasch J & Cadle SH (1989) Atmospheric carbon particle in the
    Detroit urban area: Wintertime sources and sinks. Aerosol Sci Technol,
    10: 236-248.

    Muhle H, Creutzenberg O, Bellmann B, Heinrich U, & Mermelstein R
    (1990) Dust overloading of lungs: Investigations of various materials,
    species differences, and irreversibility of effects. J Aerosol Med,
    3: 111-128.

    Muhle H, Bellmann B, & Creutzenberg O (1994) Toxicokinetics of solid
    particles in chronic rat studies using diesel soot, carbon black,
    toner, titanium dioxide, and quartz. In: Mohr U, Dungworth DL,
    Mauderly JL, & Oberdörster G, eds, Toxic and carcinogenic effects of
    solid particles in the respiratory tract. Washington DC, International
    Life Sciences Institute Press, pp 29-41.

    Muralidhara KMK, Krishnakumari HP, Ramesh HP, & Majumder SK (1982)
    Toxicity of some petroleum fractions used in pesticidal emulsions to
    albino rats. J Food Sci Technol, 19: 260-262.

    Muranaka M, Suzuki S, Koizumi K, Takafuji S, Miyamoto T, Ikemori R, &
    Tokiwa H (1986) Adjuvant activity of diesel-exhaust particulates for
    the production of IgE antibody in mice. J Allergy Immunol,
    77: 616-623.

    Neff JM, Cox BA, Dixit D, & Anderson JW (1976a) Accumulation and
    release of petroleum-derived aromatic hydrocarbons by four species of
    marine animals. Mar Biol, 38: 279-289.

    Neff JM, Anderson JW, Cox BA, Laughlin RB Jr, Rossi SS, & Tatem HE
    (1976b) Effects of petroleum on survival, respiration and growth of
    marine animals. In: Sources, effects and sinks of hydrocarbons in the
    aquatic environment. Proceedings of the symposium, American
    University, Washington DC, 9-11 August 1976. Washington DC, American
    Institute of Biological Sciences, pp 516-539.

    Nesnow S, Triplett LL, & Slaga TJ (1982a) Comparative tumor-initiating
    activity of complex mixtures from environmental particulate emissions
    on SENCAR mouse skin. J Natl Cancer Inst, 68: 829-834.

    Nesnow S, Evans C, Stead A, Creason J, Slaga TJ, & Triplett LL (1982b)
    Skin carcinogenesis studies of emission extracts. In: Lewtas J, ed.,
    Toxicological effects of emissions from diesel engines. Amsterdam,
    Elsevier, pp 295-320.

    Nesnow S, Triplett LL, & Slaga TJ (1983) Mouse skin tumor
    initiation-promotion and complete carcinogenesis bioassays: Mechanisms
    and biological activities of emission samples. Environ Health
    Perspectives, 47: 255-268.

    Netterstrom B (1988) Cancer incidence among urban bus drivers in
    Denmark. Int Arch Occup Environ Health, 61: 217-221.

    Newton JM, Rothman BS, & Walker FA (1991) Separation and determination
    of diesel contaminants in various fish products by capillary gas
    chromatography. J Assoc Off Anal Chem, 74: 986-990.

    Nikula KJ Snipes MB Barr EB Griffith WC Henderson RF & Mauderly JL
    (1994) Influence of particle-associated organic compounds on the
    carcinogenicity of diesel exhaust. In: Mohr U, Dungworth DL, Mauderly
    JL, & Oberdörster G, eds, Toxic and carcinogenic effects of solid
    particles in the respiratory tract. Washington DC, International Life
    Science Institute Press, pp 565-568.

    Notani PN, Shah P, Jayant K, & Balakrishnan V (1993) Occupation and
    cancers of the lung and bladder: A case-control study in Bombay.
    Int J Epidemiol, 22: 185-191.

    Novakov T (1984) The role of soot and primary oxidants in atmospheric
    chemistry. Sci Total Environ, 36: 1-10.

    Oberdörster G (1988) Lung clearence of inhaled insoluble and soluble
    particles. J Aerosol Med, 1: 289-330.

    Oberdörster G & Yu CP (1990) The carcinogenic potential of inhaled
    diesel exhaust: A particle effect? J Aerosol Sci, 21: 397-401.

    Oberdörster G, Baggs R, Gelein R, Corson N, Mercer P, Nguyen K, &
    Pepelko W (1995) Pulmonary effects of inhaled carbon black in rats.
    Toxicologist, 15: 46.

    Obländer K (1991) [Possibility of pollutant risk reduction in diesel
    engines.] In: [Carcinogenic substances in the environment: Origin,
    measurement, risk, minimization. Meeting of the Commission for Clean
    Air of the Society of German Engineers (VDI) and the German Standards
    Institute (DIN), Mannheim, 23-25 April]. Düsseldorf, Society of German
    Engineers Press, pp 245-253 (in German).

    Oehme M, Larssen S, & Brevik EM (1991) Emisson factors of PCDD and
    PCDF for road vehicles obtained by tunnel experiment. Chemosphere,
    23: 1699-1708.

    Oelert HH & Florian T (1972) [Registration and assessment of the odour
    intensity of diesel exhaust.] Staub Reinhalt Luft, 32: 400-407
    (in German).

    Ohgaki H, Hasegawa H, Kata T, Negishi C, Sato S, & Sugimura T (1985)
    Absence of carcinogenicity of 1-nitropyrene, correction of previous
    results, and new demostration of carcinogenicity of 1,6-dinitropyrene
    in rats. Cancer Lett, 25: 239-245.

    Ong T, Whong W-Z, Xu J, Burchell B, Green FHY & Lewis T (1985)
    Genotoxicity studies of rodents exposed to coal dust and diesel
    emission particulates. Environ Res, 37: 399-409.

    Ong T, Whong W-Z, Steward J & Brockman HE (1986) Chlorophyllin: A
    potent antimutagen against environmental and dietary complex mixtures.
    Mutat Res, 173: 111-115.

    Organisation for Economic Co-operation and Development (1993) Control
    of emissions from heavy-duty vehicles (OECD Environment Monographs
    No. 55). Paris, 121 pp.

    Oviatt C, Frithsen J, Gearing J, & Gearing P (1982) Low chronic
    additions of No. 2 fuel oil: Chemical behavior, biological impact and
    recovery in a simulated estuarine environment. Mar Ecol Prog Ser,
    9: 121-136.

    Parker GA, Bogo V, & Young RW (1981) Acute toxicity of conventional
    versus shale-derived JP-5 jet fuel: Light microscopic, hematologic,
    and serum chemistry studies. Toxicol Appl Pharmacol, 57: 302-317

    Partanen T, Heikkilä P, Hernberg S, Kauppinen T, Moneta G, & Ojajärvi
    A (1991) Renal cell cancer and occupational exposure to chemical
    agents. Scand J Work Environ Health, 17: 231-239.

    Partridge PA, Shala FJ, Cernansky NP, & Suffet IH (1987)
    Characterization and analysis of diesel exhaust odor. Environ Sci
    Technol, 21: 403-408.

    Patton JD, Maher VM, & McCormick JJ (1986) Cytostatic and mutagenic
    effects of 1-nitropyrene and 1-nitrosopyrene in diploid human
    fibroblasts. Carcinogenesis, 7: 89-93.

    Pederson TJ & Siak JS (1981) The role of nitroaromatic compounds in
    the direct-acting mutagenicity of diesel particle extracts. J Appl
    Toxicol, 1: 54-60.

    Pepelko WE (1982) Effects of 28 days exposure to diesel engine
    emissions in rats. Environ Res, 27: 16-23.

    Pepelko WE & Chen C (1993) Quantitative assessment of cancer risk from
    exposure to diesel engine emissions. Regul Toxicol Pharmacol,
    17: 52-65.

    Pepelko WE & Peirano WB (1983) Health effects of exposure to diesel
    engine emissions. A summary of animal studies conducted by the US
    Environmental Protection Agency's Health Effects Research Laboratories
    at Cincinnati, Ohio. J Am Coll Toxicol, 2: 253-306.

    Pepelko WE, Mattox J, Moorman WJ, & Clark JC (1981) Pulmonary function
    evaluation of cats after one year of exposure to diesel exhaust.
    Environ Int, 5: 373-376.

    Pereira MA (1982) Genotoxicity of diesel exhaust emissions in
    laboratory animals. In: Lewtas J, ed., Toxicological effects of
    emissions from diesel engines. Amsterdam, Elsevier, pp 265-276.

    Pereira MA, McMillan L, Kaur P, Gulati DK, & Sabharwal PS (1982)
    Effect of diesel exhaust emissions, particulates, and extract on
    sister chromatid exchange in transplacentally exposed fetal hamster
    liver. Environ Mutag, 4: 215-220.

    Perez Rodriguez E, Latour Perez J, Aller Alvarez JL, Alix J, & Alix Y
    (1976) [Diesel fuel pneumonia. Presentation of a case.] Rev Clin Esp,
    143: 397-400 (in Spanish).

    Peters RW, Montemagno CD, & Shem L (1992) Surfactant screening of
    diesel-contaminated soil. Hazard Waste Mater, 9: 113-136.

    Petersen BA & Chuang CC (1982) Methodology of fractionation and
    partition of diesel exhaust particulate samples. In: Lewtas J, ed.,
    Toxicological effects of emissions from diesel engines. Amsterdam,
    Elsevier, pp 51-67.

    Piantadosi S (1994) Invited commentary: Ecologic biases. Am J
    Epidemiol, 139: 761-764.

    Pipho MJ, Ambs JL, & Kittelson DB (1986) In-cylinder measurements of
    particulate formation in an indirect injection diesel engine (SAE
    Technical Paper No. 860024). Warrendale, PA, Society of Automotive
    Engineers, 9 pp.

    Pitts JN Jr, Van Cauwenberghe KA, Grosjean D, Schmid JP, Fitz DR,
    Belser WL Jr, Knudson GB, & Hynds PM (1978) Atmospheric reactions of
    polycyclic aromatic hydrocarbons: Facile formation of mutagenic nitro
    derivatives. Science, 202: 515-519.

    Pitts JN Jr, Lokensgard DM, Harger W, Fisher TS, Mejia V, Schuler JJ,
    Scorziell GM, & Katzenstein YA (1982) Mutagens in diesel exhaust
    particulate. Identification and direct activities of
    6-nitrobenzo( a)pyrene, 9-nitroanthracene, 1-nitropyrene and
    5 H-phenanthrene(4,5- bcd)pyran-5-one. Mutat Res, 103: 241-249.

    Pitts JN Jr, Sweetman JA, Zielinska B, Winer AM, & Atkinson R (1985)
    Determination of 2-nitrofluoranthene and 2-nitropyrene in ambient
    particulate organic matter: Evidence for atmospheric reactions. Atmos
    Environ, 19: 1601-1608.

    Plopper CG, Hyde DM, & Weir AJ (1983) Centriacinar alterations in
    lungs of cats chronically exposed to diesel exhaust. Lab Invest,
    49: 391-399.

    Poirier A, Baudin Laurencin F, Bodennec G, & Quentel C (1986)
    [Experimental poisoning of the rainbow trout,  Salmo gairdneri
    Richardson, by diesel engine fuel: Haematological changes, histology].
    Aquaculture, 55: 115-137 (in French).

    Pope CA III & Dockery DW (1992) Acute health effects of PM10 pollution
    on symptomatic and asymptomatic children. Am Rev Respir Dis,
    145: 1123-1128.

    Pople A, Simpson RD, & Cairns SC (1990) An incident of southern ocean
    oil pollution: Effects of a spillage of diesel fuel on the rocky shore
    of Macquarie Island (sub-Antarctic). Aust J Mar Freshwater Res,
    41: 603-620.

    Porter HO (1990) Aviators intoxicated by inhalation of JP-5 fuel
    vapors. Aviat Space Environ Med, 61: 654-656.

    Portier C, Hedges J, & Hoel D (1986) Age-specific models of mortality
    and tumor onset for historical control animals in the National
    Toxicology Program's carcinogenicity experiments. Cancer Res,
    46: 4372-4378.

    Pott F (1991) [Diesel engine exhaust -- results of laboratory animal
    experiments relative to risk estimation]. In: [Carcinogenic substances
    in the environment: Origin, measurement, risk, minimization. Meeting
    of the Commission for Clean Air of the Society of German Engineers
    (VDI) and the German Standards Institute (DIN), Mannheim, 23-25
    April]. Düsseldorf, Society of German Engineers Press, pp 211-244
    (in German).

    Pott F & Heinrich U (1987) [Diesel engine exhaust and lung cancer.
    Data from animal experiments and their evaluation with respect to the
    hazard for humans]. Düsseldorf, Association for the promotion of
    Environmental Health and Silicosis Research, Vol. 19, pp 130-167
    (in German).

    Pott F & Roller M (1994) Relevance of non-physiologic exposure routes
    for carcinogenicity studies of solid particles. In: Mohr U, Dungworth
    DL, Mauderly JL, & Oberdörster G, eds, Toxic and carcinogenic effects
    of solid particles in the respiratory tract. Washington DC,
    International Life Sciences Institute Press, pp 109-125.

    Pott F, Heinrich U & Roller M (1993) [Diesel engine emissions.] In:
    Wichmann HE, Schlipköter HW, & Fülgraff G, eds, [Handbook of
    environmental medicine]. Landsberg, Lech, Ecomed mbH, 77 pp
    (in German).

    Pott F, Dungworth DL, Heinrich U, Muhle H, Kamino K, Germann PG,
    Roller M, Rippe RM, & Mohr U (1994) Lung tumours in rats after
    intratracheal instillation of dusts. Ann Occup Hyg,
    38 (Suppl 1): 357-363.

    Prakash CB, Rideout GB, & Kirshenblatt M (1992) Exhaust emissions from
    diesel powered urban transit buses (Paper No. IU-17A). In: Proceedings
    of the 9th World Clean Air Congress and Exhibition, Montreal, 30
    August-4 September 1992, Ottawa, Environment Canada, Vol. 2, 22 pp.

    Pryor P (1983) Health Hazard Evaluation. Trailways Bus System, Denver,
    CO (PB85-102952, Report No HETA 81-416-1334). Cincinnati, OH, National
    Institute for Occupational Safety and Health, 11 pp.

    Pulich WM Jr, Winters K, & Van Baalen C (1974) The effects of a No. 2
    fuel oil and two crude oils on the growth and photosynthesis of
    microalgae. Mar Biol, 28: 87-94.

    Purdham JI, Holness DL, & Pilger CW (1987) Environmental and medical
    assessment of stevedores employed in ferry operations. Appl Ind Hyg,
    2: 133- 139.

    Quality of Urban Air Review Group (1993) Diesel vehicle emissions and
    urban air quality. Birmingham, 88 pp.

    Quinto I & De Marinis E (1984) Sperm abnormalities in mice exposed to
    diesel particulate. Mutat Res, 130: 242.

    Raabe OG, Al-Bayati MA, Teague SV, & Rasolt A (1988) Regional
    deposition of inhaled monodisperse coarse and fine aerosol particles
    in small laboratory animals. Ann Occup Hyg, 32 (Suppl 1): 53-63.

    Rasmussen RE (1990) Effect of fuel properties on mutagenic activity in
    extracts of heavy-duty diesel exhaust particulate. J Air Waste Manage
    Assoc, 40: 1391-1396.

    Reger R, Hancock J, Hankinson J, Hearl F, & Merchant J (1982) Coal
    miners exposed to diesel exhaust emissions. Ann Occup Hyg,
    26: 799-815.

    Regnier ZR & Scott BF (1975) Evaporation rates of oil components.
    Environ Sci Technol, 9: 469-472.

    Reidenberg MM, Powers DV, Sevy RW, & Bello CT (1964) Acute renal
    failure due to nephrotoxins. Am J Med Sci, 247: 25-29.

    Rice SD, Short JW, & Karinen JF (1976) Toxicity of Cook Inlet crude
    oil and No. 2 fuel oil to several Alaskan marine fishes and
    invertebrates. In: Sources, effects and sinks of hydrocarbons in the
    aquatic environment. Washington DC, American Institute of Biological
    Sciences, pp 395-406.

    Rice SD, Short JW, & Karinen JF (1977a) Comparative oil toxicity and
    comparative animal sensitivity. In: Wolfe DA, Anderson JW, Button DK,
    Malins, DC, Roubal T, & Varanasi U, eds, Fate and effects of petroleum
    hydrocarbons in marine organisms and ecosystems. Oxford, Pergamon
    Press, pp 78-94.

    Rice SD, Thomas RE, & Short JW (1977b) Effect of petroleum
    hydrocarbons on breathing and coughing rates and hydrocarbons
    uptake-depuration in pink salmon fry. In: Verberg JF, Calabrese A,
    Thurberg FP, & Vernberg WB, eds, Physiological responses of marine
    biota to pollutants. New York, Academic Press, pp 259-277.

    Risch HA, Burch JD, Miller AB, Hill GB, Steele R, & Howe GR (1988)
    Occupational factors and the incidence of cancer of the bladder in
    Canada. Br J Ind Med, 45: 361-367.

    Rogge WF, Hildemann LM, Mazurek MA, Cass GR & Simonelt BRT (1993)
    Sources of fine organic aerosol 2. Noncatalyst and catalyst-equipped
    automobiles and heavy-duty diesel trucks. Environ Sci Technol,
    27: 636-651.

    Roller M & Pott F (1994) Use of experimental data for risk assessment
    of fibrous dusts and other aerosols. Inf Biom Epidemiol Med Biol,
    25: 301-311.

    Rossi SS (1977) Bioavailability of petroleum hydrocarbons from water,
    sediments and detritus to the marine annelid  Neanthes arenaceodentata.
    In: Proceedings, 1977 Oil Spill Conference. Washington DC, American
    Petroleum Institute, pp 621-625.

    Rossi SS & Anderson JW (1976) Toxicity of water-soluble fractions of
    No. 2 fuel oil and South Louisiana crude oil to selected stages in the
    life history of the polychaete,  Neanthes arenaceodentata. Bull Environ
    Contam Toxicol, 16: 18-24.

    Rossi SS & Anderson JW (1978) Effects of No. 2 fuel oil
    water-soluble-fractions on growth and reproduction in  Neanthes
     arenaceodentata (Polychaeta: Annelida). Water Air Soil Pollut,
    9: 155-170.

    Rossi SS, Anderson JW, & Ward GS (1976) Toxicity of water-soluble
    fractions of four test oils for the polychaetous annelids  Neanthes
     arenaceodentata and  Capitella capitata. Environ Pollut, 10: 9-18.

    Rudell B, Marqvardsen MD, Lindström O, & Kolmodin-Hedman B (1989) A
    method for controlled exposure to diesel exhaust from a continuously
    idling diesel engine. J Aerosol Sci, 20: 1281-1284.

    Rudell B, Sandström T, Stjernberg N, & Kolmodin-Hedman B (1990)
    Controlled diesel exhaust exposure in an exposure chamber: Pulmonary
    effects investigated with bronchoalveolar lavage. J Aerosol Sci,
    21: 411-414.

    Ruder AM, Fine LJ, & Sundin DS (1990) National estimates of
    occupational exposure to animal bladder-tumorigens. J Occup Med,
    32: 797-805.

    Rushton L, Alderson MR, & Nagarajah CR (1983) Epidemiological survey
    of maintenance workers in London transport executive bus garages and
    Chiswick Works. Br J Ind Med, 40: 340-345.

    Russell TJ (1989) Additives influencing diesel fuel combustion. In:
    Owen K, ed., Gasoline and diesel fuel additives (Critical reports on
    applied chemistry, Vol 25). Chichester, John Wiley & Sons, pp 65-104.

    Rykowski RA & Brochu AJ (1986) Diesel particulate emissions: The
    United States experience, with extrapolations to Europe and Japan. In:
    Lee SD, Schneider I, Grant LD, & Verkerk PJ, eds, Aerosols. Chelsea,
    MI, Lewis Publishers Inc., pp443-457.

    Sagai M, Saito H, Ichinose T, Kodama M, & Mori Y (1993) Biological
    effects of diesel exhaust particles. I. In vitro production of
    superoxide and  in vivo toxicity in mouse. Free Radicals Biol Med,
    14: 37-47.

    Salmeen IT, Pero AM, Zator R, Schuetzle D, & Riley TL (1984) Ames
    assay chromatograms and the identification of mutagens in diesel
    particle extracts. Environ Sci Technol, 18: 375-382.

    Salmeen IT, Gorse Jr RA, & Pierson WR (1985) Ames assay chromatograms
    of extracts of diesel exhaust particles from heavy-duty trucks on the
    road and from passenger cars on a dynamometer. Environ Sci Technol,
    19: 270-273.

    Sandmeyer EE (1981) Aliphatic hydrocarbons. In: Clayton GD & Clayton
    FE, eds, Patty's industrial hygiene and toxicology, Vol. 2B,
    Toxicology. New York, John Wiley & Sons, pp 3175-3431.

    Sarojini R, Khan AK, & Nagabhushanam R (1989) Effect of petroleum
    hydrocarbons (petrol and diesel) on the physiology of the crab,
     Barytelphusa cunicularis I -- oxygen consumption. J Environ Biol,
    10: 363-365.

    Scheepers PTJ & Bos RP (1992a) Combustion of diesel fuel from a
    toxicological perspective. I. Origin of incomplete combustion
    products. Int Arch Occup Environ Health, 64: 149-161.

    Scheepers PTJ & Bos RP (1992b) Combustion of diesel fuel from a
    toxicological perspective. II. Toxicity. Int Arch Occup Environ
    Health, 64: 163-177.

    Scheepers PTJ, Velders DD, Martens MHJ, Noordhoek J, & Bos RP (1994a)
    Gas chromatographic-mass spectrometric determination of nitro
    polycyclic aromatic hydrocarbons in airborne particulate matter from
    workplace atmospheres contaminated with diesel exhaust. J Chromatogr,
    677: 107-121.

    Scheepers PTJ, Thuis HJTM, Martens MHJ, & Bos RP (1994b) Assessment of
    occupational exposure to diesel exhaust. The use of an immunoassay for
    the determination of urinary metabolites of nitroarenes and polycyclic
    aromatic hydrocarbons. Toxicol Lett, 72: 191-198.

    Schenker MB & Speizer FE (1979) A retrospective cohort study of diesel
    exhaust exposure in railroad workers: Study design and methodologic
    issues. In: Pepelko WE, Danner RB, & Clark NA, eds, Health effects of
    diesel engine emissions (EPA-600/9-80-057b). Cincinnati, OH,
    Environmental Protection Agency.

    Schenker MB, Smith T, Muñoz A, Woskie S, & Speizer FE (1984) Diesel
    exposure and mortality among railway workers: Results of a pilot
    study. Br J Ind Med, 41: 320-327.

    Schenker MB, Kado NY, Hammond SK, Samuels SJ, Woskie SR, & Smith TJ
    (1992) Urinary mutagenic activity in workers exposed to diesel
    exhaust. Environ Res, 57: 133-148.

    Schiffmann D & Henschler D (1992) [Studies of diesel engine exhaust
    fractions on the genotoxic and cell-transforming properties with the
    model of Syrian hamster embryo fibroblasts and with lung cells]. In:
    [Effects of diesel engine exhausts on health]. Munich, GSF Research
    Centre for Environment and Health Inc. pp 39-42 (in German).

    Schuetzle D (1983) Sampling of vehicle emissions for chemical analysis
    and biological testing. Environ Health Perspectives, 47: 65-80.

    Schuetzle D & Frazier JA (1986) Factors influencing the emission of
    vapor and particulate phase components from diesel engines. In:
    Ishinishi N, Koizumi A, McClellan RO, & Stöber W, eds, Carcinogenic
    and mutagenic effects of diesel engine exhaust. Amsterdam, Elsevier,
    pp 41-63.

    Schuetzle D & Perez JM (1981) A CRC cooperative comparison of
    extraction and HPLC techniques for diesel particulate emissions (Paper
    81-564). In: Proceedings of the 74th Annual Meeting of the Air
    Pollution Control Association, 16-21 June 1981, Philadelphia.
    Philadelphia, PA, Air Pollution Control Association, pp 1-32.

    Schuetzle D & Perez JM (1983) Factors influencing the emissions of
    nitrated-polynuclear aromatic hydrocarbons (nitro-PAH) from diesel
    engines. J Air Pollut Control Assoc, 33: 751-755.

    Schuetzle D, Riley TL, Prater TJ, Harvey TM, & Hunt DF (1982) Analysis
    of nitrated polycyclic aromatic hydrocarbons in diesel particulates.
    Anal Chem, 54: 265-271.

    Schuetzle D, Jensen TE, & Ball JC (1985) Polar polynuclear aromatic
    hydrocarbon derivatives in extracts of particulates: Biological
    characterization and techniques for chemical analysis. Environ Int,
    11: 169-181.

    Schuler RL & Niemeier RW (1981) A study of diesel emissions on
    Drosophila. Environ Int, 5: 431-434.

    Schultz T, Witschi H, Smith L, Haschek W, Holland J, Epler J, Fry R,
    Rao T, Larimer F, & Dumont J (1981) Health effects research in oil
    shale development (ORNL/TM-8034). Oak Ridge, TN, Oak Ridge National
    Laboratories, 56 pp.

    Schwartz J (1993) Air pollution and daily mortality in Birmingham,
    Alabama. Am J Epidemiol, 136: 1136-1147.

    Schwartz J, Slater D, Larson TV, Pierson WE, & Koenig JO (1993)
    Particulate air pollution and hospital emergency room visits for
    asthma in Seattle. Am Rev Respir Dis, 147: 826-831.

    Schwind KH, Thoma H, Hutzinger O, Dawidowsky N, Weberuss, Hagenmaier
    H, Bühler U, Greiner R, Essers U, & Bessey E (1991) [Emission of
    halogenated dibenzodioxins (PXDD) and dibenzofurans (PXDF) from
    internal combustion engines]. Z Umweltchem Ökotox 3: 291-298
    (in German).

    Shefner AM, Collins BR, Dooley L, Fiks A, Graf JL, & Preache MM (1982)
    Respiratory carcinogenicity of diesel fuel emissions. Interim results.
    In: Lewtas J, ed., Toxicological effects of emissions from diesel
    engines. Amsterdam, Elsevier, pp 329-350.

    Siemiatycki J, Dewar R, Nadon L, Gérin M, Richardson L, & Wacholder S
    (1987) Associations between several sites of cancer and twelve
    petroleum-derived liquids: Results from a case-referent study in
    Montreal. Scand J Work Environ Health, 13: 493-504.

    Siemiatycki J, Gérin M, Stewart P, Nadon L, Dewar R, & Richardson L
    (1988) Associations between several sites of cancer and ten types of
    exhaust and combustion products. Scand J Environ Health, 14: 79-90.

    Sienicki EJ & Mago RS (1992) Re-evaluation of diesel engine
    particulate emission inventories. In: Toxic air pollutants from mobile
    sources. Pittsburgh, PA, Air and Waste Management Association,
    pp 151-164.

    Silverman DT, Hoover RN, Albert S, & Graff KM (1983) Occupation and
    cancer of the lower urinary tract in Detroit. J Natl Cancer Inst,
    70: 237-245.

    Silverman DT, Hoover RN, Mason TJ, & Swanson GM (1986) Motor
    exhaust-related occupations and bladder cancer. Cancer Res,
    46: 2113-2116.

    Singh AK, & Gaur JP (1990) Effects of petroleum oils and their
    paraffinic, asphaltic, and aromatic fractions on photosynthesis and
    respiration of microalgae. Ecotoxicol Environ Saf, 19: 8-16.

    Snipes MB & McClellan RO (1985) Retention of 14C-labeled
    aluminosilicate particles inhaled by dogs and guinea pigs -- simulation
    model projections for humans. In: Inhalation Toxicology Research
    Institute Annual report 1984-1985, Albuquerque, NM, pp 96-99.

    Snipes MB, Boecker BB, & McClellan RO (1983) Retention of monodisperse
    or polydisperse aluminosilicate particle inhaled by dogs, rats, and
    mice. Toxicol Appl Pharmacol, 69: 345-362.

    Song H-G, Wang X, & Bartha R (1990) Bioremediation potential of
    terrestrial fuel spills. Appl Environ Microbiol, 56: 652-656.

    Standardization Board in Sweden (1991) Swedish Standard SS 15 54 35:
    Diesel fuel oil for high-speed diesel engines. Stockholm.

    State Committee for Immission Protection (1992) [Diesel motor
    emissions (DME)]. In: [Cancer risk through polluted air. Development
    of 'Evaluation standards for cancerogenic air pollution' commissioned
    by the Conference for Ministers of the Environment], Düsseldorf,
    pp 89-93 (in German).

    Steenland K, Burnett C, & Osorio AM (1987) A case-control study of
    bladder cancer using city directories as a source of occupational
    data. Am J Epidemiol, 126: 247-257.

    Steenland NK, Silverman DT, & Hornung RW (1990) Case-control study of
    lung cancer and truck driving in the Teamsters Union. Am J Public
    Health, 80: 670- 674.

    Steenland K, Silverman D, & Zaebst D (1992) Exposure to diesel exhaust
    in the trucking industry and possible relationships with lung cancer.
    Am J Ind Med, 21: 887-890.

    Steineck G, Plato N, Gerhardsson M, Norell SE, & Hogstedt C (1990)
    Increased risk of urothelial cancer in Stockholm during 1985-87 after
    exposure to benzene and exhausts. Int J Cancer, 45: 1012-1017.

    Stenberg U, Alsberg T, & Westerholm W (1983) Emission of carcinogenic
    components with automobile exhausts. Environ Health Perspectives,
    47: 53-56.

    Stern FB, Curtis RA, & Lemen RA (1981) Exposure of motor vehicle
    examiners to carbon monoxide: A historical prospective mortality
    study. Arch Environ Health, 36: 59-66.

    Stirling HP (1977) Effects of a spill of marine diesel oil on the
    rocky shore fauna of Lamma Island, Hong Kong. Environ Pollut,
    12: 93-117.

    Stöber W (1986) Experimental induction of tumors in hamsters, mice and
    rats after long-term inhalation of filtered and unfiltered diesel
    engine exhaust. In: Ishinishi N, Koizumi A, McClellan RO, & Stöber W,
    eds, Carcinogenic and mutagenic effects of diesel engine exhaust.
    Amsterdam, Elsevier, pp 421-439.

    Stone WA (1991) Assessing health risks associated with diesel
    contaminated soils and groundwater. In: Calabrese EJ & Kostecki PT,
    eds, Hydrocarbon contaminated soils, Vol. 1, Remediation techniques,
    environmental fate, risk assessment, analytical methodologies,
    regulatory considerations. Chelsea, MI, Lewis Publishers, pp 167-179.

    Strom KA, Chan TL, & Johnson JT (1988) Pulmonary retention of inhaled
    submicron particles in rats: Diesel exhaust exposures and lung
    retention model. Ann Occup Hyg, 32: 645-657.

    Strom KA, Garg BD, Johnson JT, D'Arcy JB, & Smiler KL (1990) Inhaled
    particle retention in rats receiving low exposures of diesel exhaust.
    J Toxicol Environ Health, 29: 377-398.

    Stromgren T & Nielsen MV (1991) Spawning frequency, growth and
    mortality of  Mytilus edulis larvae, exposed to copper and diesel
    oil. Aquat Toxicol, 21: 171-180.

    Stromgren T & Reiersen LO. (1988) A new method for testing toxicity of
    drilling fluid; effect on growth of mussels. Oil Chem Pollut,
    4: 127-138.

    Stromgren T, Nielsen MV, & Ueland K (1986) Short-term effect of
    microencapsulated hydrocarbons on shell growth of  Mytilus edilus.
    Mar Biol, 91: 33-39.

    Sun JD, Wolff RD, & Kanapilly GM (1982) Deposition, retention, and
    biological fate of inhaled benzo[ a]pyrene adsorbed onto ultrafine
    particles and as a pure aerosol. Toxicol Appl Pharmacol,
    65: 231-244.

    Sun JD, Wolff RK, Kanapilly GM, & McClellan RO (1984) Lung retention
    and metabolic fate of inhaled benzo( a)pyrene associated with diesel
    exhaust particles. Toxicol Appl Pharmacol, 73: 48-59.

    Suzuki T, Nakajima T, Maejima K, Kato A, Takaki Y, Kuwabara N,
    Hisanaga A, & Ishinishi N (1990) [Long-term inhalation experiments on
    the health effects of diesel exhaust]. Taiki Osen Gatukai-shi,
    25: 192-205 (in Japanese, with English summary).

    Swanson GM, Lin C-S, & Burns PB (1993) Diversity in the association
    between occupation and lung cancer among black and white men. Cancer
    Epidemiol Biomarkers Prev, 2: 313-320.

    Swarin SJ & Williams RL (1980) Liquid chromatographic determination of
    benzo[ a]pyrene in diesel exhaust particulate: Verification of the
    collection and analytical methods. In: Björseth A & Dennis AJ, eds,
    Polynuclear aromatic hydrocarbons, 4th International Symposium:
    Chemistry and biological effects. Columbus, OH, Battelle Press,
    pp 771-806.

    Takafuji S, Suzuki S, Koizumi K, Tadokoro K, Miyamoto T, Ikemori R, &
    Muranaka M (1987) Diesel-exhaust particulates inoculated by the
    intranasal route have an adjuvant activity for IgE production in mice.
    J Allergy Clin Immunol, 79: 639-645.

    Takafuji S, Suzuki S, Muranaka M, & Miyamoto T (1989) Influence of
    environmental factors on IgE production. In: Mast cells and the
    allergic response (Ciba Foundation Symposium 147). Chichester, John
    Wiley & Sons, pp 188-204.

    Takaki Y, Kitamura S, Kuwabara N, & Fukuda Y (1989) Long-term
    inhalation studies of exhaust from the diesel engine in F-344 rats:
    The quantitative relationship between pulmonary hyperplasia and
    anthracosis. Exp Pathol, 37: 56-61.

    Takemoto K, Yoshimura H, & Katayama H (1986) Effects of chronic
    inhalation exposure to diesel exhaust on the development of lung
    tumors in di-isopropanol-nitrosamine-treated F344 rats and newborn
    C57BL and ICR mice. In: Ishinishi N, Koizumi A, McClellan RO, & Stöber
    W, eds, Carcinogenic and mutagenic effects of diesel engine exhaust.
    Amsterdam, Elsevier, pp 311-327.

    Tan W & Chen C. (1992) A nonhomogeneous stochastic model of
    carcinogenesis and its applications. In: Arino O, Axelrod D, & Kimmel
    M, eds, Mathematical population dynamics 2. Winnipeg, Manitoba, Wuerz
    Publishing Co., pp 49-67.

    Tarshis IB (1981) Uptake and depuration of petroleum hydrocarbons by
    crayfish. Arch Environ Contam Toxicol, 10: 79-86.

    Tatem HE (1977) Accumulation of naphthalenes by grass shrimp: Effects
    on respiration, hatching, and larval growth. In: Wolfe DA, Anderson
    JW, Button DK, Malins DC, Roubel T, & Varanasi U, eds, Fate and effect
    of petroleum hydrocarbons in marine ecosystems and organisms. Oxford,
    Pergamon Press, pp 201-209.

    Tedengren M & Kautsky N (1987) Comparative stress response to diesel
    oil and salinity changes of the blue mussel,  Mytilus edilus, from
    the Baltic and North seas. Ophelia, 28: 1-9.

    Tedengren M, Arnér M, & Kautsky N (1988) Ecophysiology and stress
    response of marine and brackish water Gammarus species (Crustacea
    amphipoda) to changes in salinity and exposure to cadmium and
    diesel-oil. Mar Ecol Prog Ser, 47: 107-116.

    Thomas RG (1990) Volatilization from water. In: Lyman WJ, Reehl WF, &
    Rosenblatt DH, eds, Handbook of chemical property estimation methods.
    Environmental behaviour of organic compounds. New York, McGraw-Hill
    Book Co., pp 15:1-15:34.

    Törnqvist M, Kautiainen A, Gatz RN, & Ehrenberg L (1988) Hemoglobin
    adducts in animals exposed to gasoline and diesel exhausts. I.
    Alkenes. J Appl Toxicol, 8: 159-170.

    Tucker JD, Xu J, Stewart J, Baciu PC, & Ong T (1986) Detection of
    sister chromatid exchanges induced by volatile genotoxicants. Teratog
    Carcinog Mutag, 6: 15-21.

    Ulfvarson U & Alexandersson R (1990) Reduction in adverse effect on
    pulmonary function after exposure to filtered diesel exhaust. Am J Ind
    Med, 17: 341-347.

    Ulfvarson U, Alexandersson R, Aringer L, Svensson E, Hedenstierna G,
    Hogstedt C, Holmberg B, Rosén G, & Sorsa M (1987) Effects of exposure
    to vehicle exhaust on health. Scand J Work Environ Health,
    13: 505-512.

    Ullman TL, Hare CT, & Baines TM (1984) Influence of maladjustment on
    emissions from two heavy-duty diesel bus engines (SAE Paper
    No. 840416). Warrendale, PA, Society of Automotive Engineers, 11 pp.

    Ungers LJ (1984) Measurement of exhaust emissions from diesel-powered
    forklifts during operations in ammunition storage magazines (phase I)
    (Report No. AD-A141 792). Cincinnati, OH, PEDCO Environmental Inc.,
    57 pp.

    Ungers LJ (1985) Measurement of exhaust emissions from diesel-powered
    forklifts during operations in ammunition storage magazines (phase
    II). Final report for the period July 23, 1984 to March 15, 1985
    (Report No. AD-A153-092). Cincinnati, OH, PEI Associates, 47 pp.

    Upreti RK, Das M, & Shanker R (1989) Dermal exposure to kerosene. Vet
    Hum Toxicol, 31: 16-20.

    US Department of Energy (1988) The motor fuel consumption model
    fourteenth periodical report (DOE/OR/21400 -- H12, DE 89-011358).
    Arlingston, VA, Energy and Environmental Analysis Inc., 16 pp.

    US Environmental Protection Agency (1986) The risk assessment
    guidelines of 1986 (EPA/600/8-87/045). Washington DC, Office of Health
    and Environmental Assessment.

    US Environmental Protection Agency (1991) Alpha -globulin:
    Association with chemically induced renal toxicity and neoplasia in
    the male rat (Risk Assessment Forum Series) (EPA/625/3-91/019F).
    Washington DC, 118 pp.

    US Environmental Protection Agency (1992a) Integrated risk information
    system (IRIS). Reference concentration (RfC) file for inhalation
    exposure for diesel engine emissions. Online (Verification date
    6/25/92). Cincinnati, OH, Office of Health and Environmental
    Assessment, 25 pp.

    US Environmental Protection Agency (1992b) Regulation of fuels and
    fuel additives: Standards for highway diesel fuel quality-sulfur
    content. Final rule. Fed Reg, 57: 19535-19537.

    US Environmental Protection Agency (1993) Motor vehicle-related air
    toxics study (EPA 420-R-93-005). Ann Arbor, MI, Office of Mobile
    Sources, pp 4.1-4.17, 9.1-9.62.

    US National Institute for Occupational Safety and Health (1989a)
    Health Hazard Evaluation Report. Consolidated Freightways Pocono
    Summit, Pennsylvania (HETA 87-232-1948). Cincinnatti, OH, US
    Department of Health and Human Services, 20 pp.

    US National Institute for Occupational Safety and Health (1989b)
    Health Hazard Evaluation Report. Consolidated Freightways Peru,
    Illinois (HETA 88-077-1969). Cincinnatti, OH, US Department of Health
    and Human Services, 31 pp.

    US National Institute for Occupational Safety and Health (1990a) Final
    report of the NIOSH health hazard evaluation (HHE) conducted at
    Benedict Enterprises Dayton, Ohio (October 2 and November 5, 1987)
    (HETA 87-237). Cincinnatti, OH, US Department of Health and Human
    Services, 33 pp.

    US National Institute for Occupational Safety and Health (1990b) Final
    report of the NIOSH health hazard evaluation (HHE) conducted at
    Madison Metro's Transit System's Maintenance and Administration
    Facility (January 10-13, 1989) (HETA 88-218). Cincinnatti, OH, US
    Department of Health and Human Services, 38 pp.

    US National Institute for Occupational Safety and Health (1990c)
    Health Hazard Evaluation Report. Yellow Freight System, Inc.,
    Columbus, Ohio, Maybrook, New York (HETA 90-088-2110). Cincinnatti,
    OH, US Department of Health and Human Services, 127 pp.

    US National Institute for Occupational Safety and Health (1991)
    Research plan. A sampling and analytical method for airborne
    diesel-exhaust particels. Cincinnatti, OH, Centers for Disease
    Control, National Institute for Occupational and Health, pp 1-34.

    US National Research Council (1983) Risk assessment in the Federal
    Government: Managing the process. Committee on the Institutional Means
    for Assessment of Risks to Public Health. Washington DC, National
    Academy Press, pp 1-50.

    US National Research Council (1985) Oil in the sea. Inputs, fates, and
    effects. Washington DC, National Academy Press, 587 pp.

    US National Toxicology Program (1986) Toxicology and carcinogenesis
    studies of marine diesel fuel and JP-5 navy fuel (CAS-No 8008-20-6) in
    B6C3F1 mice (dermal studies) (Technical report series No. 310).
    Research Triangle Park, NC, 206 pp.

    Vallyathan V, Virmani R, Rochlani S, Green FHY, & Lewis T (1986)
    Effect of diesel emissions and coal dust inhalation on heart and
    pulmonary arteries of rats. J Toxicol Environ Health, 19: 33-41.

    Van Beckhoven LC (1991) Effects of fuel properties on diesel engine
    emissions -- A review of information available to EEC-MVEG group (SAE
    Technical Paper No. 910608). Warrendale, PA, Society of Automotive
    Engineers, pp 237-249.

    Vandermeulen JH & Lee RW (1986) Lack of mutagenic activity of crude
    and refined oils in the unicellular alga  Chlamydomonas reinhardtii.
    Bull Environ Contam Toxicol, 36: 250-253.

    Vandermeulen JH, Foda A, & Stuttard C (1985) Toxicity vs mutagenicity
    of some crude oils, distillates and their water soluble fractions.
    Water Res, 19: 1283-1289.

    Van Vleet ES & Quinn JG (1978) Contribution of chronic petroleum
    inputs to Narragansett Bay and Rhode Island Sound sediments. J Fish
    Resour Board Can, 35: 536-543.

    Vinegar A, Carson A, & Pepelko WE (1981) Pulmonary function changes in
    Chinese hamster exposed 6 months to diesel exhaust. Environ Int,
    5: 369-371.

    Vineis P & Magnani C (1985) Occupation and bladder cancer in males: A
    case-control study. Int J Cancer, 35: 599-606.

    Volkswagen AG (1989) Unregulated motor vehicle exhaust gas components.
    Wolfsburg, Research and Development, pp 1-128.

    Vostal JJ (1986) Factors limiting the evidence for chemical
    carcinogenicity of diesel emissions in long-term inhalation
    experiments. In: Ishinishi N, Koizumi A, McClellan RO, & Stöber W,
    eds, Carcinogenic and mutagenic effects of diesel engine exhaust.
    Amsterdam, Elsevier, pp. 381-396.

    Wade WR & Jones CM (1984) Current and future light duty diesel engines
    and their fuels (Paper No. 840105). In: International congress and
    exposition, Detroit, Michigan, 27 February-2 March 1984. Warrendale,
    PA, Society of Automotive Engineers.

    Wade JF III & Newman LS (1993) Diesel asthma. Reactive airways disease
    following overexposure to locomotive exhaust. J Occup Med,
    35: 149-154.

    Wade TL & Quinn JG (1980) Incorporation, distribution, and fate of
    saturated petroleum hydrocarbons in sediments from a controlled marine
    ecosystem. Mar Environ Res, 3: 15-34.

    Wallace WE, Keane MJ, Hill CA, Xu J, & Ong T (1987) Mutagenicity of
    diesel exhaust particles and oil shale particles dispersed in lecithin
    surfactant. J Toxicol Environ Health, 21: 163-171.

    Wallace WE, Keane MJ, Xing S, & Ong T (1990) Mutagenicity of diesel
    exhaust soot dispersed in phospholipid surfactants. In: Seemayer NH &
    Hadnagy W, eds, Environmental hygiene II. Berlin, Springer, pp 7-10.

    Waller RE (1981) Trends in lung cancer in London in relation to
    exposure to diesel fumes. Environ Int, 5: 479-483

    Waller RE, Commins BT, & Lawther PJ (1965) Air pollution in a city
    street. Br J Ind Med, 22: 128-138.

    Waller RE, Hampton L, & Lawther PJ (1985) A further study of air
    pollution in diesel bus garages. Br J Ind Med, 42: 824-830.

    Walrath J, Rogot E, Murray J, & Blair A (1985) Mortality patterns
    among US veterans by occupation and smoking status, Vol. 1 (NIH
    Publication No. 85-2756). Washington DC, US Government Printing
    Office, p 206.

    Wan MT, Watts RG, & Moul DJ (1990) Acute toxicity to juvenile Pacific
    salmonids and rainbow trout of butoxyethyl esters of 2,4-D, 2,4-DP and
    their formulated product: Weedone CB and its carrier. Bull Environ
    Contam Toxicol, 45: 604-611.

    Wang X & Bartha R (1990) Effects of bioremediation on residues,
    activity and toxicity in soil contaminated by fuel spills. Soil Biol
    Biochem, 22: 501-505.

    Waxweiler RJ, Wagoner JK, & Archer VE (1973) Mortality of potash
    workers. J Occup Med, 15: 486-489.

    Weidmann K, Menrad H, Reders K, & Hutcheson RC (1988) Diesel fuel
    quality effects on exhaust emissions (SAE Technical Paper No. 881649).
    Warrendale, PA, Society of Automotive Engineers, 11 pp.

    Westerholm R (1987) Inorganic and organic compounds in emissions from
    diesel powered vehicles. A literature survey (Report No; 3389).
    Stockholm, National Swedish Environmental Protection Board.

    Westerholm R & Egebäck KE (1991) Impact of fuels on diesel exhaust
    emissions. A chemical and biological characterization (Report
    No. 3968). Solna, National Swedish Environmental Protection Agency,
    61 pp.

    Westerholm R, Alsberg T, Strandell M, Frommelin A, Grigoriadis V,
    Hantzaridou A, Maitra G, Winquist L, Rannug U, Egebäck KE, &
    Bertilsson (1986) Chemical analysis and biological testing of
    emissions from a heavy duty diesel truck with and without two
    different particulate traps (SAE Technical Paper No. 860014).
    Warrendale, PA, Society of Automotive Engineers, pp 73-83.

    Wheeler RW, Hearl FJ, & McCawley M (1981) An industrial hygiene
    characterization of exposure to diesel emissions in an underground
    coal mine. Environ Int, 5: 485-488.

    White HJ & Garg BD (1981) Early pulmonary response of the rat lung to
    inhalation of high concentration of diesel particles. J Appl Toxicol,
    1: 104-110.

    WHO (1987) Air quality guidelines for Europe (WHO Regional
    Publications, European Series No. 23). Copenhagen, Regional Office for
    Europe.

    WHO (1990) Revised guidelines for the preparation of Environmental
    Health Criteria monographs (PCS/90.69). Geneva.

    Whong W-Z, Wen Y-F, Stewart J, & Ong T-M (1986) Validation of the
    SOS/umu test with mutagenic complex mixtures. Mutat Res, 175: 139-144.

    Wichmann HE & Brüske-Hohlfeld I (1991) [Epidemiological findings on
    the cancer risk due to diesel engine exhausts]. In: [Carcinogenic
    substances in the environment: origin, measurement, risk,
    minimization. Meeting of the Commission for Clean Air of the Society
    of German Engineers (VDI) and the German Standards Institute (DIN),
    Mannheim, 23-25 April]. Düsseldorf, Society of German Engineers Press,
    pp 171-209 (in German).

    Widdows J, Donkin P, & Evans SV (1985) Recovery of  Mytilus edulis L.
    from chronic oil exposure. Mar Environ Res, 17: 250-253.

    Wiester MJ, Iltis R, & Moore W (1980) Altered function and histology
    in guinea pigs after inhalation of diesel exhaust. Environ Res,
    22: 285-297.

    Willems MI, de Raat WK, Wesstra JA, Bakker GL, Dubois G, & van Dokkum
    W (1989) Urinary and faecal mutagenicity in car mechanics exposed to
    diesel exhaust and in unexposed office workers. Mutat Res,
    222: 375-391.

    Williams RR Stegens NL & Goldsmith JR (1977) Associations of cancer
    site and type with occupation and industry from the Third National
    Cancer Survey Interview. J Natl Cancer Inst, 59: 1147-1185.

    Williams PT, Andrews GE, & Bartle KD (1987) The role of lubricating
    oil in diesel particulate and particulate PAH emissions (SAE Technical
    Paper No. 872084). Warrendale, PA, Society of Automotive Engineers,
    pp 1-9.

    Williams DJ, Milne JW, Quigley SM, Roberts DB, & Kimberlee MC (1989)
    Particulate emissions from 'in-use' motor vehicles. II. Diesel
    vehicles. Atmos Environ, 23: 2647-2661.

    Winters K, Van Baalen C, & Nicol JAC (1977) Water soluble extractives
    from petroleum oils: Chemical characterization and effects on
    microalgae and marine animals. In: McIntyre AD & Whittle KJ, eds,
    Petroleum hydrocarbons in the marine environment (Rapports et
    procés-verbaux des réunions, Vol. 171). Copenhagen, International
    Council for Exporation of the Sea, pp 166-174.

    Witschi HP, Smith LH, Frome EL, Pequet-Goad ME, Griest WH, Ho CH, &
    Guerin MR (1987) Skin tumorigenic potential of crude and refined coal
    liquids and analogous petroleum products. Fundam Appl Toxicol,
    9: 297-303.

    Wolff RK, Griffis LC, Hobbs CH, & McClellan RO (1982) Deposition and
    retention of 0.1 µm 67Ga2 O3 aggregate aerosols in rats
    following whole body exposures. Fundam Appl Toxicol, 2: 195-200.

    Wolff RK, Henderson RF, Snipes MB, Griffith WC, Mauderly JL, Cuddihy
    RG, & McClellan RO (1987) Alterations in particle accumulation and
    clearance in lungs of rats chronically exposed to diesel exhaust.
    Fundam Appl Toxicol, 9: 154-166.

    Wolff RK, Bond JA, Sun JD, Henderson RF, Harkema JR, Griffith WC,
    Mauderly JL, & McClellan RO (1989) Effects of adsorption of
    benzo[ a]pyrene onto carbon black particles on levels of DNA adducts
    in lungs of rats exposed by inhalation. Toxicol Appl Pharmacol,
    97: 289-299.

    Wolff RK, Bond JA, Henderson RF, Harkema JR, & Mauderly JL (1990)
    Pulmonary inflammation and DNA adducts in rats inhaling diesel exhaust
    or carbon black. Inhal Toxicol, 2: 241-254.

    Wong O, Mitchell LE, Wolff RK, Mauderly JL, & Jeffrey AM (1986)
    Identification of DNA damage as a result of exposure of rats to diesel
    engine exhaust. Carcinogenesis, 7: 1595-1597.

    Woodin SA, Nyblade CF, & Chia F-S (1972) Effect of diesel oil spill on
    invertebrates. Mar Pollut Bull, 3: 139-143.

    Wormald AP (1976) Effects of a spill of marine diesel oil on the
    meiofauna of a sandy beach at Picnic Bay, Hong Kong. Environ Pollut,
    11: 117-130.

    Woskie SR, Smith TJ, Hammond SK, Schenker MB, Garshick E, & Speizer FE
    (1988a) Estimation of the diesel exhaust exposures of railroad
    workers: I. Current exposures. Am J Ind Med, 13: 381-394.

    Woskie SR, Smith TJ, Hammond SK, Schenker MB, Garshick E, & Speizer FE
    (1988b) Estimation of the diesel exhaust exposures of railroad
    workers: II. National and historical exposures. Am J Ind Med,
    13: 395-404.

    Wright ES (1986) Effects of short-term exposure to diesel exhaust on
    lung cell proliferation and phospholipid metabolism. Exp Lung Res,
    10: 39-55.

    Wynder EL, Dieck GS, Hall NEL, & Lahti H (1985) A case-control study
    of diesel exhaust exposure and bladder cancer. Environ Res,
    37: 475-489.

    Yaron B, Sutherland P, Galin T, & Acher AJ (1989) Soil pollution by
    petroleum products: II. Adsorption-desorption of 'kerosene' vapors on
    soils. J Contam Hydrol, 4: 347-358.

    Yu CP & Xu GB (1987) Predictive models for depositon of inhaled diesel
    exhaust particles in humans and laboratory species (Report No. 10).
    Cambridge, MA, Health Effects Institute, 27 pp.

    Yu CP & Yoon KJ (1990) Retention modeling of diesel exhaust particles
    in rats and humans (Report No. 40). Cambridge, MA, Health Effects
    Institute, 33 pp.

    Yu CP, Yoon KJ, & Chen YK (1991) Retention modeling of diesel exhaust
    particles in rats and humans. J Aerosol Med, 4: 79-115.

    Zaebst DD, Clapp DE, Blade LM, Marlow DA, Steenland K, Hornung RW,
    Scheutzle D, & Butler J (1991) Quantitative determination of trucking
    industry workers' exposures to diesel exhaust particles. Am Ind Hyg
    Assoc J, 52: 529-541.

    PARTIE A  COMBUSTIBLE DIESEL

    A1.  RESUME

    A1.1  Identité, propriétés physiques et chimiques et méthodes
          d'analyse

         Le combustible diesel est un mélange complexe d'alcanes normaux
    ramifiés ou cycliques (de 60 à plus de 90% de volume; longueur de la
    chaîne hydrocarbonée, généralement comprise entre C9 et C30), de
    composés aromatiques, surtout des alkyl benzènes, (5 à 40% en volume)
    et de petites quantités d'alcènes (0 à 10% en volume) obtenu lors de
    la distillation fractionnée du pétrole, à partir du distillat moyen
    qui correspond à la fraction gazole. Le combustible diesel peut
    contenir, à des concentrations de quelques parties par million, du
    benzène, du toluène, de l'éthylbenzène, du xylène et des hydrocarbures
    aromatiques polycycliques (HAP), en particulier du naphtalène et
    certains de ses dérivés méthylés. La teneur en soufre du combustible
    diesel dépend de l'origine du pétrole et du procédé de raffinage
    utilisé. Dans un certain nombre de pays, elle est soumise à
    réglementation et se situe en général entre 0,05 et 0,5% en poids. On
    utilise un certain nombre d'additifs pour modifier la viscosité, la
    conservation et la combustion, pour différencier les différents
    produits et également pour satisfaire aux spécifications commerciales.
    A la température ambiante, le combustible diesel est en général
    modérément volatile, légèrement visqueux, inflammable et se présente
    sous la forme d'un liquide brun à odeur de kérosène. L'intervalle
    d'ébullition est habituellement de 140°C à 385°C (plus de 588°C pour
    le carburant destiné aux moteurs marins); à 20°C, la densité est de
    0,87-1,0 g/cm3 et la solubilité dans l'eau de 0,2-5 mg/litre. La
    qualité et la composition du combustible diesel influent
    considérablement sur l'émission de polluants par les moteurs diesel.
    Les conditions d'allumage (exprimées au moyen de l'indice de cétane),
    la densité, la viscosité et la teneur en soufre sont des variables
    importantes. Les spécifications du combustible diesel commercial
    varient fortement d'un pays à l'autre.

         Les différents types de mazout destinés au chauffage ou de
    kérosène pour avions à réaction produits lors du raffinage du pétrole
    peuvent avoir une composition analogue à celle du combustible diesel,
    mais avec des additifs différents. Les données biologiques relatives à
    ces mélanges ont donc été prises en considération lors de l'évaluation
    toxicologique et écotoxicologique.

         En raison de la complexité du mélange, il n'existe pas de méthode
    d'analyse spécifique pour le combustible diesel et les techniques
    utilisées pour la plupart des études d'impact dur l'environnement ne
    se prêtent qu'au dosage des hydrocarbures totaux. Elles consistent
    tout d'abord à effectuer une extraction par solvant, puis à éliminer
    les hydrocarbures d'origine naturelle et enfin à procéder au dosage
    par gravimétrie, spectrophotométrie infrarouge ou chromatographie en
    phase gazeuse. Ni la méthode gravimétrique, ni la spectrophotométrie
    infrarouge ne donnent de renseignements qualitatifs ou quantitatifs
    utiles et ne peuvent donc être utilisées que pour un premier tri. La
    chromatographie en phase gazeuse couplée à des techniques de détection
    telles que l'ionisation de flamme ou la spectrométrie de masse, est la
    méthode classique d'analyse des échantillons prélevés dans
    l'environnement. Il existe de nombreuses autres méthodes pour la
    recherche et le dosage des divers hydrocarbures présents dans le
    combustible diesel.

    A1.2  Sources d'exposition humaine et environnementale

         Le combustible diesel est produit par raffinage du pétrole brut.
    Afin de satisfaire aux spécifications techniques relatives au
    rendement des moteurs, ces combustibles sont généralement mélangés;
    l'adjonction ultérieure d'additifs en améliore également l'adaptation
    à diverses utilisations particulières. Ces combustibles sont largement
    utilisés dans les transports. Les plus volatils d'entre eux, qui
    présentent une faible viscosité, sont destinés aux moteurs tournant à
    haut régime, les combustibles plus lourds étant réservés aux
    transports ferroviaires et maritimes. Une grande partie des véhicules
    lourds de transport routier sont mus par des moteurs diesel. Les cars
    à moteur diesel destinés au transport des voyageurs se répandent de
    plus en plus en Europe et au Japon (10 à 25%) alors qu'en Amérique du
    Nord, la proportion de cars assurant les transports en commun est
    d'environ 1 à 2%, et a même une légère tendance à se tasser. On
    utilise également le combustible diesel pour alimenter les chaudières
    et les moteurs fixes, qu'il s'agisse de moteurs à pistons, de turbines
    à gaz, de pompes d'oléoducs, de compresseurs, de générateurs de vapeur
    pour centrales thermiques, de brûleurs, et d'installations de
    chauffage industriel par convection ou circulation d'eau.

         La demande mondiale de combustible diesel a régulièrement
    progressé au cours des 5 dernières années. En 1985, les chiffres de
    consommation étaient les suivants: environ 170 000 kilotonnes par an
    en Amérique du Nord, environ 160 000 kilotonnes par an, gazole y
    compris, dans l'Union européenne, environ 46 000 kilotonnes par an en
    Australie, au Japon et en Nouvelle-Zélande, soit l'équivalent de
    1062 kilotonnes par jour. En 1990, la demande mondiale a été estimée à
    environ 1110 kilotonnes par jour.

         On ne dispose d'aucun renseignement sur les émissions qui se
    produisent lors de la production des combustibles diesel; toutefois,
    il semble que cette source soit d'importance secondaire, étant donné
    que le raffinage s'effectue en vase clos. Si des émissions doivent se
    produire, c'est surtout au cours du stockage et du transport. Ces
    combustibles peuvent également être répandus dans l'environnement par
    suite de déversements accidentels ou dans des stations service lorsque
    l'on fait de plein des véhicules. L'atmosphère et l'hydrosphère sont
    les compartiments du milieu les plus fortement affectés par ces
    décharges accidentelles. Il peut y avoir contamination du sol par du
    combustible diesel à la faveur d'accidents et ce type de contamination
    pose également un problème dans les gares de triage.

         Parmi les nombreuses techniques utilisables pour nettoyer les
    sols contaminés par du combustible diesel, on peut citer l'excavation,
    les méthodes biologiques et le confinement.

    A1.3  Transport, distribution et transformation dans
          l'environnement

         On ne dispose que très rares données sur la destinée du
    combustible diesel dans l'environnement mais on suppose que son mode
    de distribution et de transformation est comparable à celui des huiles
    lourdes destinées au chauffage comme le No 2 qui a été bien étudié. En
    cas de déversement dans l'eau, il se forme presque immédiatement une
    nappe de mazout. Les constituants polaires et ceux dont la masse
    moléculaire est relativement faible se dissolvent et s'éliminent de la
    nappe par lessivage; les constituants volatils s'évaporent en surface
    et il y a un début de dégradation microbienne. Par attaque chimique et
    biologique, la composition de la nappe se modifie. Ces processus
    dépendent de la température; les déversements qui se produisent en
    milieu arctique conduisent à des nappes plus durables que sous les
    climats tempérés. En milieu marin, la plupart des composés aromatiques
    de masse moléculaire relative faible se dissolvent dans la phase
    aqueuse, mais les alcanes normaux, les alcanes ramifiés, les
    cycloalcanes et les composés aromatiques restantes peuvent demeurer
    dans les sédiments pendant plus d'un an.

         Bien qu'on ne dispose d'aucune donnée sur la photo-oxydation du
    combustible diesel dans l'air et l'eau, on sait que les composants qui
    s'évaporent subissent une décomposition photochimique. Ainsi, on a
    montré que l'huile lourde No 2 subissait une photo-oxydation rapide en
    milieu aqueux dans les conditions naturelles.

         Les différents constituants du combustible diesel sont
    intrinsèquement biodégradables, mais à des degrés et à des
    vitesses variables. Les alcanes normaux ainsi que les dérivés
     n-alkylaromatiques et les molécules aromatiques simples en
    C10-C22, sont les plus facilement dégradables. Les petites

    molécules se métabolisent en général rapidement. Les  n-alcanes à
    longue chaîne sont plus lentement dégradés en raison de leur
    hydrophobicité et du fait qu'ils sont visqueux ou solides à la
    température ambiante. Les alcanes ramifiés et les cycloalcanes sont
    relativement résistants à la décomposition biologique et les
    hydrocarbures aromatiques polycycliques, franchement résistants. La
    vitesse globale de dégradation des hydrocarbures est limitée par la
    température, la teneur en eau et en oxygène, le pH, la présence de
    nutriments inorganiques et la versatilité métabolique microbienne.

         Les algues unicellulaires peuvent fixer et métaboliser les
    hydrocarbures aliphatiques et aromatiques mais on connaît mal
    l'ampleur de ce phénomène. Contrairement aux microorganismes qui
    utilisent les hydrocarbures du pétrole comme source de carbone, le
    métabolisme animal a généralement tendance à oxyder et à conjuguer les
    produits pour les transformer en substances plus solubles et donc plus
    faciles à excréter. Toutes les espèces animales étudiées sont capables
    de fixer des hydrocarbures du pétrole. On sait que les hydrocarbures
    aromatiques polycycliques, le pétrole brut et les produits pétroliers
    raffinés induisent les enzymes du cytochrome P450 et chez de
    nombreuses espèces de poissons de mer et d'eau douce, on constate
    qu'il y a accroissement du métabolisme des hydrocarbures.

         On n'a que peu de données sur la bioaccumulation du combustible
    diesel dans les conditions du laboratoire mais il est largement
    prouvé, par l'étude des nappes de mazout et d'autres études de
    laboratoire consacrées à ce genre de produits, en particulier le No 2,
    que les organismes aquatiques concentrent les hydrocarbures. Le
    coefficient de partage du combustible diesel entre le  n-octanol et
    l'eau est égal à 3,3-7,06, ce qui incite à penser que son potentiel de
    bioaccumulation est élevé; quoi qu'il en soit, de nombreux composés de
    faible masse moléculaire relative sont facilement métabolisés et la
    bioaccumulation effective des produits de masse moléculaire relative
    plus élevée est limitée par leur faible solubilité dans l'eau et les
    dimensions importantes de leur molécule. Il en résulte donc que la
    bioaccumulation peut en réalité être faible.

         On a constaté qu'après des déversements de combustible diesel, le
    poisson pouvait devenir inconsommable. On ne dispose d'aucune donnée
    sur la bioamplification du combustible diesel.

         On ne possède non plus aucune donnée expérimentale sur le
    déplacement du combustible diesel à travers le sol, encore qu'il ait
    été avancé qu'il existait une corrélation directe entre ce déplacement
    et la viscosité cinématique du produit. Le déplacement du kérosène
    dans un sol dépend de la teneur en eau et de la nature de ce sol.

    A1.4  Concentrations dans l'environnement et exposition humaine

         Les divers types de combustible diesel étant constitués de
    mélanges complexes, on n'en n'a pas mesuré la concentration dans
    l'environnement. On peut mettre en évidence la présence de leurs
    divers constituants dans presque tous les compartiments du milieu,
    même s'il n'est pas possible d'en vérifier l'origine. Lorsqu'il y a
    exposition de la population générale, elle se produit dans les
    stations service ou par suite de déversements.

         Il peut y avoir exposition professionnelle au combustible diesel
    à la faveur d'un grand nombre d'activités. Du fait de leur faible
    volatilité, ces combustibles ne devraient produire qu'une vapeur assez
    ténue à la température normale mais si l'on opère à température
    élevée, la concentration peut augmenter sensiblement.

    A1.5  Effets sur les mammifères de laboratoire et les systèmes
          d'épreuve in vitro

         Les combustibles diesel présentent une faible toxicité aiguë
    après administration par voie orale, percutanée ou respiratoire. Chez
    toutes les espèce étudiées (souris, lapin, rat, cobaye) on a obtenu
    pour la DL50 par voie orale une valeur > 5000 mg/kg de poids
    corporel. En application cutanée, on obtient une valeur de la DL50
    également > 5000 mg/kg de poids corporel chez la souris et le lapin,
    mais des valeurs > 2000 mg/kg de poids corporel ont été relevées pour
    certains types de kérosène et de distillats moyens, selon le protocole
    d'application et la limite inférieure de la dose. Chez des rats
    exposés par la voie respiratoire, on a obtenu une valeur de la CL50
    d'environ 5 mg/litre, sauf dans le cas d'un distillat moyen de
    première distillation pour lequel on a obtenu une valeur de
    1,8 mg/litre.

         Chez des lapins badigeonnés avec du combustible diesel à des
    doses allant jusqu'à 8000 µl/kg de poids corporel et par jour et chez
    des souris traitées dans les mêmes conditions avec des doses
    quotidiennes atteignant 40 000 mg/kg de poids corporel, on a observé
    une acanthose et une hyperkératose dues à une grave irritation. Les
    lapins se sont révélés plus sensibles que les souris. Chez la souris,
    l'inhalation de combustible diesel a provoqué des effets
    neurodépresseurs à des concentrations allant jusqu'à 0,2 mg/litre; en
    revanche ces effets n'ont pas été observés chez des rats exposés à des
    doses allant jusqu'à 6 mg/litre. Chez les rats, on a observé une
    réduction du poids corporel et du poids du foie.

         Après inhalation de doses de combustible diesel allant jusqu'à
    1,5 mg/litre dans des conditions de subchronicité, des souris, des
    rats et des chiens n'ont pas présenté de signes sensibles de toxicité
    cumulative. Le syndrome néphropathique spécifique observé chez les
    rats mâles est lié à l'accumulation intrinsèque d'inclusions hyalines
    dans les tubules rénaux.

         Les seuls effets de l'exposition à long terme ont été des
    ulcérations après application cutanée de combustible diesel à des
    souris (doses quotidiennes: 250 ou 500 mg/kg de poids corporel) et une
    modification importante du poids des organes après inhalation par des
    rats de ce même produit à la dose de 1 ou 5 mg/litre. Dans les deux
    études, on a constaté une réduction du poids corporel moyen.

         Les divers types de combustible diesel se révèlent légèrement à
    fortement irritants pour la peau du lapin. Ils ne produisent pas
    d'irritation oculaire mais dans le cas d'un certain nombre de
    kérosènes, on a fait état d'un léger effet irritant. Il n'y a pas de
    sensibilisation cutanée.

         Le combustible diesel et les carburéacteurs (kérosène) ne se sont
    révélés ni embryotoxiques, ni tératogènes lors de deux études
    effectuées sur des rats à qui l'on avait fait inhaler ces produits aux
    doses de 100 ou 400 ppm ainsi que dans une autre étude au cours de
    laquelle des rats avaient reçu par gavage des doses quotidiennes de
    ces produits allant jusqu'à 2000 mg/kg de poids corporel. Dans la
    dernière étude, on a observé une réduction du poids des foetus.

         Les épreuves effectuées sur Salmonella typhimurium n'ont pas
    permis de prouver de manière nette l'existence d'une activité
    mutagène. Des résultats positifs ont bien été observés chez
     S. typhimurium ainsi que sur des cellules lymphomateuses de souris
    mais leur caractère contradictoire les a fait considérer comme
    équivoques. Les tests de génotoxicité effectués sur des souris
     in vivo (induction de la formation de micronoyaux ou d'aberrations
    chromosomiques) ont également donné des résultats équivoques ou
    négatifs.

         Les combustibles diesel présentent un faible pouvoir cancérogène
    au niveau cutané. Dans l'état actuel de la recherche, on ne peut
    déterminer si l'activité cancérogène de ces produits est due à leur
    génotoxicité ou aux lésions chroniques qu'ils induisent dans le derme.

    A1.6  Effets sur l'homme

         Il peut y avoir exposition non professionnelle aux combustibles
    diesel lors du remplissage manuel de citernes. C'est principalement
    lors de déversements accidentels que la peau peut se trouver fortement
    exposée à ces produits, encore que pour une brève durée.

         A la suite de contacts directs avec le peau on a observé de
    l'anurie, une insuffisance rénale, des symptômes gastro-intestinaux
    ainsi qu'une hyperkératose cutanée. Des pneumopathies d'aspiration
    d'origine toxique ont été observées par suite de l'ingestion
    accidentelle de combustible diesel. Après inhalation, on peut observer

    une toux grasse persistante. Lors d'une étude cas-témoins portant sur
    des hommes exposés à du combustible diesel, on a constaté qu'ils
    courraient un risque accru de cancers du poumon, autres que les
    adénocarcinomes; on a également observé une association positive avec
    le cancer de la prostate, encore que le risque ait été plus important
    dans le groupe soumis à une exposition 'non substantielle' que dans
    celui qui était soumis à une exposition 'substantielle'.

         Lors d'une étude transversale portant sur des ouvriers d'une
    usine exposés à des carburéacteurs, on a constaté qu'ils étaient plus
    fréquemment sujets à des étourdissements, des maux de tête, des
    nausées, des palpitations, une sensation d'oppression thoracique et
    une irritation oculaire que les témoins non exposés. La concentration
    moyenne pondérée par rapport au temps de la vapeur dégagée par le
    combustible au niveau de la zone de respiration avait été estimée à
    128-423 mg/m3.

    A1.7  Effets sur les autres êtres vivants au laboratoire et dans
          leur milieu naturel

         Le combustible diesel est plus toxique que le pétrole brut pour
    les animaux et les plantes aquatiques. L'écotoxicité du combustible
    diesel est généralement attribuée à la présence de composés
    aromatiques solubles, mais les hydrocarbures aliphatiques insolubles
    peuvent également jouer un rôle. Parmi les composés aromatiques, les
    dérivés monocycliques sont les moins toxiques, leur toxicité aiguë
    augmentant avec la masse moléculaire jusqu'aux composés tétra- ou
    pentacycliques, encore que ces derniers soient peu solubles dans l'eau
    de mer. Certains animaux, comme les poissons et les oiseaux, peuvent
    avoir le corps enduit de produit, avec des effets toxiques parfois
    mortels.

         On a étudié en laboratoire le combustible diesel et notamment ses
    fractions solubles dans l'eau, les dispersions huile-eau et le pétrole
    microencapsulé. Il apparaît que le combustible diesel ne réduit pas
    sensiblement la croissance des cultures d'algues vertes du genre
     Euglena gracilis, mais à faible concentration (0,1%), il inhibe
    presque complètement la croissance de  Scenedesmus quadricauda. Le
    diesel léger (0,05%) stimule la croissance, la photosynthèse et la
    synthèse de la chlorophylle a chez  Chlorella salina, mais il en
    inhibe légèrement la respiration. A concentration plus élevée, le taux
    de croissance et la photosynthèse sont fortement réduits. Une
    exposition de longue durée inhibe la croissance des algues benthiques
    comme  Ascophyllum nodosum et  Laminaria digitata. Chez les algues
    bleues, la photosynthèse est réduite par les fractions aromatiques et
    asphaltiques, mais pas par la fraction aliphatique.

         Le combustible diesel est fortement toxique pour les daphnies,
    pour les larves de chironomidés et pour le mollusque  Viviparus
     bengalensis (un gastéropode). A la concentration de 0,1 ml/litre, il
    a provoqué la mort de copépodes du genre  Tigriopus californicus,
    en l'espace de cinq jours.

         Les moules du genre  Mytilus edulis, accumulent le combustible
    diesel, avec pour conséquence une réduction marquée de leur rythme
    d'alimentation et de croissance; leur reproduction souffre également
    d'une exposition de longue durée à ce produit. La CE50 relative au
    frai des moules exposées pendant 30 jours à du combustible diesel, a
    été trouvée égale à 800 µg/litre. La CL50 de gazole microencapsulé a
    été égale à 5000 µg/litre pour des moules en maturation exposées
    pendant 30 jours. Le gazole s'est révélé plus toxique pour les larves
    que pour les jeunes moules: il avait un effet nocif sur la croissance
    des larves à la dose de 10 µg/litre.

         Des crabes d'eau douce  (Barytelphusa cunicularis) exposés à des
    concentrations sublétales de combustible diesel pendant des durées
    allant jusqu'à 96 heures, ont généralement réduit leur consommation
    d'oxygène, en particulier aux concentrations les plus faibles et pour
    des durées d'exposition allant jusqu'à 8 heures. Lorsque l'exposition
    se prolongeait, la consommation d'oxygène était égale ou supérieure à
    celle des animaux témoins.

         Lors d'épreuves de 96 heures visant à évaluer la toxicité aiguë
    du produit sur des alevins de salmonidés dans des conditions
    statiques, on a constaté que le combustible diesel était plus toxique
    pour  Onchorhychus gorbuscha (CL50: 32-123 mg/litre), que pour
     O. kisutch (CL50: 2186-3017mg/litre) ou pour la truite arc-en-ciel
     O. mykiss (CL50: 3333-33 216 mg/litre) quel que soit le type
    d'eau.

         Le seuil de détection des réactions comportementales de la morue
    ( Gadus marhua L.), exposée à du combustible diesel dans de l'eau de
    mer, a été trouvé égal à 100-400 ng/litre. Un poisson de
    l'antarctique,  Pagothenia borchgrevinki, a résisté pendant des
    durées allant jusqu'à 72 heures à la fraction hydrosoluble non diluée
    du combustible diesel, tout en présentant cependant des signes de
    stress.

         Dans le cas des oiseaux, il peut y avoir contamination externe ou
    interne par les produits pétroliers. Le combustible diesel supprime
    l'hydrophobicité du plumage et peut être ingéré lorsque l'oiseau lisse
    ses plumes. Du combustible diesel et du mazout administrés par gavage
    à des canards à la dose de 2 ml/kg de poids corporel, a provoqué au
    bout de 24 heures une stéatose, une inflammation extrême des poumons,
    une infiltration graisseuse du foie et une dégénérescence hépatique.

    L'administration de combustible diesel ou de mazout à la dose de
    1 ml/kg a également entraîné une grave irritation des voies digestives
    et une néphrose toxique. A doses plus élevées, on a observé une
    hypertrophie des surrénales (due principalement à une hyperplasie du
    tissu cortical), une chute du taux de la cholinestérase plasmatique,
    de l'ataxie et des tremblements. Jusqu'à 20 ml/kg de poids corporel,
    la contamination n'a pas été mortelle pour les oiseaux en bonne santé
    mais la DL50 s'abaissait à 3-4 ml/kg de poids corporel lorsque le
    combustible diesel ou le mazout était administré à des oiseaux
    stressés.

         En cas de déversement de combustible diesel, le zooplancton se
    révèle extrêmement vulnérable aux constituants dispersés ou dissous du
    pétrole, mais moins aux nappes d'huile flottante. Pour les organismes
    aquatiques, la nocivité de ces produits peut se manifester de diverses
    manières: mortalité directe (oeufs de poisson, copépodes et plancton),
    contamination externe (chorions des oeufs de poisson ou cuticules et
    appendices buccaux des crustacés), contamination tissulaire par des
    constituants aromatiques, développement anormal des embryons de
    poisson et perturbation du métabolisme.

    A1.8  Evaluation des risques pour la santé humaine

         Il peut y avoir exposition de la population générale au
    combustible diesel et à d'autres distillats moyens sur les lieux et
    dans les circonstances suivants: dans les stations service, lors de
    déversements accidentels, lors de la manipulation de ces combustibles
    et lors de l'utilisation de pétrole lampant pour la cuisine ou le
    chauffage. Il peut y avoir exposition des travailleurs à ces produits
    lors de la manipulation et du transvasement du combustible dans les
    terminaux, les citernes et les stations services; lors de la
    fabrication, de la réparation, de l'entretien et de l'essai des
    moteurs diesel et autres matériels; lors de l'utilisation de
    combustible diesel pour le nettoyage ou comme solvant ou encore lors
    de la manipulation des prélèvements de routine au laboratoire. En
    raison de sa faible volatilité, le produit ne devrait pas émettre de
    vapeurs très denses à la température ambiante, encore que dans un
    espace confiné et à température élevée, la vapeur puisse être plus
    concentrée.

         Lors de la manipulation normale du combustible diesel,
    l'exposition aux vapeurs est minime. L'effet le plus probable sur la
    santé humaine consiste en une dermatite de contact. Le combustible
    diesel est irritant pour la peau mais il ne semble pas irriter la
    muqueuse oculaire. Des effets toxiques aigus au niveau rénal peuvent
    s'observer après exposition cutanée, mais on ignore quels peuvent être
    les effets à long terme d'une absorption percutanée de faibles
    concentrations.

         L'ingestion de combustible diesel peut avoir des effets toxiques
    qui se traduisent quelquefois par une régurgitation et une aspiration
    pouvant provoquer une pneumonie chimique, comme dans le cas de tout
    hydrocarbure entre certaines limites de viscosité.

         Chez des rongeurs exposés par la voie respiratoire à du
    combustible diesel à des concentrations allant jusqu'à 0,2 mg/litre,
    on a observé un effet neurodépresseur dans le cas de souris, mais pas
    chez des rats, mêmes aux concentrations les plus élevées. Chez des
    rats mâles, une exposition subchronique par la voie respiratoire à
    divers distillats a produit une néphropathie spécifique à
    alpha2-microglobulines; cette observation n'est pas considérée comme
    transposable à l'homme.

         Les combustibles diesel ne se sont révélés ni embryotoxiques ni
    tératogènes chez les animaux exposés par la voie orale ou par la voie
    respiratoire.

         Rien n'indique de façon nette qu'il existe une activité mutagène
    chez les bactéries, mais les résultats d'autres tests de génotoxicité
     in vitro et  in vivo sont plutôt équivoques.

         Un étude cas-témoins portant sur des ouvriers exposés à du
    combustible diesel incitent à penser qu'il existe un risque accru de
    cancers du poumon n'appartenant pas au type épithélium glandulaire
    ainsi que de ce cancer de la prostate. Il n'a pas été possible
    d'établir de relation dose-réponse pour l'un ou l'autre de ces cas. En
    raison du petit nombre d'études disponibles, du petit nombre de cas et
    par voie de conséquence, de l'étendue de l'intervalle de confiance, il
    n'est pas possible de tirer la moindre conclusion de ces données au
    sujet de la cancérogénicité du produit pour l'homme. Chez la souris,
    du combustible diesel administré par voie intradermique a présenté une
    faible activité cancérogène. En raison de l'absence d'une activité
    génotoxique avérée, il est possible d'invoquer un mécanisme non
    génotoxique pour ces cancers; il pourrait s'agir par exemple d'une
    irritation dermique chronique caractérisée par la répétition de
    lésions cutanées entraînant une hyperplasie de l'épiderme.

    A1.9  Evaluation des effets sur l'environnement

         Il peut y avoir pollution de l'environnement par libération
    accidentelle à grande échelle de combustible diesel, comme cela peut
    se produire lors catastrophes affectant des réservoirs ou encore en
    cas de fuites d'oléoduc ou, à plus petite échelle, lorsqu'il y a
    contamination du sol aux alentours d'une usine ou d'un garage. Dans
    l'eau, le combustible diesel s'étale presque immédiatement, les
    composants polaires de masse moléculaire relative peu élevée se
    dissolvent et disparaissent pas lessivage, les constituants volatils
    s'évaporent et la décomposition microbienne s'amorece. L'élimination

    des différents constituants dépend de la température et des conditions
    climatiques. La composition chimique de la nappe varie dans le temps:
    une fois répandues dans l'eau, certaines fractions s'évaporent et
    subissent une décomposition photochimique; les fractions lourdes,
    transportées par les particules qui se déposent, parviennent jusqu'aux
    sédiments du fond; dans le sol les constituants du combustible diesel
    migrent à des vitesses variables, en fonction de la nature de ce sol.

         Les divers constituants du combustible diesel sont
    intrinsèquement biodégradables mais leur vitesse de biodécomposition
    dépend largement des conditions physiques et climatiques et de la
    composition microbiologique du milieu.

         Des organismes aquatiques et en particulier les mollusques,
    accumulent les hydrocarbures à des degrés divers mais ceux-ci peuvent
    être éliminés par passage dans l'eau claire. Le combustible diesel
    peut subir une bioaccumulation; en revanche on ne dispose d'aucune
    donnée sur une bioamplification éventuelle.

         Tout déversement de combustible diesel a un effet nocif immédiat
    sur l'environnement, qui se traduit par une mortalité notable pour la
    faune et la flore. Il peut y avoir recolonisation au bout d'une année,
    selon l'espèce animale ou végétale en cause et selon la composition
    chimique et physique des résidus de produits pétroliers.

         Les organismes aquatiques qui survivent aux déversements de
    combustible diesel peuvent cependant subir une contamination externe
    par ces produits et en accumuler dans leurs tissus: le stress qui en
    résulte se traduit par un développement anormal et une modification du
    métabolisme

    PARTIE B  GAZ D'ECHAPPEMENT DES MOTEURS DIESEL

    B1.  RESUME

    B1.1  Identité, propriétés physiques et chimiques et méthodes
          d'analyse

         Les gaz d'échappement des moteurs diesel contiennent des
    centaines de composés chimiques qui sont émis en partie dans la phase
    gazeuse proprement dite et en partie dans la phase particulaire. Les
    principaux produits gazeux de la combustion sont le dioxyde de
    carbone, l'oxygène, l'azote et la vapeur d'eau: on trouve également du
    monoxyde de carbone dixoyde de soufre, des oxydes d'azote et des
    hydrocarbures et leurs dérivés. La phase gazeuse de la fraction
    hydrocarbonée contient également un faible pourcentage en poids de
    benzène et de toluène. Les autres constituants des gaz d'échappement
    sont des hydrocarbures aromatiques polycycliques (HAP) de faible masse
    moléculaire relative.

         L'échappement des moteurs diesel est principalement caractérisé
    par l'émission de particules dans une proportion environ 20 fois
    supérieure à celle des moteurs à essence. Ces particules sont
    composées de carbone élémentaire, de dérivés organiques adsorbés
    provenant du combustible et de l'huile lubrifiante, de sulfates formés
    à partir du soufre contenu dans le combustible et de dérivés
    métalliques à l'état de traces. La granulométrie globale de ces
    particules les situe en dessous du micron, entre 0,02 et 0,5 µm. Le
    vieillissement peut conduire à une agglomération qui les amène à un
    diamètre maximal de 30 µm. Les particules émises ont une aire
    superficielle importante. Les composés organiques constituent
    généralement 10 à 30% des particules totales mais avec des moteurs mal
    conçus et mal entretenus cette proportion peut atteindre 90%. Dans
    cette fraction on trouve également des hydrocarbures aromatiques
    polycycliques de masse moléculaire élevée, entre autres, sous forme
    oxygénée et nitrée, à des concentrations de l'ordre de plusieurs
    parties par million.

         Pour mesurer les émissions des véhicules on travaille en régime
    transitoire ou stationnaire. L'échantillonnage peut se faire à partir
    de gaz d'échappement dilués ou non dilués. Il est difficile d'obtenir
    des échantillons débarrassés de tout artéfact car les constituants
    peuvent subir des réactions chimiques, être adsorbés ou désorbés, ou
    encore soumis à une condensation ou à une diffusion. Les hydrocarbures
    aromatiques polycycliques d'importance toxicologique qui sont présents
    dans les particules émises par les moteurs diesel sont généralement
    dosées selon la procédure suivante: extraction au Soxhlet,
    purification et fractionnement puis analyse finale par chromatographie
    liquide à haute performance ou chromatographie en phase gazeuse
    couplée à la spectométrie de masse.

    B1.2  Sources d'exposition humaine et environnementale

         Les gaz d'échappement des moteurs diesel proviennent
    essentiellement des véhicules à moteur; parmi les autres source on
    peut citer les installations fixes, le matériel de traction
    ferroviaire et les machines de navires. Les émissions de ces véhicules
    à moteurs diesel ont été bien décrites mais les résultats obtenus pour
    chaque catégorie particulière ne sont souvent pas comparables en
    raison de différences concernant certains paramètres comme le cycle de
    fonctionnement, le type de moteur et la composition du combustible.
    Les divers constituants sont émis dans les proportions suivantes:
    dioxyde de carbone, environ 1 kg/km; monoxyde de carbone, oxydes
    d'azote, hydrocarbures gazeux totaux et particules 0,1-20 g/km;
    dérivés aliphatiques, alcools, aldédhydes, hydrocarbures aromatiques
    légers et hydrocarbures aromatiques polycycliques, quelques
    microgrammes/km. Dans un certain nombre de pays, il y a réglementation
    des émissions de monoxyde de carbone, d'oxyde d'azote, d'hydrocarbures
    gazeux totaux et de particules.

         En principe, il n'y a aucune différence entre les moteurs de
    faible ou de forte puissance pour ce qui est de la nature et de la
    quantité des émissions, encore que les véhicules lourds aient tendance
    à dégager une quantité relativement plus importante de particules. Les
    émissions dépendent du cycle de fonctionnement du moteur (régime
    transitoire ou régime stationnaire), du type et de l'état du moteur
    (injection ou aspiration, entretien, kilométrage total) et de la
    composition du combustible (teneur en soufre, teneur en composés
    aromatiques, volatilité); le réglage du moteur jour un rôle très
    important.

         Le dégagement de particules augmente en raison inverse du rapport
    air/combustible, mais en raison directe de la charge et de la
    température. Les moteurs vieillissants, soumis à une utilisation
    intense, libèrent plus de particules que les moteurs récents, qui ont
    peu de kilométrage, probablement du fait de la plus grande
    consommation d'huile lubrifiante. Dans le cas des véhicules légers à
    moteur diesel, il y a également une corrélation entre l'émission de
    particules et la teneur en soufre du combustible, car la formation de
    sulfates métalliques accroît la masse des particules. Dans le cas des
    véhicules lourds, ce type de corrélation n'a pas encore été établi.
    Entre outre, plus la teneur du combustible en dérivés aromatiques est
    élevée, plus le moteur dégage de particules.

         Les hydrocarbures aromatiques polycycliques, oxygénés ou non,
    émis par les moteurs diesel et les moteurs à allumage commandé, sont
    de nature similaire. Des hydrocarbures aromatiques polycycliques
    oxygénés ou nitrés sont émis à raison de quelques microgrammes/km,
    mais la concentration effective de ces composés n'est pas connue avec
    certitude car il peut y avoir décomposition ou formation lors de
    l'échantillonnage. Les émissions de HAP augmentent avec la charge et

    la température ainsi qu'avec l'âge du moteur, probablement du fait
    d'une consommation plus importante d'huile lubrifiante. Les émissions
    d'HAP dépendent également de la technique d'injection utilisée dans le
    moteur: ces émissions sont proportionnelles au rapport air/combustible
    dans les moteurs à injection directe mais la situation est inversée
    dans le cas des moteurs à injection indirecte. La teneur en dérivés
    aromatiques et la volatilité du combustible sont directement liées aux
    émissions de HAP. En cas de mauvais fonctionnement de certains
    éléments du moteur, en particulier des injecteurs, il y a augmentation
    de l'émission des principaux constituants des gaz d'échappement. On ne
    possède guère de données sur la contribution des moteurs diesel aux
    émissions totales de produits de combustion d'origine artificielle.

         Les émissions des moteurs diesel peuvent être réduites moyennant
    une meilleure conception du moteur, la pose de pièges à particules
    (pièges oxydants) et de convertisseurs catalytiques. Alors que les
    pièges à particules éliminent la suie et les composés organiques
    solubles adsorbés sur les particules, les convertisseurs catalytiques
    réduisent principalement la teneur en oxyde de carbone et
    hydrocarbures gazeux. En pratique, il est difficile de régénérer les
    pièges à particules. Les convertisseurs catalytiques nécessitent
    l'utilisation de combustibles à faible teneur en soufre car le soufre
    empoisonne les centres actifs du catalyseur.

    B1.3  Transport, distribution et transformation dans l'environnement

         C'est principalement l'atmosphère qui est affectée par les
    émissions de moteurs diesel. L'hydrosphère et la géosphère peuvent
    être contaminées indirectement par des dépôts secs ou humides. La
    destinée environnementale des divers constituants des gaz
    d'échappement des moteurs diesel est en général bien connue: les
    particules se comportent comme les molécules de gaz non réactives pour
    ce qui est de leur transport mécanique dans l'atmosphère; elles
    peuvent être transportées sur de longues distances et même pénétrer
    dans la stratosphère. On pense que leur vitesse d'élimination est
    faible, ce qui fait qu'elles séjournent plusieurs jours dans
    l'atmosphère. Avec le temps, ces particules peuvent s'agglomérer, et
    leur vitesse de chute augmentant, la quantité restante aéroportée
    diminue. Le carbone élémentaire présent dans des particules émises par
    les moteurs diesel peut catalyser la formation d'acide sulfurique par
    oxydation du dioxyde de soufre. Les constituants organiques adsorbés à
    la surface des particules de carbone élémentaire peuvent subir un
    certain nombre de transformations et de rections chimiques avec
    d'autres composés atmosphériques ou par exposition à la lumière
    solaire.

    B1.4  Concentrations dans l'environnement et exposition humaine

         Etant donné que les gaz d'échappement des moteurs diesel sont des
    mélanges complexes contenant des composés très divers, on ne peut pas
    définir de 'concentration dans l'environnement'. On devrait pouvoir
    mettre en évidence la présence de ces divers constituants dans tous
    les compartiments de l'environnement, encore qu'il ne soit pas
    généralement possible d'en déterminer l'origine. On connaît la
    concentration dans l'environnement de la plupart de ces constituants.

         C'est le plus probablement dans les rues très passantes ou dans
    les parkings, en particulier souterrains, que la population générale a
    le plus de chances d'être exposée aux émissions des moteurs diesel. Il
    est difficile d'identifier les sources de pollution et on calcule
    généralement la contribution des gaz d'échappement des moteurs diesel
    à la pollution totale due à la circulation des véhicules à moteur, sur
    la base des facteurs d'émission et du pourcentage de véhicules à
    moteur diesel dans le parc automobile général. Le niveau d'exposition
    de la population générale et des travailleurs aux particules émises
    par les moteurs diesel est toxicologiquement significatif.

         Les concentrations quotidiennes moyennes ambiantes de particules
    à proximité des voies de communication sont de 8-42 µg/m3. On a
    calculé que les concentrations annuelles moyennes de particules
    étaient de 5 à 10 µg/m3 dans les zones urbaines et < 1,5 µg/m3
    dans les zones rurales d'Allemagne et de 1 à 2 µg/m3 dans les zones
    urbaines et 0,6-1 µg/m3 dans les zones rurales des Etats-Unis
    d'Amérique. Ces concentrations sont directement corrélées à la densité
    de la circulation automobile et diminuent à mesure qu'on s'éloigne des
    routes.

         Il peut être également difficile de déterminer l'origine de la
    pollution sur les lieux de travail, en particulier dans les mines où
    la quantité totale de poussière est élevée. Le dosage du carbone
    particulaire a permis de déterminer l'exposition spécifique aux
    émissions de moteurs diesel sur les lieux de travail et il en ressort
    que les travailleurs seraient exposés à des concentrations de
    particules de l'ordre de 0,04 à 0,134 mg/m3 pour les camionneurs et
    de 0,004 à 0,192 mg/m3 pour les cheminots. On a également eu recours
    au dosage des particules respirables totales et des particules totales
    en suspension pour évaluer l'exposition professionnelle.

    B1.5  Cinétique et métabolisme chez les animaux de laboratoire et
          l'homme

    B1.5.1  Dépôt

         Les particules émises par les moteurs diesel, dont le diamètre
    aérodynamique médian massique est égal à environ 0,2 µm, sont quelque
    peu filtrés par le nez et leur aptitude au dépôt dans les poumons
    n'est que légèrement supérieure à la valeur minimale trouvée pour les

    particules de diamètre aérodynamique médian massique égal à environ
    0,5 µm. Autrement dit, lorsque les particules de suie émises par les
    moteurs diesel sont inhalées, 10 à 15% d'entre elles se déposent au
    niveau des alvéoles chez le rat et le cobaye, le dépôt au niveau des
    alvéoles étant d'environ 10% chez l'homme.

    B1.5.2  Rétention et élimination des particules

         L'élimination des particules par l'ascenseur muco-ciliaire est
    pratiquement complète au bout de 24 heures. L'élimination à long terme
    des particules déposées dans la zone alvéolaire a fait l'objet de
    plusieurs études sur des rats qui avaient été exposés par la voie
    respiratoire à deux types de particules: les particules émises par un
    moteur diesel et les particules servant de référence. Dans le cas des
    témoins chez lesquels la charge pulmonaire était faible (< 1 mg par
    poumon), on a obtenu des temps de demi-élimination de l'ordre de 60 à
    100 jours, alors que chez les rats présentant une charge pulmonaire
    allant de 1 à 60 mg/poumon, le temps de demi-élimination était de
    l'ordre de 100 à 600 jours. Dans plusieurs études, on a constaté que
    les effets observés étaient dus à une surcharge en particules,
    laquelle a été décrite chez diverses espèces et pour un certain nombre
    de matériaux particulaires. Ce phénomène s'observe généralement
    lorsque la vitesse de dépôt des particules de faible solubilité et de
    faible toxicité aiguë reste supérieure à leur vitesse d'élimination
    pendant une très longue durée. Chez l'homme, le temps de
    demi-élimination alvéolaire normal est de plusieurs centaines de
    jours, c'est-à-dire plus long que chez le rat.

         On a mis au point un modèle pulmonaire dosimétrique sur la base
    des données de dépôt et de rétention des particules diesel chez le rat
    après inhalation de longue durée et des mêmes données chez l'homme. Ce
    modèle peut être utilisé pour prévoir la rétention des particules
    diesel et des composés organiques adsorbés dans les poumons de
    personnes d'âges divers. On ne dispose d'aucune donnée sur la manière
    dont la rétention de différents composés évolue après une exposition
    prolongée aux gaz d'échappement de moteurs diesel.

    B1.5.3  Rétention et élimination des hydrocarbures aromatiques
            polycycliques adsorbés sur la suie de moteurs diesel

         Les hydrocarbures aromatiques polycycliques présents dans la suie
    émise par les moteurs diesel adhèrent fortement à la surface des
    particules. Environ 50% des HAP adsorbés sur les particules sont
    éliminés par les poumons dans l'espace d'une journée, mais le temps de
    demi-rétention de la fraction restante s'est révélé égal à 18-36
    jours. Des études portant sur du 3H-benzo[ a]pyrène et du
    14C-nitropyrène ont montré que lorsqu'ils sont associés à des
    particules, les HAP sont sensiblement plus longs à être éliminés des
    poumons que leurs homologues libres.

    B1.5.4  Métabolisme

         Déposé sur des particules émises par un moteur diesel, du
    benzo[ a]pyrène a subi une métabolisation oxydative au niveau du
    poumon et dans des cultures cellulaires de macrophages pulmonaires
    aboutissant à la formation de phénols, de diols et de quinones
    substituant le noyau benzo[ a]pyrène. Du nitropyrène adsorbé sur des
    particules diesel a été métabolisé en acétylaminopyrène-phénol après
    inhalation. Chez des rats exposés à des émissions de moteur diesel on
    a constaté, dans les poumons et les cellules de type II, un
    accroissement des adduits de l'ADN par rapport aux témoins. Le
    métabolisme oxydatif de certains composés organiques qui conduit à des
    époxydes pourrait être responsable de la formation des adduits, mais
    il est vrai qu'on ne les trouve qu'après exposition à des particules.
    Certains dérivés organiques présents dans les émissions de moteur
    diesel se sont révélés capables de former des adduits avec l'ADN;
    toutefois le noyau carboné lui-même (sans composés organiques
    extractibles) peut également conduire à la formation d'adduits, en
    provoquant des lésions chroniques au niveau des cellules épithéliales.

    B1.6  Effets sur les mammifères de laboratoire et les systèmes
          d'épreuve in vitro

         Les quelques données disponibles incitent à penser que les
    émissions de moteur diesel ne présentent qu'une faible toxicité aiguë.
    Des souris auxquelles on avait administré par voie intratrachéale, des
    particules émises par un moteur diesel, sont mortes des suites d'une
    oedème du poumon. La DL50 était égale à environ 20 mg/kg de poids
    corporel. Des particules du même type, extraites au méthanol, n'ont
    entraîné aucune mortalité jusqu'à une concentration d'environ 33 mg/kg
    de poids corporel. Chez le hamster, on a obtenu une DL50 de
    1280 mg/kg de poids corporel après administration par voie
    intrapéritonéale. On ne dispose d'aucune donnée concernant
    l'exposition par la voie respiratoire.

         Après exposition de rats, de cobayes et de chats pendant environ
    quatre semaines à des émissions de moteur diesel ayant une teneur en
    particules de 6 mg/m3, on a observé une altération de la fonction
    pulmonaire, et notamment une augmentation de 35% de la résistance
    ventilatoire chez les cobayes et une diminution de 10% de la capacité
    vitale chez les chats. Du point de vue histopathologique, on a observé
    un épaississement focal des parois alvéolaires, un accroissement
    significatif de l'indice de marquage des cellules de type II et des
    accumulations de macrophages chargés de particules. Ces accumulations
    se situaient au niveau des bronchioles terminales et elles ont
    augmenté de volume lors de la période de récupération postérieure à
    l'expérience en raison de la fixation des macrophages (séquestration).

         Après avoir inhalé pendant de longues périodes des émissions de
    moteur diesel à des concentrations atteignant 4 mg/m3, des rats, des
    souris, des hamsters, des chats et des singes n'ont pas présenté de
    chute spectaculaire de leur poids corporel ni de réduction de leur
    longévité. Les effets toxiques liés à la dose qui ont été observés
    chez toutes les espèces après une longue période d'inhalation étaient
    les suivants: augmentation du poids des poumons pouvant aller jusqu'à
    400%; inflammation pulmonaire objectivée par des paramètres
    biochimiques (marqueurs enzymatiques cytoplasmiques, collagène) et
    cytologiques (augmentation des neutrophiles polynucléaires);
    perturbation de la mécanique pulmonaire; accroissement du nombre de
    macrophages chargés de particules avec accumulations focales
    (séquestration) dans des conditions de surcharge; enfin, altérations
    prolifératives des cellules épithéliales et début de fibrose.

         D'après les données limitées dont on dispose sur la toxicité des
    émissions de moteurs diesel pour la fonction de reproduction et le
    développement, on peut penser qu'il n'y a pas d'effets toxiques
    déterminants. Dans la plupart des expériences, on n'a pas constaté
    d'effets chez des souris, des rats, des hamsters, des lapins ou des
    singes; toutefois, après injection par voie intrapéritonéale, on a
    observé chez des souris qui avaient reçu de particules émises pas des
    moteurs diesel, des anomalies touchant les spermatozoïdes et constaté
    une certaine embryotoxicité chez des hamsters auxquels on avait
    administré des extraits d'émissions de ces mêmes moteurs.

         La plupart des épreuves de génotoxicité  in vitro ont été
    réalisées avec des extraits d'émissions de moteurs diesel plutôt
    qu'avec les émissions totales elles-mêmes et l'on a obtenu des
    réactions positives en l'absence d'activation métabolique. Autrement
    dit, il semble que ces effets génotoxiques soient indépendants de la
    présence d'HAP. Environ 50% des études  in vivo ont donné des
    résultats négatifs; les seuls résultats positifs obtenus consistaient
    en échanges de chromatides soeurs dans le cas des émissions totales et
    des extraits organiques, et dans la présence de micronoyaux dans le
    cas des extraits organiques.

         On n'a généralement pas observé d'effets immunotoxiques après
    inhalation d'émissions de moteurs diesel; toutefois on a observé dans
    une étude une augmentation du titre des immunoglobulines
    anti-ovalbumine et deux autres expériences ont mis en évidence un
    accroissement de la sensibilité aux infections chez des souris.

         Des études effectuées sur des rats incitent à penser que
    l'inhalation de gaz d'échappement de moteurs diesel affecte l'état
    comportemental et neurophysiologique.

         Lors d'études sur la cancérogénicité des émissions de moteur
    diesel, au cours desquelles des rats étaient exposés à ces émissions
    par la voie respiratoire, on a constaté que la phase gazeuse
    (c'est-à-dire dépourvue de particules) n'était pas cancérogène. Toutes
    les études sur des rats qui ont été validées ont montré que les
    émissions de moteurs diesel avaient un effet cancérogène lorsque la
    concentration en particules était supérieure à 2 mg/m3, soit
    l'équivalent d'une exposition continue à une concentration d'environ
    1 mg/m3. Aucun effet n'a été observé en revanche chez les hamsters
    et les souris. Lors d'études au cours desquelles on a pratiqué une
    instillation intratrachéale, on a constaté que les particules
    présentes dans les émissions de moteurs diesel et le noir de fumée
    produisaient des tumeurs; en outre on a observé l'existence d'une
    corrélation entre l'aire superficielle des particules carbonées et
    leur activité tumorigène.

         L'inhalation, pendant une longue période, de noir de fumée
    pratiquement dépourvu d'hydrocarbures aromatiques polycycliques à ces
    concentrations, a également provoqué des tumeurs pulmonaires chez le
    rat.

         On ne sait pas avec certitude si la cancérogénicité des émissions
    de moteurs diesel est due à un mécanisme faisant intervenir ou non des
    réactions au niveau de l'ADN (ou à un mécanisme mixte). Différents
    modèles ont été proposés pour élucider la cancérogénicité des
    émissions de moteurs diesel.

    B1.7  Effets sur l'homme

         Les émissions de moteurs diesel contribuent à la pollution
    globale de l'air. Les études toxicologiques, qu'elles cherchent à
    mettre en évidence des effets aigus ou des effets chroniques, ne
    permettent pas d'assigner un rôle particulier à ces particules mais il
    apparaît qu'elles pourraient être en partie responsables d'un certain
    nombre d'effets attribués à la pollution de l'air.

         Les émissions de moteur diesel ont des caractéristiques qu'un
    certain nombre de personnes trouvent agressives, en particulier à
    fortes concentrations. Les symptômes observés après exposition de
    brève ou de longue durée à ce type d'émissions ont été décrits dans un
    certain nombre d'études et de rapports concernant des personnes
    exposées de par leur profession. Les effets aigus consistent en
    irritation des muqueuses oculaires et nasales et l'on a observé dans
    les cohortes professionnelles une augmentation de la fréquence des
    symptômes respiratoires; toutefois on ignore qu'elle est la
    contribution exacte des particules émises par les moteurs diesel à ce
    type de symptômes. Aucun effet à court terme n'a été régulièrement
    observé au niveau de fonction pulmonaire, mais on a signalé des crises
    d'asthme.

         Lors d'une étude contrôlée au cours de laquelle huit volontaires
    non fumeurs, en bonne santé, ont été exposés à des émissions diluées
    de moteurs diesel, pendant 60 minutes dans une chambre fermée, on a
    constaté une réduction du taux de phagocytose chez des macrophages
    alvéolaires recueillis par lavage broncho-alvéolaire.

         Un certain nombre d'études transversales et longitudinales
    portant sur des ouvriers longtemps exposés de part leur profession à
    des émissions de moteur diesel, ont révélé une altération de la
    fonction pulmonaire et une augmentation de la prévalence des symptômes
    respiratoires, toutefois la brièveté des épisodes d'exposition limite
    la portée de ces études. Aucune surmortalité n'a été constatée lors
    d'études de cohorte au cours desquelles on s'est efforcé d'étudier des
    décès par maladie cardio-vasculaire ou accident vasculaire cérébral,
    imputables à l'exposition des émissions de moteurs diesel.

         Lors d'un certain nombre d'études épidémiologiques on a tenté
    d'établir des relations entre certains cancers du poumon et de la
    vessie et une exposition professionnelle à des émissions de moteurs
    diesel. Seules les études jugées utiles pour l'évaluation des effets
    cancérogènes de ces émissions ont été prises en considération dans la
    présente monographie. Celles qui sont les plus intéressantes pour ce
    qui est du cancer du poumon, concernent des cheminots, des personnes
    travaillant dans des garages de cars et des débardeurs qui constituent
    des cohortes dont les membres sont effectivement exposés aux émissions
    de moteur diesel. Les quatre études les plus informatives font toutes
    état d'un risque accru de cancer du poumon, avec un risque relatif
    allant de 1,4 pour les cheminots, et de 1,3-2,4 pour les ouvriers des
    garages (selon le type d'exposition), jusqu'à un risque trois à six
    fois plus élevé pour les débardeurs (en fonction du mode d'évaluation
    de l'exposition, mais avec un large intervalle de confiance). Il a été
    possible de tenir compte du tabagisme dans une étude cas-témoins sur
    des cheminots et dans une étude sur des débardeurs. Dans les deux cas,
    cette correction pour tenir compte du tabagisme n'a eu aucune
    influence sur l'effet de l'exposition aux émissions diesel. Dans les
    trois études pour lesquelles on n'a pas pu tenir compte du tabagisme,
    l'analyse était basée sur des comparaisons entre les différents
    sous-groupes de ces cohortes, de sorte que l'effet de confusion créé
    par le tabagisme avait moins de chance d'être gênant qu'en cas de
    comparaison avec des groupes extérieurs.

         Plusieurs études cas-témoins ont été menées afin d'examiner la
    relation pouvant exister entre les cancers de la vessie et une
    exposition supposée à des émissions de moteurs diesel. On a constaté
    un accroissement du risque, en particulier pour le chauffeurs de
    camion; toutefois toutes ces études souffrent d'une caractérisation
    insuffisante de l'exposition. De la sorte, il n'est pas possible
    d'affirmer qu'une exposition aux émissions de moteurs diesel entraîne
    un risque accru de cancer vésical.

    B1.8  Effets sur les autres êtres vivants au laboratoire et dans
          leur milieu naturel

         Les effets des émissions de moteur diesel n'ont été abordés que
    dans une seule étude, portant sur des algues vertes.

    B1.9  Evaluation des risques pour la santé humaine

         L'évaluation des risques de cancer et de maladies non malignes a
    été menée conformément au modèle d'évaluation des risques établi par
    l'Académie nationale des sciences des Etats-Unis (National Research
    Council, 1983). Ce modèle d'évaluation comporte quatre étapes: (1)
    identification du danger; (2) évaluation de la relation dose-réponse;
    (3) évaluation de l'exposition et (4) caractérisation du risque.

         On estime que les travaux les plus intéressants sont les études
    épidémiologiques de longue durée dans lesquelles l'exposition est bien
    définie et le suivi supérieur à 20 ans. Quatre études portant sur le
    cancer du poumon chez des personnes professionnellement exposées
    satisfont à ces critères. Il apparait que le risque relatif de cancer
    du poumon lié à une exposition à des émissions de moteur diesel est
    généralement faible et qu'il est influencé par les circonstances, les
    effets des facteurs de confusion non évalués et la difficulté de tenir
    suffisamment compte des facteurs de confusion reconnus. D'autres
    études, où l'exposition est définie avec une moindre précision,
    corroborent les conclusions des études précédentes. Globalement, on
    estime que les émissions de moteurs diesel sont probablement
    cancérogènes pour l'homme; toutefois on ne dispose d'aucune donnée
    quantitative permettant d'évaluer le risque.

    B.1.9.1  Effets non néoplasiques

         Pour caractériser le risque, on a procédé de deux manières:
    premièrement une concentration sans effets nocifs observables que l'on
    divise par un coefficient d'incertitude; deuxièmement, une
    concentration de référence. Dans les deux cas, on a utilisé un modèle
    dosimétrique élaboré qui permet de réduire l'imprécision due à
    l'extrapolation interspécifique des doses.

         La dose de particules sans effet pour l'homme a été estimée à
    0,139 mg/m3. La valeur guide pour la population générale calculée
    d'après le modèle dosimétrique a été établie à 5,6 µg/m3, la valeur
    calculée sans recours au modèle étant de 2,3 µg/m3.

         La méthode basée sur l'utilisation d'une concentration de
    référence prend en compte l'ensemble de la relation exposition-réponse
    de préférence aux données ponctuelles fournies par les études
    d'inhalation, comme c'est le cas dans la méthode de la dose sans
    effets observables. On a identifié trois paramètres biologiques
    sensibles: l'inflammation alvéolaire chronique, la diminution

    d'efficacité de l'ascenseur mucociliaire et les lésions pulmonaires
    hyperplasiques. Les concentrations de référence calculées à partir du
    même modèle dosimétrique que dans le cas de la méthode de la dose sans
    effets observables, étaient de 0,9 à 2 µg/m