EPA-600/3-78-012
January 1978
Ecological Research Series
                     THIRD  ANNUAL  CATALYST RESEARCH
                                             PROGRAM  REPORT
                                                 Effects Research Laboratory
                                            Office of Research and Development
                                           U.S. Environmental Protection Agency
                                     Research Triangle Park, North Carolina 27711

-------
                RESEARCH REPORTING SERIES

Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology.  Elimination  of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:

      1.   Environmental  Health Effects Research
      2.   Environmental  Protection Technology
      3.   Ecological Research
      4.   Environmental  Monitoring
      5.   Socioeconomic Environmental Studies
      6.   Scientific and Technical  Assessment Reports (STAR)
      7.   Interagency  Energy-Environment Research and Development
      8.   "Special"  Reports
      9.   Miscellaneous Reports

This report has been assigned to the ECOLOGICAL RESEARCH series. This series
describes research on the effects of pollution on humans, plant and animal spe-
cies, and materials.  Problems  are assessed for their long- and short-term influ-
ences. Investigations include formation, transport, and pathway studies to deter-
mine the fate of pollutants and their effects. This work provides the technical basis
for setting standards to minimize undesirable changes in living organisms in the
aquatic, terrestrial, and atmospheric environments.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.

-------
                                         EPA-600/3-78-012
                                         January  1978
THIRD ANNUAL CATALYST RESEARCH PROGRAM REPORT
                     by
     Health Effects Research Laboratory
     Research Triangle Park, N.C. 27711
    U.S. ENVIRONMENTAL PROTECTION AGENCY
     OFFICE OF RESEARCH AND DEVELOPMENT
     HEALTH EFFECTS RESEARCH LABORATORY
     RESEARCH TRIANGLE PARK, N.C. 27711

-------
                              DISCLAIMER

     This report has been reviewed by the Health Effects Research
Laboratory, U.S. Environmental Protection Agency, and approved for
publication.  Mention of trade names or commercial products does not
constitute endorsement or recommendation for use.

-------
                               FOREWORD


     The many benefits of our modern, developing, industrial society are
accompanied by certain hazards.  Careful assessment of the relative risk
of existing and new man-made environmental hazards is necessary for the
establishment of sound regulatory policy.  These requlations serve to
enhance the quality of our environment in order to promote the public
health and welfare and the productive capacity of our Nation's population.

     The Health Effects Research Laboratory, Research Triangle Park,
conducts a coordinated environmental health research program in toxicology,
epidemiology, and clinical studies using human volunteer subjects.  These
studies address problems in air pollution, non-ionizing radiation, environ-
mental carcinogenesis and the toxicology of pesticides as well as other
chemical pollutants.  The Laboratory develops and revises air quality
criteria documents on pollutants for which national ambient air quality
standards exist or are proposed, provides the data for registration of new
pesticides or proposed suspension of those already in use, conducts research
on hazardous and toxic materials, and is preparing the health basis for
non-ionizing radiation standards.  Direct support to the regulatory function
of the Agency is provided in the form of expert testimony and preparation
of affidavits as well as expert advice to the Administrator to assure the
adequacy of health care and surveillance of persons having suffered
imminent and substantial endangerment of their health.

     New technologies for controlling emissions of pollutants to the
atmosphere are always a welcome advance in the pursuit of a cleaner
environment through research.  A thorough study of these new technologies
is in order, however, to assure that the net effect on public health is
beneficial.  The Catalyst Research Program, in its investigation of the
automotive oxidation catalyst, provides a sound base upon which the EPA
can make a responsible assessment of the effect on public health of this
advanced emission control technology.
                                       John H. Knelson, M.D.
                                            Director
                                Health Effects Research Laboratory
                                   i* •
                                   11

-------
                                PREFACE
     The Catalyst Research Program, initiated in FY 1975, is a broad
multidisciplinary research effort to provide a sound technical basis
for evaluating the public health issues related both to the benefits
associated with the reduction of regulated emissions and to the
potential health hazards associated with any unregulated emissions
generated by control systems.  This report covers work performed
during the third year of the Catalyst Research Program plus related
work performed outside the CRP.
                                  IV

-------
                               ABSTRACT
     This report constitutes the third annual report by the Office of
Research and Development on the Catalyst Research Program (CRP) required
by the Administrator as noted in his testimony before the Senate Public
Works Committee on November 6, 1973.  It includes all research aspects
of the broad multidisciplinary program including:  emissions characteri-
zation, measurement method development, monitoring, fuel analysis, toxi-
cology, biology, epidemiology, human studies, and unregulated emissions.
Principal focus is on catalyst-generated sulfates and other nonregulated
emissions, including particulate emissions from light-duty diesels.

     Highlights in this report are:

     •    Results of continued tests of sulfuric acid emissions from
          catalyst-equipped vehicles are such that these emissions may
          be considered much less of an air quality problem than was
          previously hypothesized.

     •    With freeway density of catalyst-equipped cars rising to 30
          percent in 1976, roadside sulfate readings rose only 10
          percent while a 25 percent decrease was observed in carbon
          monoxide, lead, and total suspended particulates.

     t    Tests conducted to date show very high particulate emissions
          from light-duty diesel vehicles, about 25 times the values
          obtained from catalyst-equipped gasoline vehicles.

-------
                           TABLE OF CONTENTS
                                                                    Page
Foreword	  i i i
Preface	   iv
Abstract	    v

     I.   Introduction	          1

    II.   Overview of Component Programs

          A.  Environmental Monitoring and Support Laboratory/RTP..    4
          B.  Environmental Sciences Research Laboratory/RTP	    5
          C.  Health Effects Research Laboratory/Cincinnati	    7
          D.  Health Effects Research Laboratory/RTP	   11
          E.  Office of Mobile Source Air Pollution Control/
              Washington, D.C	   13

   III.   Discussion of Program Highlights

          A.  Sulfates	   17
          B.  Hydrogen Cyanide	   18
          C.  Manganese Oxide	   19
          D.  Polynuclear Aromati c Compounds	   19
          E.  Ammonia	   19
          F.  Ruthenium	   19
          6.  Platinum	   20
          H.  Diesel Emissions	   20
          I.  Health Effects of Exposure to Whole Exhaust	   20

Appendices

          A.  Environmental Monitoring and Support Laboratpry/RTP..   21
          B.  Environmental Sciences Research Laboratory/RTP	   85
          C.  Health Effects Research Laboratory/Cincinnati	  129
          D.  Health Effects Research Laboratory/RTP	  160
          E.  Office of Mobile Source Air Pollution Control
              Washington, D.C	  219
                                 VI

-------
I.   INTRODUCTION

     Section 202 of the Clean Air Act of 1970, as amended, required
substantial reduction in certain specified emission products from
automobiles.  The automotive industry, to achieve these reductions,
chose the oxidation catalytic converter as a primary method of emission
control for model year 1975.  Subsequent to this decision, EPA intensi-
fied its research program to determine what, if any, new pollutants
might be emitted into the atmosphere as a result of the application of
this technology.  The effects of fuel composition and fuel additives on
these, as well as regulated, emissions were also studied.  Results from
the EPA research program indicated that though emissions of hydrocarbons,
carbon monoxide, and certain organics would be dramatically lowered,
sulfuric acid aerosol emissions would increase, and slight emissions of
platinum, palladium, and alumina might also be expected.

     Because existing ambient concentrations of sulfates in many areas of
the country have been shown to be at levels sufficient to cause concern,
and because very little information existed regarding health effects of
platinum or palladium as air pollutants, EPA initiated a broad research
program to examine the public health impact of catalyst-emitted sulfates,
platinum, and palladium.

     In testimony to the Public Works Committee of the U.S. Senate, Adminis-
trator Train specified EPA's planned program of catalyst-related research:

     •    Accelerate work on development of a reliable test procedure for
          automotive sulfate emission measurement.

     •    Consider all feasible alternatives for automotive sulfate
          emission control.

     •    Improve the Agency's ability to estimate the public health
          impact of sulfate and other automotive emissions.

     •    Improve understanding of the atmospheric chemistry involved
          in these emissions and initiate an appropriate air monitoring
          program.

     From these broad objectives defined in Administrator Train's testimony
of November 1973, an extensive interdisciplinary research program was
developed utilizing the resources of EPA's technical  staff in the Office
of Research and Development and the Office of Air and Waste Management
as well as extramural (contracts and grants) programs with the scientific
community.  The program is now in its third year.  The results of the
first two years' work were summarized in the first and second annual
reports published as EPA reports (EPA-600/3-75-010a  through j, dated
September 1976, and EPA-600/3-77-008, dated January 1977).   This report is
the third in a series planned for annual issuance through at least 1978.

-------
     The third report, like the second, presents the contents arranged
according to the laboratory organization under whose manganement the
projects were performed.  This organization is intended to facilitate
reporting requirements and to aid in the identification of responsible
program management in order to fulfill the requests of report users for
further information.

     This year, for the first time the individual program reports which
were published as separate supplements in the first and second editions,
are included as section IV of this document.  These reports present the
work performed as part of the Catalyst research Program by each of the
following groups:  health Effects Research Laboratory, Research Triangle
Park, N.C.; Environmental Monitoring and Support Laboratory, Research
Triangle Park, N.C.; Environmental Sciences Research Laboratory, Research
Triangle Park, N.C.; Health Effects Research Laboratory, Cincinnati, Ohio;
and Office of Mobile Source Air Pollution Control, Washington, D.C.  An
overview of the  organization's work is included in each report.  In
addition, the figure entitled "Organization of the Catalyst Research Program"
presents a graphic overview of the research responsibilities of the five
organizations participating in the Catalyst Program.

     Immediately following the overviews in the summary report, "Program
Highlights" focuses on particular areas of research and is organized
according to the various pollutants studied rather than according to
laboratory.  This section provides a brief discussion of the highlights
of CRP research in 1976 on sulfates, hydrogen cyanide, manganese, poly-
nuclear aromatic compounds (PNA), ammonia, ruthenium, platinum, and on
diesel and whole catalyst emissions.

-------
                                PARTICIPATING LABORATORY
                              ENVIRONMENTAL MONITORING
                                AND SUPPORT LABORATORY
                                        RTP,NC
                                ENVIRONMENTAL SCIENCES
                                 RESEARCH LABORATORY
                                         RTP,NC
                                AREAS OF RESPONSIBILITY

                                o   FUEL COLLECTION AND ANALYSIS

                                o   MONITORING (LACS)



                                o   EMISSION CHARACTERIZATION

                                o   MEASUREMENT METHODOLOGY

                                o   ATMOSPHERIC CHEMISTRY

                                o   METEOROLOGY/MODELING
 CATALYST PROGRAM
COORDINATION OFFICE
(OFFICE OF HEALTH AND
ECOLOGICAL EFFECTS)
HEALTH EFFECTS
RESEARCH LABORATORY ,
CINCINNATI.OH

                                o   INHALATION TOXICOLOGICAL EVALUATION
                                    OF WHOLE ENGINE EXHAUST

                                o   TOXICITY TESTING OF EMISSION PRODUCTS1
                                    HEALTH EFFECTS
                                 RESEARCH LABORATORY
                                        RTP,NC
                                o   IN VIVO AND IN VITRO TOXICITY TESTING
                                    OF EMISSION PRODUCTS

                                o   INHALATION TOXICOLOGY1

                                o   HUMAN STUDIES
          |	
OFFICE OF MOBILE SOURCE
 AIR POLLUTION CONTROL
       WASH, DC
o   EMISSION FACTORS

o   CONTROL TECHNOLOGY
                                  Organization of the Catalyst Research Program

-------
II.  OVERVIEW OF COMPONENT PROGRAMS

     A.   Environmental Monitoring and Support Laboratory/Research
          Triangle Park

          The work related to the Catalyst Research Program performed by
the Environmental Monitoring and Support Laboratory, Research Triangle
Park (EMSL/RTP) in 1976 involved fuel collection and analysis, monitoring,
and quality assurance.

          1.   Fuel Collection and Analysis

               The advent of the catalyst-equipped vehicle requiring
unleaded gasoline initiated studies of gasoline as part of a National
Fuels Surveillance Network.  There is interest not only in levels of
lead and phosphorus, which are limited to 0.05 grams lead and 0.005
grams phosphorus per U.S. gallon, but also in such other factors as
sulfur, manganese, and aromatic content.  Analysis of samples of gasoline
indicated that the sulfur content of unleaded gasoline in Southern Cali-
fornia is approximately twice as high as that in New York.  Sulfur content
in all 3 grades of gasoline is consistantly higher in Los Angeles than in
San Francisco.  Percent of aromatic content is increasing and is highest
in unleaded gasoline.

          2.   Monitoring

               The EMSL is responsible for all study-related functions of
the Los Angeles Catalyst Study (LACS), initiated in June 1974, to develop
information on mobile source-related pollution in the ambient air adjacent
to the San Diego Freeway, before and after the introduction of catalytic
converters in 1975.  It is estimated that by the end of 1976 the percentage
of catalyst-equipped cars on the freeway may be as high as 30 percent, and
data are being analyzed to determine the catalytic converter's impact on
sulfate, carbon monoxide, and lead levels.  Studies indicate that the use
of the catalyst and unleaded fuel has resulted in a 25 percent decrease in
carbon monoxide, lead, and total suspended particulates, while the background
levels remain essentially consistent.  (Lead has been found to be 60 percent
on weekends despite a 20% reduction in traffic).  Freeway contributions of
sulfates have increased less than 10 percent, while background levels decreased.
Oxides of nitrogen have increased 50 percent over background levels.

          3.   Quality Assurance Activities

               To verify exposure levels, EMSL conducted audits on air
pollutants used in the animal exposure chambers at HERL/Cincinnati and on
those used in human exposure chambers at HERL/RTP.  Within EMSL, the LACS
monitoring program is audited continuously through blind samples and
sample splits, programs initiated in 1974 to determine:  (1) the accuracy

-------
achieved by the Rockwell Air Monitoring Center (RAMC); (2) the compara-
bility between analysis performed by RAMC and EMSL; and (3) analytical
variability on real air samples.  Both programs have changed to accomodate
fluctuations in the number of samples collected, variability of the analy-
tical results, and EMSL's growing experience in conducting external
audits.  Through this quality assurance program, an extensive body of
ambient roadside air quality data of known quality has been collected.

     B.   Environmental Sciences Research Laboratory/Research Triangle Park

          The research efforts of the Environmental Sciences Research
Laboratory during the 1976 calendar year have been mainly focused on 4 areas:
measurement methods, emissions characterization, in-roadway studies, and
modeling.

          1•   Measurement Methods

               Several developments grew out of the studies in measurement
methodology.  A sulfuric acid sampler/analyzer was developed, tested, and
modified.  A sulfuric acid generator to deliver particles in the 0.02 - 0.5
ym size range was developed and was coupled to an animal  exposure chamber
at HERL.  Projects were initiated to develop a flame photometric detector
for continuous in situ measurement of total particulate sulfur and certain
individual sulfur compounds and to develop methodology to selectively
measure sulfuric acid in the range of 0.25 to 50 ym/m3.  The latter project
has identified 3 reagents which can be selectively analyzed in the presence
of sulfate, demonstrated the formation of adducts, and has determined their
likely composition.  Yet another study, a gas chromatographic system, has
been created which will permit the analysis of part-per-billion concentrations
of NH3.  A portable dichotomous aerosol collection system to collect atmos-
pheric sulfuric acid in particles smaller than 3.5 ym was developed, and an
evaluation was conducted of potential interferences.  Results have indicated
the strongest potential interference is from gaseous ammonia.  However, it
was shown that by using an ammonia denuder this is no longer a problem.  In
addition, if the sampling period is kept to 1 hour or less, sulfuric acid
interaction with other particles on the filter is minimal.  Prototype
instrumentation for S02, NH3, NOX» and H2S was evaluated using a mobile van
equipped with a mapping and data acquisition system.

          2.   Emissions Characterization

               The year's work on emissions concentrates  on two issues which
have not yet been studied adequately.  The first involves a study of sulfuric
acid emission factors for 101 in-use California cars.  The study revealed that
sulfuric acid emission decreases rapidly with increasing mileage, indicating
that catalysts lose their sulfate-forming activity more readily than their
capacity for hydrocarbons and carbon monoxide.  The study has shown that HC
and CO remain low for the first year in California cars, that NOx emissions
rates for '75 and '76 model years are mostly greater than California standards,

-------
and that California fuel sulfur levels are about the same as the national
average.  These samples were collected from the test car tanks; data obtained
from samples collected at individual service stations (see page 17) indicate
that sulfur levels are higher than the national average.  Reasons for this
apparent discrepancy are unclear at this time.

               The second sought to determine H2S, COS, NHs, HCN.^O,
H2SCM, and particulate emissions from vehicles with rhodium-containing
catalysts under standard and malfunction conditions.  Results indicate that
catalyst-equipped cars operating under normal conditions yield little HCN
and that rich malfunctions increase HCN and CO emissions.

          3.   In-Roadway Study

               An in-roadway study involving scientists from the University
of Minnesota, Washington University of St. Louis, and the California
Institute of Technology, with support from Rockwell International, was conducted
in October 1976, to determine sulfate levels in the roadway and to determine
the feasibility of making in-roadway measurements with instruments installed
in moving vehicles.  The measurements showed that:  (a) the sulfuric acid
emitted by catalyst-equipped cars is in the nuclei mode (0.1 ym) particle
size range; (b) the emitted sulfuric acid is dispersed and may be neutralized
in such a way that the population primarily at risk is the vehicle occupants
in the roadways; (c) the highest freeway concentrations of sulfates due to
vehicle emissions observed in the two-week experiment during October of 1976
were below 2 yg/m3; and (d) the ammonia concentrations appeared to be great
enough to neutralize the sulfuric acid emitted from the vehicles.

               During the spring of 1978, an in-roadway sampling program with
measurements both on the roadway and at background sites will be conducted
for approximately 30 days using improved H2S04 sampling and analytical methods
currently under development.  The experimental sampling program will be
statistically designed, based on models prepared during 1977 so as to ascertain
the exposure risks that can be anticipated in the future.  Since the test
vehicles will also be monitoring in-roadway N02, the statistical analysis of
the measurements may also provide an estimate of the N02 exposure to
vehicle occupants.

     At the inclusion of the in-roadway Los Angeles study, an experiment
will be conducted in Research Triangle Park, North Carolina, to measure
the H2S04 inside vehicles driving immediately behind a cluster of several
high-sulfate emitting automobiles.

          4.   Modeling

               Due to the complexity and great number of variables such
as atmospheric turbulence, wind changes, pressure/temperature differentials,
and turbulence from vehicle motion, the progress in developing acceptable
models of in-roadway conditions has been tedious.  To explain how H^SO  mist

-------
reaches the exposed public from catalyst-equipped vehicles, the basic theory
of dispersion and aerosol coagulation has been examined in models.  A
Gaussian model of a line-source roadway has been used with data from
General Motors proving-ground-track experiments.  Both short-time-scale
(second to second) and half-hour meteorological data were applied to the
model.  Results indicate that assigning Gaussian dispersion parameters on
the basis of atmospheric stability alone leads to overestimation of
half-hour sulfuric acid concentrations.  To improve the predictive
capability of this simple theory, best-fit Gaussian distribution para-
meters were extracted from the GM short-time-scale meteorological data.
Analysis of these data will continue into early FY-78.  The predictive
error found in this approach may be due to the inability to account
for the short-term wind meander under low wind conditions or to non-
Gaussian concentration profiles induced by cars.

               Models incorporating greater physical detail to improve
predictive capability are being tested.  Numerical integration of the
conservation of mass equations may produce a better picture of the roadway
as a line source.  Techniques using computer programs to perform calcu-
lations have been developed and tests are projected for future work.  Although
GM data are valuable and useful, they have been shown to be limited.  More
extensive data are being obtained in an ESRL-supported study on the Long
Island expressway.  These data may allow more extensive evaluation on
various models and identify problem areas in prediction.

               Changes in automobile and fuel technology are occurring
rapidly in response to pollution control and energy conservation challenges.
These induce new requirements for quick assessment of environmental impact
which engender the development of new models to apply to new problems.

               More detailed discussions of each project together with
tabular data is available in the Environmental Sciences Research Laboratory
report (Appendix B).

     C.   Health Effects Research Laboratory (HERL)/Cincinnati

          The Catalyst Research Program (CRP) at the HERL/Cincinnati
location began studies in 1973 to provide data on the toxicology of whole
automotive emissions with and without the catalytic converter, and on
certain individual pollutants which comprise whole exhaust.  A new facility
for collecting data on animals exposed to automotive emissions was con-
structed and fitted with new equipment to replace the previously used
Laidlaw Avenue installation.  The new facility is able to generate either
gasoline or diesel exhaust atmospheres for a broader range of tests.  It
allows the assessment of effects on inhalation of irradiated (artificial
sunlight) and non-irradiated exhaust on biological parameters related to
general health and immunity.  The relocation, installation, and calibration
on the new facilities at the EPA Center Hill Road laboratory delayed the
initiation of certain segments of the 1976 research program in Cincinnati.

-------
          The CRP effort has been divided into two major segments:

          •    Assessing biological effects of total exhaust which has
               been processed by an oxidative catalytic converter.

          •    Toxicology of single pollutants associated with the
               oxidation catalyst.

          The HERL in Cincinnati works to gather information to
determine the biomedical consequences of exposure to automotive
emissions which may lead to an improved understanding of those
consequences.

          The National Research Council document on Fuels and Fuel
Additives for Highway Vehicles and Their Combustion Products points
out that hundreds of compounds are present in vehicle exhausts which
are practically impossible to detect and measure for toxicology
studies.  Rather than try to create individual components synthetically
and then mix them for research, it may be more realistic to use whole
emissions from actual working models to achieve more representative
results.  Biological hazards from interaction of the components—synergism
or potentiation--can be just as important as the toxic effects of single
isolated components.  The CRP provides data on the toxic effects of animal
exposure to whole emissions of engines with or without catalytic conver-
ters.

          A secondary aspect of the program is to study metabolic and
biological effects of platinum, palladium, and other nonregulated
pollutants (such as H2S04) associated with the use of catalytic converters.

          Since the use of catalytic converters in auto exhausts, aero-
metric data show a decrease in carbon monoxide and hydrocarbon concentra-
tion.  As sulfuric acid, a product of the catalysts, is a greater irritant
than sulfur dioxide ($62), its effects could offset the benefits provided
by catalytic converters.  An additional study, therefore, was initiated
in 1976 in which an engine-catalyst combination producing high levels of
sulfuric acid would be used with gasoline containing increased levels of
sulfur.  This was to assess the effects on the respiratory system.  The
study was also designed to examine animal adaptation to the test atmosphere.
The results of this study will be available in 1977.

          The CRP at HERL/Cincinnati has four major areas of research
all involving assessment of biological effects in laboratory animals
exposed to automobile emissions:  (1) automotive catalytic emissions,
(2) diesel emissions, (3) ultrafine sulfuric acid and metal sulfates,
and (4) manganese compounds associated with the use of the additive
methylcyclopentadienyl manganese tricarbonyl (MMT) in gasoline.

-------
          1.   Physical Facilities

               Automobile engines used were selected on the basis of
wide distribution in the real environment, and thereby functioned as
representative sources of emission.  Two 1975 automobile gasoline engines
and one automobile diesel engine are used in the models for this research.
The gasoline engines are equipped with pelletized noble-metal oxidation
catalytic converters.  To simulate reality more closely, the emissions
are channeled into mixing chambers and frradiation (artificial sunlight)
chambers before reaching the animal exposure areas which are monitored
continuously for gases and particulates of interest to the study by
various appropriate methods.  Typical gasoline and diesel fuels were
selected as representative samples.

          2.   Automotive Catalytic Emissions Study

               Using a 1975 Chevrolet engine with a catalyst, experiments
were begun in 1976 to study effects of exposure on pulmonary function
in guinea pigs, cats, and rats, and effects on blood gases in rats.  Effects
of exposure were also noted on susceptibility to respiratory infection.
There were two phases in the guinea pig experiment:  acute and chronic,
defined according to exposure time.  The acute phase exposed the experimental
group to emissions for one hour per day, compared with 16 hours per day, 7
days per week, in the chronic phase.  The concentration levels for emissions
were the same for chronic and acute phases.  Both phases began on November
15, 1976, and the chronic phase is to be completed March 25, 1977.  Measure-
ment of all animals, acute and chronic, is scheduled for completion by
June 1977, and a final report is anticipated by October 1977.

               The experiment investigating effects of inhalation exposure
to catalyst-treated exhaust using cats, focused on pulmonary function and
histopathology.  Cats, being larger animals, allowed a wider range of
pulmonary function tests, and using a variety of species of animals lends
greater accuracy in extrapolating the data to humans.  Measurements are to
be completed by April 1977, and the report is to be prepared later in 1977.

               Effects of exposure on arterial blood gases and pulmonary
function were examined in rats.  Rats were exposed to catalyst-treated
exhaust 16 hours per day for 45 or 90 days, and parameters included body
wieght, food intake, lung weight, static lung compliance, and other
respiratory function parameters and blood chemistry.  Analysis of these
data is scheduled for completion in 1977.

               Difference in susceptibility to respiratory infection as
a result of exposure to catalyst-treated exhaust were examined against
the hypothesis that exhaust exposure will enhance susceptibility to
lethal infection and thus increase mortality in the exposed group.  Princi-
pal variables in the experimental design were:  type of exhaust, exposure
duration, single versus intermittent infectious aerosol exposure, and
animal age.  Preliminary results support the hypothesis.  Final analysis
and reporting will be completed during 1977.

-------
          3.   Automotive Diesel Emissions Studies

               One of the concerns of CRP focuses on health hazards from
diesel engine exhaust.  There are increasing numbers of diesel-powered
vehicles on the highways, and the EPA has been characterizing their exhaust
components.  To date it appears that although there are lower concen-
trations of carbon monoxide and hydrocarbons than from gasoline^engines,
there are higher concentrations of nitrogen oxides in diesel emissions.
Diesel emissions are more visible to the public and particulate emissions
are 20-50 times higher than gasoline-fueled cars burning unleaded fuel.
Little biological information is available on diesel particulate deposits
in the lung, and this is a major area of emphasis in the CRP diesel study.

          A capability to study diesel engine emissions was designed into
the Center Hill Road Laboratory in Cincinnati.  Installation and calibration
of this hardware is progressing satisfactorily, and investigation of the
effects of the exhaust on animals are to begin in 1977.  These studies
include pulmonary function, histopathology and blood chemistry, biochemical
alterations in the lung, and effects on spontaneous locomotor activity.
This will be the first HERL experiment in exposing animals to diesel exhaust.
To date, there is very little information in the literature to serve as a
guide for anticipating biological effects of diesel exhaust exposure.  The
Cincinnati laboratory has extensive experience in assessing lung changes
and pulmonary function; therefore, this will be a major emphasis.

          4.   Ultrafine Sulfuric Acid and Metal Sulfates

               Automobiles using catalytic converters emit ultrafine
Studies on the effects of exposure to H2S04 have previously used H2S04
particles beyond the size of the ultrafine range because development of a
generation system for the proper size had not been completed.  In the studies
using the greater size 1^504 particles (_> 0.1 micrometers), the data suggest
an increase in pulmonary resistance in animals exposed, but these are
considered acute exposures.  Little information is available on the effects
of long-term low-level exposures.  Along with projected studies of such
exposures, the toxicity of some of the metal sulfates will also be determined.

          5.   Manganese Oxide Studies

               The Federal requirement for phasedown of lead in gasoline,
additive methylcyclopentadienyl manganese tricarbonyl  (MMT), an organic
manganese compound, was proposed as an alternate antiknock.  The possible
toxicity of airborne manganese subsequently became a concern of CRP, and
inhalation exposure studies are scheduled for 1977.  Parameters of major
interest include pulmonary changes and neurological alterations.

               Manganese is an element essential to life and in trace
amounts is beneficial.  When man is exposed to industrial amounts, two
clinical  syndromes are manifest:  manganese pneumonia and manganese
                                10

-------
poisoning affecting the central nervous system (CNS).  The CRP in HERL/
Cincinnati have designed studies to examine changes in the lung and CNS
effects, and to investigate retention, tissue distribution, and excretion
following acute inhalation exposure to 54MnS(L and ^V\npSCL aerosols.

     D.   Health Effects Research Laboratory/Research Triangle Park

          Catalyst Research Program (CRP) activities performed by the
Health Effects Research Laboratory (HERL/RTP) can be divided into three
major areas of research:  studies on sulfuric acid, studies on metals,
and characterization of in-use catalyst vehicle emissions.  Discussion
of each project appears in this volume in the HERL/RTP program report
(Appendix D).

          1.   Sulfuric Acid Research

               The CRP's interest in sulfuric acid (H2$04) stemmed from
the finding that the catalytic converter, used as an emission-control
device in American cars since 1975, can produce significant sulfuric
acid emissions.  Animal studies during 1976 have focused on the effect
of inhalation of sulfuric acid mist on the immune system.  The lymphocytes
of rabbits exposed to 1 mg/m  (0.3 pm-0.5 ym for 4 hours a day for 2 days
showed a marginally significant (p < 0.1) increase in transformation of
T cells, suggesting the possibility that sulfuric acid enhanced the blasto-
genesis of T cells or caused an alteration in the ratio of B:T cells.
(See Table 1 in the HERL/RTP laboratory report).  A statistically signi-
ficant  (p < 0.05) decrease in hematocrit occurred after a single 2-hour
exposure to the acid mist.  The number and percent of polymorphonuclear
leukocytes (PMN) increase (p < 0.05) and the number and percent of
lymphocytes decreased (p < 0.05) both in the animals exposed to acid mist
and in  the controls exposed to air.  The changes were greater (p < 0.05),
however, in the acid-exposed animals, indicating that a nonspecific stress
was enhanced by the sulfuric acid,  the experiments also revealed a
decrease  (p < 0.05) in hematocrit following exposure to the aerosol.
Table 3 in the HERL/RTP laboratory report, presents the hematocrit measure-
ments .

               One of the most interesting results of these experiment^
was the finding that when a 2-hour exposure to sulfuric acid (1.0 mg/m )
was preceded by a 3-hour exposure to sulfuric acid (1.0 mg/m3), the
mortality increased significantly over that caused by either pollutant
alone.  This apparent additive effect suggests that exposure to ozone
initiates changes in tissue resulting in increased sensitivity to sulfuric
acid.

               In addition to these animal studies on sulfuric acid, the
CRP is  initiating a project to investigate the health effects of low-level
sulfuric acid aerosols on young healthy human male volunteers.  This study
                                  11

-------
scheduled to begin early in 1977, will use sulfuric acid concentrations
ranging from 50 to 200 n9/nr with a mass median diameter ranging from
0.05 ym to 0.5 urn in an attempt to reflect the projected sulfuric acid
aerosol burden created by Increased use of the catalytic converter.  The
test parameters will be primarily measurements of pulmonary function.

               One vegetation study has been conducted as part of CRP's
interest in sulfuric acid.  The project at the University of Minnesota's
Department of Plant Pathology has investigated the effects of submicro-
metric sulfuric acid aerosol exposure on vegetation.   Hybrid poplar,
pinto bean, and soybean showed marginal and tip necrosis when exposed to
25-30 mg/m3 sulfuric acid aerosol.

          2.   Metals Research

               The metals research conducted in 1976  by HERL/RTP under the
CRP focused on manganese (Mn) and platinum (Ft).   Studies on manganese were
performed in conjunction with a laboratory-induced streptococcal Infection
in order to determine the relationship between manganese inhalation and
chronic respiratory disease in mice.  A statistically significant enhance-
ment in respiratory infections occurred after 2-hour  inhalations of mangan-
ese at >_ 2.0 ing Mn/m3 for manganous-manganic oxide atmospheres (^1304) and
2. 3.2 mg Mn/m  for manganese chloride aerosols (MnCl2-4H20).  Exposed
animals also experienced delayed clearance and an Increase in growth of
inhaled streptococci.  (See Figures 3 and 4 in the HERL/RTP laboratory
report).

               Most of the platinum studies sponsored by HERL are nearing
completion with final results expected in 1977 or early 1978.  Preliminary
results from several studies are suggesting some  of the element's possible
adverse effects.  Although research has shown that platinum does not
induce allergy, palladium, when complexed with albumin, will induce a
delayed allergy which can be transferred via spleen cells from sensitive
donors.  Platinum sulfate has been characterized  as a severe eye irritant
and may also have some cytogenetic activity.  Preliminary results of a
study of the effects of trace metals on the central nervous system suggest
that platinum sulfate has a moderately depressive effect on behavior.  It
has also been shown that, under laboratory conditions, platinum sulfate
or potassium hexachloroplatinate, like mercury, can be methylated with
methyl B^p-

               Reports associated with HERL/RTP metal studies which are
expected to be of particular value to researchers include:  (1) a series
of reports, expected in late 1977, evaluating analytical methods for
measuring trace elements in biological matrices;  and  (2) a follow-up report
to "Proceedings - Catalyst Research Program Platinum  Research Review
Conference."  Since platinum had received little  attention before the
establishment of the CRP, this report should be very  helpful to researchers
in the field.
                                  12

-------
          3.   Characterization of In-Use Vehicles

               A study of 49 in-use catalyst vehicles which is being
conducted in cooperation with New York State's Department of Environmental
Conservation has as its major objective the measurement of sulfate, parti-
culate, and regulated emissions.  Few of the vehicles meet all regulated
emission standards.  In general, the tests show low sulfate and high
carbon monoxide emissions.  In many cases, the test cars were found to
have had the idle mixture adjusted, increasing the carbon monoxide
values beyond the manufacturers' specifications.  The results reveal that
idle carbon monoxide concentration correlates with carbon monoxide mass
emission data and that sulfate emissions are inversely related to carbon
monoxide values.  Carbon monoxide and HC emissions tend to increase with
mileage, but no overall correlation of sulfate with mileage is presently
apparent.  Other conclusions reached thus far indicate (1) that air
pumps increase sulfate production and aid the converter's decrease of
carbon monoxide, (2) that large amounts of sulfur dioxide are purged from
catalysts upon deceleration to idle when idle carbon monoxide is high,
(3) that little particulate sulfate is emitted in idle periods for either
high or low carbon monoxide conditions, and (4) that emissions vary
markedly according to vehicle manufacturer.

     E.   Office of Mobile Source Air Pollution Control/Washington, D.C.

          As hydrocarbon, carbon monoxide, and nitrogen emissions are
controlled, various unregulated emissions may be altered.  It is important
that these pollutants be characterized to assure that control of regulated
emissions does not create problems in other areas.  The unregulated
emissions discussed in the following paragraphs are sulfuric acid, hydrogen
cyanide, ruthenium, ammonia, polynuclear aromatic hydrocarbons, and diesel
parti oil ates.

          1.   Sulfuric Acid

               Oxidation catalysts in exhausts oxidize a portion of the
emitted sulfur dioxide to sulfuric acid.  Sulfur dioxide is formed in
the exhaust by combustion of trace quantities of organic sulfur compounds
in gasoline.  Even if the oxidation catalyst converted all of the sulfur
dioxide in the exhaust to sulfuric acid, the resultant contribution from
vehicles would be less than 1 percent of the total sulfuric acid in the
atmosphere.  However, it is possible to have high localized levels of
sulfuric acid along heavily traveled roads.  Experiments have been
conducted by Exxon, GM, Ford, and Chrysler to determine the effect of
various parameters (catalyst composition) on sulfuric acid formation.
Also, tests have been run using sulfuric acid traps which are chemical
containers (mufflers or catalyst-like) placed in the exhaust system after
the oxidation catalyst.  Ford and GM conducted work in measuring roadside
sulfuric acid levels.
                                 13

-------
               Work during the past year confirmed earlier studies
showing that exhaust oxygen levels are the most important single para-
meter affecting sulfuric acid emissions.  Low oxygen levels (such as
those found with 3-way catalyst cars) can result in sulfuric acid emission
levels of under 5 mg/nri which is equivalent to levels found with non-
catalyst cars.  Limited test results so far show that high mileage  _
oxidation catalyst vehicles generally have lower sulfuric^acid emissions.
The  past year's roadside measurement tests of sulfuric acid showed
levels far below those predicted by EPA air quality models.  These  lower
level measurements, plus the possible reduction of sulfuric acid emissions
at higher mileages, may result in these emissions being considered mucn
less of an air quality problem than previously thought.

          2.   Hydrogen Cyanide

               In early 1976, EPA found that vehicles using 3-way catalysts
could produce hydrogen cyanide (HCN) under rich malfunction modes.  Since
Volvo and Saab were certifying 3-way catalyst systems for use in California
in 1977, EPA had to make a rapid determination on the levels of HCN that
would be acceptable from a health standpoint.  EPA was concerned primarily
about worst-case situations, i.e., localized levels that could occur in
indoor parking garages and along heavily traveled roadways.  The highest
HCN  emissions were observed under rich malfunction modes when carbon monoxide
(CO) emissions were also highest.  In the closed environment it was found
that the adverse health effects from CO would overshadow possible adverse
effects from HCN by more than two orders of magnitude.  For the worst
reasonably conceivable highway-type exposure, it was determined that the
amount of HCN present would not have unacceptable health effects.  Tests for
HCN  emissions using a variety of catalysts were conducted by GM, Ford, and
Chrysler.

               It has been found that oxidation catalyst cars emit far
less HCN than non-catalyst cars even under malfunction conditions.  HCN
emissions from catalyst cars are currently not a problem; however, HCN
measurements should be made on future emission control systems as they
are  developed.

          3.   Ruthenium

               Ruthenium is being considered by several companies for use
in catalysts because of its low cost and attractiveness for nitrogen oxide
control.   Since ruthenium oxides are considered to be very toxic, it
appears that the use of ruthenium catalysts depends on there being no
significant emission of their oxides.  EPA is not currently in a position
to quantify what levels, if any, of emissions of ruthenium or its oxides
would be considered significant.  EPA health effects studies are planned
for  investigation of the toxicity of ruthenium and ruthenium oxides.
                                     14

-------
            4.    Ammonia
                     :                        .:> „ i ,•

                 Ammonia emissions have been known to occur with,dual
  catalyst systems under rich operating conditions.  Ammonia formed over
  the reduction catalyst is frequently reoxidized^to NO over the oxida-
  tion catalyst.  Incomplete oxidation of the ammonia over the oxidation
  catalyst, however, could result in significant  ammonia emissions.  A
  malfunction of the oxidation catalyst system.(e.g.i air pump failure)
  could have similar results.  Ammonia can also,be formed with 3-way
  catalyst systems.   EPA is currently investigating possible adverse
  health effects from automotive ammonia emissions to include both entire
  air quality regions and localized situations such as parking garages and
  heavily traveled freeways.  Ford, GM, and Exxon are running tests on
  ammonia emissions  resulting from various types  of catalysts...  !i
                        .'-| ,=;-  ^                .-     \t l  •*, T •';,'•  '    ''-
,,,                Ammonia emissions may be significant from ^vehicles with
7'3-way catalysts but;only under malfunction conditions,,3 Ammonia emissions
 •.from vehicles equipped with oxidation catalysts do not appear significant
  under normal  operating conditions.  Currently,  there is inadequate infor-
  mation to permit determination of the level of  ammonia emissions that would
  be considered acceptable.

            5.    Polynuclear Aromatic Compounds

                 The polynuclear aromatic compounds (PNA's)  are a group of
  multi-ring aromatic hydrocarbon compounds with  very low volatility and
  high molecular weight and are of interest due to their carcinogenicity.
  About 90 percent of the total PNA's in the atmosphere come from stationary
  sources, and it is estimated that of the remainder, half come from mobile
  diesels and half from other motor vehicles.  PNA compounds in automotive
  exhaust have been  studied for some time.  It has been found that post-1968
  cars have about half the PNA emissions of pre-1968 models.  Tests on proto-
  type cars with advanced catalysts and thermal reactors indicate a reduction
  in PNA emissions by 95 percent or more compared to uncontrolled vehicles.
  Recent Volkswagen  tests on PNA emissions in gasoline engine vehicle exhaust
  confirm earlier Exxon work which shows only 5 percent of PNA emissions in
  catalyst-equipped  cars over those in non-catalyst cars.   VW also reported
  that their studies overall showed extremely low levels of PNA concentra-
  tions in automobile exhaust emissions.

            6.    Diesel Particulate Emissions

                 Mobile sources contribute a small, but significant, part
  of total suspended particulates in ambient air.  Prior to the introduction
  of unleaded fuel,  automobiles contributed between 2 and 13 percent of the
  total suspended particulates in urban areas. As the use of unleaded fuels
  in catalyst-equipped vehicles displaces the use of leaded fuels, the
  automotive contribution to urban air total suspended particulates will
                                      15

-------
decrease to about 0.2 to 1.3 percent-   The Introduction of large numbers
of light-duty diesels may reverse this trend; therefore, it is important
to investigate the quantity of diefel  participates emitted.

               The composition of light-duty diesel par ticulates has not
been completely determined.  A major constituent of the Pj^cuiiates is
known to be elemental carbon which has the ability to*?  ,!;  en   c,,if«tP
compounds.  Sorbed compounds that have been identified include S02, sultate,
high molecular weight organic compounds, and PNA s.
               Almost all of the w0rk to obtain emission factors on light-
duty diesels had been accomplished, by EPA either through Cftra9*°^arill
in-house testing; little data have been provided by the automobile manu-
facturers.  Tests conducted to datp show very high particulate emissions
from light-duty diesels; about 25 times the values obtained for gasoline
vehicles equipped with catalysts. rThe partieulates consist of elemental
carbon as well as various hydrocarbon compounds and small quantities or
sulfate.  The health effects, if apy, associated with these particulates
have not been determined.

-------
III. DISCUSSION OF PROGRAM HIGHLIGHTS

     During calendar year 1976, the various laboratories of EPA involved
in the CRP continued tests and studies on mobile gasoline engine exhaust
emissions, various gasoline engine catalyst systems, and the health effects
which exhaust emissions have on humans, animals, and vegetation.  Some
studies and testing programs were newly initiated in CY-76, and other
efforts were continuations of those initiated in previous years.  Because
of the imminent introduction of large numbers of light-duty diesel vehicles
to the highways of the United States, attention also began to focus on
diesel particulate emission and its effects on health.  Highlights of
research conducted under CRP and results obtained during CY-76 are presented
in the paragraphs which follow.

     A.   Sulfates

          Sulfur dioxide is formed in vehicle exhaust systems by combustion
of trace quantities of organic sulfur compounds in gasoline, and oxidation
catalysts in the systems oxidize a portion of the sulfur dioxide to sulfuric
acid.  Work during the past year confirmed earlier studies showing that
exhaust oxygen levels are the most important single parameter affecting
sulfuric acid emissions.  Low exhaust oxygen levels, such as those found
with 3-way catalyst cars, can result in low sulfuric acid emissions (under
5 mg/mi).  Limited test results so far show that high-mileage oxidation
catalyst vehicles have even lower sulfuric acid emissions (approaching 1
mg/mi).  Car manufacturers are continuing to run tests on catalyst vehicles
as they accumulate mileage.  During the past year, GM completed a major
test-tract experiment measuring roadside sulfuric acid levels.  Results
showed levels far below those predicted by EPA air quality models.  The
roadside lower levels, combined with further reduction in sulfuric acid
emissions at higher mileages, may result in sulfuric acid emissions
being considered much less of an air quality problem than previously
hypothesized.

          The main objective of the Los Angeles Catalyst Study (LACS) is
to develop data on mobile source-related pollutants in the air before and
after the introduction in 1975 of the catalytic converter.  Analysis of
data collected in CY 76 indicates that the use of the catalytic converter
and unleaded fuel has resulted in a 25 percent decrease in carbon monoxide,
lead, and total suspended particulates, while the freeway contributions
of sulfates have increased less than 10 percent.

          Under the National Fuels Surveillance Network program, analyses
of samples of commercial gasoline showed that unleaded gasoline in Southern
California contained approximately twice the amount of sulfur as New York
samples.  The content of sulfur in all three grades of gasoline was consist-
ently higher in Los Angeles than in San Francisco.
                                   17

-------
          An in-roadway study, conducted in October 1976, showed that
catalyst vehicles caused an increase in sulfur concentration levels over
background sulfate of no more than 1 pg/m3.  No larger scale study of
sulfates is planned, but there are plans to investigate the effects on
automobile occupants driving immediately behind a cluster of high-suirate-
emitting vehicles.

          During the reporting period, studies were initiated which seek to
assess the health effect(s) of exposure to sulfuric acid aerosols.  None
of the studies thus far, however, have used sulfuric acid aerosols in the
ultrafine size range, from 0.01-0.02 ym mass median diameter (MMD), which,
as reported in last year's CRP report, may be the size range of the sulfuric
acid actually emitted by automobiles equipped with catalytic converters.
The capability of generating such ultrafine aerosols should soon be achieved,
however, and studies should be underway shortly.   Plans are complete for a
human studies experiment which will  test the effects of sulfuric acid aerosol
health parameters (primarily pulmonary) of healthy young male volunteers.  The
aerosol generated will contain sulfuric acid particles ranging from 0.05 to
0.5 ym.  Results of these tests will be available in the 1977 report.

          The 1976 health effects studies of sulfuric acid have suggested
several ways in which the pollutant  having a larger MMD affects laboratory
animals.  Results of studies using a concentration of 1 mg/m3 sulfuric acid
(0.03-0.5 ym) suggest that such exposure may enhance blastogenesis of T
cells or cause alterations in the ratio of B:T cells.  Exposure also resulted
in a decrease in hematocrit, an increase in leucocytes, and a decrease in
lymphocytes, which was greater than  that experienced by controls.  It was
also found that prior exposure to ozone increased the effect of exposure to
sulfuric acid.  It seems likely that ozone initiates changes in the tissue,
making it more sensitive to sulfuric acid.

     B.   Hydrogen Cyanide

          In early 1976, EPA found that vehicles  using a 3-way catalyst
could produce hydrogen cyanide (HCN) under rich-malfunction modes.  There
was concern regarding the levels of  HCN that would be acceptable from a
health standpoint under worst-case situations, such as levels that could
occur in indoor parking garages and  along heavily traveled roadways.
In the closed environment it was found that adverse effects from carbon
monoxide would overshadow possible adverse effects from HCN by more than
two orders of magnitude.  For the worst possible  highway exposure, it
was determined that the amount of HCN present would not cause adverse
health effects.
                                 18

-------
     C.   Manganese Oxide

          An organic manganese compound, methylcyclopentadienyl manganese
tricarbonyl (MMT), was proposed for widespread use as a substitute
additive when Federal regulations required a phasedown of the use of
lead antiknock agents in gasoline.  Manganese in trace amounts is an
element essential to life, but when man is exposed to very high levels,
two clinical syndromes are manifest:  manganese pneumonia and manganese
poisoning affecting the central nervous system.  A possible increase in
airborne manganese resulting from exhaust emissions prompted EPA to design
manganese sulfate and manganese oxide inhalation studies scheduled to
begin in 1977.

     D.   Polynuclear Aromatic Compounds

          The polynuclear aromatic compounds (PNA) are a group of multi-ring
aromatic hydrocarbon compounds with very low volatility and high molecular
weight which are of interest due to their possible carcinogenic effects.
About 10 percent of the total PNA's in the atmosphere come from motor
vehicles, diesel as well as gasoline-powered.  PNA compounds have been
studied for some time, and it has been found that post-1968 cars have
about half the PNA emissions that pre-1968 models have.  Tests on prototype
cars with advanced catalysts and thermal reactors indicate a reduction in
PNA emissions by 95 percent or more compared to non-catalyst vehicles.

     E.   Ammonia

          Ammonia emissions have been known to occur with dual and 3-way
catalyst systems under malfunction operating conditions.  Currently there
is inadequate information on which to base a determination of the level
of ammonia emissions that would be considered acceptable.  EPA is now
investigating possible adverse health effects from automotive ammonia
emissions to include both entire air quality regions and localized situa-
tions, such as parking garages and heavily traveled freeways.  Elements of
the automotive industry are running tests on ammonia emissions resulting
from various types of catalysts.

     F.   Ruthenium

          Because of its low cost and attractiveness for nitrogen oxide
control, ruthenium is being considered by several companies for use in
catalysts.  Since ruthenium oxides may be very toxic, the use of ruthenium
in catalysts depends on there being no significant emission of its oxides.
EPA plans to address the subject of toxicity of ruthenium and its oxides
since EPA is not at this time in a position to quantify what levels of
their emissions would be considered significant.
                                      19

-------
     G.   Platinum

          Platinum studies have formed an important part of CRP's
research program because the metal is commonly used in the catalytic
converters for U.S. automobiles.  The degree of emphasis on platinum
is declining, however, and most platinum and other noble metal research
is scheduled for completion in the next 18 months.  A report is planned
as a follow-up to "Proceedings - Catalyst Research Program Platinum
Research Review Conference" which was published in early 1976.  Since
platinum had received little attention before the establishment of the
CRP, the report should be extremely useful to future researchers.

     H.   Diesel Emissions

          Since automobile manufacturers have indicated an intention to
increase production of diesel-powered automobiles, EPA has initiated
studies of diesel exhaust in anticipation that it may be asked to evaluate
the risks versus the benefits of this development.  Tests conducted to date
show very high particulate emissions from light-duty diesels—about 25
times the values obtained for catalyst-equipped gasoline vehicles.  The
composition of diesel particulates has not been completely determined.
It is known that the particulates consist of elemental carbon as well as
various hydrocarbon compounds and small quantities of sulfates.  Diesel
engines have lower emissions of carbon monoxide than gasoline engines but
higher concentrations of nitrogen oxides.

          Very little information is available on the health effects of
diesel emissions.  The major health effects effort is being conducted
jointly by ESRL, EMSL, and HERL.  This effort is providing basic background
information prior to whole animal testing.  Installation and calibration
of hardware in the Cincinnati Health Effects Research Laboratory to obtain
more information in this area has progressed satisfactorily, and tests
on animals are scheduled to begin in 1977.  The laboratory has extensive
experience in assessing  lung changes and pulmonary function, and the major
emphasis of the tests will be in these areas.

     I.   Health Effects of Exposure to Whole Exhaust

          A major research effort during late CY-76 was the initiation
of a 3-month exposure of laboratory animals to automotive catalytic-treated
exhaust.  The study placed major emphasis on defining the effects of the
exposure on pulmonary function.  The exposure period has been completed,
data are being analyzed, and the reports should be completed in CY-77.
After completion of this study, there are no immediate plans for additional
animal exposures to automotive catalytic emissions.
                                    20

-------
                  APPENDIX A
 THIRD ANNUAL CATALYST RESEARCH PROGRAM REPORT
                  Prepared by

Environmental Monitoring and Support Laboratory
       Research Triangle Park, N.C.  27711
     U.S. ENVIRONMENTAL PROTECTION AGENCY
      OFFICE OF RESEARCH AND DEVELOPMENT
      HEALTH EFFECTS RESEARCH LABORATORY
      RESEARCH TRIANGLE PARK, N.C. 27711

-------
                           TABLE OF CONTENTS
                                                                          Page
List of Tables	              23
List of Figures	              24
  I.   Overvi ew	              26
 11.   Background and Introducti on	              26
III.   Current Status
       A.  Fuel Collection and Analysis	              26
       B.  Monitoring - The Los Angeles Catalyst Study	              27
       C.  Quality Assurance Activities	              69
 IV.   Problem Areas
       A.  Fuel Collection and Analysis	              72
       B.  Monitoring - The Los Angeles Catalyst Study	              72
       C.  Quality Assurance Activities	              72
  V.   Plans for Future Research
       A.  Fuel Collection and Analysis	              77
       B.  Monitoring - The Los Angeles Catalyst Study	              77
       C.  Quality Assurance Activities	              78
 VI.   Conclusions
       A.  Fuel Collection and Analysis	              78
       B.  Monitoring - The Los Angeles Catalyst Study	              79
       C.  Quality Assurance Activities	              79
Appendix A-l:  Precision of LACS Sampling and Analytical
               Methods	              80
References	              83
                                  22

-------
                         LIST OF TABLES
                                                                 Page
  I    Analytical  Methods	        28
 II    Content of Sulfur, Lead, and Aromatics of Gasoline
       Collected in San Francisco and Los Angeles by
       Season	        29
111    Inventory of Integrated Data from LACS	        37
 IV    San Diego Freeway Traffic Flow - Daily Average
       Traffic Count	        63
  V    Percent Data	,	        68
 VI    Blind Audit and Split Sample Frequencies	        71
                              23

-------
                        LIST OF FIGURES
                                                                      Page
 1    LACS Site Location in Los Angeles	              31
 2    LACS Study Site Composition and Elevation	              32
 3    Annual Wind Frequency and Speed by Wind Direction..              33
 4    CO By Wind Direction, Annual Composite	              35
 5    LACS Platform Sampler Schedules for Integrated
      Samp! ing Techniques	              36
 6    Monthly 4-Hour Lead from Hi -Vol	              39
 7    Monthly 4-Hour Sulfate from Hi-Vol	              40
 8    Monthly 4-Hour Lead from Membrane Filters	              41
 9    Monthly 4-Hour Sulfate from Membrane Filters	              43
10    Monthly 24-Hour Lead from Hi-Vol	              44
11    Monthly 24-Hour Sul fate from Hi -Vol	              45
12    Monthly 24-Hour Lead from Dichotomous Samplers	              47
13    Monthly 24-Hour Sulfate from Dichotomous Samplers..              48
14    Monthly 24-Hour Total Sulfur from Dichotomous
      Sampl ers	              49
15    Cascade Impactor Results from Summer 1976, By Site               50
16    Monthly 24-Hour Lead Particles £0.7 ym as Determined
      by Cascade Impactor	              51
17    Monthly 24-Hour Sulfur Particles £0.7 ym as
      Determined by Cascade Impactor	              52
18    24-Hour Cascade Composites, Lead	              54
19    24-Hour Cascade Composites, S07	              55
                             24

-------
                                                                 Page
20   Monthly 24-Hour Sulfur Dioxide Levels	          56
21   Monthly Hourly Averages of Carbon Monoxide	          57
22   Monthly Hourly Averages of Nitric Oxide	          58
23   Monthly Hourly Averages of Nitrogen Dioxide	          59
24   Monthly Hourly Averages of Ozone	          60
25   Monthly Hourly Averages of Total Sulfur	          61
26   LACS Traffic Data:  Daily Composite for a Monday...          64
27   LACS Traffic Data:  Daily Composite for a Sunday...          65
28   Percent Catalyst Cars - Los Angeles County	          66
29   Comparison of RAMC Blind Sulfate Results with
     Spike Values	          73
30   Comparison of RAMC Blind Nitrate Results with
     Spi ke Val ues	          74
31   Comparison of RAMC Blind Lead Results with
     Spi ke Val ues	          75
32   Comparison of RAMC Blind Sulfur Dioxide Results
     with Spike Values	          76
                             25

-------
 I.   OVERVIEW

     The Environmental Monitoring and Support Laboratory's involvement
 in  the  Catalyst Research Program focuses on three major areas.  They are
 the fuels and fuels additives, the Los Angeles Catalyst Study (LACS), and
 quality assurance.

     The National Fuels Surveillance Network (NFSN) annually analyzes
 approximately 3,000 commercially purchased samples of gasoline, motor
 oil, fuel oil, and fuel additives.

     The LACS, a monitoring program, was established to assess the impact
 of  pollutants emanating from catalyst-equipped vehicles on ambient air
 quality adjacent to a major freeway.  A variety of pollutants is being
 monitored by both integrated and continuous methods.

     The quality assurance activities provide for the routine audits of
 laboratory analysis, development and distribution of standard reference
 samples and materials, measurement method evaluation, and quality assurance
 guidelines.

 II.  BACKGROUND AND INTRODUCTION

     The Catalyst Research Program, established in 1974 as an interdisci-
 plinary research program, involves three areas of the Environmental
 Monitoring and Support Laboratory, Research Triangle Park (EMSL/RTP):  fuel
 collection and analysis, monitoring, and quality assurance.  Previous
 reports (1-5) have summarized EMSL's activities prior to 1976.  This
 report summarizes all catalyst-related work performed by EMSL during or
 related to calendar year 1976.

     In the following pages each of the three aforementioned inter-
 disciplinary areas which pertains to EMSL will be discussed.

 III. CURRENT STATUS

     A.   Fuel Collection and Analysis

          The advent of the catalyst-equipped vehicle which requires the
 use of unleaded gasoline initiated the surveillance of motor vehicle
 gasoline.   Initial  interest focused on sulfur in gasoline as a possible
 contributor to atmospheric pollution from engine emissions.  Next,
 attention was directed to lead and phosphorus content of unleaded gasoline
 -- known poisons to vehicle catalysts.  Unleaded gasoline by definition has
 limited lead and phosphorus content, with an allowed maximum of 0.05 grams
 lead per U.S. gallon (g Pb/gal) and 0.005 grams phosphorus per U.S. gallon
 (g P/gal).   Presently, interest has expanded beyond these areas to include
consideration of aromatic content, manganese content and other factors.
                                   26

-------
          During the period of this report with the assistance of the
EPA Regional Offices, EMSL collected and analyzed approximately 3,000
commercially purchased samples of gasoline (premium, regular, leaded,
and unleaded), motor oil, fuel oil, and fuel additives.  In addition, 70
unleaded samples were collected from New York State in support of the
ROADS Study, a micrometeorological study conducted by the New York State
Department of Environmental Conservation.  The analyses included those
for more than one element, using both isotopic dilution spark source
mass spectrometry (ID-SSMS) and instrumental neutron activation (INAA).
Of the 3,000 samples, approximately 100 unleaded gasoline samples were
collected in Southern California by contractor and analyzed specifically
for sulfur (S), lead (Pb), manganese (Mn), and phosphorus (P).  Additional
measurements were made to characterize the fuel including viscosity
measurement, Reid vapor pressure (RVP), aromatic content, and American
Petroleum Institute (API) gravity.  Table I lists the analytical techniques
employed for these measurements.  Fuel composition data are utilized by
researchers for emission characterization, mass-balance models, surrogate
projections, and by officials concerned with mobile source enforcement
and control decisions.

          Table II compares sulfur, lead, and aromatic content of gasoline
sampled in Los Angeles and San Francisco in the summer of 1975, winter
of 1975-76, and summer of 1976.  The average sulfur content of unleaded
gasoline ranges from 0.017 wt. %S in San Francisco to 0.034 wt. %S in
Los Angeles.  This.is nearly double the level measured in 70 New York
samples (6).  The analyses of both the regular leaded and premium leaded
gasoline grades collected in Los Angeles and San Francisco indicate
higher sulfur content in Los Angeles.  Aromatic content of gasoline is
of importance to the refiner since higher aromatic content can increase
the octane number of the gasoline.  As organic metallic fuel additives
are reduced, the addition of aromatics is one way to keep the octane
number up and avoid the customer complaint of "spark knock."  Note that
aromatic content is higher in the summer than winter.

          A related activity includes analytical method development for
fuels analysis.  During the year, two methods were developed:  sulfur in
gasoline using gas chromatography with a flame photometric detector (7)
and phosphorus in gasoline using a graphite furnace screening method
(8).  Work has started on the analysis of BaP in diesel exhaust.

     B.   Monitoring — The Los Angeles Catalyst Study

          The impact of pollutant emissions from catalyst-equipped
vehicles on the ambient air has been studied since June 1974 in West Los
Angeles adjacent to the San Diego Freeway.  The study area was chosen
because of heavy traffic (approaching 200,000 cars per day), total
sulfur emissions from stationary sources are relatively low compared to
the rest of the nation, predominant perpendicular winds to the freeway,
and California's stricter standards for automotive emissions have resulted
in a larger percentage of catalyst-equipped cars on the highway.  Additional
details concerning the site and sampling methodologies are available
elsewhere (1).


                                    27'

-------
                                TABLE  I

                          Analytical Methods


  I.  Available and  Used

     Fluorescent Indicator Adsorption                       ASTM       D  -  1319
     DEPENTANIZATION                                        ASTM       Jj  '
     API  Gravity 0 60°F                                    ASTM       D  -   287
     Reid Vapor Pressure                                   AS™       "  "   J"
     DISTILLATION                                          AS™       S  "  ,o?S
     Sulfur, Lamp                                          ASTM       D  -  1266


 II. Available - Slow and Tedious

     Sulfur, Lamp                                          ASTM       D  -  1266
     Phosphorus, Colorimetric                               ASTM       D  -  3231
     Lead, Gravimetric                                     ASTM       D  -   526


III. Developed In-House

     Lead, AAS                                             ASTM       D  -  3237
     Lead, FTK                                             ASTM       D  -  3348
     Sulfur, GC/FPD
     Phosphorus, AAS-HGA
     Manganese, AAS
                                  28

-------
                                          TABLE II

                Content of Sulfur, Lead, and Aromatics of Gasoline Collected
                         in San Francisco and Los Angeles by SeaswT
San Francisco Basin

  Premium:
    Summer 75
    Winter 75-76
    Summer 76
                         Sulfur (Weight % S)
                          Hi
        Low
                                        Avg
                         0.024  0.001   0.009
                         0.021  0.002   0.014
  Lead (g Pb/gal)         Aromatics  (%}

 Hi     Low    Avg      Hi     Low    Avg
 	*
3.670  1.180  2.277
                                               37.4    18.6    28.4
  Regular:
    Summer 75
    Winter 75-76
    Summer 75
0.094  0.003  0.035
0.081  0.007  0.035
0.077  0.016  0.042
                                                 1.970  0.420  1.331
                                                 3.360  0.600  2.056
                       32.8   23.9    27.1
                       34.7   25.7    28.9
  No Lead:
    Summer 75
    Winter 75-76
    Summer 76
0.068  0.003  0.027
0.038  0.002  0.017
0.046  0.012  0.030
                                                 0.015  0.003  0.007
                                                 0.018  0.001  0.006
                       38.6   19.6   30.6
                       42.5   28.5   33.8
Los Angeles Basin

  Premium:
    Summer 75
    Winter 75-76
    Summer 76
0.062  0.002  0.024
0.107  0.018  0.043
                                                 3.830  0.790  2.237    34.8   23.4   28.8
  Regular:
    Summer 75
    Winter 75-76
    Summer 76
0.090  0.033  0.051
0.145  0.006  0.052
0.143  0.029  0.066
                                                 2.050  0.370  1.259
                                                 2.810  0.520  1.487
                       32.7   23.3   27.4
                       35.4   23.7   31.0
  i'io Lead:
    Summer 75
    Winter 75-76
    Summer 76
0.085  0.010  0.034
0.058  0.006  0.025
0.048  0.012  0.026
                                                 0.007  0.002  0.004
                                                 0.007  0.001  0.004
                       37.3   27.5   32.7
                       44.1   26.9   34.1
* "--" means no measurements taken
                                                  29

-------
           The main  objective of the Los Angeles Catalyst Study  (LACS)  is
 to  develop ambient  air data bases for sulfate  (SQ7), carbon monoxide
 (CO),  lead (Pb),  and other mobile source-related pollutants before  and
 after  introduction  of the 1975-model automobiles that employ catalytic
 converters.  The data from this study are being analyzed to determine
 whether  the  catalytic converter has significantly increased the ambient
 sulfate  levels and/or simultaneously decreased the ambient CO and Pb
 levels near  the San Diego Freeway in Los Angeles.

           The Environmental Monitoring and Support Laboratory (EMSL) is
 responsible  for all study-related functions including instrumentation,
 operation, sample analyses, quality control, and data validation and
 analyses.   However, since January 1976, the operation of instruments and
 analyses  of  samplers were performed under contract to Rockwell International
 or  by  interagency agreement with the Lawrence Berkeley Laboratory.  To
 assure the quality  of the data supplied by these two organizations, EMSL
 maintains  a  comprehensive quality assurance program in all experimental
 aspects  of the study (9).   EMSL issues periodic reports which discuss
 the trends and the  interrelationships among the various pollutant
 patterns  (1-5).  The LACS reports are available on request from the
 Office of  the Director, EMSL/RTP.

           The site  locations in Los Angeles and the site layouts in
 relation  to  the San Diego Freeway are shown in Figure 1.  By selecting
 sites with the prevailing wind perpendicular to the freeway, the cross-
 freeway  contribution to the pollutant levels can be determined using
 concurrent upwind and downwind measurements.

           Figure 2  is a horizontal-elevation view showing the vertical
 locations  of the sites relative to the freeway.  The monitoring equipment
 at  each site also is indicated in the figure.

           Three short-term studies have been performed at the LACS sites
 to  determine: (1) the repeatability of 24-hour high volume sampler data,
 (2) the validity of utilizing data at upwind Site A for background
 levels, and  (3) the repeatability and comparability of data from 4-hour
 high volume  and membrane samplers.  Results of the first two studies
were included in earlier reports.  The third study is contained with
 this report  (Appendix I).

           1.   Data Analysis

               a.    Wind Data

                    The overall patterns of wind speed and direction are
depicted in Figure 3.   This figure, based on measured data at the site
composited over the year, shows that the more prevalent wind directions
and the higher wind speeds occur near the perpendicular direction to the
                                    30

-------
SANTA MOMICA FREEWAY

-------
          MET
         TOWER
                                       PREVAILING WIND-
        10m
  SAMPLER INLETS
     1 m ABOVE
 FREEWAY SURFACE
                                             \
    SAN DIEGO
FREEWAY SURFACE
                                                               1	n
                                                   8 m"
                                    •30mr
                                                             2 m ABOVE GRADE
                                                             T
u>
SITE A

CO ANALYZER
TOTAL SULFUR (SO2) ANALYZER
NO/NO, ANALYZER
O3 ANALYZER
24-HOUR SO, BUBBLER
24-HOUR HI-VOL.
4-HOUR HI-VOL., 3-7 p.m.
4-HOUR MEMBRANE, 3-7 p.m.
24-HOUR CASCADE
MASSIVE VOL. AEROSOL SAMPLER
24-HOUR DICHOTOMOUS SAMPLER
AMBIENT TEMPERATURE AND
   DEWPOINT
WIND SPEED
WIND DIRECTION
                                                            SITE B

                                                            24-HOUR HI-VOL.
                                                            4-HOUR HI-VOL.,
                                                               3-7 p.m.
                                                            24-HOUR MEMBRANE
                SITEC

                CO ANALYZER
                TOTAL SULFUR (SO2) ANALYZER
                NO/NO: ANALYZER
                O3 ANALYZER
                24-HOUR SO, BUBBLER
                24-HOUR HI-VOL.
                4-HOUR HI-VOL., 3-7 p.m.
                4-HOUR MEMBRANE, 3-7 p.m.
                24-HOUR CASCADE
                MASSIVE VOL. AEROSOL SAMPLER
                24-HOUR DICHOTOMOUS SAMPLER
                TRAFFIC SPEED AND COUNT
                   SYSTEM
SITE D

24-HOUR HI-VOL.
4-HOUR HI-VOL., 3-7 p.m.
24-HOUR MEMBRANE
                                                   Figure 2.  LACS study site composition and elevation.

-------
10
 *   I  i  I  I  I  I  I  I  I  I
                                    I  I  I  I  I  I  I  I  III  I  I  I  I  I  I  I  I  I  I  I  I
                                                       I"*—FREEWAY PERPENDICULAR
           WIND DIRECTION
     .8 8 1 1 g §
DEGREES
                                                                     § 1
                                                                ^
                                                         AVORABLE SAMPLING SECTOR
              Figure 3. Annual wind frequency and speed by wind direction.
                                            33

-------
freeway.  Micro-meteorology at the site tends to depict two distinct
patterns, "summer" or "winter," with summer demonstrating a six-month
period from April through September when winds are predictably favorable
for measuring across freeway pollutant differences.

                    The percentage of favorable (perpendicular) wind
decreased substantially during the winter of 1976 compared with those of
previous winters.  This decrease makes interpretation of pollutant data
more difficult, since adjustments to the pollutant data are necessary to
account for the change of wind.

               b.   Effects of Wind on Gaseous Pollutants

                    Figure 4 depicts the typical effect of wind direction
on the concentration of CO obtained hourly at Sites A and C measured
over a year.  Concentrations are averages of values in 10-degree intervals
of wind direction.  The figure consists of three plots showing the
average concentrations at Sites A and C, and the average difference
(freeway contribution) between the two sites.  The marked effect of the
previously selected favorable wind direction interval on the cross-
freeway differences is shown in this figure.  Within a 10-degree tolerance,
the concentration difference between Sites A and C peak at the freeway
perpendicular direction (235 degrees) and approach zero at the parallel
directions (145. and 325 degrees) for all primary pollutants generated by
automotive emissions.

               c.   Pollutant Data

                    (1)  General - With the addition of 03 analyzers at
Sites A and C in January 1976, the increase in frequency of operation of
the 4-hour membrane and 24-hour dichotomous samplers, and elemental data
obtained by x-ray fluorescent (XRF) spectroscopy, many additional data
summaries are now available.  Figure 5 depicts the operating schedules
of the integrated sampler measurements for both the 15-to-19 and 0-to-
24-hour intervals.

                         While the number of pollutants and techniques
for presenting these pollutants are many, for the purpose of this report,
lead and sulfate have been chosen to represent the study results.
Accordingly most of the following integrated data concerns only these
two pollutants, although a number of other data have been collected.
Also the results are generally shown in figures as across-freeway differences
Site C, Site A and the difference (C-A).  Table III lists the parameters
being monitored and the status of the data base.  For more detail consult
the references.
                                   34

-------
  6.98>-





  6.28





  5.58





  4.87





  4.17





  3.47



a
a 2.76


8


  2.06




  1.36





  0.65





  -0.05





  -0.75




  -1.45
                                                                                    A.A
                                                                                    C  C
                                                      -, 6.98





                                                        6.28





                                                        5.58




                                                        4.87





                                                        4.17.





                                                        3.47





                                                        2.76"





                                                        2.06





                                                        f.36





                                                        0.65
          \ ,*
     I   I  I  T  I
                               I  I  I
                                     I  I  I J
° 2 CH PJ  * ID «0 IN CO .« S  «-'




            WIND DIRECTION
3 S S ° §  °S ?•.

 DEGREES


    K—
                                                         jj l.j.
                                                                          i   i  i  i
                                                                  S 8 g 8 || ? §
                                                        -0.05




                                                        -0.75





                                                        -1.45
                                                                                         n
                                                        FAVORABLE SAMPLING SECTOR
                     Figure 4.  Carbon monoxide by wind direction, annual composite.
                                                   35

-------
OJ
          SITES
          A&C
         SITES
         B&D

24-HOUR HI-VOl
4-HOUR HI-VOL. AND
4-HOUR MEMBRANE*
24-HOUR CASCADE
DICHOTOMOUSj
4-HOUR HI-VOU."

tA um ID IUICIUIDD A uct
24-MUUR McMBnAIMc'

«*•"•*•*.*«*"***' !_••»»•••»••. *.*•*.**.*•*•*•*•*•
	 '•'• ••-•-•-! L'^-.'.- •.••••« (•,•.•,•-•• . . .•.•.•! r. • • •^•^v.j^J
0 H K Ea ca Q ra
1 ' i i i

K3 S3 H S3 ESS ES S
'-"-'-"-"-'lilllrlillD C'.'.'.'.'.'.T.'.'f'T) • ^''vIvIvnVj

I ill
     •4-HOUR SAMPLES COLLECTED 3-7 p.m.    DAY 1
                                                        DAY 2
DAY 3
DAY 4
DAYS
                                             DAY 6
DAY 7
                                      Figure 5. LACS platform sampler schedules for integrated sampling techniques.

-------
                                            TABLE III
                             Inventory of Integrated Data from LACS
Sampling
Interval
6-10

8-12


15-19






0-24











Continu-
ous














Sampling Sampling
Method Sites
Hi-Vol B & D
Membrane A & C
Hi-Vol B & D
Membrane A & C

Hi-Vol BSD

A & C
Membrane A & C



Hi-Vol A & C
B & D

Membrane B & D



Dichotomous A & C
Cascade A & C
(Backup)

Bubbler A & C
Climet An- A
emometer
Thermistor/ A
LiCl Detector
Inductive All
Loop Lanes
System
Non-Ois- A & C
persive
Infrared
Chemilumi- A & C
nescence
Chemilumi- A & C
nescence
Flame A & C
Photometric
Sampling
Frequency
Every day
Every 3rd day
Every day
Every 3rd day

Every day

Every day
Every 3rd day


Every day
Every day
Every 3rd day

Every 3rd day



Every 2nd day
Every 3rd day


Every day
Continuous

Continuous

Continuous


Continuous


Continuous

Continuous

Continuous
~ ~j _ ft 'CM_ n .- 7*™Z

Pollutants Determined
TSP, Pb, SO!, NO!
TSP 4 3
TSP, Pb, SOa, NOZ, NHt
TSP, SO,, NO., J *
(XRF)* 4 3
TSP, Pb, so!, NO;,
NH| . : *
TSP, Pb, S04, N03, NH*
TSP
so$, NO;
(XRF)* I _ +
TSP, S04, N03, NH4, (XRF)*
TSP, Pb, SO!, NO!
NHl I 3
TSP, Pb, SO., NO,
NHJ 4 3
TS?
sot, NO;
(XftF)* 3
NH|
TSP, (XRF)*, strong acidity
TSP

(XRF)*
so2
Wind direction/speed

Temperature/dew point

Traffic count/speed


Carbon monoxide (CO)


Nitric oxide (NO)/
Nitrogen dioxide (NOJ
Ozone (0.,)

Total sulfur (S)
£ -I C« 7« rtl a nA PI
Sampling
Period
6/74-12/74
6/74-12/74
1/75- 3/76
1/75- 3/76
7/75- 3/76
6/74-present
1/75-present
3/75-present
9/74- 3/76
1/75- 3/76
7/75- 3/76
3/76-present
6/74-present
1/75-present
6/74-present
1/75-present
6/74-present
1/75-present
7/75-present
3/75-present
5/75-present
1/75-present

7/75-present
6/74-present
6/74-present

10/74-present

9/76-present


6/74-present


1/75-present

1/76-present

5/75-present

"Elemental analyses by XRF inc
                                                  37

-------
                     (2)  Four-Hour Integrated Sampler Data, 15-to-19-Hours.

                         (a)  High Volume Sampler Data.  The 15-to-19-
 hour data from the samplers indicate that the average freeway contribution
 of  total suspended participate (TSP) between Sites A and C declined
 about 40 percent since the study began (25 percent from 1975 to 1976),
 although the background TSP (average 85 yg/nr) has not significantly
 changed.  The freeway contribution of lead (Pb) during this period
 declined nearly 15 percent, although the background level (Site A) had
 not changed significantly (Figure 6).  For example, the average lead
 difference in July .975, was about 7.3 yg/nr and the corresponding value
 in  July 1976, was about 5.7, a drop of 1.6 yg/irT (t-2056).

     The sulfate (So!) data from the high volume samplers (Figure 7)
 demonstrate that the average S07 background level (Site A) decreased
 more than 25 percent from the summer of 1975 to the summer of 1976.  The
 SO, differences between Sites A and C in3the summer of 1976 averaged
 only 2.8 yg/m  as compared with_5.2 yg/m  for a similar period in 1975.
 As  a matter of interest, the SO, difference between Sites B and D, which
 operated daily like Sites A and C, remained essentially unchanged (2.8
 yg/m  in 1975).  The average values at Sites B and D also decreased more
 than 25 percent since the beginning of data collection in June 1974.

                         (b)  Membrane Sampler Data.  The membrane samplers
 use cellulose ester as the filter material as compared with glass fiber
 in  the high volume sampler filters.  For TSP, the 4-hour duplicate sample
 comparison study (10) indicated that the membrane and high volume sampler
 data exhibit a poorer correlation than was expected (^0.5).  However,
 the 15-to-19-hour membrane sampler data for TSP exhibit a decrease in
 TSP background level  similar to that indicated by the high volume
 sampler data.   Although the monthly averages of the membrane sampler
 data indicate a wider fluctuation from month to month than did the high
 volume sampler data,  the overall  monthly average at Site A is identical
 by both types  of samplers.   At Site C, the high volume sampler levels of
 TSP were approximately 10 percent higher overall than the membrane sampler
 levels.

                         The 15-to-19-hour membrane sampler data for Pb
 shown in Figure 8 correlates well (0.99) with the 15-to-19-hour high
 volume sampler lead data.   However, the membrane sampler data for Pb are
 approximately  20 percent higher than those determined by the high volume
 sampler.   Monthly levels range as high as 11 yg/m  at Site C which is
about 1.5  yg/m  higher than the maximum lead level  by the high volume
sampler.
                                 38

-------
  10.45 K
   9.41
   8.36
   7.32
 -  6.27
O  5.23
tr
H-

8  4.19

O
O
   3.14
   2.09
   1.05
   0.00
                  i  i  i  j  i
         2 .
              1974                  1975
                      A= MONTHLY AVERAGE AT SITE A
                      C= MONTHLY AVERAGE AT SITE C
                      X= MONTHLY AVERAGE DIFFERENCE, C-A
1976
                   Figure 6.  Monthly 4-hour lead from high-volume samplers.
                                                   39

-------
   29.1 L_
    26.0 -
    22.8
    19.7

U
in
(M
»'   16.6

i
z"
¥   13.5
Z*   10.4
O
U
     7.2




     4.1




     1.0
     0.0


    -2.1
                        XX.*
                                                                    X
                                                                         f.
                                                                      '*  '••
I   I  I  I   I  I  I   I   1  I  I   I  L  I   I   I  I  I   I
                                                         1111
                                                                            X

                                                                       1   I  I  '-.X
                                                                                      K >  O
                                                                                      o z  a
               1974                  1975                          1976

                       A= MONTHLY AVERAGE AT SITE A
                       C= MONTHLY AVERAGE AT SITE C
                       X= MONTHLY AVERAGE DIFFERENCE, C-A
                     Figure 7. Monthly 4-hour sulfate from high-volume samplers.
                                                40

-------
   12.32 h-
   11.08
    9.85
   8.62
u
in
CM
».*  7.39
2   6.16
Ul
U  4.93

O
U
   3.70
   2.46
   1.23
   0.00
                                             C—° X'
                                                                                        X.
          I   I  I  I   I  I  I   I   I  1   I   I  I   I   I  I   I   1  I
                                                       >
                                     I EJ s  & d ~i ti
                                                                                       Z  Q
               1974
1975
1976
                  A= MONTHLY AVERAGE AT SITE A
                  C= MONTHLY AVERAGE AT SITE C
                  X= MONTHLY AVERAGE DIFFERENCE, C-A
                       Figure 8.  Monthly 4-hour lead from membrane filters.
                                              41

-------
                         The membrane sampler data for SO* (Figure 9) at
 Site A exhibit a downward trend similar to that of the high volume
 sampler data.  The overall average of the membrane sampler data at Sites
 A  and C, however, are 30 to 39 percent lower, respectively, than those
 of the high volume sampler data.  Between Sites A and C, the overall
 concentration differences determined with the membrane samplers was 0.6
 ug/nr as compared to 3.5 yg/m  obtained with the high volume samplers.

                         The same filters which were analyzed for SOT
 by the methyl thymol blue (MTB) method were reanalyzed for total sulfur
 by XRF spectroscopy. The correlation between the two measurement methods
 is reasonably good, 0.89.  The ratio of S to SO, on a mass basis is
 0.33, which indicates that the sulfur on these filters was predominately
 in a sulfate form.

                    (3)  O-to-24-Hour Integrated Sample Data - The LACS
 samples integrated over a 24-hour period provide a poorer distinction
 than the 4-hour samples between the freeway and background pollutant
 contributions, because of the variable wind directions experienced
 during the longer sampling time.  The 24-hour values, however, do provide
 information OQ the long-term pollutant levels and data for comparison
 with other SO. studies in the Los Angeles area.

                         (a)  High Volume Sampler Data.  Freeway contri-
 bution of both the 24-hour TSP and Pb data (Figure 10) indicate decreases
 during the span of the study.  The 24-hour Pb levels at Site C decreased
 about 1.5 yg/nr (*2Q%) between 1975 and 1976 from 8.0 yg/m  to 6.5 yg/m .
 Concentration levels for 24-hour samples are approximately 25% lower
 than the levels obtained from 4-hour samples.

                         The S07 high volume sampler data on a 24-hour
 basis (Figure 11) show a downward trend of about 25% between the summer
 of  1975 and the summer of 1976.  The background level also decreased
 substantially.  The absolute values on a 24-hour basis are approximately
 25  percent lower than those of the 4-hour samples collected at the same
 site.  The freeway contribution (concentration difference between Sites
 A and C) began to show a significant positive difference of approximately
 1 yg/m  in the summer of 1976.

                         (b)  Membrane Sampler Data.  Membrane samplers
were not operated for 24 hours at Sites A or C, only at Sites B and D.
These data are discussed in another report (11).

                         (c)  Dichotomous Sampler Data.  The dichotomous
samplers exhibited aerosol  collection problems in the larger particle
ranges  and are not reported, but this caused no major problems to the
                                      42

-------
  29.9
   26.9
   23.7
   20.6
 -  17.5
O  14.4
£C
I-



O  11.3


O
O
    8.3
    S.2
    2.1
    0.0



    -1.0
                                                                                       A
                                                               X
                                                         *»     • •
                                                          •••x-  -..)c

                                                                 '-••*..*
JO.-*
I   I  I  I   I   I   I
                              I	l
                                                 I   I   1  I  I   I  I  I   I  'XL 1  I
     s  a  &'  | §  |  s I: j!  |  § § |  §  g" |  a  I s 111 §  i 11


     1974                  1975                          1976


              A- MONTHLY AVERAGE AT SITE A

              C= MONTHLY AVERAGE AT SITE C

              X- MONTHLY AVERAGE DIFFERENCE, C-A
                                                                                            a
                        Figure 9.  Monthly 4-hour sulfate from membrane filters.
                                                      43

-------
    10.06 ._
to
CM
     9.06
     8.05
     7.04
     6.04
*    5.03
z
111
O
u
     4.02
     3.02
     2.01
     1.01
     0.00
           I	i
                                                         I  II  I   l  1  l   I  1  I  I   111
                                                                                      H  >  U
                                                                                      O  O  UJ
                                                                                      O  Z  O
1974
                                     1975
                         A= MONTHLY AVERAGE AT SITE A
                         C= MONTHLY AVERAGE AT SITE C
                         X= MONTHLY AVERAGE DIFFERENCE, C-A
1976
                     Figure 10. Monthly 24-hour lead from high-volume samplers.
                                                   44

-------
   25.3
   22.9
    19.9
    17.3
 ;   14.6
O   11.9
GC
H
O
O
    8.2
    6.6
    3.9
     1.1

    0.0


     1.5
 X
/'-.X
                                         — X      X...X
J _ I  I   I   I  I   I   I  I  I   I
                                      1            1  I   I  I   I  J.  J  I  I.  <  J
               1974                  1976                          1976

                        A= MONTHLY AVERAGE AT SITE A
                        C= MONTHLY AVERAGE AT SITE C
                        X= MONTHLY AVERAGE DIFFERENCE, C-A
                   Figure 11.  Monthly 24-hour sulfate from high-volume samplers.
                                              45

-------
 study since the compounds of interest are primarily located in the small
 size range  (<2.5 ym).  The mass concentration data indicate that the
 overall fine particulate loading decreased from 1975 to 1976, and Jhe
 24-hour contribution from the freeway remained constant at 10 yg/m
 during the summer seasons.

                         The Pb data (Figure 12), determined from analysis
 of dichotomous samples, are too incomplete to judge trends.  However,
 the Pb levels in the fine size fraction are approximately two-thirds of
 the corresponding levels determined with the high volume sampler at
 Sites A and C.

                         The SO! data (Figure 13], determined with the
 dichotomous samplers, indicate rto significant S04 contribution from the
 freeway.  There is an indication that the S07 background level in the
 fine size fraction decreased slightly from 1975 to 1976.  The average
 values are approximately 30 percent less than the corresponding 24-hour
 membrane sampler value, and about 60 percent less than the 24-hour high
 volume sampler values.

                         The XRF spectroscopy data for total  sulfur
 (Figure 14) which were determined from analysis of dichotomou§ samples,
 follow the same trends as those of the high volume 24-hour SQ* data and,
 indeed, correlate well with it (0.95).  The ratio of S to SO, of the
 monthly averages is 0.28, which is not as close to the expected 0.33
 value as was exhibited by the membrane sampler data.

                         (d)  Cascade Impactor Data.   The aerosol size
 distributions are computed by season and an example is shown in Figure
 15.  The particle diameter is plotted as a function of the cumulative
 percentage less than or equal to the selected diameter.  In general,
 during the favorable summer periods, the mass median diameter (MMD) is
 smaller at Site C than that at Site A.  This relationship is to be
 expected, since automobile aerosol emissions tend to be less than 1 ym in size.

                         The backup filter, which collects the smallest
 particles (less than 0.7 ym in size), was analyzed by XRF spectroscopy
 for the same elements as were analyzed on a standard membrane sampler
 filter.   The Pb particles (Figure 16) are strongly affected by unfavorable
winds as evidenced by the relationship between Site C and A.   During  .,
months of favorable wind, Site C Pb results range between 4 and 6 yg/m
which are generally higher than Site A results which range between 0.5
and 4 yg/m .  The sylfur data (Figure 17) indicate a 0.2 yg/m  (0.6
yg/m  equivalent S0fl) contribution from the freeway traffic during the
summer seasons  in 1975 and 1976.
                                    46

-------
   6.06 i-
   5.45
   4.84
   4.24
u
ID
CM
 '.  3.63
O  3.03
1  2.42
z
O
u
   1.82
   1.21
   0.61
"A
                                            M
                                          /
            i   I   i  i  i   i  i	i  i  i   i  j  i   i  i   i  i  i   i  i  i  ,  L L
I  i  ? i i  181
             s
                                                                              s    d  S a
                                                                           ^  < M o  z a
               1974
                                    1975
                1976
                       A= MONTHLY AVERAGE AT SITE A
                       O MONTHLY AVERAGE AT SITE C
                       X= MONTHLY AVERAGE DIFFERENCE, C-A
                   Figure 12. Monthly 24-hour lead from dichotomous samplers.
                                                     47

-------
   13.2
   11.8
   10.3
u
ID
CM



i


o
Z
ID
U

O
o
   8.9
7.5
    6.1
   3.3
    1.9
0.5 -

UJ
Z
D


1 Jl A * '
* g fcVg
3 «3 uj O O
=5 < w O Z
1974


Hi 5 ui ^
0 3 u. S



of
Q.
<
19'


|
75
• (
• •
5ft V '
3 3, 3 2i U O UJ tf £
=; ^ < co oz osu


?
1
\l
19


of
Q.
<
76


<


X
UI
Z



>•
3

X
.-.
/ *-j_
i i II J.
S a" d •$' a
< w o Z-Q

                     A= MONTHLY AVERAGE AT SITE A

                     C= MONTHLY AVERAGE AT SITE C

                     X= MONTHLY AVERAGE DIFFERENCE, C-A
                  Figure 13.  Monthly 24-hour sulfate from dichotomous samplers.
                                            48

-------
   3.55
   3.19
   2.83
   2.47
o

8  2.11
O
GC


u
U

O
U
   1.38
   1.02
   0.66
   0.30
  -0.07
                                                                               *•    ••
                                             i*Xr i  i
                                                                       J  1   L  L  L

            i §  si t £  a  1 8 5  E 5 I  S g t  H > o 2
                                                                            -»  < 'co O  Z Q
               1974                  1975


                       A= MONTHLY AVERAGE AT SITE A"   	
                       C= MONTHLY AVERAGE AT SITE C
                       X' MONTHLY AVERAGE DIFFERENCE, C-A
                                                                  1976
                Figure 14. Monthly 24-hour total sulfur from dichotomous samplers.
                                              49

-------
         SUMMER 1976
             SITE A
             SITE C
                                                                ill
0.1
   12      5     10      20   30  40  50   60  70   80      90    95    98   99
                MASS < PARTICLE DIAMETER, cumulative percent
          Figure 15. Cascade impactor results for summer I976, by site.
                                            50

-------
   6.30 p
   5.67
   5.04
   4.41
o
in
CM
 - 3.78
z

2  3.25
tr
t-
z

O  2.62


O
O
    1.99
   1.26
   0.63
                                                      9. A     A
0.001   I   I  I   I   I   I   I  I   I   I   I  I  I   I   I  I  I   I   I  I  I   I   I  I   I   I   I  I  I


      1  §  §  £  d §' a  | 2 S  £  <  § § i  £  d  &*'  a i  ss  S  c ^  1 §  S  ^^  d
      -9  =3  <  V)  OZQ-4U.S  <  S  -5 -5 <  MOZ  Q-»U.S
-------
o
U)
CM
   2.53 i-
   2.27  -
   2.00
   1.73
   1.46
z
O
P  1.19
<
cc.
1-
z
Ul
O
U
   0.92
   0.65
   0.39
0.12
0.00
-0.15
~
X
1 I 1 1 1 1 1 I 1 1 1 I 1 1 1 1 1 1 I 1 1 1 1 1 I 1 I 1 1 1 1
               1974                  1975


                        A= MONTHLY AVERAGE AT SITE A

                        C= MONTHLY AVERAGE AT SITE C

                        X= MONTHLY AVERAGE DIFFERENCE, C-A
1976
        Figure 17.  Monthly 24-hour sulfur particles ^0.7 /zm as determined by cascade impactor.
                                                    52

-------
                         To examine the distribution of S07 and Pb
across the cascade size range, composite analyses were performed com-
bining stages for a number of samples.  To minimize the sample size, 35
cascade sets in which all stx stages contained valid data were selected
from the data for each of Sites A and C for the summer of 1975.  From
the analyses of these samples, weighted averages were calculated.  From
these data, cumulative percentages were determined and size distributions
were plotted.

                         The Pb data (Figure 18) indicate similar distri-
butions at both sites, but the particles collected at Site C were smaller
than those obtained at Site A.  The plots of S07 data (Figure 19) collected
at both sites again indicate a similar curvature as that determined
previously and, within the experimental uncertainty of the method, do
not indicate a difference in median size.

                         (e)  SOp Bubbler Sampler Data.   Figure 20
presents 24-hour S02 data obtained with bubbler samplers and indicates
that a significant change did not occur in the SCL background level and
only a slight decrease was experienced in the freeway contribution.  Few
monthly averages exceed 26 vig/m.

                    (4)  Hourly Averages of Continuous Data - Hourly
averages were derived from measurements of the pollutants -- CO, NO,
N02, 03, and total sulfur -- and the ancillary measurements of wind
speed and direction, ambient temperature and relative humidity, and
traffic count and speed in each lane.  The wind data have already been
discussed.

                         (a)  Pollutant Data.  The summary data by month
for the continuously monitored pollutants are presented in Figures 21
through 25.  Because of the interrelationship between the wind and
pollutants, the pollutant data were analyzed according to the two seasons,
winter and summer.

                         The CO plot indicates no apparent change in
background level, but a decrease of approximately 20 percent in the
cross-freeway difference between Sites A and C from the summer of 1975
to the summer of 1976. The NO data shown in Figure 22 is difficult to
interpret because of the erratic nature of the plot, but the dramatic
increase of NO at Site C in September 1976 is apparent and unexplainable.
According to quality control records, the instrument used to monitor for
NO was operating properly during this month.  In Figure 23, the NO^ data
collected at Site C during September also shows a similar peak, but
neither the CO nor the Oo data (Figure 25) suggest any reason for the
peak readings. The N02 plot is also erratic and somewhat difficult to
                                         53

-------
   10.0


    8.0


    6.0




    4.0
£   2.0
a.

tc
LU

LU
5


5   1.0
LU

U   0.8
tr
<
0.
    0.6
    0.4
    0.2
    0.1
         —    SUMMER 1975
                              SITE A Pb
                                                               SITE C Pb
        12     5   10    20   30 40  50 60  70  80    90   95   98  99


                       MASS .£. PARTICLE DIAMETER, cumulative percent




                        Figure 18. 24-hour cascade composites, lead.
                                                                                99.9
                                            54

-------
10.0


 8.0



 6.0




 4.0
E
3.

OC
UJ

UJ
5


5
UJ
_l
U

OC
 zo
 1.0
 0.8
 0.6
 0.4
 0.2
 0.1
        	    SUMMER 1975
                                              SITE A S04
                   SITE C SO4
        I     I     I      I    I    I   I   I   I    I     I    I     I   I     I
     12    5    10    20  30  40 50 60 70   80   90  95    98  99



                    MASS < PARTICLE DIAMETER, cumulative percent
                                                                               99.9
                    Figure 19.  24-hour cascade composites, sulfate.
                                             55

-------
o
in
tM
   40
   36
   32
   28
   24
z
O  20
III
o  16

O
O
   12
r\
             1974                1975                        1976

                    A= MONTHLY AVERAGE AT SITE A
                    C= MONTHLY AVERAGE AT SITE C
                    X= MONTHLY AVERAGE DIFFERENCE, C-A
                        Figure 20. Monthly 24-hour sulfur dioxide levels.
                                              56

-------
   6.2 r-
	

o
   5.6
   5.0
   4.3
   3.7
   3.1
LU
u

I  2-4
   1.9
   1.2
   0.6
   0.0
   -0.1
I   I
         UJ >  •
         2 J "
         3 3 3
              1974                  1975

                      A= MONTHLY AVERAGE AT SITE A
                      O MONTHLY AVERAGE AT SITE C
                      X= MONTHLY AVERAGE DIFFERENCE, C-A
                                                                  1976
                        Figure 21.  Monthly hourly averages of carbon monoxide.
                                                 57

-------
   0.72
   0.65
   0.57
   0.50
a  0.43


5
a  0.36
ai
U


8  0.29
   0.21
   0.14
   0.06
0.00
-0.0
                       i TT  r
JL
                                                        '   '
               S!
            <  w

             1974
                       O  uJ
                       z  a
as  = of >
ui  S °- 3


      1975
s
                      a


                                                                              §
                                                                      1976
                       A= MONTHLY AVERAGE AT SITE A

                       C= MONTHLY AVERAGE AT SITE C

                       X= MONTHLY AVERAGE DIFFERENCE, C-A
                             Figure 22. Monthly hourly averages of nitric oxide.
                                                                                           a
                                                       58

-------
   0.11
   0.09
   0.08
   0.07
   0.06
o

<  0.05
cc


ui
O  0.04
U
   0.03
   0.02
   0.00


1 1 1 1 1 1 1
Z 5 g t £ > 0
3 3 •< w O Z Q
1974
* •
X
I 1 1


I 1 I I


1 1 1
2" s < 5 < i 5 3 S s
^ u. 5

A= MONTHLY
C= MONTHLY
X= MONTHLY
s < 2 ^ ^
1975
< W O

.
•
X | | | i 1 i i
5 a i s 5 £^ I !
ZQ^U.S 0
3 3 Si 0 0 lB
3 < w o z a

AVERAGE AT SITE A
AVERAGE AT SITE C
AVERAGE DIFFERENCE
,C-A

                             Figure 23.  Monthly hourly averages of nitrogen dioxide.
                                                          59

-------
0.053 r-
0.044  -
0.035 -
0.026 -
0.017 -
                                                                           A  A
CONCENTRATION, ppm
0.008
0.001
-0.012
-0.021
\/ C V c^c
c c-° CN:
X
X •
'••x •
                                                                     •x
-0.030 -
-0.039 -
-0.048
        J	I
                 I	I
                   J	I
                                                                I  1
                                  I   I  X  I   I
ui  >
Z  -I
D  r>
C  H >
ui  U O
CO  O Z
                              OQ
                              UJ
5  a
I  5
,.  ui >.
>;  z j
£33
|  §  |
                                                 CO
a  I
Q  3
5of>:zjollH>cj
^iS333wa2"i
                                                                                   CO  O  Z  Q
      1974
                                    1975
                                                                  1976
                      A= MONTHLY AVERAGE AT SITE A

                      C= MONTHLY AVERAGE AT SITE C

                      X= MONTHLY AVERAGE DIFFERENCE, C-A
                           Figure 24. Monthly hourly averages of ozone.
                                                 60

-------
   64
   57
   51
   44
 I  37
Z
2  31
O  24

O
O
   19
   11
          J	i
                                                              c_c
                                                                               •x
                             111
                I  I
                                     I  I
i
<
1974
d  S a     s'
o  z D

s. s <  s
     1975
                i

                                                  d g  tf |  s 1 c  <      S
                                                  OZQ^U!S d
O ui
z o
                     A= MONTHLY AVERAGE AT SITE A
                     C= MONTHLY AVERAGE AT SITE C
                     X= MONTHLY AVERAGE DIFFERENCE, C-A
                        Figure 25. Monthly hourly averages of total sulfur.
                                                 61

-------
 interpret, but, in general, it indicates that the freeway contribution
 increased from the summer of 1975 to the summer of 1976, while the
 background level remained unchanged.  Since CL monitoring did not begin
 until January 1976, sufficient data are not available to describe a
 trend.

                         The total sulfur data shown in Figure 25 indicate
 a decrease from 1975 to 1976.  The agreement between monthly averages of
 the total sulfur and S0? bubbler sampler data is poor, partly because of
 the low concentrations and, also, because of sampler and calibration
 problems with the total sulfur analyzer at Site C during 1976.

                         (b)  Traffic Data.  Data of traffic speed and
 number of vehicles on each of the eight lanes of the freeway have been
 collected since September 1976.  After examining the traffic data, the
 University of Wisconsin, under Contract Number 68-02-2261 (12), suggested
 that the traffic count (in cars per hour) be divided by the average
 speed (in miles per hour) to obtain the traffic density or number of
 vehicles per mile, which can be compared with published data.

                         Figures 26 and 27 show the traffic count composited
 over the 4 lanes for a typical week- and weekend-day by direction (four
 lanes each way).  The corresponding traffic densities are shown also.
 On any week day, although the traffic count southbound during the morning
 rush hour was approximately equal to the traffic count northbound in the
 evening rush hour, the northbound traffic density was nearly twice as
 great as that southbound. This is caused in part by the slight upgrade
 in the northbound direction, which reduced the average speed substantially.
 The weekend plot indicates a much lower traffic volume with little
 indication of peak traffic counts or densities.   To assess the change in
 traffic flow over the previous years, data were obtained from two traffic
 counting sites on the San Diego Freeway near the LACS traffic counting
 site.  These data are shown in Table IV.  The former two sites were
 checked only for traffic count by the California Transportation Department.
 At the Wilshire Boulevard site, traffic was counted routinely.  However,
 at the Montana Avenue site, traffic was counted randomly and the yearly
 total was estimated using the random data.  In view of these data, the
 total traffic volume evidently has not changed significantly since 1972.
 Due to the constancy of these data, the limited LACS traffic data were
 extrapolated and assumed to represent the hourly and daily traffic
 patterns since the beginning of the study.

                         In Los Angeles County, the percentage of registered
automobiles equipped with catalytic converters has been compiled quarterly
by Rockwell under Contract Numbers 68-02-1766 and 68-02-2292.  According
to these data (Figure 28), 15 percent of the automobiles registered by
the end of 1976 were equipped with catalytic converters.  A recent draft
report prepared by TRW (13) for the California Air Resources Board proposes
                                          62

-------
                               TABLE IV

    San  Diego  Freeway Traffic Flow - Dally Average Traffic Count2
Counter .
Location '
Wilshire Blvd.
Montana Ave.
Year
1972
191,000
172,000
1973
192,000
173,000
1974
189,000
171,000
1975
195,000
170,000
1976
189,000
177,000
JData collected by the California Department of Transportation


3LACS site is located approximately half-way between Wilshire Blvd.  and
 Montana Ave. on San Diego Freeway


^Wilshire Blvd. traffic numbers are actual counts made at a count station
 just south of the interchange.  Montana Ave. numbers are estimates.
                                       63

-------
 0.00   2.00   4.00   6.00
aOO   10.00   12.00  14.00  16.00  18.00   20.00  22.00  24.00
        TIME, MONDAY
0.00   2.00   4.00   6.00    8.00   10.00   12.00  14.00  16.00  18.00   20.00  22.00  24.00
                                   TIME, MONDAY
      •SOUTH-BOUND LANES
       NORTH-BOUND LANES

                 Figure 26. LACS traffic data: daily composite for a Monday.
                                         64

-------
   8
   o'
   ix

   8
   §'

o  §
I §
co 6-
ui •*
o
I g
SJ *

t g
CO

UI
Q
                             TRAFFIC DENSITY
     0.00  2.00   4.00   6:00   8:00
      10.00  12.00   14:00.

        TIME, SUNDAY
16.00  18.00  20.00  22.00 24.00
                             TRAFFIC COUNT
    0.00   2.00   4.00   6.00
8.00  10.00   12.00  14.00   16.00  18.00  20.00  22.00 24.00

        TIME, SUNDAY
                SOUTH-BOUND LANES

               • NORTH-BOUND LANES


                  Figure 27.  LACS traffic data: daily composite for a Sunday.
                                                   65

-------
    B - ADJUSTED FOR ESTIMATED MILES DRIVEN

    X - REGISTRATION DATA
                                            \
                                              h
                                                 h
                                                   V
                                                     V
                                                       \
                                                                     -'  JAN, 1976
                                                         \
                                                           \
                                                             \
                                                               \     -
      I I I  I  I  I  i i  i  i  i  I  i  i  i I i  I  i  i  I  i  i i i  i  i  i  I  i i  i  i  i  (
              30       25        20        15       10        5

                              % CATALYST CARS
                                                                          JAN. 1977
                                                                          JUNE 1976
- .  JUNE 1975
                                                                      - -  JAN. 1975
                                                                          SEPT. 1974
Figure 28.  Percent of registered automobiles with catalytic converters in Los Angeles County, 1976.
                                                66

-------
areater thanP?hP  rt^ ?lleskdr1ven  by  new  automobiles  is substantially
              H.n    Ja1  K-Tber  W0uld 1ndic*te.   For instance, in 1975,
              Jh  }  autoi?3blle  was  estimated to have been driven 2.05
             ™an  *" °lder  automob1le-   A  revised plot based on the
            Q7*W  est1lmates>  also s^wn  in Figure  28, indicates that at
«,  «        Seafly one-third of  the vehicle miles traveled were with
further   converters-   The validity  of  the  TRW estimate will be investigated


                     (5)   Ancillary Data

,      .                   (a)   Dustfall  Data.  The collection of dustfall  data
began in late 1975  and has resulted  in  information on the larger particles
(>100 yin) found at  the four  sites.   The mass and  the Pb, in general, follow
the spatial patterns  of the  high volume sampler monthly averages.   However,
the Pb data show  higher values at  Site  C  as compared with those at
Site B than does  the  data for  mass.  Presumably,  traffic-generated particles
larger than 100 ym  in size would not travel  substantial distances, and pre-
vailing wind direction together  would account for the observed trends of
mass.  The SO., NOZ,  and NH. data  show  larger values at Site A compared with
the other sites.  This is probably attributable to the dustiness of the open
field, which was  plowed routinely  for weed  control near Site A.

                          (b)   Percentage  of Collected Data.   The percentages
of valid data by  sampler type  are  presented in Table V,  These percentages
have not changed  substantially since the  beginning of the study.

               d.    Discussion

                     The validated  LACS data set is voluminous and  contains
types of data which permit a number  of complex data analyses, some of
which may continue  for several years.   In terms of trend analyses  on a
year-to-year basis, however, only  3  years of data or three data points
for each of two seasons  are  available.  Hence, trend predictions  are
presently somewhat  uncertain,  although there appears to be significant
changes occurring in  some of the pollutant  levels.  The University of
Wisconsin, under  Contract Number 68-02-2261, is examining the LACS data to
identify and quantify data trends.

                     Of the three monitored  pollutants that should  be
directly affected by  the catalyst  — CO,  Pb, and SOT — all  show some
indication of a change in freeway  contribution.  Th? levels of CO  and Pb
decreased approximately 25 percent since  1974, with most of the decreases
occurring between 1975 and 1976.   These decreases are consistent with
the estimated increase in the  number of automobiles equipped with  catalytic
converters and the  estimated corresponding  increase in unleaded fuel
consumption.  The three  integrated sampling methods that used high
volume, membrane, and  dichotomous  samplers  have shown that the S04
background level decreased on  a  24-§our basis.  The 24-hour background
measurements decreased about 1 yg/m3 each year since 1974  but the
corresponding 4-hour  measurements  decreased  about 1.5 yg/m  during the
same period.
                                        67

-------
                                                                              TABLE V


                                                                           Percent Data

Method
Interval
June 1974
July
August
September
October
November
December
Jan. 1975
February
March
April
May
June
July
August
September
October
November
December
Jan. 1976
February
March
April
May
June
July
August
September
October
November
December
Overall
Wind
Speed
Hourly
99.4
33.3
86.8
98.5
90.7
93.8
88.7
100.0
97.9
94.5
98.1
99.3
90.4
100.0
99.7
88.7
69.8
83.5


67.2
98.7
98.1
98.1
100.0
83.9
93.5
82.3
96.8
96.7
2.4
87.3
Wind
Direction
Hourly
100.0
58.1
93.3
98.3
99.9
96.2
73.1
100.0
99.9
93.0
97.6
99.5
97.1
97.3
99.7
93.8
70.8
86.9


65.5
100.0
97.9
98.3
99.9
83.9
92.6
81.1
92.7
97.6
43.3
90.3

Hi-Vol
4-Hour
96.7
89.5
100.0
91.7
96.0
100.0
100.0
94.2
76.2
92.5
92.2
94.6
97.2
98.9
94.6
92.8
96.2
100.0

35.5
97.1
83.9
63.9
61.8
62.8
63.4
64.0
62.8
62.4
64.4
32.8
80.2

Hi-Vol
24-Hour
97.5
92.7
96.3
92.5
100.0
84.4
77.1
100.0
89.2
97.6
91-2
95.1
92.5
100.0
90.2
90.0
100.0
100.0

32.9
90.8
100.0
90.0
91.5
97.5
91.5
97.6
96.2
95.1
91.2
47.6
90.4
i 	 1
Membrane
	 _, 	 __
4-Hour
	

20.0
97.5
95.0
34.4
79.2
100.0
94.4
85.0
100.0
100.0
95.0
100.0
100.0
100.0
100.0
100.0


50.0
100.0
100.0
100.0
100.0
100.0
100.0
95.0
100.0
100.0
67.5
100.0

Membrane
24-Hour
80.0
90.0
10.0
20.0
60.0
87.5
83.3
100.0
100.0
100.0
95.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0


56.6
100.0
100.0
100.0
95.0
100.0
100.0
100.0
95.0
90.0
35.0
88.7
S02
Bubbler
24-Hour
56.7
98.4
90.3
91.7
98.4
100.0
61.1
100.0
89.3
96.8
100.0
90.3
98.3
87.1
98.4
93.3
98.4
100.0

37.1
86.2
95.2
98.3
100.0
90.0
93.5
93.5
96.7
83.9
96.7
50.0
89.3

CO
Hourly
73.5
82.7
fc3.6
97.8
86.6
86.7
73.7
100.0
79.5
62.2
96.0
90.3
94.7
98.1
97.6
93.1
61.0
96.1

61.5
87.1
95.0
95.8
95.6
61.7
31.6
72.2
96.0
95.2
99.9
45.8
82.9
[ 	
NO
Hourly







18.2
37.6
96.8
85.8
82.2
90.3
95.9
57.9
96.5
69.0
100.0

58.9
91.7
97.0
95.1
94.3
71.7
31.5
96.6
96.7
100.0
98.3
51.3
79.6

-N02
	 *. 	
Hourly







77.7
77.0
98.1
87.8
81.6
90.6
95.3
89.0
95.3
69.4
100.0

60.1
92.2
97.4
94.9
93.8
71.5
31.4
97.2
93.1
98.7
100.0
51.3
84.4

S
Hourly











89.0
83.7
84.7
85.8
92.5
51.2
26.0


77.6
45.4
84.8
47.5
57.8
31.9
96.4
83.1
96.8
96.7
52.3
71.8

Ozone
Hourly



















41.0
94.0
96.0
96.0
97.2
35.0
30.0
90.0
93.0
97.0
98.0
57.0
76.0
ON
CO
            NOTE:   Percent Data = # Actual Valid Readings/* Possible Readings x 100

-------
                    Since the summer of_1975, the 4-hour membrane and
24-hour high volume sampler data for SOI show a small  (<1.0 yg/m3)
contribution from the freeway traffic. HThe 4-hour high volume sampler
data show a consistent 3 to 5 wg/m  freeway contribution over the same
period, except for the last six months of 1976 when the freeway contribution
decreased inexplicably to zero.  The diochotomous sampler data obtained
by the MTB method does not exhibit any change across the freeway.
Excluding the 4-hour high volume sampler measurements, the data obtained
by other methods indicate that sulfate was generated by the freeway
traffic.g This freeway contribution is estimated to be between zero and
0.5 iag/m  on a 24-hour basis and 0.5 and 0.7 yg/m  on a 4-hour basis.

                    The elevated values on the 4-hour high volume sampler
filters suggest artificial SOT formation of the glass fiber media.
Extensive filter pretreatment was performed on all LACS filters to
minimize undesirable reactions.  This, in conjunction with the low
levels (<26 yg/m ) of S02 present at the LACS site is believed to=provide
the best data obtainable with glass fiber filters.  The 4-hour S04
background levels obtained with the high volume samplers, however, are
45 percent greater than those determined with the membrane samplers, but
the 24-hour SO^ high volume sampler averages are only 21 percent greater
than the comparable 24-hour membrane sampler averages.  Although the
membrane and high volume sampler data do not correlate well on a mass
basis, a comparison of data for Pb pollutant, which is composed mostly
of small particles similar to those of SO*, shows fairly good correlation
between the methods.  This would indicate that the two sampler types are
reasonably equivalent in terms of aerosol sampling capabilities for
small particles.  Further details are available elsewhere (11,14,15).

     C.   Quality Assurance Activities

          EMSL serves as an external auditor of monitoring activities
performed within the Catalyst Research Program.  Within EMSL, the LACS
monitoring program is continuously audited through blind samples and
sample splits.  During the year, intermittent audits were conducted of
CO, S09, THC, NO , and ozone used in the animal exposure chambers used
by HERC/Cincinnati.  Similar audits were performed of S02, ozone, and CO
at HERL/RTP as used in their human exposure chambers,  me purpose of
these audits is to verify exposure levels.

          A split sample analysis quality assurance program was initiated in
July of 1974 when collection of air samples began in LACS.  Aliquots of
the air samples are analyzed by the contractor, Rockwell Air Monitoring
Center (RAMC), and EMSL.  The split sample program was begun to determine
the comparability between analyses performed by RAMC and EMSL and to
determine analytical variability on real air samples.
                                      69

-------
          In November 1974, a blind audit quality assurance program was
initiated for RAMC by EMSL.  The blind audit program was initiated to
investigate and document the precision and accuracy achieved by RMAC
during the analysis of ambient air samples collected in LACS.

          Both the blind audit and split sample program have changed
since their beginnings in 1974 to accomodate fluctuations in the number
of samples collected in LACS, variability of the analytical results, and
EMSL's growing experience in conducting external audit programs.  Table
VI gives the pollutants, types of audit sample and frequencies of both
the blind audit and split sample programs currently followed.

          Sulfate and nitrate are spiked on the same filter for the
blind audit samples; lead is spiked on a separate filter.  RAMC then
extracts and analyzes these samples as they would a regular sample.
Sulfur dioxide blind samples are special freeze-dried tetrachloromer-
curate-sulfite solutions.  RAMC thoroughly rinses the audit sample from
a sealed vial with absorbing reagent before analyzing as a normal  50 ml
bubbler sample.

          The main use of the blind audit results is to document Chrono-
logically the precision and accuracy achieved by RAMC.   A second major
use of the blind audit data is to determine if any portions of the
analytical methods employed by RAMC need improvement.  The third use of
the blind audit results is to point out a large bias or variability that
might occur across all levels for a pollutant during an analytical
period.  If this situation occurs, both the EPA project officer and RAMC
are immediately notified.  This situation has never occurred during the
blind audit program.

          Since the blind audit samples are not real air samples,  but
synthetic samples, the split sample program serves to show variabilities
encountered on analysis of real samples.  Secondly, the split sample
results serve as a measure of the reproducibility on real samples  between
two laboratories which are using the same analytical methods.  Initially,
the split sample program was conducted at a much higher analysis rate
than is shown in Table VI but the larger time delay between analyses,
the information obtained in the blind audit results, and the continuing
comparability between EMSL and RAMC, allowed EMSL to cut back to the
current split sample rate.

          Summary reports for both blind and split sample analysis are
periodically made to the EPA project officer and to RAMC (16,17).   Split
sample results are reported at the end of each quarter, while blind
audit results are reported at twenty-five week intervals.
                                      70

-------
                               TABLE VI
               Blind Audit and Split Sample Frequencies
Pollutant
Sulfate
Nitrate
Lead
Sulfur Dioxide
Sample Type
Hi-Vol Filter
Hi-Vol Filter
Hi-Vol Filter
Bubbler
                                                     Audit Frequency
Blind Audit      Split Sample
  10/week
  10/week
   7/week
6/month
6/month
6/month
  10/bi-weekly     None
                                         71

-------
          The external quality assurance program consisting of split and
blind sample analyses has been useful 1n improving, verifying, and quanti-
fying the quality of data obtained in LACS.  Both the lead and sulfate
analytical procedures have been improved because of the results obtained 1n
the external audit program.  Detection and correction of analytical problems
has been timely enough to avoid invalidating large blocks of analytical
data.  Besides upgrading analytical  methods, a chronological history of
accuracy and precision is available  from the start of this project.  Figures
29 through 32 show the agreement of  RAMC with the spike result for sulfate,
nitrate, lead, and SO,,, respectively.  With the exception of lead, RAMC
agreed with the snike value +5%.  In the case of lead, however, agreement
for 1976 was only 87% of the spiked  values.  For the split sample results,
sulfate and nitrate were within +5%  and lead ±10% (18).


IV.  PROBLEM AREAS

     A.   Fuel Collection and Analysis

          1.   Fuel samples are collected by the Regional Offices.  We
have experienced problems in obtaining fuel samples from Region IX, and
have not received samples since the  summer of 1976.  We will remedy this
situation in 1977 by hiring a contractor to obtain the samples.

          2.   There has been a reoccurring problem when attempting to
relate the analytical results of the fuels samples to the amount of fuel
consumed in the area of concern -- in particular, Southern California.
This link is vital to the emission modeling program, reviewing trends of
pollutant concentration affected by  the catalyst and interpreting impact
of control technology.

     B.   Monitoring — LACS

          Without a specific sulfuric acid monitor, inferences about
sulfuric acid emissions are based on indirect measurements, e.g., sulfate.
This problem will be solved when the state-of-the-art H2SQ. analyzers
become available.  Similarly, an ammonia monitor is neeaed, since no
satisfactory instrumentation is available.

     C.   Quality Assurance

          Problems identified through the auditing program have been
found and corrected.
                                    72

-------
    7000  T
   (1.5%)
ui
                                  SPIKE,/^/SAMPLE
           Figure 29. Comparison of RAMC blind sulfate results with spike values.
   *RESULTS EXPRESSED AS COEFFICIENT OF VARIATION IN PERCENT OF AVERAGE RESULT
                                                73

-------
 2000
(1.5%)
                                                                               2000
                                                                             (10/Ug/m3)
                                   SPIKE, pa/SAMPLE
              Figure 30. Comparison of RAMC blind nitrate results with spike values.
     * RESULTS EXPRESSED AS COEFFICIENT OF VARIATION IN PERCENT OF AVERAGE RESULT
                                            74

-------
    2000
   (2.0%)'
a.

<
K
Ul
U
<

-------
         T
     50  --
   (2.5%)
LU

0.


I
=!   25
J2  (4.0%)
oc
ui
(3
K
UJ
>
    10
  (10.0%)
                                           SPIKE, fig/SAMPLE
                 Figure 32. Comparison of RAMC blind sulfur dioxide results with spike values.
            •RESULTS EXPRESSED AS COEFFICIENT OF VARIATION IN PERCENT OF AVERAGE RESULT
                                                  76

-------
V.   PLANS FOR FUTURE RESEARCH

     A.   Fuel Collection  and Analysis

          Since manganese  (Mn)  is  being  used  by  refiners  to  replace lead
as an anti-knock  compound, effects to  the emission  control device as
well as analytical  screening  techniques  to detect presence of Mn in the
gasoline need further study.  Progress has already  started within EMSL
to measure Mn by  atomic  absorption spectroscopy  (AAS)  and steps to
standardize  the procedure  are being considered by the  American Society
of Testing and Materials (ASTM).

          Other method development and fuel data include  polycyclic
organic matter, such as  BaP and benzene.

     B.   Monitoring —  LACS

          The following  items  concern  the future research program for
LACS:

           1.  There is  a need  to continue to evaluate the impact of
catalysts  on air  quality.   This will be  increasingly important as new
catalyst-type vehicles  and additional  diesel  traffic are  added to the
traffic mix. Our present data  base and  site  are extremely well suited
for assessing effects of transportation  control  strategies on air quality
data.  Only  through a long-term monitoring study under appropriate
quality control  can theoretical and laboratory findings be verified from
which  regulatory  decisions should be based.  At least two more summer
seasons of data  are necessary to establish a  reliable trend  using present
sampling  technology.

           2     The addition of a sensitive, reliable continuous sulfuric
acid analyzer will be made when available to  verify our off-roadway
estimates of ambient H9SOd levels.  Assuming  an acceptable IJ2S04 ana yzer
will be  available in the rtear future, two year's of summer data should
be gathered  to  demonstrate the relationship between sulfate  measurements
 and sulfuric acid aerosol.

           3    There is a need for additional in-roadway measurements
 taken at the fixed site collected  in the median as  well as mobile  sampling
 abound the  LACS area.   Knowledge of spatial variability together with
 our temperal variability  studies will provide us the means  from which
 theoretical  population  exposure estimates can be verified   In  addition,
 in-vehicular concentrations should be determined as needed.

           4   There is a definite  need  to quantify the  amount of sulfate
 artifact formation  on filters  used  to monitor sulfate at the LACS site,
 especially  for 4-hour measurements.
                                      77

-------
          5.   It is intended that time series models will be developed
to concurrently employ data from all sites to estimate spatial and
temporal variability across the freeway.   This will include fixed in-roadway
and mobile monitoring data.

          6.   The present data base contains information in examining a
variety of other equally important issues associated with transportation
control; e.g. (a) establishment of the proposed Pb standard; (b) input
to the formulation of an oxidant control  strategy; (c) examination of
the cross-freeway relationships of the oxides of nitrogen and ozone; and
(d) verification of the relationships between auto emission measurements
and actual ambient levels at the roadside.  It should be noted that an
additional" two summer seasons of data will greatly augment this data base.

          7.   At present, EPA is concerned with environmental insult
due to toxic substances, some of which are emitted from mobile sources,
such as polynuclear aromatics, benzene, and BaP.  Maintaining surveillance
of air quality in proximity of major roadways could serve to identify
and determine trends of air quality due to pollutants from mobile sources
for the purpose of determining the need for and effect of control measures.

     C.   Quality Assurance

          Ongoing research will be continued.  Auditing of animal
studies conducted at HERL/RTP will begin  in 1977.


IV.  CONCLUSIONS

     A.   Fuel Collection and Analysis

          The following observations are  made based on fuels analysis
research supported by the catalyst program.

          1.   The content of sulfur in unleaded gasoline 1n Southern
California is approximately twice that found in New York samples and the
content of sulfur in the 3 grades of gasoline is consistently higher in
Los Angeles than San Francisco.

          2.   Sulfur content in Los Angeles appears to be decreasing
within grades, but the range of measured  levels is narrowing.

          3.   Percent aromatic content is increasing, with no lead
gasoline grade higher than other grades.

          4.   Method development and standardization for analysis of
gasoline for a number of constituents including sulfur, phosphorus and
lead has been realized.
                                      78

-------
     B.   Monitoring — LACS

          The following observations are made following the analysis of
the body of information collected from th.e monitoring effort at LACS.

          1.   Based on the across freeway difference, the use of the
catalyst and unleaded fuel have resulted in substantial decreases (25%) in
CO, Pb, and TSP while the background levels remained essentially consistent.

          2.   The freeway contributions of so! have Increased less than
10% while the background levels decreased.

          3.   Oxides of nitrogen over background levels have increased
50% from 38 yg/nT to 56 pg/nT from 1975 to 1976.

          4.   Lead is significantly higher on the weekends (^60%)
despite a reduction of traffic volume (^20%).

          5.   The 4-hour evening peak traffic period from 1500 to 1900
hours is a favorable period in which to detect cross-freeway differences.

          6.   The sulfate loading_on 4-hour high volume samples is
strongly influenced by artifact SOI formation.

          7.   Traffic patterns are markedly different between weekdays
and weekends.  Friday has the highest volume of traffic (^210,000 vehicles/
day compared to ^190,000 vehicles/day during other weekdays).

          8.   Traffic density (count/speed) is a better descriptor of
emissions for continuous pollutant empirical modelling purposes.

          9.   At the end of 1976, the estimated percentage of catalyst-
equipped cars may be as high as 30%.

     C.   Quality Assurance

          Through a carefully controlled quality assurance program consisting
of external audits and split samples, an extensive body of ambient roadside
data of known quality has been generated.
                                    79

-------
                  APPENDIX A-l
PRECISION OF LACS SAMPLING AND ANALYTICAL METHODS
                         80

-------
      Two special studies have been conducted at the Los Angeles Catalyst
 Study sites fn an effort to identify and quantify the major sources of
 error inherent In the sampling and analytical methods employed.*

      First, co-located high volume samplers were operated concurrently
 for twelve 24-hour periods at a location 700 feet upwind from the freeway.
 Following equilibration and weighing, each exposed filter was analyzed in
 duplicate (using center and end strips) for concentrations of lead,
 ammonium ion, nitrate ion and sulfate ion by each of two laboratories.
 The results were used to construct estimates of the amount of variability
 (i.e., "variance component") between samplers, between laboratories and
 within laboratory.  In addition, effects of strip position (center versus
 end) and the time lag between sample collection and chemical  analysis  on
 reported pollutant concentration were estimated.

      The second study consisted of operating five samplers (two high
 volume samplers with glass fiber filters, two membrane samplers with
 cellulose ester filters and a single membrane sampler with teflon filters)
 at each of two sites (upwind and downwind) on a 4-hour interval (3-7 p.m.)
 for twelve days.  In addition to final weight, each filter was analyzed
 in duplicate for sulfate ion concentration by each of two laboratories.
 Estimates were obtained for the between sampler, between laboratory and
 within laboratory components of variance.  Finally, the effects of site
 and method on reported pollutant concentrations were investigated.

      The major results drawn for the 24-hour high volume sampler study
 are as follows:

       (1)  Sample collection from co-located instruments is not a
           significant source of error (<5%).

       (2)  Laboratory variability (same filter) is significant except
           for SO^ (0%)**:  Pb (7.5%), NH^ (38.2%), and H0~3 (3.8%).

       (3)  Variability among filter strips is the largest source of
           error for the above pollutants:  Pb (17.2%), NH4 (11.2%),
           N0~ (6.2%), and SO^ (3.1%).

       (4)  Variability among chemical analyses within a laboratory is
           relatively minor for all pollutants:  Pb (2.4%), NH4 (3.2%),
           N0~ (1.7%), and SO^ (2.3%).

       (5)  Strips cut from the center of filters tend to be slightly
           (<5%) higher in concentration than those cut from the ends.
 *A report is available from the Office of the Director, EMSL/RTP
**Percentages are based on coefficient of variation.
                                      81

-------
     (6)  The time lag between sample collection and chemical analysis
          1s a serious source of error in the determination of NH^.

     (7)  Total measurement variability is 18.9%, 7.5%, 3.9%, for Pb,
          NO^, and SO^, respectively.


     The major conclusions drawn from the 4-hour high volume and
membrane sampler study are as follows:

     (1)  The overall  TSP precision for 4-hour high volume samples is
          only slightly greater than that for 24-hour samples (5.6% versus
          5.2%).

     (2)  Though sample collection and laboratory are significant sources
          of error for SQ7 determination from 4-hour high volume samples,
          the overall  SOT precision is only slightly greater than that
          for 24-hour samples (5.9% versus 3.9%).

     (3)  The overall  TSP precision for 4-hour membrane samples  (cellulose
          ester filters) is approximately three times that for 4-hour
          high volume samples (14.9% versus 5.1
     (4)  The overall  SOT precision for 4-hour membrane samples (cellulose
          ester filters)  is approximately three times  that for 4-hour
          high volume  samples (17.6% versus  5.!
     (5)  TSP concentrations appear to be dependent upon filter material
          employed (i.e., cellulose ester or teflon),  though  not
          necessarily upon sampler type (i.e.,  high volume  or membrane).

     (6)  SOT concentrations are very dependent upon filter material
          employed with the following ranking in the order  of descending
          concentration:   high volume glass  fiber,  membrane cellulose
          ester, membrane teflon.
                                    82

-------
                              REFERENCES

 1.   Los Angeles Catalyst Study, Quarterly Report, August 1974.  Internal
     Report, Environmental Monitoring and Support Laboratory, Research
     Triangle Park, North Carolina  27711

 2.   Ibid, December 1974.

 3.   Ibid, April 1975.

 4.   Ibid, June 1974 - December 1975.

 5.   Ibid, June 1974 - December 1976.

 6.   Jungers, R.H.  "Fuels Surveillance in Southern California —  Analytical
     Results and Methodology" in Proceedings of the Los Angeles Catalyst
     Study Symposium, April 12-13, 1977.  EPA-600/4-77-034.

 7.   Clay, D.A., C.H. Rogers, and R.H. Jungers.  "Determination of Total
     Sulfur in Gasoline by Gas Chromatography with a Flame Photometric
     Detector," Analytical Chemistry 49, page 126, January 1977.

 8.   Driscoll, D., D. Clay, and R. Jungers.  "Graphite Furnace Screening
     Method for Determination of Phosphorous in Gasoline.   Internal  Report,
     Source Fuels and Molecular Chemistry Section, Analytical Chemistry
     Branch, Environmental Monitoring and Support Laboratory, Research
     Triangle Park, North Carolina  27711.

 9.   Rodes, C.E., B.E. Edmonds, and M.C. Wilkins.  "The Quality Assurance
     Program Employed During Sampling" in Proceedings of the Los Angeles
     Catalyst Study Symposium, April 12-13, 1977.  EPA-600/4-77-034.

10.   Evans, Gary.  "Precision of Sampling and Analytical Methods"  in
     Proceedings of the Los Angeles Catalyst Study Symposium, April  12-13,
     1977.  EPA-600/4-77-034.

11.   Rodes, C.E. and G. Evans.  "Summary of Integrated Pollutant Data" in
     Proceedings of the Los Angeles Catalyst Study Symposium, April  12-13,
     1977.  EPA-600/4-77-034.

12.   Tiao, G.C. and S.C. Hi Timer.  "Statistical Analysis of the Los  Angeles
     Catalyst Study Data -- Rationale and Findings" in Proceedings of the
     Los Angeles Catalyst Study Symposium, April 12-13, 1977.  EPA-600/4-77-034.

13   Parry, E.P., R.A. Meyer, and C.E. Rodes.  "Determination of Percentage
     of Diesel Trucks and Catalyst-Equipped Cars" in Proceedings of the
     Los Angeles Catalyst Study Symposium, April 12-13, 1977.  EPA-600/4-77-034.
                                     83

-------
14.   Rodes, C.E.  and G.  Evans.   "Summary of Continuous  Pollutant Data"
     in Proceedings of the Los  Angeles  Catalyst Study Symposium, April
     12-13, 1977.   EPA-600/4-77-034.

15.   Rodes, C.E.   "The Los Angeles  Catalyst Study and the Related Fuels
     Surveillance  Program in Southern California" in  Proceedings of the
     Catalyst Research Program  Sulfuric Acid Research Review Conference
     January 31-February 3, 1977, Health Effects Research Laboratory,
     Research Triangle Park, N.C.   27711.

16.   U.S.  Environmental  Protection  Agency,  "Second Annual  Catalyst
     Research Program Report, Supplement II,"   Research  Triangle Park,
     N.C.   27711,  January 1977.

17.   Evans, E.G.   "Split Sample Results of  the  Los Angeles Catalyst Study
     (LACS) and the Community Health  Air Monitoring Program (CHAMP),
     January through August 1976,"  Internal  Report, Environmental
     Monitoring and Support Laboratory, Research Triangle Park,  N.C.   27711

18.   Puzak, J.C.,  T.A.  Clark, and E.G.  Evans.   "External  Laboratory
     Quality Assurance for the  Los  Angeles  Catalyst Study"  in  Proceedings
     of the Los Angeles  Catalyst Study  Symposium, April  12-13, 1977.
     EPA-600/4-77-034.
                                84

-------
                 APPENDIX B
THIRD ANNUAL CATALYST RESEARCH PROGRAM REPORT
                 Prepared by

 Environmental Sciences Research Laboratory
     Research Triangle Park, N.C.  27711
     U.S.  ENVIRONMENTAL PROTECTION AGENCY
      OFFICE OF RESEARCH AND DEVELOPMENT
      HEALTH EFFECTS RESEARCH LABORATORY
      RESEARCH  TRIANGLE PARK, N.C. 27711

-------
                           TABLE OF CONTENTS

                                                                       Page


List of Tables	     87

List of Figures	     88

  I. Introduction	     89

 II. Measurement Methods

     A.   Ambient Air Measurement Methods	     92
     B.   Mobile Source Measurement Methods	     95


III. Emissions Characterization	     95

 IV. In-Roadway Study	,	     96

     A.   California Institute of Technology Measurements	     98
     B.   University of Minnesota Measurements	     98
     C.   Washington University Measurements	    100
     D.   EPA-Rockwell  International Measurements	    102
     E.   Summary of Results	    104
     F.   Conclusions and Planned Work	    104

  V. Model i ng	    107


Appendix B-l:  Monitoring Program for Catalyst-Generated Sulfuric
               Acid	    109

Appendix B-2:  Estimate of Probability Distribution of One-Hour
               Average Exposures to Automobile-Generated Sulfuric
               Acid for Morning Peak Period Travelers in Los
               Angeles	    115
                                   86

-------
                            LIST OF TABLES
1    Summary of ESRL Catalyst Program Measurement Methodology	    90
2    Summary of Sulfate Emissions Factors by Manufacturer	    91
3    HCN Formation - 1977 California-Standard Certification Cars...    97
4    Instruments in the Caltech Van	    99
5    Nomenclature and Description of Instruments in the University
     of Mi nnesota Van	   101
6    Average Daily Net Roadway Sulfur from Two-Mass Sampler in the
     Washington University Car	   103
7    Small Particle Sulfate Differences Daily Averages of Data from
     the LBL Dichotomous Samplers	   105
8    Freeway-Background Sulfate Differences Observed on October 21,
     1976	   106
     Appendix B-2
1    Features of the  Distribution of Travel Time During the Hour
     Following  the  Start  of  a Morning  Peak Period Trip in
     Los Angeles	   121
2    Features of the  Distribution of Concentrations of Automobile-
     Generated  SuIfuric Acid on  Streets  and Freeways in
     Los Angeles	   122
3    Features of the  Distribution of One-Hour Average Exposures to
     Automobile-Generated Sulfuric  Acid  During  the Hour Following
     the Start  of a Peak  Period  Trip in  Los Angeles	   123
                                  87

-------
                            LIST OF FIGURES
                                                                    Page
1    Histograms of Freeway and City Street Travel Times	      124
2    Histogram of City Street Sulfuric Acid Concentration,
     Increment Over Background	      125
3    Histogram of Freeway Sulfuric Acid Concentration,
     Increment Over Background	      126
4    Histogram of EXP	      127
                              88

-------
I.    INTRODUCTION

      The  Environmental  Sciences Research  Laboratory/North Carolina has
primary responsibility  within EPA for the conduct  of research pertaining
to the atmospheric  transformation and transport  of airborne pollutants.
Essential  related research responsibilities  pertain to  (1) the character-
ization of emissions  to the atmosphere, and  (2)  the development of measure-
ment  methodology for  such  characterization and for determining the nature
and distribution of pollution in the  ambient environment.  Consequently,
the Laboratory was  prepared to play an important role in the Agency's
research  effort to  assess  the potential impact of  automotive catalyst-
generated "sulfate" when this issue was surfaced and accepted research
assignments  in the  following areas:

      1.    Develop measurement methodology for the  qualitative and
           quantitative  characterization of emissions emanating from
           catalyst-equipped vehicles  and  of  the  air affected by such
           emissions.  (Measurement Methods)

      2.    Ascertain the nature and magnitude  of  sulfur-bearing and related
           emissions under  typical  vehicle operating and malfunction
           conditions.   (Emission Characterization)

      3.    Apply measurement methodology to characterize the pattern of
           human exposure within the localized area.  (In-Roadway Study)

      4.    Establish the relationships  between emissions and exposure
           of the public at risk and develop models for predicting the
           potential impact of such emissions  on  the quality of the air
           to which  the  population  at maximum  risk  is likely to be
           exposed.  (Modeling)

      The  attached report documents the research  efforts of the Environmental
Sciences  Research Laboratory during the 1976  calendar year, organized
according  to the laboratory branch primarily  responsible for its conduct.
A short commentary  on the  catalyst program is offered here, as an
overview  of  the subject.

      The  issue that is  being addressed by the catalyst research program
belongs to a growing class of problems—a second generation of issues
that  have  their origin  in  otherwise successful solutions to initially-
perceived  problems.  It underscores the importance of "technological
assessment"  wherever technology is to  be  applied to broad-based societal
problems.  It calls for a  new attitude in research and development, one
that  looks beyond the short-range  results to  the long-range consequences,
with  a view  to bringing these in balance.  The catalyst, which promises to
be an effective deterrent  to the traditional  automotive pollutants, is
inherently capable  of generating other pollutants  of great potential
concern.
                                     89

-------
                                                                               TABLE 1

                                                       SUMMARY  OF  ESRL CATALYST PROGRAM MEASUREMENT METHODOLOGY

                             Task Number                                                                                          Funding ($K)
                               (FY-77)       Description                                         Milestones	            FY76	FY77

                             AMBIENT AIR MEASUREMENT  METHODS

                             CA-23          Sulfuric  Acid Sampler/Analyzer                   5/77 Begin Delivery                  75         35
                                            Cabot Corp. Contract 68-02-2402                  2/78 Complete Delivery
                                                                                            8/78 Final Report

                             CA-31          Sulfuric  Acid Generator  for  Ultrafine Particles  9/77 Deliver UM Generator            71         50
                             8E-37          University of Minnesota  Grant 80-4600            9/77 Report on Northrop Generator
                                            Northrop  Contract 68-02-2566

                             CA-24          Development of a  Gas Chromatographic System      7/77 Interim Report                  18          0
                                            For NH3                                         8/78 Final Report
                                            University of North Carolina-Charlotte
                                            Grant 80-4413

                             CA-21          Continuous Monitor  for Particulate  Sulfur        10/77 Assemble Instrument            30         35
                                            Compounds
o                                          Oregon Graduate Center Grant 80-4750
                             CA-11           Development of Methodology for Sulfuric Acid     6/77 Report on Interferences         64         10
                                            Atlantic Research Contract 68-02-2247

                             CA-12           Development of a  Portable  Collection System      6/78 Final Report                    64         50
                                            for Atmospheric Sulfuric Acid
                                            Southern Research Inst. Contract 68-02-2234

                             CA-32           Rural  and Urban Measurements                     6/77 Final Report                     9         20
                                            Environmental  Measurements,  Inc. Contract 68-02-2484

                             MOBILE SOURCE  MEASUREMENT METHODS

                             CA-15           S02 Compliance Test  Instrument                   10/77 Deliver Prototype              60         14
                                            Aeronutronic-Ford Contract 68-02-2448

                             CA-16           SO!, Compliance Test  Instrument                   10/77 Deliver Prototype              110           0
                                            Ford Motor Co.  Res.  Lab. Contract                2/78  Final Report

-------
                         TABLE 2
  Summary of Sulfate Emissions Factors by Manufacturer
(Olson Study - 1975 and 1976 Model  Year California  Cars)
Manufacturer
No.
of Vehicles
Test Average
No. Mileage
SOU
FTP1
mean
mg/mi 1 e
1st CFS2

HFET3

Fleet
General Motors
Ford
Chrysler
AMC
Foreign
(Datsun, Toyota
VW)
Notes:
1. FTP =
2. CFS =
3. HFET =
101
44
26
20
5
6
»
1 3395
1 3752
2 7428
3 11274
1 3352
2 7614
3 12730
1 3039
2 7617
3 11537
1 2879
2 5189
3 7906
1 2706
2 5377
3 11525
1975 Federal Test Procedure
Crowded Freeway Simulation (
Highway Fuel Economy Test (
4.3
1.96
3.03
1.10
8.36
3.40
2.64
5.17
3.04
3.41
3.26
0.58
1.43
2.01
1.23
1.01
- Urban Driving
35.0 mph)
49.9 mph)
17.1
9.55
8.86
3.74
27.69
9.91
6.73
24.26
15.31
8.03
12.09
4.35
7.12
6.84
3.56
2.51
Cycle (19.9
22.3
17.55
13.25
6.39
26.85
14.51
7.72
27.86
24.70
15.96
9.36
9.70
3.19
13.48
4.73
2.28
mph)
                            91

-------
     The technological assessment approach as applied by the Environmental
Sciences Research Laboratory recognizes the critical importance of
defining the future consequences of general usage of catalysts in
automotive pollution control.  This is causing a change in emphasis  in
the Laboratory from retrospective research to more prospective research.
Major efforts are underway to ascertain the definitive chemical and
physical structure of the atmospheric pollution that is likely to be
associated with the automobile of the future, independent of and combined
with other traditional pollutant sources.  Predictive models are being
developed and tested to estimate the related pollutant dosages that the
public is likely to be exposed to in various risk situations and, in
particular, the situations of maximum risk.  The composition of the pollu-
tion and the potential dosages will  depend to a large degree on the "design"
of the catalyst that will find broad future use.  Accordingly, it will
depend on the changing performance of the catalyst with changing operational
requirements imposed on it through various driving patterns and engine
modes.  It will also depend on the changing performance with age of the
catalyst and changing composition of future fuels, lubricants, and additives.
It will depend on changing fuel mileage requirements in the future.  These
emission characteristics of the catalyst are being explored to the extent
that the state-of-the-art, resources, and manpower permit.   It calls for
the development and acquisition of innovative sampling and  analytical
equipment that will not alter the complex physical dimensions and finite
chemical composition of the emissions being analyzed.  The  report that
follows addresses the status of these efforts, presents preliminary findings,
and indicates future directions and  plans.


II.  MEASUREMENT METHODS

     A.   Ambient Air Measurement Methods

          1.   Sulfuric Acid Sampler/Analyzer
               Cabot Corporation/Fowler (PI)
               Contract 68-02-2402

          In 1974 ESRL initiated a program with Cabot Corporation to
develop an automated aerosol sampler/sulfuric acid monitor.  As a result
of the GM/EPA Sulfate Dispersion Experiments in which the Cabot sulfuric
acid monitor was successfully employed, Cabot was further funded to
fabricate and deliver several coupled automated aerosol sampler/sulfuric
acid analyzers to EPA.  The acid aerosol  is sampled through a NH~ denuder
into a dichotcmous sampler so that only the fine aerosol-fraction is
collected on a Teflon filter.  The filter is then automatically moved into
a volatilization zone where sulfuric acid is selectively volatilized from
the filter into a flame photometric  detector.
                                    92

-------
anal Jer wlrJ^lW*?  of Jhertauto^ted aerosol sampler/sulfuric acid
analyzer was  tested  during the  October  1976 Los Angeles In-Roadwav

andSanerna vi±f  ^ th6SS  t6StS  a nu"*er of^alfunctJonfoccurre
                          68111  Ca
th                   Cabot was requested to  make modifications
  J re ^ability of the analyzer and to separate the sampler from
thp     v «
Thl  mnH-f  tJ.s1mPllfy  operation  and maintenance of the two components.
These modifications  caused  a  6-month delay in delivery of the samplers and
analyzers.

          2.   Sulfuric Acid Generator for Ultrafine Particles
               University  of Minnesota/Liu (PI)
               Grant  R804600

          The GM/EPA  Sulfate Dispersion Experiment indicated that ultrafine
particles should be used in exposure studies designed to determine the
effects of sulfuric acid on animals.  A grant was awarded to the University
of Minnesota to fabricate  a sulfuric add generator that will  deliver
particles in the 0.02 to 0.5 ytn size range.  This generator has  been
coupled to an animal  exposure chamber for use by the Health Effects
Research Laboratory.   Northrop Services, Inc. is also fabricating a
generator for use with animal exposure studies.

          3.   Continuous Monitor for Particulate Sulfur Compounds
               Oregon Graduate Center/Huntzicker (PI)
               Grant  R804759

          A program was initiated to develop a flame photometric detector
for continuous, in situ measurement of total  particulate sulfur  and certain
individual sulfur compounds.  The technique is based on coupling a
diffusion S0? denuder to a flame photometric detector.   The denuder removes
the gaseous sulfur compounds and only permits the particulate  sulfur to
enter the photometric detector.  Modulated heating of the aerosol upstream
of the denuder permits differentiation between sulfuric acid and ammonium
sulfate.  To optimize the  operation of the heater-denuder, the flash
volatilization of several  sulfate compounds was studied.  Several different
types of denuders are being fabricated to permit the removal of  S02, HLS,
CH3SH, and ChLSSCH.,.   A prototype system, suitable for use in  fiela stodies,
will be assembled and evaluated.

          4.   Development of a Gas Chromatographic System for NhL
               University  of North Carolina at Charlotte/Cooke (PI)
               Grant  R804413

          The University of North Carolina-Charlotte was awarded a grant
to (a) develop a gas  chromatographic system which will  permit  the analysis
of sub-part-per-billion concentrations of NFL, and (b)  develop improved
chemical converters for chemi luminescent NO  and NH3 monitors.  For the
q*s c^romatographic system, a column was developed which permits passage
of low ppbNH, concentrations with near ideal peak symmetry.  The gas
             0
                                   93

-------
chromatographic system will be coupled to a chenriluminescent detector,
being developed by ESRL, and used in field studies to measure ambient
concentrations of NH- in rural and urban locations across the United
States.             J

          5.   Development of Methodology for Sulfuric Acid
               Atlantic Research/Valentine (PI)
               Contract 68-02-2247

          The objective of this task was to develop methodology to selectively
measure sulfuric acid in the range of 0.25 to 50 vg/m .  The project was
directed toward identifying reagents which would react with sulfuric acid
aerosol forming adducts which could be selectively analyzed in the presence
of sulfate.  The current status and achievements to date are as follows:

          •    Three candidate reagents for the formation of adducts
               were identified.
          §    The formation of adducts by the reagents was demonstrated.
          •    The adducts were isolated and their likely composition
               was determined.
          •    S02 is released by decomposing the adducts at 200°C.

          Major problem areas exist with regard to the specificity of the
reaction adduct and the decomposition efficiency of the adducts.

          6.   Development of a Portable Collection System for
                 Atmospheric Sulfuric Acid
               Southern Research Institute/Barrett (PI)
               Contract 68-02-2234

          As a first step in development of a collection system for
sulfuric acid, an evaluation of the interferences that may occur when
samples containing sulfuric acid are collected on a filter was conducted.

          Results indicate that the collected sulfuric acid aerosol is
subject to interference by alkaline gases and by other particulate materials
in the sample air.  Strong interference is caused by gaseous ammonia,
particulate calcium carbonate, and ambient particulate matter from an urban
oxide and a silicate clay soil; and relatively little or no interference
is caused by nitrogen dioxide, sulfur dioxide, fly ash, soot, and the
vapors of pyridine and phenol.  These conclusions suggest that research
on the formation of sulfuric acid-adducts which could be selectively
analyzed should be continued and expanded.

          7.   Rural  and Urban Measurements
               Environmental Measurements, Inc.
               Contract 68-02-2484

          Prototype instrumentation for S02, NH3, NO , and hLS was evaluated
using a mobile van equipped with a mapping and data acquisition system.  The
                                       94

-------
       X  X   i°?  -h  a5jcomP1!shed 1" a cross-country  run  from  Denver,
         ' ^^S19^ N°rth  Caroltna> ^  August  1976.  Measurements were
    * H T 1 3J  ^lu3"  concentrations along  the route.   Data  from this
study demonstrated  the operation  characteristics of an NH0 denuder and
capability of a new prototype  Meloy  SA-285  flame photometric detector
to measure sub-part-per-btllion concentrations of S02 and H2S.

     B-   Mobile  Source  Measurement  Methods

          Several successful analytical techniques for S09 and SO. mobile
source emissions  have been reported  in previous  annual reports.   While
in-house efforts  to standardize these analytical  techniques are continuing,
no further in-house developments  are planned  under the catalyst proaram.

          Contract  programs  for routine SCL and  H-SO,, mobile source monitors
are currently active  at  Aeronutronics-Ford and at Ford Motor Company
Scientific Research Laboratory.   However, there  is no experimental perform-
ance data available for  either the UV-correlation spectrometer for S09
or the microcoulometric  sulfate monitor at this  time.


III. EMISSIONS CHARACTERIZATION

     Since Catalyst Research Program-supported efforts in previous years
have adequately defined  H2S04  emissions factors  for low-mileage vehicles
in the 1975 and 1976  vehicle years,  as well as for many prototype auto-
mobiles aimed at  future  HC,  CO, and  NO emissions standards, the current
year's programs concentrate  on two issues which  have not yet been adequately
studied.  The first program  involves  a study  of  sulfuric acid emissions
factors for 101 in-use California cars performed by Olson Laboratories, Inc.
While initial emissions  factors for  this fleet were reported in the 1976
Catalyst Research Program report, currently available data from this
recently-completed  project shows  a rapid decrease in H2S04 with increasing
mileage.  The second  program (in-house) aimed at characterization of both
gaseous and particulate  pollutants for a group of 24 California-standard
cars.  The majority of these cars were European  imports, utilizing at
least a measure of  catalytic NO   reduction; all were tested under both
certification tune  and under simulated rich malfunction.  Significant levels
of HCN and NH, were found with both  dual-bed  and three-way catalyst
configurations.   Additional major findings from  both studies are given below.

     A summary of the Olson data  is  given in  Table 2.  While the initial
fleet average SO, emission rate of 17.1 mg/mile was surprisingly low, the
observed decrease to  10  mg/mile at 7,000 miles and to 5 mg/mile at 11,000
miles is remarkable.   This decrease  is about  evenly distributed throughout
the major manufacturer subsets of this fleet.  Thus, ft appears that
catalysts lose their  sulfate-forming  activity more rapidly than their
ordinary capacity for HC and CO.
                                     95

-------
     Other major conclusions include:

     1.   HC and CO remain low and mostly within Federal standards
          in the first year of life for California cars.

     2.   NO  emissions rates are mostly greater than applicable
          California standards for 1975 and 1976 model years.

     3.   Average California fuel sulfur levels are about the same
          as the national average.


     Goals of the in-house characterization study of California cars
 included determination of H^SO* and other particle emissions under a
 variety of driving cycles and determination of HpS, COS, NH.,, and N^O
 emissions under both standard and malfunction conditions.  HCN and ammonia
 emissions were traced to rhodium-containing catalysts specifically.
 Previous experience with such catalysts indicated rather low activity
 for sulfuric acid formation.  Subsequently, all 4,000 mile emissions
 data cars in the 1977 certification program were tested over urban,
 expressway, and highway driving cycles.  HCN, NFL, and SO. data for the
 highest HCN cars are given in Table 3.  HCN yields from catalyst-equipped
 cars are usually at least one order of magnitude below those for non-
 catalyst automobiles.  When three-way or dual-bed systems fail rich of
 stoichiometry, HCN emission rates increase dramatically as a product of
 the catalytic reduction of NO.   However, the maximum value found so far
 has been about 0.15 grams/mile; rich malfunction also accelerates CO
 emissions in the order of 20-70 grams/mile.  Therefore, CO will  always
 be 200-500 times greater in concentration than HCN.  Since OHEE has indi-
 cated the health effects of HCN are similar, apparently the CO formed
 poses a greater health threat than HCN.  Ammonia toxicity at present is
 less well-defined.  HCN and NH3 monitoring will be continued for future
 catalyst designs, but no significant problem is anticipated.

     Future in-house work will  concentrate on measurement of trace
 components such as low molecular weight amines, cyanogen, and nitro-
 methane from catalyst cars.  Sulfuric acid emissions monitoring for 1978
 model year Federal-standard cars will also be continued as an adjunct
 to specific characterization programs designed for this group.


 IV.  IN-ROADWAY STUDY

     Studies carried out during the previous year, particularly the
General Motors/Environmental Protection Agency test roadway sulfate
 dispersion experiment, indicated that off-roadway levels of sulfate
 emitted from catalyst-equipped  cars would be near background and not
 constitute an air pollution hazard.  Attention was then turned to the
 roadway itself, since this was  the only place where sulfate levels from
 catalyst-equipped cars might become high enough to be hazardous.  In this
event, occupants of vehicles in the roadway could be affected.
                                   96

-------
                                                      TABLE  3

                                                  HCN Formation
                                    1977 California-Standard Certification Cars

         Vehicle             Control        Driving      	Pollutant emission rate, gr/mile
                             System          Cycle            Malfunction mode          Standard tune
                                                        HO    CO    NHa.    N0£      HCN    CO    SOU
       Volvo, auto,  trans.  3-way         urban         0.046  22.4  0.518  0.56     0.001  2.7   10"4   0.33
       Volvo, manual tr.    3-way         urban         0.033  18.3  0.400  0.42     0.001  2.7   IO"4   0.52
       Saab  auto,  trans.   3-way         expressway    0.078  22.0  0.705  0.87     0.003  1.0   10"4   0.53
       Mercedes 280        dual  bed       urban         0.045  65.7  0.251  0.60     IO"4   1.7   0.008  1.33
^      AUDI TOOLS          ox.  cat.       urban         0.006  9.9   0.028  1.15     0.003  1.9   0.001   1.24
       1976 Chrysler       lean burn     urban           -            ...      o.Oll  4.5   0.001   2.56

-------
     Sulfate levels in the roadway would, of course, vary considerably
depending upon the mix of vehicle emission factors, traffic density, and
meteorological conditions.  Also, the exposure would depend on the
length of time the vehicle occupants would be in the roadway.  Horowitz and
Papetti of EPA developed statistical curves showing the probability of
exposure to various levels of sulfate by vehicle occupants (see Appendix B-l).
These curves were developed using a substantial data base of vehicle usage
during the 6 to 9 a.m. period on weekdays and the best available information
on predicted roadway levels of sulfate.  This statistical analysis served
to define the type of in-roadway sampling program that would be needed to
obtain an estimate of the exposure hazard of vehicle occupants.

     To implement this study, a contract was awarded to Rockwell International
Air Monitoring Center to participate in and support a field program that
would involve aerosol scientists from the University of Minnesota,
Washington University at St. Louis, and the California Institute of Technology.
The study was conducted in Los Angeles during the last two weeks of October
in 1976 and was designed to demonstrate the feasibility of making in-roadway
measurements with instruments installed in moving vehicles.  If feasibility
could be demonstrated a larger scale in-roadway measurement program could
be conducted in the following year.

     A.   California Institute of Technology (Caltech) Measurements

          The group from the California Institute of Technology which parti-
cipated in this study was associated with Professor S.K.  Friedlander and
was primarily organized and led by Dr. Susanne Hering.  Caltech served as
host for this study and provided invaluable laboratory space and services.
Caltech also operated a van which made measurements at the background site
selected for each day.  The instruments that were in the van are listed in
Table 4 along with the name of the organization responsible for analyzing
the filter aerosol samples collected.

          The low pressure impactor (LPI) used by Caltech is a single-jet,
eight-stage impactor with 50% collection efficiency at cut sizes of 0.02,
0.05, 0.11, 0.25, 0.5, 1.0, 2.0, and 4.0 urn aerodynamic diameter.  The
0.25 and 0.11 ym cutoff stages were calibrated with polystyrene latex
spheres and the 0.05 and 0.02 ym cutoffs are based on theoretical calcula-
tions.  Comparisons with the electrical aerosol analyzer (EAA) for
laboratory-generated ammonium sulfate and sulfuric acid aerosols is
reasonably consistent with the theoretical calculations.   The impaction
surfaces above 0.5 ym are coated with Vaseline to minimize particle
bounce-off to the lower stages.  The lower stages were analyzed for sulfur
content by flash volatilization in a flame photometric detector.

     B.   University of Minnesota Measurements

          The University of Minnesota participants in this study were
associated with Professor K.T. Whitby, and a key role in the field operations
                                 98

-------
                      TABLE 4
          Instruments in the Caltech Van

Manual Dichotomous Sampler (Caltech-AMC)
Low Pressure Impactor (Caltech)
Two Mass Sampler  (Washington University)
Ammonia Sampler (AMC)
Meteorological Measurements
                            99

-------
and data analysis was entrusted to Bruce Cantrell.  The main activity
of the University of Minnesota was to operate a mobile laboratory in
freeway traffic and at background sites.  For most experiments, air
samples were collected through tubes which had inlets just above the
headlights of the van.  Table 5 gives a listing of the equipment in the
van.  Because the aerosol-measuring equipment requires several minutes
to determine a particle size spectrum, bag samples were collected about
every four minutes, and the size distribution of the aerosol in the
bag was determined.  Data from the trace gas measuring instruments were
recorded every half second, and then averaged over the time of each
bag fill.  The aerosol and gas data resulting from one bag fill was
computed and printed in tabular form for each run on board the van.

          A trichotomous sampler designed and built at the University
of Minnesota was used in this field study for the first time.  Its purpose
was to collect size fractionated aerosol samples on Teflon filters and to
include in the fractionation, a cut between the nuclei and accumulation
mode size range.  A dichotomous sampler splits the aerosol into a flow
containing the large particles collected on filter A, and the particles
smaller than 3.5 y.n collected on filter B.  A portion of the flow to
filter B is drawn to an EAA, where it is further classified.  The nuclei
mode particles are drawn in the drift tube toward the central collector
rod, and are collected on filter D.  The larger particles remain the
outer portion of the flow and are collected on filter C.   These filters
were designed to fit into the Cabot sulfuric acid analyzer (SAM).

          An automated dichotomous sampler was also operated in the Univer-
sity of Minnesota van and samples collected were Analyzed for sulfuric
acid concentration with SAM or for hydrogen ion H , and sulfate (SO. )
concentrations by AMC.

          An observer aboard the University of Minnesota van made sample
counts of catalyst-equipped cars on October 26, 27, and 28 and estimated
the number of catalyst cars to be approximately 28% of the total.

     C.   Washington University Measurements

          The Washington University participants in the October field
program were associated with Professor E.S. Macias.  Their main activity
was to operate an instrumented automobile on the freeways and at the
background sites.  The Washington University car and the University of
Minnesota van operated as a team and generally followed the same sampling
schedules and itineraries.

          The instrumentation in the automobile included a Two-Mass
sampler, an LBL dichotomous sampler, and an aerosol charger which included
the capability of adding ammonia to the sampler air flow and passing the
sample through a temperature-programmed heat tube.  The LBL sampler used
an air intake outside the car from October 19 to 23 and sampled from inside
the car with the windows open from October 24 to 30.  The Two-Mass sampler
and the charger always sampled outside air.
                                    100

-------
                                 TABLE  5
            Nomenclature  and  Description  of  Instruments in
                     the University  of  Minnesota Van
Meteorology
WDIR, WD
WSPD, WSP
TOUT
TIN
TEAA
DEWPT, TOP
RH, RELHUM
Gas Measurements
so2
°3
CO
NO, NOV
Wind direction, clockwise degrees from north
Wind speed, kilometers/hour
Outside temperature, °C
Inside temperature, °C
EAA classifier temperature, °C
Dew point, °C
Relative humidity, percent
Sulfur analyzer, Monitor Labs, Model  8450
Ozone monitor, Dasibi Corp., Model  1003-AH
Carbon monoxide monitor, Energetics,  Sci.  Inc.
Chemiluminescent NO-NO -N0? analyzer, Thermo-
Electron Corp., Model  14
Aerosol Measurements
EAA
CNC
Royco 220, R220
LBL
TCS
LPI
Electrical aerosol analyzer; Thermo-Systems,  Model  3030
Condensation nuclei counter; Environment One,
Model RICH 1000
Optical particle size counter, Royco Model 220
Automated dichotomous sampler, Lawrence Berkeley Labs.
Trichotomous sampler, University of Minnesota
Low pressure impactor, Kech Lab., Caltech
                                       101

-------
          The Two-Mass  sampler separates particles  into two size fractions,
Coarce particles are  impacted on a glass fiber filter; the remaining
particles are collected on an identical high-efficiency glass fiber
filter.  Samples were collected on a portion  (0.3 cm) of a light-weight
(1.2 mg/cm  ) low pressure-drop glass fiber filter (Railflex 70) with a
consistent  and low sulfur blank of 0.3 to 0.5 ug/cm  .  The fine particle
cutoff is designed to have a 50% collection efficiency at 3.5 \im.

          Analysis of the water-soluble particulate matter collected in
2-hr intervals with the Two-Mass sampler was performed using a flash
vaporization/flame photometric system.  The analyzing system consisted
of a flash  vaporization vessel, flame photometric detector, integrator
and a capacitor discharge across a tungsten boat, resulting in resistance
heating to  100°C.  A water extract of^each sample is prepared by punching
out a circular filter segment (0.6 cm) containing the aerosol deposit.
An aliquot  of the extract is evaporated on the tungsten strip and then
vaporized into the FPD.

          The roadway sulfur values as measured with the Two-Mass and
FPD systems are calculated by taking the difference between the data
obtained with two different samplers in two different vehicles.  The mean
of all these differences is -0.04 ug/m  sulfur., and the standard deviation
of the differences about the mean is 0.55 ug/m  sulfur.
          Table 6 is a comparison of the in-roadway sulfur with sulfur
me?surements taken at the background sampling site.

     D-   EPA - Rockwell International Measurements

          A portion of the filter samples collected in the October field
program were returned to Thomas Dzubay, EPA, Research Triangle Park, N.C.,
for analysis by X-ray fluorescence (XRF).  The data obtained for lead on
filters from the trichotomous sampler in the University of Minnesota van
shows clearly that large amounts of lead (> 10 ug/m  in some cases) are
present on the freeway in the nuclei mode size range.  These data also
provide a confirmation of the proper operation of the trichotomous sampler.

          The main role of the Rockwell International Air Monitoring Center
(AMC) in the field program was to provide analytical chemistry services for
the rapid turn-around analysis of filter samples.  Over 400 wet chemical
determinations were performed on 76 filters and 71 filters were analyzed
at Caltech by the Cabot SAM.

          Aerosol samples were collected on 1 urn fluoropore filters by
various sampling devices in the three vehicles described previously and
then were returned to the laboratory at Caltech.  There the filters were
placed in desiccators lined with glass filters coated with oxalic acid.
The humidity in the desiccators was generally near ambient.
                                  102

-------
                    TABLE 6

Average Daily Net Roadway Sulfur from Two-Mass
   Sampler in the Washington University Car
                Aerosol Sulfur Concentration, ug/m
Date -
October
19
20
21
22
24
25
26
27 a.m.
27 p.m.
28

Average
Freeway
6.3
4.68
3.04
3.14
2.96
2.64
0.34
0.58
0.55
0.46

Average
Background
9.04
4.88
2.27
2.59
2.92
2.75
0.11
0.28
0.29
0.57
Measure
Standard Error
Difference
F-B
-2.75
-0.21
0.77
0.55
0.04
-0.11
0.23
0.30
0.26
-0.11
-0.11
0.98
                         103

-------
          The filters were returned daily to the AMC where they were
ultrasonically extracted and titrated for strong acid by the Gran procedure.
The ammonium icn concentration was determined by an ion selective electrode/
and sulfate, bromide and nitrate determined with an ion chromatograph.

          Table 7 summarizes the sulfate results using the LBL dichotomous'
samplers.  It can be seen that as a general rule there is no significant
difference in the total small particle sulfate numbers on and off the
freeway.  This was true even on October 26 when the Santa Ana winds.,were
blowing and the background sulfate concentration was about 0.5 yg/m .  In
contrast, the bromide is clearly traffic-related.

          No acidity was detected on the filters by Gran titration.  Studies
on techniques to preserve the acid on filters revealed that neutralization
could have occurred during transport or handling prior to the acid measure-
ments.  Procedures to protect aerosol samples from being neutralized during
handling have now been perfected.

          The sulfuric acid monitor (SAM) from Cabot Corporation had just
been constructed a few days before the study began and required several major
modifications.  The sampler and analyzer combination proved impractical
and it was decided to operate the SAM as an analyzer only.   Samples would
be collected with the LBL and the trichotomous samplers.  Duetto day-to-day
variations in calibrating the SAM, uncertainties of  0.5 ug/m  sulfuric
acid were typical with this prototype instrument.  The best SAM data were
obtained from the LBL dichotomous samplers, but these filters were submitted
to the SAM only early in the program.  The mean daily difference between
the background sulfuric acid and the in-roadway LBL samples was 0.12
Sulfuric acid measurements on samples collected with the trichotomous
sampler also showed about the same differences between background and
in-roadway.

     E.   Summary of Results

          The measurements made during these studies in Los Angeles show
that the sulfate concentrations on roadways due to catalyst-equipped
automobiles are small enough in comparison with background sulfate concen-
trations that they are generally obscured by the scatter in the analytical
data.  On a clear day (low background sulfate) with good ventilation, the
sulfate increase due to catalyst cars was still not clearly seen, but was
almost certainly below 1 ug/m .  Table 8 is a summary of measurements made
on October 21, 1976, showing the sulfate differences with all the measurement
methods and is typical of the results obtained during the study.

     F.   Conclusions and Planned Work

          This study has demonstrated the feasibility of determining
in-roadway levels of vehicle-emitted sulfate by making measurements using
sampling and analysis equipment installed in vehicles moving on the roadway.
There is no reason to believe that this technique would not be applicable
to other aerosol or gaseous pollutants.
                                     104

-------
                                TABLE 7

                  Small  Particle Sulfate Differences
       Daily  Averages of Data from the LBL Dichotomous Samplers
                   Average
Average
Date -
October
1976
19
20
21
21
22
24
25
26
27(c)a.m.
27(c)
Freeway
Concentration
uq/m3
13.16
15.23
6.71
8.77(b)
8.21
8.22
8.81
0.40
1.50
0.98
Background
Concentration
ug/m3
13.19
17.03
8.37
8.37
7.50
11.73
11.97
0.67
1.02
0.71
Difference
ug/m3
-0.03
-1.80
-1.66
+0.40
+0.72
-3.51
-2.16
-0.27
+0-48
+0.27
(a)   Average includes filter 202 with 0.5 ug/m3 sulfate.

(b)   Average omits filter 202.

(c)   Only one freeway and one background sample analyzed.
                                 105

-------
                                         TABLE 8
                          Freeway-Background Sulfate Differences
                              Observed on October 21, 1976
Collection Analytical
Method Method
LBL Dichot.
Two-Mass (WU)
Two-Mass (WU-CIT)
LBL Dichot.
LPI (UM)(a)
LPI (UM)^b^
LPI (UM-CIT)^

SAM
FV-FPD^
FV-FPD
Ion Chrom.
FV-FPD
FV-FPD
FV-FPD

Daily
Freeway
1.24
3.04
3.04
8.77
2.20
0.29
0.29

Average
Background
0.91
2.27
2.67
8.37
2.31
0.16
0.20

Sulfate
Difference Difference
F-B Units ug/m3
0.33
0.77
0.38
0.40
-0.10
0.13
0.09
Standard
ug/m3 H2SOi+
ug/m3 S
ug/m3 S
ug/m3 SO.T
ug/m3 S
ug/m3 S
ug/m3 S
Mean
Deviation
0.3
2.3
1.1
0.4
-0.3
0.4
0.3
0.6
0.8
(a)  All  stages of the low pressure impactor.
(b)  Stages 7 and 8, which nominally collect particles  0.02 to  0.11  urn in  diameter
(c)  FV-FPD = Flash Volatilization - Flame Photometric  Detector
                                           106

-------
        fro. I?*,?   ,were obtained for only a few days, the levels of
        from catalyst-equipped vehicles (estimated to be 30% of all
vehicles) were so low that there seems to be no reason for conducting
^o9f^Ca]e-StUdy nex? y^r.  The lower-than-expected sulfate emission
factors found in a surveillance study of in-use vehicles reported in
Section IV. B.» above, supports this conclusion.

          It is planned to conduct a small-scale in-roadway study to
assess an extreme worst-case situation.  This is called the "cluster
experiment.   It is  designed to determine the levels of sulfate to which
vehicle occupants would be exposed if they were driving immediately behind
a cluster of several high-sulfate-emitting vehicles.  Such high-sulfate-
emitting vehicles have been shown in the surveillance study cited above
to constitute a small percentage of the in-use vehicle population.  If
this experiment indicates high levels of sulfate from the "cluster" we
will ask the statisticians to estimate the probability of occurrence of
this worst case.
V.   MODELING

     Development of theory adequate to quantitatively describe the transport
of suIfuric acid mist from catalyst-car tailpipes to the exposed population
has been concentrated during the past year on dispersion and aerosol
coagulation processes.

     The Meteorology and Assessment Division of ESRL has conducted a  variety
of tests of the current Gaussian plume model of a line-source roadway (the
HIWAY model) using long-time-scale data from the GM proving ground experiment.
The tests so far clearly indicate the current procedure -- that of assigning
overall Gaussian distribution parameters on the basis of wind stability
alone — leads to overestimation of ambient air sulfuric acid mist concen-
trations under conditions of poor atmospheric dispersion.  Follow-on  efforts,
which seek to improve the predictive capability of this simple theory,
involve extraction of best-fit Gaussian distribution parameters from  the
short-time-scale GM data.  However, the data have only recently been  received
by EPA and these calculations will continue into early FY-78.  At this  point,
the mechanical source of predictive error might be either inability to
account for short-term wind meander under conditions of low wind speed  or
to non-Gaussian concentration profiles induced by motion of the cars.

     More complicated models are also being tested in the hope that
inclusion of greater physical detail will improve predictive capability.
One such approach involves numerical integration of point sources to
produce a better description of the roadway as a line source.  The techni-
ques and computer programs for performing these calculations have been
developed during the past year.  Testing of this procedure will be accom-
plished in future work.
                                    107

-------
     Another, still more complicated, method summarizes overall roadway
dispersion by following the mechanical and chemical transformations of
individual parcels of air.  A contract to develop this Lagrangian model
has been initiated in FY-77.  Still another approach involves handling
the mathematical expressions describing the variety of mechanical influences
on air motion (wind effects, advection, automobile-induced turbulence) as
essentially additive.  Integration of these effects by the finite
difference approximation can then produce an overall view of the roadway.
This procedure has been tested using the long-time-scale GM data base
during'the past year and does represent a significant improvement over
the simpler HIWAY mcdel.

     While the GM data base has proven valuable in preliminary testing of
dispersion models, it is very limited.  Only a few conditions of wind
speed, direction, and stability occurred in the atmosphere during the
short time of the experiment.  A much more extensive set of data is being
obtained in an ESRL-supported study on the Long Island expressway
cooperative with the New York State Department of Environmental  Conservation.
Nine months of SFg dispersion data are now available from the study and
workup of wind field, energy balance, and aerosol composition data is in
progress.  It is expected that this new data will allow much more accurate
and complete evaluation of various models and determination of specific
problem areas in prediction.

     An important addition to our ability to assess pollution problems
associated with roadway-produced aerosols is the development of theoretical
procedures sufficient to determine effects of aerosol  growth, deposition,
and chemical reactions on exposures.  Recent work carried out by ESRL-
Aerosol Research Branch has shown that the ultrafine sulfisric acid aerosols
found in the GM track experiment must persist in the size range for
significant time intervals as sulfuric acid.  Only insignificant neutral-
ization, growth can occur over a roadway.  Hence, motorists must be exposed
primarily to sulfuric acid.

     The value of the predictive ability represented by these physical
and chemical mathematical procedures is very great.  Changes'in automotive
and fuel technology are occurring rapidly in response to pollution control
and energy conservation challenges.  Since these changes can frequently
induce new requirements for quick assessment of environmental impact, we
expect models developed here to be applied to new problems with some
regularity in the next few years.
                                    108

-------
                              APPENDIX B-l

        MONITORING PROGRAM  FOR CATALYST-GENERATED SULFURIC ACID


INTRODUCTION

*   MSS™ consecJuence of the deferral of the sulfuric acid standard
for MY-79 cars, a surveillance program has been proposed to monitor
sulfuric acid concentrations on and near roadways.  The purpose of
this surveillance is to measure concentrations of automobile-generated
sulfuric acid to determine whether it is necessary to take action to
prevent sulfuric acid concentrations from reaching levels that would
constitute a hazard to public health.  California has been chosen as
the most suitable area for  surveillance since projected sulfuric acid
concentrations there should precede those in major urban areas in the
rest of the U.S. by two or  three years.

     Past experience dictates that the highest and therefore controlling
concentrations will occur on freeways.  Therefore, emphasis should be
placed on measurement of sulfuric acid concentrations on or immediately
adjacent to such roads with high vehicle densities.

     Two general measurement approaches have been considered.  The first
employs measurements on the road with monitors situated in vehicles
moving along a road segment (e.g., measurements on the Milford test track);
the second uses fixed sites adjacent to roads and measures differences in
concentrations across the roads.  These differences are then attributed
to sources on the roads (e.g., the LACS program).  Each approach has its
advantages and disadvantages, and an abundance of proponents to express
their views in the forthcoming meetings which will address the details
of a suitable surveillance  program.

     While considerable technical discussion is taking place relative to
suitable measurement techniques to detect catalyst-generated sulfuric
acid in ambient air, an equally important subject has received less atten-
tion.  This is the design of a combined sampling and simulation program
that produces estimates of  population exposure to catalyst sulfates.
Whatever the nature of the  roadway measurements, they do not necessarily
convey information on the distributions of concentrations and exposures.
They do so only when they are employed within a sampling and/or simulation
framework that permits inferences concerning the distributions of human
exposures.  This point seems to have been somewhat neglected to date.

     To underscore this aspect of the proposed surveillance program,
we would like to briefly discuss three ways of inferring distributional
information that have emerged from our own thinking and from discussions
at NPRM committee meetings.
                                    109

-------
     These are:

     (1)  Estimation of the n-th highest concentration over the
          roadway network of Interest.  This method is analogous
          to the procedure used to establish emission increments
          or decrements for AQCR's in SIP's.

     (2)  Estimation of the probability distribution of sulfuric
          acid concentrations on a roadway network.

     (3)  Estimation of the probability distribution of sulfuric
          acid exposures to travelers on a roadway network.

(1)  n-th Highest Concentration Method

     In this method, we pick the roadway network of interest (e.g., all
Los Angeles freeways).  Decisons on whether to control further growth
in sulfuric acid emissions would be determined by comparing the measured
n-th highest concentrations on the network (e.g., the second highest)
to a reference concentration that could not be exceeded.*  Such an approach
would be analogous to the approach currently used in SIP's for criteria
pollutants.

     There are two difficulties with this approach — one theoretical and
one practical.  The theoretical difficulty is that the concept of a
threshold value not to be exceeded more than n-1 times per year has no
statistical meaning.  The number of times a threshold value is exceeded
is a random variable that depends on such things as traffic and meteorological
conditions.  There exists no fixed concentration that is certain to be
exceeded exactly n-1 times per year.  Any control decision based on the n-th
highest value, therefore, is certain to be either too stringent or too
lenient.

     The theoretical difficulties of the n-th highest concentration
approach have not been acknowledged formally by EPA. owing in part to
the disruptions in the SIP process that such an acknowledgement might
cause.   However, the theoretical problems are recognized informally.
They are the reason for many prolonged and inclusive discussions over
how to select a base year for rollback modeling (e.g., how one knows
that pollutant concentrations in a particular year are not unusually  high
or unusually low.
*This, of course, assumes that the health effects people will be able
to define a concentration that corresponds to the onset of significant
adverse health effects prior to the time such concentrations will be
achieved at frequencies of concern.
                                  110

-------
                   P^blem with the n-th highest concentration approach
                fu*  ]S3e-number of samPl™9 sites at which this value
                th sufficient knowledge of the relevant conditions, one
               d ell™nate m<*t of these sites as being unlikely locations
rnnrtTenf °f Jhc n'th hi9hest values> and monitoring could be
conducted only at a  relatively small number of "high-concentration" sites.
However, we believe  that great practical difficulties would be encountered
in attempting to identify a suitable set of high-concentration sites within
the freeway network.  Complicating factors of wind angle relative to the
road, and highly variable vehicle-average emission factors (and total
emission rates) associated with local driving conditions and vehicle
flowrates could make the identification of these sites extremely elusive.
Moreover, the shorter the averaging time chosen, the greater the difficulty
that will be encountered in identifying the n-th highest concentration.
Both the cost and time period required to identify suitable monitoring
sites could be substantial.  It would be necessary to acquire sufficient
monitoring data to choose a small number of sites out of a much larger
set of candidate sites.  It is likely that annual frequency distributions
of sulfuric acid concentrations at many potential sites (e.g., tens or
hundreds of sites) could be needed to establish the monitoring network.

     The problem of  selecting suitable monitoring sites, like the problem
of the meaningless n-th highest value threshold, has been recognized
informally in discussions of SIP's.  In the SIP's, the siting problem is
the cause of questions about whether an existing monitoring network
corresponds to pollution "hot spots."  It also is the source of the fre-
quent observation that when a monitoring network in an AQCR is expanded,
measured concentrations of some pollutants increase.

(2)  Estimation of the Probability Distribution of Sulfuric Acid
     Concentrations  on a Roadway Network

     In this method, we pick a roadway network of interest and construct
an estimate of the cumulative probability distribution of sulfuric acid
concentrations over  the entire network.

     The decision variable which can be obtained from this distribution
is a specified percent!le concentration, representative of upper bound
on concentrations which would be experienced by a large fraction of
travelers on the network.  This high percentile concentration (as in the
case of the n-th highest concentration) is compared to a given threshold
concentration to determine whether further growth in sulfuric acid
emissions must be controlled.

     There are two methods for obtaining an estimate of the probability
distribution of roadway concentrations on the network.  Each calls for
a substantially different monitoring program.
                                     Ill

-------
     (a)  Direct Measurement Method

          In this method, sulfuric acid concentrations would be measured
at randomly selected sites and times on the network of interest.  The
measured concentrations would be used to estimate the probability distri-
bution of sulfuric acid concentrations on the network.  There would be no
need for an air quality simulation model and statistically meaningful
estimates of the features of the probability distribution of sulfuric
acid concentrations would be obtained.  The method would require a care-
fully conducted random sampling program.  Monitors would have to be
sufficiently mobile to obtain random samples over the entire freeway/
highway network.

          The accuracy with which the features of the distribution of
sulfuric acid concentrations could be determined through sampling depends
on which specific features are of interest, the size of the sample, and,
in some cases, the probability distribution function of sulfuric acid
concentrations.  For example, suppose the sample size is 1000 per year
and one is interested in percentiles of the distribution of sulfuric acid
concentrations on freeways.  The computations with the HIWAY model
suggest that the distribution of sulfuric acid concentrations on freeways
is approximately exponential.  With an exponential distribution, a random
sample of size 1000 would establish the 99th percentile with an accuracy
of about ±5 yg/m  at the 95 percent confidence level.  For the 95th
percentile, the accuracy would be ±2 yg/m .  For the median it would be
±.2 yg/m .

          An alternative approach would be to use sampling to estimate
the probability that freeway sulfuric acid concentrations exceed some
threshold value.  A sample size of 1000 would establish such a probability
with an accuracy no worse than ±.03 with 95 percent confidence.  The
accuracy would be better at threshold values that have relatively low
probabilities of being exceeded.

     (b)  Combined Measurement and Simulation

          In this method, average sulfuric acid emissions factors would
be estimated from measurements of roadway sulfuric acid concentrations.
These emissions factors then would be used in connection with an air
quality simulation model to develop freouency distributions of sulfuric
acid concentrations.  In effect, this method would recompute the existing
concentration estimates using a better estimate of the emission factor.

          The principal advantage of this approach is that it might
require a smaller sample than the direct measurement approach, because
meteorological variables would be handled with the simulation model,
                                     112

-------
                  Variab11ity  1" the data.  However, the approach would
                up?*r  bound1n9 arguments, approximations, and issues
^  Estimation of the Probability Distribution of Sulfuric Acid
     Exposures to  Travelers on a Roadway Network -

     In this method, either of the concentration distributions of the
preceding method would be used simultaneously with travel time data on
the same freeway/highway network to construct cumulative probability
distributions of traveler exposures on the freeway/highway network.

     The decision variable from either of the two exposure distributions
would be a specified percentile exposure measured as a time-averaged
sulfuric acid concentration.  Again, either of these exposure parameters
could be compared to a preassigned threshold value of exposure to deter-
mine when further growth in sulfuric acid emissions must be controlled.

     (a)  Direct Measurement Method for Exposure Distributions

          In this method we would construct a concentration probability
distribution by method (2-a).  We would then place this distribution
in the current algorithm which uses it along with travel time data to
produce a cumulative probability distribution of traveler exposures.  A
preassigned percentile exposure can then be compared directly with a
threshold value of exposure to determine whether control actions should
be taken.

     (b)  Combined Measurement and Simulation Method for Exposure
          Distributions

          In this method we would construct a concentration probability
distribution by method (2-b), and proceed as in method (3-a).


CONCLUSIONS

     We consider the direct measurement approach using random sampling
Imethods (2-a) or (3-a)l to be superior to the other two monitoring
approaches described here.  The recommended approach:

     1.   Relies on statistically meaningful quantities.

     2.   Is relatively free of problems of site selection.

     3.   Does not reouire the use of air quality simulation models.

     4.   Produces data which can be used to form a variety of
          policy-oriented indicators of sulfuric acid concentrations
          and exposures.
                                    113

-------
In contrast, the n-th highest concentration approach does not have
advantages 1 or 2 above, and the measurement/simulation approach
[methods (2-b) or (3-b)] does not have advantage 3.

     If, for some reason, the recommended approach cannot be adopted,
then we believe the combined measurement/simulation approach should
be used.  Next to the recommended approach, the measurement/simulation
approach would have the fewest disadvantages -- particularly if an
improved air quality simulation model is developed.  We would advise
against using the n-th highest concentration approach.  It cannot
produce statistically meaningful results; it cannot produce results that
can be compared with the existing exposure/concentration estimates; and
it has severe practical problems.

     If the combined measurement and simulation approach is used in
either methods (2) or (3), it may be a good idea to replace the
HIWAY model by an Air Quality Simulation Model (AQSM) which more
accurately simulates the physical processes of dispersion and convection
at wind speeds, wind angles, and stability conditions leading to the
higher pollutant concentrations.  Results of the Milford test track
experiment support previous evaluations of the HIWAY model.   These
results show that the HIWAY model performs poorly (or at best indifferently)
at low wind speeds, at low atmospheric diffusivity and when the wind is
nearly parallel to the road.  Each of these factors separately contributes
to increased concentrations.  When they exist simultaneously they produce
the highest concentrations at a fixed emission density and roadway geometry.
But the high concentration ranges predicted by the AQSM must be accurately
represented since they constitute the most likely regions for use in
decisions to initiate control actions.

     One final note on model development and validation is in order, since
there appears to be some confusion on its role relative to the monitoring
program discussed above.  New model  development, calibration and valida-
tion would not employ the data collected in the measurement program on
the freeway and highway networks mentioned above.  Use of that data for
the development of concentration frequencies on the freeway and highway
networks requires the existence of an accurate air quality simulation model,
which has been calibrated and validated against data sets developed inde-
pendent of the freeway/highway network measurements.   This does not mean
that the validation data could not be collected on the same network --
in fact, that would probably be the best place to collect it.  However,
the validation data-gathering exercise here would constitute an entirely
independent activity.
                                     114

-------
                               APPENDIX B-2
                   ,mc PROBABILITY  DISTRIBUTION OF ONE-HOUR
             MniP?SURES T0 AUTOMOBILE-GENERATED SULFURIC ACID  FOR
             MORNING PEAK PERIOD TRAVELERS  IN  LOS ANGELES


INTRODUCTION

     One  of  the  problems involved in  the development of a sulfuric acid
emissions standard for new cars is  the estimation of the sulfuric acid
exposures that would result from various levels of automobile emissions
ot sulfuric  acid.   This problem has been addressed in previous EPA papers
(Ref.  1,  2,  3),  and estimates  of "typical" and "worst case" exposures to
automobile-generated sulfuric  acid  for various emissions scenarios have
been presented.  These estimates place bounds on the range of sulfuric
acid exposures likely to be experienced as a  result of sulfuric acid
emissions from cars.   However,  the  estimates  provide no information about
the probability  of occurrence  of various levels of sulfuric acid exposure.
thus,  for example, the estimates cannot be used to answer questions such
as:  wh^t is the probability that 1-hour average exposures will exceed
15 ug/m   if  cars emit sulfuric  acid at a specified rate?

     The  failure of the previous estimates to answer questions of this
kind significantly constrains  their usefulness.  The "worst case" estimates
are based on the simultaneous  occurrence of various rare events (e.g.,
prolonged periods  of travel  on  high-volume freeways under meteorological
conditions conducive to the formation of unusually high concentrations
of pollutants) and, therefore,  may  represent exposure levels that are too
unlikely  to  be of  practical  importance.  On the other hand, the "typical
case"  exposure estimates provide no information about the reasonable
extremes  of  exposure and,  thus,  cannot be used in developing policies
to protect people  from these extremes.

     This paper  provides an estimate of the probability distribution of
short-term (1-hour average)  exposures to automobile-generated sulfuric
acid for  a particular urban sub-population, namely weekday morning peak
period travelers in Los Angeles.  The choice of Los Angeles as the city
for which the probability  estimates were made was dictated by data
availability; Los  Angeles  is the only city in the U.S. for which the
required  data were readily available.  The estimates were made for the
weekday morning  peak period (6-9 a.m.) because the traffic and meteoro-
logical conditions that occur during the period tend to cause 1-hour
average concentrations of  automobile-generated sulfuric acid to be
higher during this period  than  at other times (Ref. 4).  Thus, weekday
morning peak period 1 -hour-average  exposures to automobile-generated
sulfuric  acid are  likely to be  higher than 1-hour exposures experienced
at other  times.  The probability estimates were made for morning peak-
period travelers only.   This sub-population includes roughly 30 percent
of the total  Los Angeles population.  The excluded sub-population, persons
who do not travel  in  the morning peak, will be addressed in a subsequent
paper.
                                   115

-------
EMISSIONS FACTORS

     At present, techniques for measuring sulfuric acid emissions from
automobiles are not fully developed.  Consequently, sulfuric acid
emissions from current catalyst cars cannot be estimated with precision.
Therefore, the exposure estimates presented in this paper are based not
on current measured sulfuric acid emissions by catalyst cars, but on a
set of plausible (but hypothetical) sulfuric acid emissions factors.
The assumed emissions factors are 0.010 g/mi for cars on city streets and
0.035 g/mi for cars on freeways.  Although these emissions rates are
hypothetical, it is likely that they could be achieved by cars equipped
with catalysts if emissions standards were established at the appropriate
level?.  The exposure ~stimates assume that all cars in use emit sulfuric
acid at these rates.

     Data

     Two types of data were used in contructing exposure estimates:
travel data and meteorological data.  The travel data were obtained from
the 1967 Los Angeles Regional Transportation Study (LARTS).  These data
consisted of records of 36,733 trips made during the weekday morning peak
period (6-9 a.m.) in the Los Angeles area and estimates of the time
spent on freeways and city streets during each trip.  1967 is the only
year for which these data are available.

     The meteorological data consisted of hourly observations of wind
speed, wind angle, and stability class at Los Angeles International
Airport during the 6-9 a.m. period of each day during 1964.  1964 is the
only year and Los Angeles International Airport the only site for which
suitable meteorological data were available.  Meteorological conditions
tend to vary somewhat from year to year; in some years, conditions are
conducive to higher concentrations of pollutants than in other years.
Conditions in 1964 were roughly in the middle of this range and were not
conducive to forming either unusually high or unusually low concentrations
of pollutants (Ref. 4, 5).

     Methodology

     The travel data were used to construct the joint frequency distri-
bution of travel time on city streets and freeways during the hour
following the start of a morning peak period trip in the Los Angeles area.
Some features of this travel time distribution are shown in Table 1.  The
travel time distributions are skewed to the right.  This means that travel
times are relatively short for most travelers but are long for a few
travelers.  For example, 55 percent of peak period travelers do not use
freeways at all during the hour following the start of a morning peak
period trip.   However, one percent of travelers are on freeways for at
least 56 minutes during the hour following the start of a trip.
                                   116

-------
      ?L? f?r!  9lcal data were used together with the HIWAY diffusion
      i ™  6> t0,.9°nstruct the joint frequency distribution of 1-hour
        co"cftratlons of automobile-generated sulfuric acid en city
      r     Ire?ways  in the Los Angeles area during the 6-9 a.m. period.
      co.ncentrations  represent increments over the general background
      of automobile-generated sulfuric acid.  In performing this calcu-
i? nnn*  £* tlc volumes of 320° vehicles per hour on city streets and
i/,uuu vehicles per hour on freeways were assumed.  These volumes correspond
to observed peak-hour traffic flows on heavily-traveled streets and
freeways in Los Angeles (Ref. 7).  The sulfuric acid concentrations
obtained from HIWAY were divided by two to correct for the observed
tendency of HIWAY to  over-predict concentrations by roughly a factor
of two on the average (Ref. 8).  It was assumed that sulfuric acid
emitted by cars is not removed from the atmosphere by chemical reactions
or other means during the time period being modeled.

     Some features of the frequency distribution of concentrations are
displayed in Table 2.  Histograms of the marginal distributions are
shown in Figures 2 and 3.  The concentration distributions, like the
travel time distributions, are skewed to the right.  Thus, the peak period
one-hour average sulfuric acid concentrations on freeways can be as high
as 290 gg/m , but 99  percent of the time it is less than or equal  to
40 ug/m3.  On city streets, the sulfuric acid concentration can reach  3
22 yg/m , bt't 99 percent of the time it is less than or equal to 4 yg/m .*

     For fixed values of city street and freeway travel times and sulfuric
acid concentrations,  sulfuric acid exposure during the hour following the
start of a peak period trip was computed as follows:

     (1)  EXP = Cf Tf y Cs Ts y B

          where:

          EXP = One-hour average exposure to sulfuric acid
          Cf  = One-hour average sulfuric acid concentration on
                freeways (increment over background)

          C   = One-hour average sulfuric acid concentration on
           s    city  streets (increment over background)

          Tf  = Time  spent on freeways (in hours) during the hour
                following the start of a trip
          T   = Time  spent on city streets (in hours) during the
           s    hour  following the start of a trip
          B   = Sulfuric acid background due to sulfuric acid
                emissions from cars


*These concentrations represent increments over the general background
 level of automobile-generated sulfuric acid.
                                      117

-------
     The background sulfuric acid concentration, B, was estimated
using a simple box model and was found to be less than 1 yg/m  . o
However, for computational reasons, background was set at 1 yg/m
when performing the exposure computations.

     The probability distribution of EXP was computed from Equation
(1) by allowing Cf, C , Tf, and T  to be random variables with the
previously discussed probability Distributions for concentrations and
travel times.

RESULTS

     EXP estimates the one-hour average exposure to automobile-generated
sulfuric acid during the hour following the start of a weekday morningg
peak period trip in the Los Angeles area.  There were approximately 10
such trip starts in Los Angeles in 1967, the year of the LARTS survey.

     Significant features of the probability distribution of EXP are
displayed in Table 3.  Figure 4 shows a histogram of EXP.  The distri-
bution of EXP is skewed right and has a very long tail.  EXP can reach
a level of 2903iag/m  (the worst case).  However, the 99th percentile of
EXP is 12 vg/m , and the median value of EXP is less than 2 jjg/m .

     The values of EXP shown in Table 3 and Figure 4 as well as the
concentrations in Table 2 and Figures 2 and 3 were computed for a specific
set of sulfuric acid emissions factors, namely 0.010 g/mi on city streets
and 0.035 g/mi on freeways.  Exposures and concentrations for other sets
of emissions factors with the same ratio of freeway to city street
emissions can be obtained by linear scaling.  Thus, the exposures and
concentrations would be double those shown in emissions were 0.020 g/mi
on streets on 0.070 g/mi on freeways.  The concentrations and exposures
would be half those shown in emissions were 0.005 g/mi on streets and
0.0175 g/mi on freeways.

SOURCES OF ERROR IN THE ESTIMATES

     The exposure estimates are subject to two types of error:  systematic
biases and random errors.  Systematic biases create errors in the
estimated exposures to be mostly too high or mostly too low.  Random
errors cause the estimated exposures to be sometimes too high and sometimes
too low but correct on the average.

     The principal sources of systematic bias in the results presented
here are the street and freeway traffic volumes used to estimate sulfuric
acid concentrations.  These traffic volumes tend to cause the estimated
sulfuric acid concentrations and exposures to be too high, because
traffic volumes on most street and freeway segments are not as great as
                                     118

-------
the volumes used in developing the estimated concentrations.  Another
possible source of systematic bias is the absence of sulfuric acid
removal mechanisms in the HIWAY model.  If chemical reactions with
atmospheric ammonia or other mechanisms remove significant quantities
or sulfuric acid from the atmosphere over the period of one hour, the
estimated exposures again would tend to be too high.

     The principal source of random error in the estimates is the random
error of the HIWAY diffusion model.  This model, like other diffusion
models, produces concentration estimates that can be too high or too low
by up to a factor of roughly two.  This factor of two is distinct from
the HIWAY model's tendency to systematically overestimate concentrations
by a factor of two.  The latter source of error was removed in the
estimation procedure.  In addition, all of the travel and meteorological
data used in developing the exposure estimates are subject to random
error.  The most significant source of random error in these data may be
the use of instantaneous hourly meteorological conditions at Los Angeles
International Airport to represent hourly average conditions over the
entire Los Angeles area.

     It is not possible to estimate the magnitudes of the various errors
in the exposure calculations.  As a guess, it is possible that the
systematic biases might cause the exposure estimates to be too high by
as much as a factor of two and that the random errors might cause the
estimates to be too high or t©o low by a factor of two to three.

COMPARISON WITH PREVIOUS RESULTS

A previous EPA summary and evaluation of sulfuric acid exposure
levels (Ref. 3) concluded that if cars emitted 0.030 g/mi of sulfuric
acid on freeways, then peak 1-hour exposures of urban commuters to   3
automobile-generated sulfuric acid "are not likely to exceed 127 yg/m 3
except under very rare conditions, when exposures could reach 208 yg/m ."
When adjusted to reflect the 0.035 g/mi freeway emissions factor used
in the present paper, the earlier paper's conclusion becowes:  1 hour
sulfuric acid exposures are not likely to exceed 1483yg/m  except under
rare conditions, when exposures could reach 243 yg/m .  As explained in
the introduction to the present paper, the prior exposure estimates did
not include quantitative estimates of the likelihood of various exposure
levels.  The probability distribution developed in this paper indicates
that the^probability of a 1-hour exposure equal to or greater than
148 yg/m  is roughly 3.5 x 10" .  The probability of an exposure equal
to or greater than 243 yg/nr is roughly 10" .
                                      119

-------
CONCLUSIONS

     The probability distribution of 1-hour exposures to automobile-
generated sulfuric acid for weekday morning peak period travelers in
the Los Angeles area has been estimated.  The estimate is based on
travel and meteorological data for Los Angeles.  The results suggest
that if cars emit 0.010 g/mi of sulfuric acid on city streets and 0.035
g/mi on freeways, then:
                                                                 3
     •    Median 1-hour average exposure will be less than 2 yg/m .

     •    Average 1-hour exposure will be 2 pg/m .
                                                                 3
     t    The 99th percentile of 1-hour exposures will  be 12 yg/m .
                                                      3
     •    The maximum 1-hour exposure will be 290 yg/m  and will  occur
          with probability 6 x 10"'.
                                      120

-------
                                TABLE 1
          FEATURES OF THE DISTRIBUTION OF TRAVEL TIME DURING
            THE HOUR FOLLOWING THE START OF A MORNING PEAK
                     PERIOD TRIP IN LOS ANGELES1*J
Average
Median
95th Percent!le
99th Percent!le
Maximum
Probability of Maximum
Time on
Streets
11
9
29
43
60
0.001
Time on
Freeways
6
0(b)
32
56
60
0.008
Total
Travel Time
17
13
52
60
60
0.03
 (a) Travel  times  are  in  minutes.
 (b) 55  Percent of peak period  travellers do not use freeways during
    the hour following the start  of  a  trip.
                                      121

-------
                                TABLE 2
           FEATURES OF THE DISTRIBUTION OF CONCENTRATIONS OF
                 AUTOMOBILE-GENERATED SULFURIC ACID ON,
                 STREETS AND FREEWAYS IN LOS ANGELESia;
                                   Concentration on     Concentration on
                                       Streets              Freeways
Averaqe                                    1                    7
Median                             less than  2^              5
95th Percentile                    less than  2^             19
99th Percentile                            4                   40
Maximum                                   22                  290
Probability of Maximum                 8 x 10"5            8 x 10"5
(a) Concentrations represent increments  over background  in  yg/m  -
    one-hour average.
(b) 2 ug/m3 is 98th percentile.
                                   122

-------
                                TABLE 3
            FEATURES OF THE DISTRIBUTION OF ONE-HOUR AVERAGE
EXPOSURES TO AUTOMOBILE-GENERATED SULFURIC ACID DURING THE HOUR FOLLOWING
             THE START OF A PEAK PERIOD TRIP IN LOS ANGELES
                                             Exposure^
               Average                            2
               Median                        Less than
               95th Percentile                    6
               99th Percentile                   12
               Maximum                          290
               Probability of Maximum         6 x 10
 (a)  Exposure  in  yg/m
           o
 (b)  2  yg/m is 67th  percent!le.
                                    123

-------
   0.15 -
ui

-------
                 1.0-
                 0.8 —
                 0.6-
               ui

               O
               ui
               GC
               U.
                 0.4-
                 0.2-
                                      PROBABILITY OF CONCENTRATION

                                        GREATER THAN 6jug/ma Is 0.002
0246


     CONCENTRATION, fjg/m3
                                                     T
                                                     8
Figure 2. Histogram of city street sulfuric acid concentration, increment over background.
                                       125

-------
  0.25-
  0.20-
ui
3 0.15
in
oc
  0.10-
  0.05 -
PROBABILITY OF CONCENTRATION
  GREATER THAN 40(xg/m3 is 0.01
                                                                  -=*•
                       10
           I
           20
30
               40
                             CONCENTRATION, fig/m3
  Figure 3.  Histogram of freeway sulfuric acid concentration, increment over background.
                                       126

-------
  0.70-,
  0.60
  0.50 -
  0.40 -
CD
2
o
K
o.
  0.30 -
                                         PROBABILITY OF EXPOSURE
                                          GREATER THAN 14 /Lig/m3
                                                  is 0.006
  0.20  -
  0.10 -
458
     EXPOSURE, jifl/m3

Figure 4. Histogram of EXP.
                                                                14
                                 127

-------
                              REFERENCES
1.   Estimated Changes in Human Exposure to Suspended Sulfate
     Attributable to Equipping Light-Duty Motor Vehicles with
     Oxidation Catalysts, Environmental Protection Agency,
     January 11, 1974.

2.   Issue Paper:  Estimated Public Health Impact as a Result of
     Equipping Light-Duty Motor Vehicles with Oxidation Catalysts,
     Environmental Protection Agency, January 30, 1975.

3.   Evaluation of Sulfuric Acid Exposures from LDV Emissions,
     Environmental Protection Agency, June 2, 1976.

4.   G.C. Tiao, G.E.P. Box, and W.J. Hamming.  A Statistical  Analysis
     of the Los Angeles Ambient Carbon Monoxide Data, Paper No.  74-77,
     Air Pollution Control Association, June 1974.

5.   Ralph I. Larsen and Henry W. Burke.  Ambient Carbon Monoxide
     Exposures, Paper No. 69-167, Air Pollution Control Association,
     June 1969.

6.   John R. Zimmerman and Roger S. Thompson.  HIWAY:  A Highway Air
     Pollution Model, Environmental Protection Agency, December  1973.

7.   Telephone conversations with personnel of the Los Angeles City
     Traffic Division and the California Department of Transportation.

8.   Comparison of the EPA HIWAY Model with Data Collected by
     Environmental Systems Laboratory, EPA internal memorandum from
     William B. Petersen to Robert Papetti, October 28, 1975.
                                  128

-------
                 APPENDIX C
THIRD ANNUAL CATALYST RESEARCH PROGRAM REPORT
                 Prepared by

     Health Effects Research Laboratory
           Cincinnati, Ohio  45268
    U.S. ENVIRONMENTAL PROTECTION AGENCY
     OFFICE OF RESEARCH AND DEVELOPMENT
     HEALTH EFFECTS RESEARCH LABORATORY
     RESEARCH TRIANGLE PARK, N.C. 27711

-------
                           TABLE OF CONTENTS

                                                                      Page


List of Tables	   130

Li st of Figures	   130

  I. Overview

     A.   Facilities	   132
     B.   Research	   132

 II. Background and Introduction	   134

III. Current Status

     A.   Physical Facilities	   138
     B.   Automotive Emissions Study	   145
     C.   Diesel Emissions Study	   149
     D.   Ultrafine Sulfuric Acid and Metal Sulfates	   149
     E.   Manganese Oxide Studies	   152

 IV. Problem Areas	   153

  V. Plans for Future Research

     A.   Automotive Catalytic Emissions Study	   154
     B.   Automotive Diesel Emissions Study	   154
     C.   Ultrafine Sulfuric Acid and Metallic Sulfates	   155
     D.   Manganese Oxide Studies	   157

 VI. Conclusions	   158
                                   130

-------
                            LIST OF TABLES


                                                                      Page

1    Analytical Methods for Measurement of Atmospheric
     Components	   142


2    Modified California Cycle of Engine Operation	   144
                            LIST OF FIGURES   '


1    Toxicological Assessment of Diesel Emissions	   139
                                   131

-------
I.   OVERVIEW

     The Catalyst Research Program at the Health Effects Research
Laboratory-Cincinnati, has been designed to provide EPA with biolog-
ical data on the toxicologic effects of automotive emissions with and
without the catalytic converter and on related single pollutants.  In
1973, the Health Effects Research Laboratory-Cincinnati, began a
series of studies in response to EPA's need for inhalation toxicology
data on regulated and nonregulated emissions from oxidation catalysts.
The HERL research effort was divided into two major segments:  (1)
assessment of the biological effects of automotive emissions which
passed through the oxidative catalytic converter; (2) toxicological
studies on single pollutants (noble metals, sulfates) associated with
the oxidation catalyst.  Data from these studies have been submitted in
the annual Catalyst Research Program reports and have been published
in the scientific literature.  In the program for CY 1976, an additional
study on the biological effects of catalytic emissions was planned in
order to verify the findings of the previous catalytic emission study
and to initiate investigations on the biological effects of ultrafine
sulfuric acid.  The initiation of the program for CY 1976 was delayed
because of relocation to new facilities.  Much time and effort was
spent in moving and setting up new facilities and laboratory equipment.

     A.   Facilities

          During CY '76, a new mobile emissions animal exposure facility
was constructed at the EPA Center Hill Road laboratory.  This facility
replaces the Laidlaw facility which was used in previous studies.  In the
new facility, either gasoline or diesel exhaust can be generated independ-
ently and simultaneously.  The facility will serve as a center for toxico-
logic screening, in relevant animal models, of potentially hazardous
emissions derived from either gasoline or diesel automotive emissions.  It
thus provides for the assessment of the effects of the inhalation of non-
irradiated and irradiated (artificial sunlight) automotive exhaust upon
those biological parameters associated with general health and resistance
to disease.

     B.   Research

          The CRP plans at HERL-Cincinnati, can be categorized into
four major areas of research.  The areas involve the assessment of bio-
logical effects in laboratory animals following exposure to:  (1) auto-
motive catalytic emissions; (2) diesel emissions; (3) ultrafine sulfuric
acid and metal sulfates; and (4) Mn compounds associated with the use of
the additive MMT in gasoline.  During CY '76, research was initiated in
three of these areas—automotive catalytic emissions, diesel emissions,
and metal sulfates.
                                    132

-------
          1.   Automotive Catalytic Emissions Study

               A major research effort during late CY  '76 was the initia-
tion of a three-month exposure of laboratory animals to automotive
catalytic-treated exhaust.  Major emphasis was placed on defining the
effects on pulmonary function.  The exposure period has been completed
and data are being analyzed.  The reports should be completed in CY '77.
After completion of this study, there are no immediate plans for addi-
tional animal exposures to automotive catalytic emissions.

          2.   Diesel Emissions Study

               During the construction of the new Center Hill Road
laboratory, arrangements were made to include a diesel capability in
the facility.  Installation and calibration of all of the necessary
hardware required to operate the diesel engine is progressing satis-
factorily and animal exposures to the exhaust will begin in CY '77.
Major biological parameters to be investigated include effects on pul-
monary function, histopathology and blood chemistry, biochemical  altera-
tions in the lung, and effects on spontaneous locomotor activity.

          3.   Ultrafine Sulfuric Acid and Metal Sulfates

               Because of their presence in exhaust emissions of vehicles
with catalytic converters, respirable aerosols of sulfuric acid and metal
sulfates (aluminum sulfate, platinum sulfate) were tested using various
biological parameters.  Work on the metal sulfates is continuing.  The
exposures to sulfuric acid were done using particle sizes larger than
those in the ultraflne range.  Experiments in the ultrafine range await
development of a generation system capable of producing ultrafine particles,
A generation system should become available during CY  '77.

          4.   Manganese Oxide Studies

               The possible toxicity of airborne manganese has become
an issue of concern because of the potential widespread use of an organic
manganese compound (methylcyclopentadienyl manganese tricarbonyl) as an
antiknock agent in gasoline.  Studies involving inhalation exposure to
manganese will be initiated during CY  '77.  Major biological parameters
of interest include pulmonary changes and neurological alterations.   One
problem that must be solved before initiation of the experiments involves
verification of the chemical form of manganese found in automotive exhaust.

               It should be noted that most of the work conducted in
CY '76 was in-house; however, there was some contract research which
supplemented in-house work on the biological effects following exposure
to automotive catalytic-treated emissions and the effects of exposure
to sulfuric acid.  This work has been incorporated into the report.
                                   133

-------
II.  BACKGROUND AND INTRODUCTION

     The Health Effects Research Laboratory-Cincinnati feels that is
is important to maintain a research program directed toward obtaining
information that will lead to an improved understanding of the biomedical
consequences of exposure to mobile source emissions.  An additional stimu-
lant for research on mobile source emissions is the recognition that the
Agency needs a research system which allows exposure to total emissions of
alternative power and emission control devices for determining potential
health hazards.

     Mobile emissions (automotive, diesel, etc.) are comprised of a complex
mixture of interacting chemical substances of varying toxicities.  The
National Research Council document on Fuels and Fuel Additives for Highway
Vehicles and Their Combustion Products states that hundreds of compounds
are present in the emissions of fuel combustion and it is practically
impossible to detect and measure each one for the purpose of toxicologic
experiments.  A number of the components which may pose hazards to human
health have not yet been fully investigated.  We believe that the genera-
tion of automotive or diesel emissions in the laboratory setting is a far
better and more feasible simulation of the "real world" than can be
achieved "synthetically" by mixing combinations of pure (single pollutant)
compounds.  The  generation of whole exhaust requires careful control and
characterization but it is more practical in this sense than the monstrous
task of generating the number of single pollutants and combinations of
these.  It is a mistake to ignore the interactions of various components
which affect toxicity and which cannot be assessed in single pollutant
studies.  The potential biological hazards which may result from potentia-
tion, synergism and the various other interactions among all the exhaust
components can only be assessed by exposure to the total emission.
Whole exhaust studies should be regarded as a biological screen utilizing
various animal models for detection of biohazards of complex mobile emissions.
This is not to detract from single pollutant studies which should be
emphasized and coordinated with whole emission studies.

     When one considers the magnitude of human exposure and the size of
the automotive industry, it seems rather logical to suggest that the
Agency should have or support a mobile source emissions generation
capability which permits exposure of laboratory animals to total emissions.
It would appear that data on the effects of exposure to total emissions, as
well as data from exposure to single components of the emissions, would be
extremely useful in replying to inquiries from interested groups.

     The Health Effects Research Laboratory-Cincinnati has the physical
facilities for the inhalation exposure of laboratory animals to total
automotive gasoline or diesel engine emissions and also to single pollu-
tants.  The automotive emissions generation and exposure system which had
                                      134

-------
been previously located at the Laidlaw Avenue building was moved in
January 1976, to a laboratory on Center Hill Road and the exposure
chambers used for single pollutant studies were moved to the Environ-
mental Research Center.  The installation of all equipment has been
completed and all chambers are operational.

Current Research Activities and Rationale

     !•   Automotive Catalytic Emission Studies

          Since the inception of the CRP, the major research emphasis of
this laboratory has been to provide EPA with biological data on the
toxicological effects of animal exposure to whole emissions from automotive
engines equipped with and without the catalytic converter.  A secondary
aspect of the program involved studies on the metabolism and biological
effects of noble metals (Pt and Pd) and other nonregulated pollutants (H2S04)
associated with the use of the catalyst.

          Aerometry data from the use of the catalyst has shown a major
decrease in the concentrations of carbon monoxide and total hydrocarbons
in the exhaust.  The use of the catalyst also caused a decrease in sulfur
dioxide accompanied by an increase in sulfuric acid concentrations.  Previous
work conducted at HERL has shown that animals exposed to catalytically
treated auto exhaust fared better healthwise than did analogous aniamls
exposed to noncatalytic exhaust.  However, since sulfuric acid has been
shown to have a much greater irritant potency than sulfur dioxide, its
effects could offset the benefit gained by a decrease in other pollutants.
Thus an additional study was planned for CY '76 in which an engine-catalyst
combination (California hardware package) known to produce high concentra-
tions of sulfuric acid would be used along with increased levels of sulfur
in the gasoline.  The primary purpose of this study was the assessment of
the effects on the respiratory system.  In addition, the study was designed
to study information on the possibility of animal adaptation to the test
atmospheres.  The study was initiated in CY '76 and results will be availa-
ble in CY '77.

     2.   Automotive Diesel Emissions Study

          As the initial concerns of the use of the catalyst have been
delineated, some of the activities of the CRP have been focused into
related areas of research.  One of these areas involves studies on the
possible hazards of exposure to diesel exhaust.  Automotive manufacturers
have indicated the intention of increasing the production of diesel-powered
automobiles.  With the increasing number of diesel automobiles, EPA,
certainly will be asked about the hazards to health and well-being of man
exposed to these emissions and may be required at some later date to develop
health effects criteria for evaluating risk versus benefit.  The biological
data based on the health effects of diesel emissions is extremely limited
and inadequate for assessing possible hazards associated with increased
                                    135

-------
exposure.  The EPA has been engaged in the identification and charac-
terization of the diesel exhaust components.  With respect to emissions
of regulated pollutants, diesel engines have lower emissions of carbon
monoxide and hydrocarbons than gasoline engines.  However, this is not
the case with nitrogen oxides, where high combustion pressures inherent
in the diesel result in high nitrogen oxide emissions.

     The aspects of pollution most usually associated with diesel engines
by the public include visible smoke and odors, of which the former has
been regulated since 1970 engines built for service in heavy duty vehicles.
Participates emitted by diesel engines are largely carbonaceous material,
a portion of which falls within the range of particle sizes visible to
the human eye as smoke.  Smoke is influenced strongly by the size distri-
bution of the particles as well as their total mass; in fact, some approaches
to smoke reduction (metallic fuel additives) will actually increase total
particulate emissions on a mass basis.  Particulate emissions from diesel-
powered passenger cars are approximately 20-50 times higher than comparable
passenger cars burning unleaded gasoline.  Data also indicate that various
organics are absorbed on the surfaces of these carbon particles.  The
toxicity of these organics will be influenced by the sites of deposition
and clearance of the particulate.  Little biological information is
available on the deposition of diesel particulate in the lungs and this
is a major area of emphasis in the diesel study.

     The contribution of diesel automobiles to the emissions of sulfates,
or sulfuric acid, are of interest in view of the considerable attention
given this subject since the advent of catalyst cars.  The range for
sulfate emissions for diesel cars falls between the average catalyst car
without air injection and the average catalyst car with air injection.
Another fact that should be pointed out is that, whereas the sulfate
emitted by catalyst cars is essentially all sulfuric acid, it is not known
what types of sulfates are emitted by diesels, nor how much of it is sulfuric
acid.

     Initially, the major emphasis of the research approach will be in
evaluating the major target organ--the lungs.  The laboratory has extensive
experience in assessing lung changes by pulmonary function, biochemical
and histopathological approaches.  Other biological parameters such as
teratogenicity, mutagenicity, collagen formation, etc., have been considered
and will be evaluated through extramural support.  During the last quarter
of CY '77, the first diesel emission exposures will be initiated.  The
primary objective of the first diesel emission study is to gather sufficient
information for:  (a) preliminary toxicological assessment of the atmospheres;
and (b) development of a second more specialized protocol.
                                    136

-------
     3.   Ultrafine H2$0a and Metallic Sulfates

          The next major area of research involves the assessment of the
biological effects of exposure to ultrafine HpSCh and metallic sulfates.
Automobiles equipped with catalytic converters emit ultrafine sulfuric
acid. In order for EPA to evaluate the possible health risks associated
with ultrafine H2S04, animal exposure studies to ultrafine sulfuric acid
are needed in oraer to compare with results of exposures to larger size
HpSO* particles.  Amdur has used changes in pulmonary resistance of
gainla pigs as a measure of irritant potency of sulfate salts.  All of
her studies were done using particle sizes of at least 0.1 y or greater.
The trend of her data suggests an increase in resistance with decreasing
particle size (0.1 ym - .43% increase in resistance/ygS04/m  - 1.0 ym -
.20% inc. in resistance/ygSO.m ).  All of these studies involved acute
exposures and very little information is available on the effects of
long-term low-level exposures.  Questions arise as to the pathophysiological
significance of the acute bronchoconstrictive response and whether
adaptation occurs following chronic exposure.  What are the long-term
sequelae of low-level H2S04 exposure?

          This area of research involves exposures of laboratory animals
for a longer duration (for example - 3 months) to sulfuric acid and measure-
ment of pulmonary function, histopathological and biochemical changes.  The
toxicity of some of the metal sulfates will also be determined.

     4.   Manganese Oxide Studies

          The possible toxicity of airborne manganese has become an issue
of concern because of the potential widespread use of an organic manganese
compound (methylcyclopentadienyl manganese tricarbonyl - MMT) as an antiknock
agent in gasoline.  In addition to its use in gasoline, MMT is used as an addi-
tive to fuel oil for inhibiting smoke formation and improving combustion in
turbines.  The widespread use of MMT would result in an increase in the
concentration of Mn in the ambient air although the degree of environmental
impact is not precisely known.  Major contributors to atmospheric Mn include
ferromanganese blast furnaces, incinerators, coal-fired power plants, and
the burning of residual fuel oil.  For six types of coal-fired plants, the
concentrations of Mn emitted ranged from 60 to 400 yg/m .

          Manganese is an essential trace element in all living things and
is quite ubiquitous in its distribution in nature.  While in trace amounts,
Mn is beneficial, industrial exposure of man to various Mn compounds has been
associated with two different clinical syndromes:  manganous pneumonia and
chronic Mn poisoning affecting the CNS.  Manganese poisoning in man has been
primarily associated with mining and processing of Mn ores and in use of Mn
alloys in the steel and chemical industries.  This, coupled with poor
ventilation, provides the setting for most cases of chronic Mn poisoning
in the literature.  The lungs have been incriminated as the major portal
of entry in cases of Mn intoxication in man.  However, there is evidence
that much of the inhaled Mn is transferred by the mucociliary ladder to
the gastrointestinal tract from which it is either absorbed or eliminated.
                                    137

-------
          Our studies are designed to examine changes in the lung and CNS
effects.  A major portion of the Mn experiments will consist of long-term
(6 months to 1 year) inhalation exposure of laboratory animals to the
appropriate oxide of manganese.  The effects on the lungs will be assessed
histopathologically, biochemically, and by using pulmonary function
measurements.  Any neurological impairment will be determined through
behavioral tests and measurement of biochemical changes in neurotransmitters.
Many of the symptoms of chronic Mn poisoning are similar to Parkinson's
disease and both have as the basis of symptomatology abnormal functioning
of the extrapyramidal system.


III. CURRENT STATUS

     The program for CY '76 called for additional studies on the biological
effects of inhalation exposure to automotive emissions and to initiate in-
vestigations on the biological effects on inhalation exposure to ultrafine
sulfuric acid and metal sulfates.  However, the initiation of the program
for 1976 was delayed because of relocation to the new Environmental
Protection Agency center and to the Center Hill laboratory.  The construction
and testing of the new mobile emissions animal exposure system was completed
in mid-summer 1976.

     A.   Physical Facilities

          Briefly, the entire engine exhaust is mixed with filtered and
conditioned air in a dilution tube.  Diluted exhaust from the tube enters
a large volume mixing chamber and a portion passes through dynamic flow
irradiation chambers (to simulate sunlight) and is then conducted to
animal exposure chambers.  The system also provides nonirradiated exhaust
in the same concentration (directly from the mixing chamber) and filtered,
conditioned, ambient air for control animal exposure.  Figure 1 illustrates
the flow and essential equipment in the diesel emission system.  The
gasoline emission system is duplicated except for the common sharing of the
control air purifier and monitoring instrumentation.  The major physical
components of the facility are:  air purifiers, engine dynamometers,
mixing chambers, irradiation chamber, animal exposure chambers, and
monitoring instrumentation.

          1.   Air Purifiers

      -        There are three identical air purifiers and each supplies
21.2 m /min. (750 CFM) of CBR (chemical, biological, radiological)
filtered and conditioned air to maintain the required temperature and
relative humidity in the animal exposure chambers.  Three units are
required:  one for dilution air for the gasoline engine exhaust, another
for dilution air for the diesel engine exhaust, and one for the clean
air (control) animal exposure chambers.  The pneumatic controls are
completely automatic once the outlet temperature and relative humidity
changes when the hot and moist engine exhaust enters the system.
                                  138

-------
                                             AUTOMATIC
                                                CYCLE
                                             CONTROLLER
                                                 I
                                                DIESEL
                                                ENGINE
                                            DYNAMOMETER
                                                  I
AIR
PURIFIER


DILUTION
SYSTEM
VO
                                                                  IRRADIATION
                                                                   CHAMBER
      IRRADIATED!
        EXHAUST  [
                                                                                           NON-
                                                                                                   I
      IRRADIATED   I  	
                                                                                         EXHAUST
                           AIR
                         PURIFIER:
       CONTROL!
          AIR   i
                                                                                                                   INSTRUMENT
                                                                                                                    ANALYSIS
                                                                                                                    I
DATA
ACQUISITION
SYSTEM

                   AIR CONDITIONING
                   AND PURIFICATION
POLLUTANT j
GENERATION!
1    ANIMAL EXPOSURE!
{       CHAMBERS    [
I
I
MONITORING AND!
   RECORDING   I
                                         Figure 1. TADE project flow diagram: toxicological assessment of diesel emissions.

-------
          2.   Engine Dynamometers

               Mounted on one base plate are two 1975 General  Motors
350 CID 8-cylinder gasoline engines equipped with automatic transmissions
and pelletized noble metal oxidation type catalytic converters.  A dyna-
mometer power absorption unit is located between the transmissions which
can be coupled, one at a time, to the double-ended dynamometer.  The unit
is an air gap, eddy current, dual rotational type that employs water
cooling and has a power absorption rating of 50 horsepower in  a speed
range of 1600 to 6000 rpm.  A 450-lb fly wheel is directly coupled to
the dynamometer to provide the rotational energy to simulate the inertia!
mass of a moving 4000 Ib vehicle.  The dynamometer field current can be
set at constant value or made dependent on the flywheel speed.

               Another base plate has a Nissan CN6-33 diesel engine
coupled to a Chrysler Torque-Flite automatic transmission (Model A-727).
The engine is a six-cylinder unit, rated at 92 brake horsepower at 4000
rpm, and the displacement is 3.2 liters (198 CID).  The transmission is
connected to the same type dynamometer as for the gasoline engine and has
a 300-lb flywheel to provide an inertia mass of 3400 Ibs.

          3.   Mixing Chambers

               The reasons for the use of a mixing chamber in  the system
are to reduce concentration peaks to the nonirradiated exposure chambers
and to avoid problems associated with stream splitting and proportional
sampling.  The purpose of two mixing chambers is to allow simultaneous
but independent studies of either gasoline or diesel exhaust emissions.
Each chamber is 7.17m (23.5 ft) long b« 2.44m (8 ft) wide, and 1.22m (4 ft)
high, with a volume of 19.34m  (683 ft ).  The construction is a framework
of aluminum structural members with welded aluminum sheet metal panels.
Pipes from the mixing chamber supply the exhaust either directly to the
animal exposure chambers or to irradiation chambers enroute to exposure
chambers.  An adjustable pressure sensor controls a motorized  damper in
the vent line to the external atmosphere which maintains a positive
pressure in the chamber.

          4-   Irradiation Chambers

               The photochemical reactions that result from the exposure
of diluted raw exhaust to artificial sunlight take place in eight irradia-
tion chambers.  Fluorescent lighting panels composed of blue lamps,
black lamps, and sun lamps outside the chamber pass intense ultraviolet
radiation through windows of clear teflon FEP fluorocarbon film.  Two
irradiation chambers are needed to provide the atmosphere for each large
animal exposure chamber so that the average irradiation time of the
atmosphere is one hour.  Normal flow through the irradiation chamber is3
0.71 nr/min (25 cfm).  Each irradiation chamber has a volume of 19.34 m
(683 ft ) and they are stacked in pairs to conserve floor space.
                                   140

-------
          5.   Animal Exposure Chamber Rooms

               There are three animal3exposure chamber rooms and each
contains eight of the 2.83 m  (100 ft ) volume animal chambers.  Each
chamber room is individually air conditioned by a 5-ton, self-contained,
air-cooled unit located outdoors.  It is necessary to maintain controlled
temperature in the exposure rooms in order to prevent heat transfer
through the walls of the chambers and to avoid stressing the animals
when they are removed for examination or chamber cleaning.  The animal
exposure chambers are constructed of 14-gauge type 304 stainless steel
and are cubicle  with formed funnels at the top and bottom of each        3
chamber.  The normal air circulation rate is 15 volumes per hour or 0.71 m /
min (25 cfm).

          6.   Monitoring Instrumentation

               Sixteen exposure chambers are monitored continuously
(approximately once per hour) for carbon monoxide, total hydrocarbons,
nitric oxide, nitrogen dioxide, carbon dioxide, and sulfur dioxide.
Concentration measurements are collected via an electronic interface
between the analyzers and a Hewlett-Packard 3050-B automatic data
acquisition system with an attached calculator.

               Gas chromatographic measurements of methane, ethane,
ethylene, and carbon monoxide are performed periodically.  Aldehydes,
ammonia, and ozone are also measured occasionally by appropriate methods.

               Particulate is measured daily and analyzed for mass, and
hydrogen, ammonium, and sulfate ion concentrations.  Aerosol size distri-
bution, temperature and relative humidity are also determined daily.  The
analytical methods for measurement of the atmospheric components in the
animal exposure chambers are listed in Table 1.

          7.   Engine Selection Criteria

               In order to meet the needs of program planning under the
Clean Air Act Amendments, it was necessary that the engine selected for
the study be widely distributed and thereby representative of a typical
source of emission.  The General Motors Corporation is the largest
manufacturer of automotive engines and Its Chevrolet model has the greatest
distribution of all automobile engines.  Furthermore, the Chevrolet 350-CID
is the standard engine used as a baseline for emissions development research.
Since this engine was thought logically to represent the typical power
plant that other sectors of industry and government will be using, it was
selected for our studies to provide comparability and relevance.
                                   141

-------
                                TABLE 1

                 ANALYTICAL METHODS FOR MEASUREMENT OF
                        ATMOSPHERIC COMPONENTS
Component

Carbon Monoxide (CO)

Total Hydrocarbons (THC)

Nitrogen Oxides (NO )
                   X
     Nitric Oxide (NO)
     Nitrogen Dioxide (N02)

Carbon Dioxide (C02)

Sulfur Dioxide (S02)


Specific Hydrocarbons

     Saturated (C,-C,>)
     Olefin (C2)  '  *

Aldehydes (-CHO)

Ammonia  (NH-)

Ozone (03)

Parti cul ate

     Total Mass (M)
     Size           +
     Hydrogen Ion (H )
     Ammonium (NH.+)
     Sulfate
Chamber Conditions

     Temperature (T» °C)
     Humidity (% R.H.)
Primary Method of Analysis

Non-Dispersive Infra-Red (NDIR)
Flame-Ionization Detector (FID)


Chemi1umi nescence
Chemiluminescence
Non-Dispersive Infra-Red (NDIR)

Flame Photometric Detection (FPD)
of Pulsed UV Fluorescence


Gas Chromatography
Gas Chromatography
Colorimetry (MBTH)
Colorimetry (Phenol-Hypochlorite)

Chemiluminescence
Filter Sample Gravimetry
Electrical Precipitation (Mobility)
pH Measurement
Colorimetry (Phenol-Hypochlorite)
Thorin Titration (Optical End-Point)
Thermistor
Wet Bulb-Dry Bulb Relationship
                                   142

-------
               The diesel engine selected is made in Japan by the
Nissan Motor Company (makers of Datsun automobiles).  It is also the
same small engine which the International Harvester Company Introduced
in 1976 as an option on its new line of Scout II sport-utility vehicles,
and on Traveler station wagons and Terra ptck-up trucks.  For long-term
animaPexposure studies it is necessary to have automatic cycle control,
and this in turn requires an automatic transmission.  Thfs feature of
the Nissan diesel was a decisive factor in selecting it over the French
Peugot, which comes equipped only with a manual transmission.

          8.   Gasoline Engine Cycle

               The dynamometer driving schedule for the Chevrolet engines
consists of a repetitive series of idle, acceleration, cruise, and decelera-
tion modes of fixed time sequences and rates.  Table 2 shows this repetitive
series of a modified "California" cycle used in present fuel emissions
studies.

          9.   Fuel Selection and Storage

               The gasoline selected for use as a standard reference is
American Unleaded 91 Octane Test Fuel, Intermediate Grade ("indolene
clear").  As a reference, it is important that it be of quite precise
and reproducible composition and character.  This includes an absence of
lead and other additives (except as specifically noted), and should also
represent high volume "regular" market gasoline.  This gasoline has been
used for emissions studies in research and development by industry and
other agencies.

               There are numerous Federal military and commercial diesel
fuel specifications and classifications used by industry.  For automobiles,
highway trucks, tractors, marine, and off-road equipment, No. 2 diesel is
the typical fuel requirement and therefore was selected for the Nissan
engine.  A fair comparison between the statutory emission levels and
those of the engine under test should be made with a fuel that is repre-
sentative of that in actual commercial use.

               There are two underground fuel storage tanks of 2200 gallon
capacity, with gasoline stored in one tank and No. 2 diesel fuel in the
other.  Each tank is equipped with an electric fuel pump to deliver fuel
from the main storage supply to a smaller companion tank located outside
the building near the engine area.  These tanks are provided with a special
connection to permit the introduction of fuel additives so that 50 gallon
batches can be prepared without contaminating the main tank.
                                     143

-------
                                TABLE 2
             MODIFIED CALIFORNIA CYCLE OF ENGINE OPERATION
Mode                          Speed (M.P.H.)            Time (Seconds)
Idle                               0*                       20
Acceleration                    0 to 30                     14
Cruise                            30                        15
Deceleration                   30 to 15                     11
Cruise                            15                        15
Acceleration                   15 to 48                     29

Peak                           48 to 50                      1.5
Deceleration                   50 to  0                     31.5
                                             Total          137
*Drive shaft 100 rpm
                                     144

-------
     B.   Automotive Emissions Study

          After installation of the physical facilities, a series of
animal exposures involving different species were initiated for the
purpose of assessing biological effects on the respiratory system.
The automotive exhaust generation system consisted of a 1975 Chevrolet
engine equipped with the California hardware package, converter, and
associated equipment as described under Physical Facilities.  The
engine was run 16 hours daily and cycled continuously using the 7-mode
cycle.  The fuel consisted of unleaded 91 octane indolene with thiopene
added to produce a high sulfur fuel (0.10% S by weight).

     The study consisted of the following experiments which were initiated
in CY  '76:  (1) effects of exposure on pulmonary function in guinea pigs;
(2) effects of exposure on pulmonary function in cats; (3) effects of blood
gases and pulmonary function in rats; and (4) effects of exposure on suscep-
tibility to respiratory infection.

          1.   Effects of Exposure on Pulmonary Function in Guinea Pigs

               The goal of the experiment was to determine the effects of
irradiated and raw catalytically altered exhaust on guinea pig growth rate,
cardiac and pulmonary function and respiratory system histology.  The design
of the investigation included two interrelated parts—acute and chronic
exposures.

               a.   Acute Study - Three groups of guinea pigs were measured
in the study.  Animals were young and of the same weight as the average
six-week-old guinea pig (400 ± 50g).  All animals were treated identically
with the exception of the exposure to the test gas; one group was exposed
to clean air, one to raw catalytically altered auto exhaust and one to
irradiated catalytically altered auto exhaust.  Exposure time to each test
gas was one hour (concentrations were the same as used for the chronic study).
Animals were randomly assigned to each group.

                    Sulfur dioxide (SO, has been reported to be a mild
respiratory irritant, the inhalation of which results in bronchoconstriction.
One-hour exposures have demonstrated an increase in airway resistance in
guinea pigs, cats, dogs, and man.  Sulfuric acid mist has been shown to have
a much greater irritant potency.  Both contaminants are found in measurable
quantities in the catalytically altered emission atmospheres.  We were
interested in quantitatively comparing data from animals exposed to emissions
with data from those exposed to a known irritant.  For this reason, we
conducted a parallel study using S02 alone.

               b.   Chronic Study - As in the acute study, three types of
test gas (clean air, raw catalytically altered auto exhaust and irradiated
catalytically altered auto exhaust) were used.  A total of sixty guinea
pigs were exposed to each gas type for consecutive periods of six and twelve
                                  145

-------
weeks (16 hr/day, 7 days/week) starting the first week of life.  Thirty
animals from each group were exposed for each time period.  The sixty
guinea pigs of each group were divided equally into three exposure
chambers (20/chamber) where they lived and were exposed to the test gas
for 16 hours each day; there were three different chambers for each
group.  Exhaust concentration for each group was maintained at a constant
dilution and characterized throughout the time of the experiment.  Animals
were staged into and out of the exposure periods.  The schedule was
arranged so that baby animals were randomly selected from litters each
week, and distributed between the three test atmospheres.  Each animal
was numbered, half were tested after six weeks of exposure and half after
twelve weeks of exposure.

               c.   Methods and Equipment - Physiological methods for
evaluation of pulmonary function of guinea pigs have been developed and
described by Amdur and Mead.  Three simultaneously recorded measurements
are necessary—intraplural pressure, tidal volume, and air flow rates
into and out of the lungs.  Certain modifications of their techniques
were made by our laboratory to eliminate the need for anesthesia and
surgery on the day of measurement.  A body plethysmograph and single
animal exposure chamber was designed and constructed that allowed for
physiological measurement of unanesthetized guinea pigs during exposure to
test atmospheres.  Pressure, flow, and concentration of each test atmos-
phere was controlled in this chamber.  A Statham P23Db pressure transducer
was used for intraplural pressure measurements.  The transducer was in
direct communication with a saline-filled silastic catheter inserted into
the guinea pig intraplural space.  A Statham PM5 differential pressure
transducer, connected to the body plethysmograph was used to measure
tidal volume.  Flow was measured by electrical differentiation of the
tidal volume signal with respect to time.  Pulmonary flow resistances were
computed at maximal flow rates and expressed in cmHpO/ml/sec.  Dynamic
compliance was obtained at maximum tidal volume and expressed as ml/cmhUO.

                    Both phases of the experiment (acute and chronic)
started on November 15, 1976.  Guinea pigs were placed in exposure chambers
for the chronic study and measurements on acute study animals were initiated.
Infant guinea pigs were continually placed in the chambers until December
29, 1976.  Chronic exposure of guinea pigs was scheduled for completion
March 25, 1977, and physiological measurement of all animals, acute as well
as chronic, will be implemented by June 1977; a final report on this work
is anticipated by October 1977.

          2.   Effects of Exposure on Pulmonary Function in Cats

               This experiment was designed to determine the effects on
inhalation exposure to catalytically treated exhaust upon pulmonary
function and histopathology.  Cats were used as experimental animals because
they are a larger species and this facilitates the use of a wider range of
pulmonary function tests.  Also, comparison of the pulmonary effects in
several species would aid in the extrapolation to humans.
                                      146

-------
               The cats were divided into two groups with one group
exposed to exhaust and the other group to clear air.  The animals were
exposed for 16 hours per day for 90 days.  After the completion of the
exposure period, the cats were removed individually from the exposure
chambers, anesthetized, and prepared for pulmonary function measurements.
Pulmonary function measurements with the exception of helium washout
were conducted with the aid of an animal plethysmograph.  Following
anesthesia and tracheotomy, the cat was placed on its back in the plethys-
mograph.  The door was sealed, the tracheal cannula connected to the
pneumotachometer, and the intraplural catheter connected to the pressure
transducer.  Intraplural pressure, respiratory volume and flow were
recorded.  The cat was then automatically respired by cyclic pressure
changes with a range of ±20 cmhLO pressure at rates of 40, 70, 110, 140,
and 160 cycles/minute. Dynamic Compliance and resistance were calculated.
Expiratory flow volume measurements were then determined by suddenly
decreasing the pressure in the plethysmograph to -60 cmH^O pressure
followed immediately by a sudden increase in pressure to 60 cmhLO.  At
least four maximal expirations were recorded with two plotted as volume
and flow versus time and two plotted as flow versus volume.  The cat was
then removed from the plethysmograph and connected to the Howard respiration
pump.  Stroke volume was set at 15% of vital capacity.  The cat was
hyperventilated for several minutes to inhibit respiratory efforts.  Ten
percent helium was then inspired until the expired helium concentration
was stable.  The animal was then switched to room air.  Helium concentration
was monitored and the number of breaths required to reduce the expired
He to 1% of the original concentration was recorded.  The measurements
were repeated at 25, 50, 75, and 100 breaths per minute.  After completion
of the pulmonary function measurements, the cats were sacrificed and
tissues taken for histopathology.

               Measurements will be completed in April 1977, and a
report prepared during CY  '77.

          3.   Effects of Exposure on Arterial Blood Gases and
               Pulmonary Function in Rats

               Groups of 20 rats each were exposed 16 hours per day for
45 or 90 days to the automotive catalytically treated exhaust.  Parameters
measured include body weight, food intake, lung weights, static lung
compliance, functional residual capacity, hematocrit, hemoglobin, Po2,
and Pco2-

               Following completion of exposure, the rats were removed
individually from the chamber, briefly anesthetized and catheter implanted
in the caudal artery.  After regaining consciousness, an arterial blood
sample was collected.  The rat was then placed in a small chamber with the
arterial catheter and a rectal temperature probe ducted to the exterior.
The chamber was flushed with 100% oxygen and a blood sample collected after
10 minutes.  The chamber was then flushed with 9% oxygen for an additional
                                   147

-------
10 minutes and a final arterial blood sample collected.  Analysis of
arterial Pop and Pco, were carried out on all three blood samples.
Hematocrit and hemoglobin content were measured on the first sample.
After completion of blood collection, the rats were reanesthetized and
a cannula placed in the trachea.  Functional residual capacity (FRC)
was measured by injecting a known volume of gas containing 10% helium
into the lungs at the end of a normal expiration.  An equal volume was
withdrawn 10 second later and the helium content analyzed and the func-
tional residual capacity calculated.  After measurement of FRC, the rat
was administered pure oxygen for 5 minutes to flush out all nitrogen.
The animal was then asphyxiated, the chest wall incised and the aorta
and vena cava severed.  The lungs were filled with air to a pressure of
30 cm HpO.  Volume was recorded at each pressure with the final recording
at ambient pressure.  Deflation curves of transpulmonary pressure plotted
against percent of total lung volume were made for each animal.  After
compliance was measured, the lungs were removed, weighed, and fixed for
histopathology.  Analysis of data from these animals will be completed in
CY '77.

          4.   Effects of Exposure on Susceptibility to Respiratory
               Infection

               As part of the multidisciplinary effort to evaluate the
biological effects of inhalation exposure to automotive catalytically
treated exhaust, a series of experiments were performed in mice to assess
the effects on susceptibility to infection by pathogenic microbes.  The
basic experimental design used in the infectivity studies consisted of:
(1) a control group which was exposed to clean air, the removed and exposed
in a separate chamber to an atmosphere containing suspended aerosol of
pathogenic bacterial broth culture (challenge) followed by observation
for morbidity and mortality; and (2) the experimental group was exposed
to diluted exhaust atmosphere followed by exposure to the infections
aerosol challenge, followed by observation for morbidity and mortality.
Experimental group data (mortality and body weight) were then compared
with control data to detect the relative difference reflecting exhauust-
exposure effect.  The experimental hypothesis asserts that exhaust exposure
will  enhance susceptibility to lethal infection, so that mortality in the
exposed group will exceed the control group.  Principal variations within
the basic design framework included:  (1) type of exhaust: (2) duration
of test exposure; (3) single versus multiple intermittent infectious
aerosol exposure; and (4) age of animal.

               Preliminary results indicate a significant enhancement of
susceptibility to lethal infection following exposure to catalytic auto
exhaust.  There was a significantly greater enhancement effect in mice exposed
to irradiated exhaust.  Early indications suggest that the greater irradiated
exhaust effect may be attributable, at least in part, to a higher concentra-
tion  of nitrogen dioxide.  Further interpretative efforts in relation to
exhaust components must await finalization of aerometry data.  There was
                                 148

-------
no clear evidence of a dose-effect relationship with respect to duration
of exposure.  The effects of enhanced susceptibility to fnfection was,
in most respects, as great, or nearly so, in the acute as In the more
prolonged experiments.  Analysis of the data will be completed in CY
'77.

     C.   Diesel Emissions Study

          Installation of much of the hardware for the diesel study was
completed in CY  '76 as described in Physical Facilities.  Protocols have
been developed for the operation and measurement of the emission atmos-
pheres.  After completion of the automotive emissions study in CY '77,
the diesel engine will be put into operation and the development of the
engine cycle will be tested along with characterization of the emissions
in various locations within the system.  Following completion of engine
testing and emissions characterization, biological exposures will be
initiated.

     D.   Ultrafine Sulfuric Acid and Metal Sulfates

          The current status of investigations of the biological effects
of exposure to HpSO,, and other metal sulfates are described below.  The
work was conducted Both in-house and extramurally.  The exposures to
sulfuric acid did not include the ultrafine range because development of
the generation system has not been completed.

          1•   Pulmonary Effects of Inhaled Aluminum Sulfate and
               Sulfuric Acid

               Aluminum sulfate and sulfuric acid were selected for use
in a study designed to establish whether toxicity of sulfates is due to
the sulfate anion (S04=) or to the associated cation metal or proton.
The aluminum salt was chosen because (a) the base material of certain
converters is alumina, (b) the presence of this metal in airborne particles
is almost inevitably due to its abundance in the earth's crust, (c) the
burning of coal may increase atmospheric levels of aluminum, and (c) it
is considered to be relatively non-toxic.

               Pregnant Sprague-Dawley rats, with a mean body weight of
200 g., were divided into three groups and placed in inhalation chambers.
The control group received clean aiic (C); the aluminum-sulfate-exposed
group  (AL) received 2.4 ± 1.07 mg/m  A12(S04)3, the mean particle size
was 1.6 ±30.25 y; the sulfuric acid-exposed group (SA) received 4.05 ±
1.63 mg/m , for the first 2 weeks, thereafter the concentration was
adjusted to 4.80 ± 0.51 mg/m , with mean particle size of 2.75 ± 1.25 y.
The lengths of exposure for the dams (D) were 67, 66, and 65 days for D-
C, D-AL, and D-SA groups, respectively. At the end of exposure, 10 dams
and 10 pups, chosen at random, were killed and studied for pulmonary
effects.
                                  149

-------
               The parameters monitored in this experiment were primarily
those involving the alveolar macrophage,  population and morphology, the
level of lysozyme spilled in the alveolar space and the permeability,3ff
the lung wall as measured by the leakage  of i.v.-injected albumin     I
into the alveolar fluids.  The selection  of these parameters  is based on
the hypothesis that injury to macrophages by the inhaled toxicant with
consequent spillage of lysosomal enzymes  in the alveolar fluid will
initiate an inflammatory process and produce an alteration of the permeability
of the alveolar wall.  We also measured serum trypsim inhibitory capacity,
serum proteins and enzymes such as GOT, LDH, and GPT.

               The results indicated a significant increase in the number
of alveolar macrophages, as collected by  lung lavage, was observed in both
AL and SA groups; however, the morphology of the macrophages  from AL group
presented abnormalities such as swelling  and granulation while the macrophages
from SA group appeared normal.  Lysozyme  activity in the cell-free lavage
was shown by the radioactivity of the lung lavage was increasedo5-fold
in the D-AL group but remained normal in  the DSA group.  When    I counts in
the lavage were plotted against lysozyme  activity, we obtained a high
correlation coefficient (r ± 0.904, n = 29) with less than 10% relative
standard deviation of the slope.

               We also obtained the electrophoretic patterns  of serum
proteins in pups.  The data showed significant alteration in  albumin,
a-,, ou, & and  y-globulins.  Serum albumin levels were reduced below
control values by both sulfuric acid and  aluminum sulfate inhalation.
Aluminum caused a reduction below controls in levels of a -globulin,
a2-globulin and y-globulin, while sulfuric acid caused a reduction in the
a2-globulin level but raised the Y-globulin level considerably.

               These data have shown that:  (a) the irritating effects of
sulfuric acid and aluminum sulfate on lungs were reflected in the number
of the alveolar macrophages collected in  the lung lavage; (b) the morpho-
logical changes of the collected macrophages in the aluminum sulfate-exposed
animals, which did not occur in the sulfuric acid-exposed groups, suggest
that the cation is responsible for this effect; (c) the spillage of lysosomal
enzymes, as shown by lysozyme activity, may be due to the deleterious effect
of Al   on the macrophage and other luncucells; (d) the change in the lung
wall permeability, as shown by albumin-   I counts in the lung lavage, was
related to the high concentration of lysozyme in the cell-free lung fluid; and
(e) sulfate by itself, in the form of sulfuric acid and at the concentration
used, did not seem to produce pronounced  effects on the lung as did aluminum
sulfate, and serum protein profiles were  disturbed by both sulfuric acid
and aluminum sulfate, but in different ways to to different degrees.
                               150

-------
     From these results, it seems to follow that aluminum cation is more
irritating to rat lungs than is either the hydrogen ion or sulfate anion
(SA) and that much more attention and concern should be given to the
health effects of cations in the environment than to the small amount of
sulfuric acid produced by the catalytic converter.

          2.   Effects of Exposure to Sulfuric Acid and Platinum Sulfate
               Aerosols on Susceptibility to Respiratory Infection from
               Inhaled Bacterial Aerosol

               Sulfuric acid and platinum sulfate aerosols were tested for
their relative toxicity with respect to influencing susceptibility to
inhaled pathogenic bacteria.  Female mice aged 3-4 weeks were exposed to
groups of 10.  Each experiment constituted a 2 x 2 factional design and
included exposure to:  (a) neither test material nor infectious challenge
(controls); (b) infectious aerosol challenge only; (c) test material
aerosol only; or (d) test material aerosol plus infectious aerosol  challenge.

       -       Two concentrations of sulfuric acid aerosol (242 and 26.7 mg
H2S04/m , respectively; particle size estimated at 0.82 and 0.48 unuMMAD,
respectively) and one of platinum sulfate (13.1 mg Pt (S04)2.4H20/m  or 5.6
mg Pt/m ; particle size estimated at 0.4-0.5 ym MMAD) were tested using an
exposure duration of 3.62 hours.  Exposures were followed by infectious
challenge.  The mice were observed for mortality, body weight change, and
overt toxicity or illness for at least 15 days following challenge.

               Lethal response to infective aerosol exposure was signifi-
cantly (p<0.05) in the mice exposed to platinum sulfate plus infective
aerosol relative to the corresponding infection-only group.  Apart from
apparent irritant responses during exposure, no gross signs of overtly
toxic effects were obvious in mice exposed just to pollutant aerosols.

               These results indicate that inhalation of the high,  but
not the low, level of sulfuric acid aerosol, and of the platinum sulfate
aerosol at a relatively low level, resulted in enhanced susceptibility
(impaired resistance) to disease induced by inhaled pathogens, and further,
that the metallic sulfate aerosol was comparatively much more toxic in
this respect than the sulfuric acid on a mass dose basis.  These findings
are consistent with other work indicating greater toxicity of certain
metallic sulfates than of sulfuric acid.

          3.   Effects of Exposure of Rats to Sulfuric Acid. Sulfur
               Dioxide, or Aluminum Sul fate"

               Rats were exposed continuously for periods ranging from 6
to 14 weeks to sulfuric acid, sulfur dioxide or aluminum sulfate.  Pulmo-
nary function studies indicated that the aluminum sulfate was most
detrimental.  Exposure to aluminum sulfate at a concentration of 2.04 mg/nr
and having a mass median aerodynamic diameter (MMAD) of 2.0 pm resulted in
increased static deflation volumes in juvenile rats but exposure decreased
                                151

-------
the static deflation volume in the guinea pig.  Exposure to either
sulfuric acid (4.05 mg/m , MMAD of 0.5 ym) or aluminum sulfate caused
an increased pulmonary resistance and respiratory rate in adult rats.
Spontaneous locomotor activity during exposure and treadmill performance
for up to 40 days post-exposure were significantly decreased by exposure
to aluminum sulfate having a MMAD of 1.4 ym and a concentration of 2.59
mg/m .

          4.   The Direct Measurement of Sulfuric Acid Mist in the Lung

               The object of this study was to directly measure the depth
of penetration of sulfuric acid particles in the lungs of beagle dogs.
This was to be carried out by implanting a small catheter containing a
filter in an airway no larger than 1-2 mm in diameter.  The sulfuric acid
particles would be collected on the filter and visualized directly with the
aid on an electron microscope.

               Numerous delays and difficulties arose from attempting to
collect and characterize sulfuric acid particles on the filter.  The
contract effective date was October 15, 1975, with a six-month performance
period.  This was eventually increased to fourteen months with a final
date of December 15, 1976.  The investigators were finally successful in
generating sulfuric acid mist of the correct particle size, collecting
the mist on a filter and counting the particles.  When an attempt was made
to use this system in the dog, however, no sulfuric acid particles could
be collected.  This was true even if the tube and filter were placed at
the upper end of the trachea near the source of the mist.

               At this point, time had run out and funds were depleted.
Although possible reasons for failure were discussed, an adequate explana-
tion was not achieved.  A decision was made not to fund the project further
since we had no assurance that it would be carried out successfully.


     E.   Manganese Oxide Studies

          In October 1976, it was decided to reprogram some of the efforts
in the Catalyst Research Program to address the problem of manganese
toxicity following inhalation exposure.  The lung has been incriminated
as the major portal of entry in case of Mn intoxication in man.  However,
there is evidence that much of the inhaled Mn is transferred by the
mucociliary ladder to the gastrointestinal tract, from whichgit is either
absorbed or eliminated.  Mena, et al. conducted a series of   Mn exposures
in humans using oral and inhalation routes.  From these exposures, the
investigators concluded that the gastrointestinal tract was the portal of
entry for most of the inhaled Mn.  This observation is in conflict somewhat
with those of other investigators.  In the case of lead, it has been
estimated that 30 to 50% of inspired Pb is absorbed by the lungs and 10%
swallowed Pb by the gastrointestinal tract.
                                   152

-------
               Preliminary experiments in,-this laboratory have indicated
a marked difference in the absorption of   MnSO, when administered orally
and intratracheally.  However, intratracheal administration has a number
of disadvantages which may affect retention.

               Current experiments are designed to investigate the reten-
tiog. tissue distribution, and excretion following acute inhalation exposure
to   MnSO, and   Mn203 aerosols.  Comparisons will be made as to the effect
of chemical form on the retention of manganese.


IV.  PROBLEM AREAS

     Conducting inhalation research is expensive, especially if the tasks
involve the investigation of biological effects of exposure to whole emissions
from automotive engines.  The Health Effects Research Laboratory-Cincinnati
has the physical facilities for inhalation exposure of laboratory animals
to total automotive and/or diesel emissions and to single pollutants.  There
are 25 large exposure chambers associated with the automotive exhaust genera-
tion system and 46 exposure chambers available for single pollutant studies.
The current funding is not sufficiently large to permit full utilization of
the exposure chamber space.  The current budget is barely adequate to meet
salary and operation costs.  In-house funds per man-year were $27.18K which
is certainly minimal for this type of work.  Existing contracts were funded
from funds available in past years and there are no funds available for
renewal of extramural support.

     Other problems involve technical aspects of individual studies.  In
the manganese studies, the determination of the appropriate chemical form
of Mn for use in the inhalation studies is of concern.  Based on data
supplied primarily by industry, it has been assumed that the combustion
product of methyl cy cl open tad ienyl manganese tricarbonyl was MrioO*.  However,
recent data has indicated that Mn304 was also found in dilution tunnel
sweepings.  Further, in a memorandum from Mr. Thomas Murphy, Deputy Assistant
Administrator for Air, Land, and Water Use, EPA, it is stated that the probable
species is Mn203 (actually MnO(OH)).  The oxidation state of Mn is important
to determine Because some investigators have postulated that the lower
oxidative states are more toxic than higher ones.  This has often been
reported but not experimentally documented.  Indeed, dose-response relation-
ships have not been done to determine if the toxicity of manganese is
influenced by its physical form.  Initiation of the ultrafine sulfuric acid
study awaits the arrival of a generation system being developed through a
contract by EPA-Research Triangle Park.  However, HERL-Cincinnati, through
modification of an existing generation system, possibly has an ultrafine
sulfuric acid generation system that, according to preliminary measurements,
is producing an acid aerosol of about 0.02 ym mmd over the range of 10-100
yg/m  in a test chamber.  Current work involves confirmation of concentration
levels.
                                 153

-------
V.   PLANS FOR FUTURE RESEARCH

     Future research plans can be broken down into four major areas:  (1)
completion of current automotive catalytic emissions study; (2) studies on
the biological effects of exposure to diesel emissions; (3) biological
effects of ultrafine sulfuric acid and metallfc sulfates; and (4) toxicity
of airborne manganese oxides associated with the combustion of methyl-
cyclopentadienyl manganese tricarbonyl (MMT).

     A.   Automotive Catalytic Emissions Study

          The current exposures will be completed in April 1977.  Data
analysis and preparation of reports will be finalized in CY '77.

     B.   Automotive Diesel Emissions Study

          The Center Hill Road facilities to be used for the diesel exhaust
study are the same as have been used for the previous study of catalytic
converter exhaust emissions.  The most notable change will be the substitu-
tion of an automotive-type diesel engine.  Future plans for engineering and
aerometry in the diesel study during CY '77 require cycle testing and
characterization of the entire exhaust generation and animal exposure chamber
system using several steady state speeds before attempting to measure the
average emissions under automatic cycle control.  Since both the magni-
tude and nature of the losses will vary with engine speed and dilution
factor, the optimum operating conditions will be determined before the
animal exposures are started.  Material balances for carbon and sulfur
will require both tailpipe- and dilute-particulate sampling at steady
state speeds.

          The guideline for maximum deterioration allowable in engine
emissions will be based on Federal Register vehicle certification procedures.
Published emissions deterioration factors allowable over 50,000 miles can
be applied as a "not to exceed" limit.  In the chronic exposure, our opera-
tion will be 16-20 hours each day of the week at approximately 20 mph average
cycle speed, therefore, mileage accumulation could exceed 100,000 miles per
year.  Since this approaches half the life of the diesel engine, a spare
engine will be obtained if the feasibility of using the Nissan engine is
found to be acceptable for biological studies.

          Aerometry procedures and instrumentation will include the best
available approaches that will satisfy research requirements.  A quality
assurance plan has been developed and quality control procedures have been
formalized.  The aerometry staff will work closely with RTP laboratories
to insure that all quality measurements are accurate as the state-of-the-art
permits.  In addition to the in-house effort, there will be independent
audits of the aerometry measurements conducted by an outside contractor.
                                     154

-------
          During CY '77 a protocol for the first biological study that will
be initiated in the assessment of toxicity following exposure to dtesel
emissions will be prepared and the actual exposures will begin in November
1977.  The primary objective of the first dlesel emission study is to gather
sufficient information for:  (a) preliminary toxicological assessment
of the atmospheres; and (b) development of a second more specialized protocol.
It should be realized that this experiment is the first attempt at exposing
animals to diesel exhaust at HERL.  There is very little information in the
literature to serve as a guide for anticipation of possible biological effects.
This is particularly true for diesel particulate.  Data on the composition of
diesel emissions suggest a strong possibility of biological effects, considering
the content of irritant and aromatic hydrocarbons and the respirable particulate.
Major emphasis of the research approach in this protocol will be in evaluating
the major target organ—the lungs.  The laboratory has extensive experience in
assessing lung changes by pulmonary function, biochemical and histopathological
approaches.  Thus, as a starting point, this protocol will assess the effects
of inhalation upon the same biological parameters as measured in the automo-
tive catalytic emissions study.  This approach will permit comparisons with
previous automotive exhaust studies.

          The parameters to be measured in the first diesel exposure study
include:  (1) pulmonary function in rats, guinea pigs, and cata; (2) suscep-
tibility to infection in mice; (3) histopathology a d clinical chemistry in
rats; (4) biochemical alterations in the lungs using rats and guinea pigs;
and  (5) spontaneous locomotor activity in rats.  The duration of exposure
will be 20 hours a day for 30 days.

          Based on the data obtained from this exposure to diesel exhaust,
a more specialized protocol will be developed for the second study.  This
will include the use of extramural contracts supported by carry-over funds
to provide the capability in (1) mutagenicity-carcinogenicity testing of
diesel exhaust condensate, (2; measurement of collagen precursors in the
lung, (3) teratogenicity testing, and (4) a three generation study.

     C.   Ultrafine Sulfuric Acid and Metallic Sulfates

          The current studies on sulfuric acid and metal sulfates will
continue with major emphasis being shifted to exposures to ultrafine acid
as generation equipment becomes available.  These experiments include
exposures of guinea pigs to ultrafine sulfuric acid, short and intermediate
term effects of exposure of rats to ultrafine sulfuric acid alone or with
ozone, biochemical alterations in rats following exposure, and effects on
susceptibility to infection.
                                     155

-------
          1.    Guinea Pig Exposures

               Unanesthetized guinea pigs will be exposed for one hour to
fine particulate (O.Olu - O.ly) sulfuric acid at a concentration of
approximately 100 yg/m .   Pulmonary function measurements will be recorded
from each animal during exposure.  Measurements include airway resistance,
lung compliance, breathing rate, tidal volume, minute volume, and ECG.  To
obtain signals for these measurements, each animal is fitted with a saline-
filled intraplural catheter and placed in a body plethysmograph.  Three
groups of guinea pigs will be tested:  (1) Group I animals will serve as
controls; (2) Group II animals will be exposed to H2S04; and (3) Group III
will consist of sensitized animals exposed to the same test atmosphere as
Group II.  Temperature and humidity of test atmospheres will be controlled.

          2.    Exposure of Rats to Ultrafine Sulfuric Acid and Ozone

               Theoretical considerations suggest that the maximum conctn-
tration of ultrafine sulfuric acid attainable is approximately 100 yg/m .
While this concentration may produce no detectable toxicological effects
alone, its activity may well be enhanced by presence of other pollutants.

               The study using rats is designed to measure the toxicological
effects of an ultrafine sulfuric acid alone and in combination with ozone.
The emphasis will be upon functional and pathological alteration of the lung.
A wide range of concentrations of pollutants will be used initially.  Specific
questions to be considered include:  (a) the toxicological properties of
ultrafine sulfuric acid mist, (b) possible enhancement of the toxicological
properties of ozone including influence on threshold levels as well as mortality,
and (c) the influence of low levels of ozone upon detection limits for biolo-
gical changes due to ultrafine sulfuric acid.

               Groups of ten to twenty rats each will be exposed to the
maximum level of sulfuric acid attainable and sacrificed after exposure
periods, ranging from 1 to 28 days.  Separate groups will be exposed to
ozone concentrations ranging from 0.2 to 4 ppm alone and in combination with
sulfuric acid for similar periods of time.  The group size may be adjusted
depending upon within-group variance of the parameters being measured.  All
exposures will be eight hours per day, five days per week.  Body weight,
food consumption, general condition, and mortality will be monitored.  Meas-
urements will include arterial blood Po2> Pco,, pH, standard bicarbonate,
hemoglobin, hematocrit, white cell counts, differential counts, total lung
capacity, vital capacity, residual volume, functional residual capacity,
static pressure volume curves and histopathological changes.

          3.    Biochemical Alterations Following Exposure

               Biochemical alterations will be assessed in animals following
acute and chronic exposures to MnSO,, A12(S04)3, and ultrafine H?S04.  The
parameters that will be studied in these exposures are primarily those
                                 156

-------
involving the alveolar macrophage (population and morphology), the level
of lysozyme sptlled in the alveolar space, and the permeability,of the
lung wall as measured by the leakage of t.v.-injected albumfn -   I into the
alveolar fluids.  Additional work is planned to study the effects of these
aerosols on the connective tissue of the lung.  Lung protein synthesis
(collagen synthesis) will be investigated.

          4.   Ultrafine Sulfuric Acid Aerosol Infectivity Study

               Assess toxicity of ultrafine sulfurtc acid with respect to
effect on susceptibility to respiratory infection.  Mice will be exposed for
varied periods to ultrafine H2S04 aerosol and to a superimposed challenge with
aerosol containing pathogenic bacteria.  Comparative infective response in
test-versus-control subjects will discern induction of impared resistance
to infection.

     D.   Manganese Oxide Studies

          Plans for the Mn investigations are designed to examine changes in
the lungs and CNS after exposure.  The effects on the lungs will be
assessed using pulmonary function techniques, the infectivity model, bio-
chemical assays, and histopathological examination.  Any neurological impair-
ment will be determined through behavioral tests and measurement of biochemical
alterations in neurotransmitters.

          1.   Effects of Manganese Oxide on Pulmonary Function and Other
               Clinical Parameters

               Disease-free cats will be exposed to MnpCL continuously for
period of six,months.  Groups of 11 cats each will be exposed to clean air,
Mn203 25 yg/m  or Mn203 yg/m .

               Measurements will include body weights and food intake.  At
the end of exposure, the cats will be anesthetized and a series of pulmonary
function measurements made.  These will include maximum expiratory flow-volume
curves, dynamic compliance, frequency dependence of compliance, helium washout,
total lung capacity, vital capacity, residual volume, etc.

               Blood samples will be collected for hematocrit, red and white
cell counts, differential counts, pH, and standard bicarbonate.  The lungs
will then be infused with 10% buffered formalin at 30 cmH20 pressure and
examined for pathology.

               Biochemical and clinical chemistry measurements will also be
made on these animals, including Mn concentrations in the various tissues.
                                     157

-------
          2.   Effects of Manganese Oxide on Neurophysiological and
               Behavioral Parameters

               Adult rats will be exposed to manganese oxide via inhala-
tion at several different levels.  General toxicity tests, as well as
specific tests, assessing neurophysiological function and behavior will
be performed (spontaneous activity, electroencephalographic changes,
visual-evoked response).  After completion of this task, neonatal exposures
will begin.  Similar tests will be performed as above; however, an attempt
will be made to assess the effect of these exposures on the development
of the central nervous system.

          3.   Manganese Oxide Aerosol Infectivity Study

               Effects of exposure to airborne particulate manganese oxide
on susceptibility to respiratory infection will be determined.  Mice will
be exposed to manganese oxide aerosol atmospheres at varied levels and
for brief to moderate durations, and to a superimposed challenge with
pathogenic bacteria.  Infective response (morbidity, mortality) between
controls and exposed animals will be compared to determine if enhancement
of susceptibility has occurred.


VI.  CONCLUSIONS

     The Catalyst Research Program at the Health Effects Research Laboratory-
Cincinnati has been designed to provide EPA with biological data on the toxi-
cologic effects of mobile emissions, emission control devices and fuel
additives.  The major effort during CY '76 was the construction of a new
mobile emissions animal exposure facility at the EPA Center Hill Road labora-
tory and the installation of animal exposure chamber at the Environmental
Research Center.  The mobile emissions animal exposure facility was completed
in the Fall and animal exposures to automotive catalytic emissions were
initiated.  These studies will not terminate until April 1977, so no conclu-
sions can be reached on the biological effects of exposure.

     Exposure of mice to catalytic auto exhaust resulted in enhancement of
susceptibility to lethal infection.  There was significantly greater
enhancement of effect in mice exposed to irradiated exhaust than in those
exposed to nonirradiated exhaust.

     Inhalation exposure of rats to aluminum sulfate and sulfuric acid
indicated that the aluminum cation was more irritating to the lungs than
is either the hydrogen ion or sulfate anion.
                                      158

-------
     In rats and guinea pigs exposed to sulfuric actd, sulfur dtoxide or
aluminum sulfate, pulmonary function indicated that aluminum sulfate was
the most detrimental.  Exposure to either sulfuric acid or aluminum sulfate
caused increased pulmonary resistance and respiratory rate in adult rats.

     Spontaneous locomotor activity during exposure and treadmill performance
for up to 40 days post-exposure were significantly decreased by exposure
to aluminum sulfate.
                                                                        o
     Inhalation exposure of mice to a high concentration (242 mg H^
of sulfuric acid for 3.6 hours resulted in increased susceptibility t
infection.  Inhalation exposure of mice to a low concentration (13.1 mg
Pt(SCL)2 . 4H20) of platinum sulfate for 3.1 hours resulted in increased
susceptibility to infection.

     The CRP for CY '77 consists of completion of automotive catalytic
emissions study and initiation of research on the biological effects of
exposure to:  (1) diesel emissions; (2) ultrafine sulfuric acid; and (3)
manganese oxides.
                                      159

-------
                APPENDIX D
THIRD ANNUAL CATALYST RESEARCH PROGRAM REPORT
                 Prepared by

     Health Effects Research Laboratory
     Research Triangle Park, N.C.  27711
    U.S. ENVIRONMENTAL PROTECTION AGENCY
     OFFICE OF RESEARCH AND DEVELOPMENT
     HEALTH EFFECTS RESEARCH LABORATORY
     RESEARCH TRIANGLE PARK, N.C. 27711

-------
                           TABLE OF CONTENTS
Li st of Tab! es	   162
List of Figures	   163

  I. Background and Introduction	   164

 II. Overview

     A.   Animal Studies on Sulfuric Acid	   164
     B.   Animal Studies on Metals	   165
     C.   Human Studies on Sulfuric Acid	   166
     D.   Vegetation Studies on Sulfuric Acid	   167
     E.   Characterization of In-Use Vehicles	   167

III. Current Status

     A.   Animal Studies	   167
     B.   Human Studies	   183
     C.   Mobile Source Emissions	   190
     D.   Vegetation Studies	   208

 IV. Problem Areas

     A.   Ultrafine Sulfuric Acid Studies	   209
     B.   Nitrogen Dioxide	   209
     C.   Manganese Exhaust Products	   209
     D.   Diesel Exhaust	   209

  V. Plans for Future Research

     A.   Animal Studies	   210
     B.   Human Studies	   211
     C.   Mobile Source Emissions	   212
     D.   Vegetation Studies	   212

 VI. Conclusions	   212

Appendix D-l:  Aerosol Generation and Monitoring, Human Exposure...   214

References	   218
                               161

-------
                            LIST OF TABLES
 1    Effect of H2SO^ on Lymphocyte Transformation ............       169
 2    Effect of H9S(L Mist on Number of Antibody-Producing
     Spleen Cells... .........................................       170
 3    Effect of H2S04 Mist on Hematocrit ......................       170
 4    Clearance of Manganese from Mouse Lungs Following Inha-
     lation of Different Atmospheres for 2 Hours ........... . .       180
 5    Blind Exposure Protocol for H2S04 Aerosol ...............       187
 6    Time Course for Each of the Three Experimental  Sessions.       188
 7    Data List for Each Emissions Test .......................       191
 8    Classification of Cars and Tests ........................       193
 9    Average Emission Results; All Tests, All  Manufacturers..       195
10    Average Emission Results; General Motors  Vehicle Tests
     Only [[[       196
11    Average Emission Results; Ford Vehicle Tests Only .......       197
12    Average Emission Results; Chrysler Vehicle Tests Only...       198
13    Average Emissions; Mileage Accumulation (gm/mi)... ......       203
14    Sulfur Dioxide Purge Fractions ..........................       204
1 5    Summary of Test Resul ts for Car 1 .......................       206

-------
                            LIST OF FIGURES

                                                                    Page

 1    Effect  of  a  2-Hour Exposure to HpS04 Mist on Percentage
     of  Circulating  Polymorphonuclear Leukocytes  (PMN's),
     Lymphocytes, and  Monocytes	        172

 2    Effect  of  a  2-Hour Exposure to H^SO, Mist on Numbers  of
     Circulating  Polymorphonuclear Leukocytes  (PMN's),
     Lymphocytes, and  Monocytes	        173

 3    Alteration in Pulmonary Susceptbility to  Streptococcol
     Infection  Following Inhalation of Various Metals	        174

 4    Effect  of  Manganese Inhalation on the Growth and
     Clearance  of Streptococci  in Mouse Lungs  Following
     Infection	        175

 5    Increase in  Mortality after Exposure to Sulfuric  Acid
     and to  Ozone	        176

 6    Manganese  Deposition Immediately After Inhalation  of
     Various Aerosol Concentrations in Mice	        179

 7    Two Distinct Patterns of SO, Emissions -  High and  Low
     Idle	        201

 8    S02 Purge  Fractions - Post 50 MPH Cruise  Idle	        202

 9    Exposure Chamber  System with Added Filter Units	        215

10    Method  for Control of the Aerosol Generator  Input
     Air Supply	        216
                                   163

-------
I.   BACKGROUND AND INTRODUCTION

     The Catalyst Research Program (CRP) was initiated in 1974 by the
Environmental Protection Agency (EPA) Administrator Russell Train,
concurrent with the Agency's decision in November 1973, to permit
the use of the noble-metal oxidation catalyst as an emission control
device on 1975 model-year passenger vehicles.  The program, which resides
in the Office of Health and Ecological Effects, encompasses a multi-
disciplinary research effort to assess the health and environmental
impact of catalytic control technology for mobile emission sources.
The program office provides overall guidance and coordination of the
projects within the Cincinnati and Research Triangle Park Health
Effects Research Laboratories (HERL's), the Environmental Monitoring
and Support Laboratory (EMSL), and the Environmental Sciences Research
Laboratory (ESRL) at the Research Triangle Park.  The responsibilities
of HERL/RTP include administrative support and the preparation of reports,
position papers, testimony, and other program documents for various
Agency needs.

     A substantial portion of research activities, as well as the program's
coordination function, also resides in HERL.  In its affiliation with the
CRP, HERL has been involved mainly in sulfate studies (focusing primarily
on sulfuric acid), studies of trace metals associated with catalyst-
equipped vehicles (focusing mainly on platinum compounds), and a study of
emissions from consumer-owned national standard catalyst-equipped vehicles.

II.  OVERVIEW

     The activities of HERL/RTP in 1976 under the auspices of the Catalyst
Research Program may be categorized in four major areas of research:  animal
studies on sulfuric acid, animal studies on metals, human studies on
sulfuric acid, and characterization of in-use vehicles.  There was, in
addition, one project managed by HERL/RTP on effects of sulfuric acid on
vegetation.

     A.   Animal Studies on Sulfuric Acid

          Studies of the effects of inhalation exposure to sulfuric acid
involved several parameters.  The first set of experiments related to
effects on the immune systems of rabbits.  Results of a single 2-hour
exposure to 1 mg/m  concentration of sulfuric acid with a particle size of
0.3 pm-0,5 ym showed no significant alterations in cell-mediated immune
function of circulating lymphocytes.  An exposure of 4 hours per day for 2
days showed a marginally significant (p <0.1) alteration.  Effects on the
                                     164

-------
humoral immune system were examined using white mice exposed for 2
hours to 1 yg/m  sulfuric acid prior to intraperitoneal (IP) immuni-
zation with sheep erythrocytes.  The acid mist at this level showed
no statistically significant effect.

          Hematological changes were also observed during the immuno-
logical investigation.  A decrease of hematocrit that occurred in rabbits
following the single 2-hour exposure was found to be statistically
significant (p <0.05).  Also, though no significant effect on the number
of leucocytes per cubic centimeter (cc) was found as a result of exposure,
an increase in the number and percent of polymorphonuclear leucocytes and
a decrease in the number and percent of lymphocytes were found in both
sulfuric acid-exposed rabbits and the control group.  These effects were
more pronounced in the exposed rabbits (p <0.05).  This may indicate that
there was some type of nonspecific stress which was enhanced by the
sulfuric acid.

          The possibility of exposure to other pollutants occurring at or
near the same time as exposure to sulfuric acid led to an experiment in
which the mouse infectivity model was employed to assess effects of
sequential exposure to sulfuric acid and ozone. In this test system mice
are exposed to a pollutant or clean air (controls) before receiving an
aerosol of viable bacteria.  An enhancement of mortality from the ensuing
pneumonia is indicative of increased susceptibility to infectious respira-
tory disease.  Results of^these experiments showed that when exposure
to first ozone (.196 mg/m  x 3 hr) and then sulfuric acid (1 mg/m3 x 2 hr)
preceded the challenge by the infectious agent, the observed mortality was
equal to the additive effect of two pollutants.  This was not the case when
the sequence was reversed and the animals were exposed to sulfuric acid,
then ozone, and then the infectious agent.  Further work with these
sequential exposures (63, then H2S04) involved measurement of trachea!
ciliary beating frequency, an indicator of upper respiratory clearance
functioning.  A two-hour sulfuric acid exposure, at 1 mg/m3, but not at
0.5 mg/m3, caused a statistically significant depression in ciliary activity.
When a three-hour acid aerosol (1 mg/m3) was administered, the observed
decrease in ciliary beating frequency was that which would be expected from
sulfuric acid alone.

     B.   Animal Studies on Metals

          1.   Manganese

               In order to clarify the relationship between manganese
inhalation and infectious respiratory disease, inhalation studies of acute
exposure of respirable manganese aerosols were performed in conjunction
with a laboratory-induced streptococcal infection.  Following a 2-hour
inhalation of manganese (Mn) at >2.0 mg Mn/m3 for the insoluble manganous-
manganic oxide and >3.2 mg Mn/m^for soluble manganese chloride aerosols,
                                     165

-------
a statistically significant enhancement in respiratory infections was
observed.  There was also delayed clearance and an increase in the
growth of inhaled streptococci in the exposed animals.

          2.   Platinum

               Several studies relating to platinum and palladium are
nearing completion, and final reports are expected in 1977-78.  Major
findings of the current research include the following:

               a.   Platinum alone or in the complexed state will not
induce any type of allergy.  Palladium alone is without effect, but when
complexed with albumin will induce a delayed allergy which can be trans-
ferred via spleen cells.

               b.   A study underway at Stanford is investigating the
effect of platinum or palladium sulfate on leucocyte metabolism and
chromosomal aberrations; alteration of platelet metabolism, immunological
response, and PMN neutrophile phagocytosis; and enzyme changes in the
liver, kidney, and heart.  Preliminary results reveal that platinum
sulfate has little or no effect on the selected parameters except as a
severe eye irritant.

               c.   Preliminary results of a study at the State University
of New York at Buffalo on the effects of platinum sulfate in the central
nervous system suggest that the metal has a moderately depressive effect
on behavior.

               d.   Work at Lawrence Livermore Laboratory has established
that platinum (as platinum sulfate or potassium hexachloroplatinate) can
be methylated, under laboratory conditions, by reaction with methyl B,^-


               The other studies on platinum, all of which are nearing
completion, have not yet begun to report results.

     C.   Human Studies on Sulfuric Acid

          Early in 1977, the Catalyst Research Program is planning to
initiate a study of human exposure to sulfuric acid.  The study will
seek to build a data base on the effects of sulfuric acid aerosol in a
concentration range of 50 to 200 pg/m  with a mass median diameter
particle size range of 0.05 to 0.5 ym; these ranges are intended to
reflect the increased sulfuric acid burden expected from the increased
use of the catalytic converter.  The study will be performed on healthy
male volunteers, and the test parameters will primarily be measurements
of pulmonary function.
                                       166

-------
     D.   Vegetation Studies on Sulfuric Acid

          When vegetation was exposed to submicrometer sulfuric acid
aerosol in a greenhouse facility at the University of Minnesota, hybrid
poplar, pinto bean, and soybean showed marginal and tip necrosis,
symptoms similar to those caused by fluoride- and chloride-induced
injury on broad!eafed plants.

     E.   Characterization of In-Use Vehicles

          In cooperation with the New York State Department of Environmental
Conservation, the EPA is conducting a research project to measure the
sulfate, particulate, and regulated emissions for a group of 49 in-use
catalyst vehicles.  Each car is subjected to a sequence of test-driving
modes:  a 1975 Federal Test Procedure; a 1-hour, 50 mph cruise followed
by a 5-minute idle; and 5 successive EPA congested Freeway Driving
Schedule Cycles, each followed by a 5-minute idle.  Few test results
meet all three regulated emissions standards, and the carbon monoxide
standard is the one exceeded most often.  The test results indicate a
correspondence between high FTP carbon monoxide emissions and high idle
carbon monoxide.  The sulfate emissions are shown to be inversely related
to the carbon monoxide emissions.  The average sulfate emission value
for the entire fleet was 2.7 mg/mi.  Sulfate emissions increase to 5.8 mg/mi
when HC, carbon monoxide, and nitrogen oxide regulations are met.  Two types
of sulfur dioxide emissions have been observed.  In vehicles with low-idle
carbon monoxide values, insignificant sulfur dioxide emissions occur during
idle; in vehicles which have a high carbon monoxide value, a short burst of
sulfur dioxide accompanied by a strong sulfur odor occurs upon deceleration.

          Results of tests to determine the effect of mileage on emissions
indicate that HC emissions increase from 1.1 gm/mile to 2.0 gin/mile after
20,000 miles and that carbon monoxide emissions increase with mileage.  But
sulfur dioxide purge behavior has not been found to be a function of mileage,
and no overall correlation between mileage and sulfate emission is presently
apparent.

          The vehicle tests have also revealed that emissions patterns vary
markedly according to vehicle manufacturer and that idle mixtures are
frequently enriched, creating idle carbon monoxide levels higher than
manufacturers' specifications.

III. CURRENT STATUS

     A.   Animal Studies
          Since its inception in 1974, the Catalyst Research Program has
sponsored work concerned with the effects of animal exposure to several
pollutants.  During the calendar year of 1976, the pollutants receiving
major emphasis were sulfuric acid, manganese, and platinum.  A summary
of the major accomplishments in the studies of these pollutants follows:
                                      167

-------
          1.   Sulfuric Acid

               The first set of experiments focused on the immune system,
investigating the effects of sulfuric acid (H2S04) mist on host defense
mechanism.   The animals were exposed to 1 mg/m3 (0.3 ym-0.5 ym) for
various time periods.

               The cell-mediated immune function of circulating lymphocytes
was examined in male New Zealand white rabbits.  The animals were assayed
pre- and post-exposure so that each animal could serve as its own control.
The animals were exposed for 2 hours per day to either clean filtered room
air or to sulfuric acid.  Using a modified microtechnique (1), the trans-
formation of T cells was measured by culturing lymphocytes for 3 days in
the presence or absence of 0.5 yg purified phytohemagglutinin (PHA)/well,
a lectin that specifically induces blastogenesis of T cells.  During the
final 24 hours, all cultures were pulsed with tritiated thymidine (0.5 ym
Ci/well).  The extracellular label was removed and acid-precipi table protein
collected on filters using a MASH III.  With a scintillation counter,
incorporation of tritiated thymidine was determined and data collected
as counts/minutes.

               For analysis, the data for each individual animal were
converted to ratios.  Table 1 gives the mean and standard error of these
ratios.  The analysis indicated that no significant alterations resulted
from a single 2-hour exposure to air or sulfuric acid or from a 2-day
exposure (4 hours/day) to air.  However, lymphocytes of animals exposed to
sulfuric acid for 4 hours/day for 2 days, showed a marginally significant
(p < 0.1) increase in transformation, an increase that was similar in
cells cultured both with and without PHA.  This finding suggests the
possibility that sulfuric acid enhanced the blastogenesis of T cells inde-
pendent of PHA stimulation, or that sulfurtc acid caused an alteration in
the ratio of B:T cells.

               To examine the humoral immune system, female Swiss white
mice (strain CD-I) were exposed either to air or to sulfuric acid mist
prior to an intraperitoneal immunization with sheep erythrocytes.  Four
days later, the spleens were removed for determination of the number of
specific antibody (IgM) producing spleen cells using a modification of
the Jerne hemolytic plaque technique (2).  The data were analyzed as
previously described.  Acid mist caused no statistically significant
effect (Table 2).

               Because hematological changes were observed during the
immunological investigation, experiments were conducted to quantitate
these responses.  One group of rabbits was exposed, for a single 2-hour
period, to acid mist; a second group was exposed for the same length of
time to air.  Using analysis of variance, the decrease in hematocrit that
occurred following exposure to sulfuric acid was found to be statistically
significant (p < 0.05).  Table 3 presents the data resulting from this
experiment.
                                        168

-------
Effect of
                                                TABLE  1
                                             on  Lymphocyte  Transformation
Exposure Regimen
Clean Air x 2 hrs
1 rug H2S04/m x 2 hrs
Clean Air x 4 hrs x
2 days
1 mg H2SOi+ x 4 hrs x
2 days
Post-/Pre- Exposure
Number of
Rabbits
13
8
6
9
0.0 PHA
1.33+0.21
0.95±0.18
1.54±0.20
2.19±0.67*
0.5 pg PHA
1.3U0.33
0.86±0.18
0.89±0.05
2.6U0.89*
Pre-
Exposuret
7.40±0.18
5.75±1.30
10.15±3.62
9.58±2.19
0.
Post-
Exposuret
6.34±97
4.70±1.30
5.24H.29
13.68±5.40
5/0.0 yg
Difference
{Post - Pre)#
-1.06±0.90
-1.056±0.87
-4.92+2.36
4.10±4.88
                                                                                                                   v£>
*Reject HQ:  X =  1 at p  <  .10
tReject H  :  X" =  1 at p  <  .05  for  all means
#Fail to reject HQ:  I = 0 for all means

-------
                             TABLE 2
Effect of H?SOU Mist on Number of Antibody-Producing Spleen Cells
Treatment (2 hr)
Air
1 mg/m3 H2S04
n
24
25
Logic Plaques/106 Cells
2.80
2.74
                             TABLE 3
               Effect of H2SOt Mist on Hematocrit
Treatment (2 hr)
Air
1 mg/m3 H2S04
n
8
16
Hematocri t
(pre- and post-exposure)
0.75 ± 0.59*
2.19 ± 0.59
*Mean  ± s.e.
                               170

-------
               Also, the number of leukocytes and their classification
were determined, using a hemocytometer and differential counting of
Wright's stained smears.  Data were analyzed with analysis of variance.
There was no statistically significant effect on the number of leucocytes
per cubic centimeter.  Figures 1 and 2 show the results obtained in a
comparison of pre- and post-exposure response in animals exposed to air
and those exposed to suIfuric acid, the number and percent of polymorpho-
nuclear leucocytes (PMN) increased (p < 0.05), while the number and percent
of lymphocytes decreased (p < 0.05).  In the animals exposed to sulfuric
acid, the decrease in percent lymphocytes and the increase in the total
number of PMN's was greater (p < 0.05) than the changes found in the
air-exposed control animals.  These data indicate the possibility that
thera was some type of nonspecific stress (as evidenced by the air-exposed
animals) which was enhanced by sulfuric acid.

               Another subject of investigation was the extent to which
the combined action of ozone and sulfuric acid aerosol affects host
susceptibility to an aerosol of viable microorganisms (3).  Observations
of mortality in this streptococcal infectivity system are thought to
represent the result of several subtle effects of pollutants on respiratory
defense systems—namely, mucociliary clearance, the alveolar macrophage,
and the immune system, suggesting the complexity of the model as observed
by increased mortality rates, occurred following 2-hour inhalations of
_> 2.0 mg Mn/m3 for the insoluble manganous-manganic oxide atmospheres and
>^ 3.2 mg Mn/m3 for soluble manganese chloride aerosols (Figure 3).  These
data were substantiated by the streptococcal clearance as shown in Figure 4.
In animals exposed to manganese, the lungs were cleared of the streptococcal
infection more slowly, and subsequently, the inhaled streptococci increased
more in the manganese-exposed animals than in the infected control
animals.  For 15 days following infection, a significant reduction in
the mean survival rate was observed in mice exposed to the particulate
atmosphere, though not in mice exposed to the soluble aerosol (>_ 5.25
mg Mn/m3).  The decrease in percent survival from the control response
at these concentrations varied from a 12.5 percent decrease for those
exposed to the soluble aerosol to 21.7 percent for those exposed to the
particulate atmosphere.  Exposure to ozone (0.196 mg/m3) was for 3
hours, while exposure to sulfuric acid lasted for 2 hours.  Neither
pollutant alone caused a significant increase in mortality compared to
controls exposed only to clean air.  Figure 5 shows the increase in
mortality for each treatment group.  In those studies involving the
sequential action of the two pollutants, the treated group experienced a
statistically significant increase (p < 0.05) in respiratory infections
over controls (indicated by percent mortality) only when the exposure to
ozone immediately preceded exposure to sulfuric acid.  When this order
of exposure was followed, the observed mortality was equal to the additive
effect of the individual pollutants.  The two combined exposure groups
were not significantly different.
                                    171

-------
UJ
DC
a.
a
z
CO
O
a.

to
   25
    20
    15
UJ
QC
3
vt
O
?;   10
uj   0
O
O
H
u.
O   -5
SS
z
                     PMN'S
                                                                 CONTROL, n=16
                                                                   mg H,SO4/m3, n=12
U
Z
u.
O
   -10
   -15
   -20
   -25
                                                               MONOCYTES
         3H2SO4 TREATMENT MEAN SIGNIFICANTLY

         DIFFERENT FROM CONTROL, p<0.05
                                            LYMPHOCYTES
 Figure 1. Effect of a 2-hour exposure to sulfuric acid mist on percentage of circulating polymorphonuclear
 leucocytes (PMN's), lymphocytes, and monocytes.
                                                172

-------
                                                           CONTROL, n-15
                                                           1 ma HiSO4/m3, n=11
                                         LYMPHOCYTES
           N,SO4 TREATMENT MEAN SIGNIFICANTLY
             DIFFERENT FROM CONTROL, p <0.05
Figure 2.  Effect of a 2-hour exposure to sulfuric acid mist on numbers of circulating polymorphonuclear
leucocytes, (PMN's), lymphocytes, and monocytes.
                                       173

-------
          30
          20
O
ce
i-

o
u


O
cc
LL
UJ
u

UI
cc
UI
       Q

       >
       CC
       O  10

       H

       ui
       U
       cc
       UI
       O.
                                                                                    MnCI2
                                                                             O
                                                                        CARBON DUST
                                                        I
                                                                            Fe20,
             0.1
                                   0.5
     1.0

mg METAL/m3
5.0
                                                                                                 10.0
Figure 3. Alteration in pulmonary susceptibility to streptococcal infection following inhalation of various metals.
                                                174

-------
                                                                    Mn-TREATED,
                                                                 Mn3O4,4.106 mgMn/nr
                                                              INFECTED CONTROL
                                                                      I
                                24                48                 72

                                     POST-INFECTION INTERVAL .hours
96
Figure 4.  Effect of manganese inhalation on the growth and clearance of streptococci in mouse lungs following
infection.
                                            175

-------
  14
  12
O 10


I
9
Q
UJ
UJ
cc
UI
u

UI
cc
UJ
UL
u.
                                                               10
            "NUMBER OF REPLICATE EXPERIMENTS
             GIVEN ATOP STANDARD ERROR BARS
            ^SIGNIFICANTLY DIFFERENT FROM ZERO, p<0.06
                               19
H,SO4
0,
                                              H2S04
                                                •«•               +
                                            _  0,            H,S04

         Figure 5.  Increase in mortality after exposure to sulfuric acid and to ozone.
                                     176

-------
               It is particularly interesting that the sequence of exposure
to these two chemicals is a factor in determining the toxicity.  It is
unlikely that the difference is a spurious result, since the number of
experimental replications and animals tested were adequate, based on the
results of previous studies using this model system.  From these data, one
can postulate that the effect on the host's pulmonary defense system or on
pulmonary function parameters is of greater magnitude when ozone is adminis-
tered prior to sulfuric acid.  A possible explanation of the differences
between the sequential exposure regimens may be related to altered frequency
and depth of respiration of the test animals.  It is more likely, though,
the exposure to ozone initiates biochemical and physiological changes in
the tissue which result in increased sensitivity to sulfuric acid.  This
may lead either to larger deposits of sulfuric acid, causing a greater acid
dose to sensitive target tissue, or to deeper penetration into the respira-
tory tract.  To determine whether sulfuric acid and ozone could also influ-
ence the mucociliary escalator, which is responsible for clearing and
conducting airways of any materials that impact upon the mucous layer,
additional studies were performed.  Hamsters were exposed to 1 mg/m
sulfuric acid for 2 hours.  Immediately, and at 24 and 48 hours post-
exposure, ciliary beating was significantly depressed (p < 0.05).  After
72 hours of recovery in clean air, beating frequency has approached
control values.  Lower concentrations (0.5 mg/m ) of sulfuric acid had
no statistically significant effects.  Hamsters exposed to 0.196 mg/m
ozone for three hours exhibited no alterations of ciliary beating.  When
the animals were exposed for three hours to .196 mg/w  ozone, immediately
prior to a two-hour aerosol of sulfuric acid (1 mg/m ), the observed
depression in ciliary beating was that which would be expected from acid
alone.

               More research involving sulfuric acid has recently been
initiated.  The objective of a study begun in the fall of 1976 is to
provide information about the way metallic sulfates, sulfur dioxide (502),
and sulfuric acid can alter respiratory defense mechanisms, particularly
the alveolar macrophage.  The work is being conducted under an EPA grant
to the School of Public Health, Harvard University.  The principal inves-
tigator is Dr. Joe Brain.  Only preliminary experiments have been conducted
thus far.

          2.   Metals

               a.   Manganese

                    The next area of emphasis in EPA's animal exposure
experiments in the CRP involved inhalation studies of acute exposures
to respirable manganese (Mn) aerosols.  These were performed in conjunction
with a laboratory-induced streptococcal infection.  Utilizing this infec-
tivity system (4) in mice, studies were conducted to elucidate the relation-
                               177

-------
ship between manganese inhalation and chronic respiratory disease.  Two
manganese sources were selected for this study:  soluble manganese chloride
(MnCl2.4H20) and insoluble manganous-manganic oxide (Mn^).

                    Manganese aerosols '(<_ 2 ym) were generated within an
aerosol exposure apparatus.  The fluid atomization generating system was
used with the manganese chloride.  The particulate aerosol generating
system was utilized with manganous-manganic oxide.

                    Mouse exposures were performed as single acute exposures
of two hours duration.  Specific exposure parameters, such as total
particle counts and temperature/relative humidity measurements, were con-
tinuously monitored throughout the exposure period.  Manganese quantisation,
both during the exposure and during tissue deposition, was performed using
atomic absorption spectrophotometry (A.A.).

                    Differences were noted in pulmonary deposition and
subsequent clearance between the soluble manganese chloride aerosols
and the insoluble, particulate manganous-manganic atmospheres (Figure 6
and Table 4).  These differences are apparently due to variations in
aerosol particle size and perhaps to the solubility of the particles.
The mass median diameter (MMD) of manganese chloride aerosols was routinely
_< 2 pm, whereas for manganous-manganic oxide the MMD was _< 1.2 pm.  These
data suggest enhanced deposition by the manganous-manganic oxide atmospheres
as was observed.

                    Carbon black and iron oxide particles did not exhibit
toxic effects in the infectivity system.  These data seem to substantiate
the occurrence of true metal toxicity rather than physical irritation by
the particulate nature of test atmospheres.

               b.   Platinum

                    As reported in the Second Annual Catalyst Research
Program Report, and in the 1975 Platinum Research Review Conference
proceedings, that portion of the Catalyst Research Program which relates
to platinum, palladium, and other noble metals is declining in emphasis
with relation to the total program effort.  Most extramural projects have
been completed or are nearing completion.  There are a few studies,
however, which are still active at this writing; a brief summary of the
status of the projects follows.

                    One study now nearing completion is an investigation
of whether a hypersensitive state (immediate and/or delayed) can be induced
in experimental animals exposed to platinum or palladium metals under
various conditions.  This work, carried out at St. Vincent College,
Latrobe, Pennsylvania, has shown that platinum alone or in the complexed
                                     178

-------
                      2.0
                                          mg Mn/m3
Figure 6. Manganese deposition immediately after inhalation of various aerosol concentrations in mice.
                                             179

-------
                                TABLE 4
     Clearance of Manganese from Mouse Lungs Following Inhalation

of Different
Manganese Atmospheres
for 2 Hours
Manganese Atmosphere
Post-
exposure
Interval ,
Hours
0
1
2
3
4
6
12
18
24
30
36
48
Control
MnsO^
(3.654 mg Mn/m3)
yg Mn/g
Dry lung
54.94
51.15
42.02
31.70
26.91
15.99
10.27
N/D*
6.78
N/D
N/D
4.32
2.15
Percent
Remaining
100.0
93.1
76.5
57.7
49.0
29.1
18.7
-
12.3
-
-
7.9
-
(7.468
yg Mn/g
Dry lung
48.90
N/D
29.21
N/D
20.83
16.95
10.53
7.30
6.39
4.28
3.86
2.71
2.53

MnCl2
mg Mn/m3)
Percent
Remaining
100.0
-
59.7
-
42.6
34.7
21.5
14.9
13.1
8.8
7.9
5.5
-
*N/D, not determined.
                                       180

-------
state (i.e., with egg or guinea pig albumin) will not induce any type of
allergy.  Palladium alone is also without effect; however, palladium,
when complexed with albumin, will induce a delayed allergy which can be
transferred to normal recipients via spleen cells from sensitive donors.
A final report on this project is scheduled for Spring of 1977.

                    Platinum or platinum sulfate [Pt(SO.)?] was the compound
selected for study in a series of experiments assessing tne use of leucocyte
metabolism as an indicator of health effects as expressed in a number of
subtle biochemical and physiological changes.  Three separate studies are
being undertaken:  alteration of leucocyte metabolism and chromosomal aber-
rations; alteration of platelet metabolism, immunological response, and PMN
neutrophile phagocytosis; and enzyme changes in the liver, kidney, and heart.
In addition, histamine and serotonin levels are monitored in blood and urine.
All three studies use rabbits exposed to a high and low dose of platinum
sulfate except in a limited portion of the in vitro chromosome work, where
a limited study of human blood cells of nonexposed individuals is included
for comparative purposes.  This effort is still underway at Stanford
Research Institute, but preliminary results indicate that platinum
sulfate appears to have little or no effect on the parameters measured,
with the exception of its activity as a severe eye irritant.  Also,
recent preliminary work suggests that the sulfate may also have some
cytogenetic activity.  A final report is expected in late 1977.

                    The biological activity of platinum, specifically
the metallic oxide (Pt02) in the induction of pulmonary neoplasia by
benzo(a)pyrene is the subject of a 3-year study nearing completion at the
State University of New York at Stoney Brook.  The actions of lead oxide
(PbO) and iron oxide (Fe203) in the same system are being examined for
comparison.  In addition; the role of platinum or lead in changes of
transpulmonary absorption, pulmonary clearance, and pulmonary response to
intrapulmonary particulates, is being examined.  The study utilized both
intratracheal instillation and cutaneous application.  Necropsy of animals
from early exposure groups is now being performed.  No results are yet
available in early 1978.

                    Platinum sulfate is one of several trace substances
being evaluated at the State University of New York at Buffalo for effects
in major metabolic pathways in the central nervous system.  Biotransfor-
mation of trace substances in a developing embryo-fetus animal model and
in an adult model is being determined as well.  The study will also examine
the correlation of these two efforts with those of selected behavior in
the adult and during postnatal development of offspring of treated mothers
in appropriate models.  The purpose of these studies is to identify bio-
chemical indicators of exposure to trace substances such as platinum.  The
                                   181

-------
resulting biochemical changes will be correlated with changes in behavior
to increase understanding of the ways in which these substances interact
with mammalian systems.  Preliminary results to date indicate that platinum
(as the sulfate) has an effect on behavior which may be characterized
as moderately depressive.  Final results on the portion of the grant
dealing with effects of platinum should be available in late 1977.

                    Work at Lawrence Livermore Laboratory has established
that platinum in the form of either platinum sulfate or potassium hexa-
chloroplatinate (K2PtCl6) can, under laboratory conditions, be methylated
by reaction with methyl 612 (methylcobalamin) in the same way in which
mercury is methylated.  Ability of other metals to be methylated by this
means has also been studied and was reported in the first and second
Annual Catalyst Research Program Reports.  The emphasis of current research
is on two areas:  (1) further chemical characterization of one MePt compound
prepared from K2PtCl6 and MeB-jg and isolation and characterization of MePt
products from the reaction of Pt(SOA)2 witn MeB]2» and (2) studies of
the cellular toxicity and mutagenicity of at least one MePt compound
as well as several related Pt-salt complexes.  In vitro biological
experiments will employ Chinese hamster ovary ceTls as a mammalian test
system.  It is anticipated that final results of this research will be
reported in late 1977 or early 1978.

                    Evaluation of analytical methods for measuring trace
elements in biological matrices (especially mammalian tissue) is the subject
of a project in progress at Stewart Research Laboratories.  Existing
analytical methodology for platinum, as well as for palladium, nickel,
manganese, chromium, cerium, and vanadium, is being reviewed through a
thorough literature survey.  A series of reports will document (1) the
various limits of detection achieved; (2) preconcentration, separation, and
extraction techniques:  (3) sources of contamination; and (4) other consid-
erations necessary in order to recommend the most promising of those methods
developed, applied, and reported to date.  A limited laboratory evaluation
of one or more of the methods recommended on the basis of the literature
is planned for late 1977 and will  be reported in early 1978.  The reports on
methods for individual metals should be available late in 1977.   It is
anticipated that this series of reports will be valuable to researchers in
the area of trace metals, especially where measurements of extremely small
amounts of metal in tissue samples are necessary for interpretation of
experimental results.

                    In 1975, the Catalyst Research Program held a review
conference on the platinum research sponsored by the program.  A report
of the conference was published in early 1976 as "Proceedings - Catalyst
Research Program Platinum Research Review Conference."  Since all research
                                  182

-------
under the CRP which is related to platinum and noble metals is scheduled
for completion within the next eighteen months, a followup report is
planned to summarize final results of all aspects of the program.  Since
platinum is an element which had received little attention either in
analytical or toxicological research prior to the establishment of the CRP,
it is expected that this summary report will become an extremely useful
reference for future researchers in this field.  The report is scheduled
for publication in late 1978.

     B.   Human Studies

          1.   Objectives

               The CRP, as part of its intensive research on health effects
of sulfuric acid, is planning to initiate early in 1977 at its Chapel Hill,
N.C., facility, a study of human exposure to sulfuric acid.  A summary of
the project and a detailed description of facilities and procedures follow.

               This proposed research seeks to investigate in young healthy
male volunteers the potential health effects of low-level sulfuric acid
aerosols in a concentration range of 50 to 200 yg/m3 with a mass median
diameter (MMD) particle size in the range of 0.05 to 0.5 micrometers.

               It is estimated by EPA that the peak urban atmospheric
burden of sulfuric acid aerosol is at present in the range of 20 yg/m3 and
that the widespread use of automobile catalytic converters might increase
this level.  The catalytic converter produces a sulfuric acid aerosol
via the generation of sulfuric acid within the converter.  There is little
definitive information presently available as to the health effects dose
response curve for sulfuric acid aerosol concentrations in this range
and size.

               Sulfuric acid aerosol refers to a suspension of sulfuric acid
droplets in air.  Under some conditions in the ambient atmosphere, sulfuric
acid aerosol has limited stability and may exist only briefly in its original
chemical and physical form.  For example, the aerosol can be neutralized by
interaction with ambient atmospheric ammonia.  Sulfuric acid particles are
hydroscopic and increase in size when the humidity is high.  Sulfuric acid
aerosol is one of several airborne parti oil ate sulfates which may be found
in ambient air.

               Particle size is a crucial factor in sulfuric acid aerosol
toxicity because it is a major determinant of the deposition site within
the respiratory system (5).  The hygroscopic and irritant potential of
sulfuric acid aerosol may alter the normal rate of regional particle depo-
sition under certain circumstances.  Since particle size influences depo-
sition and since deposition is important to the relative toxicity, size
characterization is essential in the assessment of the health impact of
                                      183

-------
of suIfuric acid aerosol exposure.  Mass median aerodynamic diameter
(MMAD) is a necessary but not a fully sufficient measure of aerosol
particle size.  In addition to MMAD, a measure of the variability of the
particles or frequency distribution, such as range or standard arithmetic
or geometric deviation, is also desirable.  Aerosol characterization has
been deficient in some studies described in the literature.

          2.   Previous Studies

               Animal toxicology, clinical studies, and epidemiologic
studies were reviewed extensively by Battigelli and Gamble (6), Lewis,
et al. (7), and by Ariey, et al.  (8).  Data on sulfuric acid toxicity
in humans are more limited than animal data, but there is some information
on short-term exposures.  Sulfuric acid is detected by human subjects at
about 500 yg/m3 (9).  At higher concentrations, there is a spectrum of
responses ranging from increased air flow resistance to acute irritation
of mucous membranes, lacrimation, and coughing (7).

               Toyama, et al. (10) have reported an increase of 18 percent
in pulmonary flow resistance in humans with H2$04 exposure as low as 10-100
yg/m3 with a particle size 1.8 ym count median diameter (CMD).  At 800-1400
yg/m3, with 4.6 ym (CMD) particles, airway resistance was increased by
36.5 percent.  In these experiments:  (1) sulfuric acid aerosol was produced
by combining hydrogen peroxide and sulfur dioxide in a mixing chamber; (2)
particle size was determined by impacting the acid aerosol on a thymol-blue
impregnated strip and microscopically measuring the size of the resulting
circular red spots; and (3) flow resistance was measured by use of the
interrupter technique.  The use of these techniques has been criticized by
independent reviewers who have stated that, as a result of the methods used,
Toyama's conclusions are of questionable value (7).

               Amdur, et. al. (11) have reported that respiratory rates
have increased by about 30 percent, depth of breathing decreased by about
28 percent, and maximum inspiratory-expiratory flow rates decreased by about
20 percent at a concentration of 350-500 yg/m3.  These concentrations are
below subjectively detectable levels.  The ventilatory parameters changed
in a nonlinear fashion with 80 percent of the changes occurring during the
first 3 minutes of exposure.  The initial phase of the recovery period
was almost a mirror image of the exposure period, i.e., nonlinear rapid
return to normal values, with the fastest return in the first 3 minutes.
As demonstrated in animal studies, such rapid and marked changes are
reflex in nature.  Although the airway resistance was not measured, its
characteristic increase can be easily observed on a number of air flow
tracings shown in Amdur's publications.  It also appears that as the H?S04
concentration increases, so do respiratory rate and air flow resistance.
                                      184

-------
               Amdur has shown that in humans the average retention of
inhaled sulfuric acid varies from 50 percent to 87 percent with exposure
to 400 to 1000 yg/m  of sulfuric acid aerosol.  As the respiratory rate
increases, which occurs at somewhat higher levels of sulfuric acid aerosol,
the percent retention is reduced, possibly indicating a protective response
at these levels (11).  All of the subjects involved in the clinical studies
reported above were healthy young adults who could easily compensate for
the increased resistance imposed upon their breathing.

               Recently, Hazucha and Bates (12) have shown synergism
between ozone and sulfur dioxide, as measured by dynamic lung function
tests in man.  They postulated that the two gases interact rapidly on
the large, wet, air-tissue interface of the lung, covering the inner
surface with sulfuric acid.  These results suggest that both peripheral and
central airways are affected, but more detailed studies are needed to
determine the mechanism responsible for these changes.

               All the above studies were performed using healthy human
subjects.  No studies on subjects with pre-existing disease have been
attempted to date.

          3.   Methodology

               The generation and monitoring of these concentrations and
size of sulfuric acid aerosol are technically difficult and are therefore
covered in an attachment which describes the design plan and operating
protocol of the facilities for the controlled exposure of humans to
sulfuric acid aerosol.

               A single blind protocol will be employed, using young healthy
male volunteers between the ages of 19 and 30.  Judging from the past
literature, it would seem possible to use a double blind protocol in that
the subjects should not sense acid aerosols at these concentrations.  This
permits study observation without the subject's awareness of the exposure
conditions and without investigator's prejudice.  The exposures are to take
place in the CSD Lexan Chamber, which has been used for the-ozone exposures.
The levels to be studied initially would be 50 and 200 yg/m  with a MMD
particle size controlled in the range of 0.05 to 0.5 ym.

               For this study, the subjects will be exposed to acid aerosol
for 2 hours.  There is some concern that adaptation may occur quickly and that
longer exposures are unnecessary.  It is expected that such a 2-hour protocol
could be accomplished in a double-blind, randomized fashion such that the
subjects could be studied over several days with the control days randomized
with the exposure day.  In this manner, as many as two to three subjects
could be studied per week, with one subject following the other in succession.
The exact design of this random protocol will depend on subject availability.
                                    185

-------
It will also be important to determine whether there are any lasting
effects which would carry over from one day to the next, but the litera-
ture does not suggest this.  The protocol is outlined in Tables 5 and 6.

               When the subject enters the chamber, baseline determina-
tion of all parameters will be made.  He will be exposed 45-60 minutes
later.  At the end of 5, 10, and 15 minutes of exposure, resting tidal
volume will be recorded.  After an additional 15 minutes of exposure,  all
of the appropriate parameters will be remeasured.  A comparison of the
readings at 15 minutes and 30 minutes of exposure will allow us to observe
whether adaptation or further effects of the exposure has occurred.  After
5 minutes of rest, the subject will begin an exercise program of approximately
15 minutes, taking him up to 50 minutes.  The exercise will be performed on
a bicycle ergometer at a workload of 700 kg/min, an exercise level  that
produces approximately a doubling of heart rate, to evaluate the effect of
stress in combination with acid aerosol exposure.  A similar 15 minutes
of exercise will be performed prior to completion of 2-hour exposure.
Immediately following the 2-hour exposure, the whole battery of tests used
during baseline measurements will be determined.

               The test parameters will primarily be measurements of pulmo-
nary function since the limited literature suggests this as the most impor-
tant area of change.  The following measurements will be taken.

                    a.   FVC   - forced vital capacity
                    b.   FEV,  - forced expiratory volume in one second
                    c.   MMFR  - mid-maximal expiratory flow rate
                    d.   FERV  - forced expiratory reserve volume
                    e.   V5Q   - peak flow at 50 percent of forced  vital
                                 capacity
                    f.   V25   - peak flow at 25 percent of forced
                                 vital capacity
                    g.   Raw   - airway resistance
                    h.   FRC   - functional  residual  capacity
                    i.   TLC   - total lung capacity computed from
                                 adding FVC to RV


               In addition to the studies of pulmonary function, a  single
lead precordial ECG will be continuously recorded while the subject is in
the exposure chamber, either on controlled days or on exposure days.  This
is part of the subject safety, but, in addition, the ECT signal provides
an indication of change in heart rate with exercise or as a result  of
exposure under resting conditions.  It also provides an opportunity to
observe any arrythmogenic potential of the exposure.   The ECG is not seen
as a primary investigative tool but as one which insures monitoring for
subject safety reasons.
                                 186

-------
                                TABLE 5
               Blind Exposure Protocol for H?S(X Aerosol

                           Experimental Sessions
Intervention Sequence I:  Control, Control. Exposure (CCE)
Time              Session 1             Session 2            Session 3
8:30 a.m.
Control (C)
Pre-exposure
Control (C)
Same
Exposure (E)
Same
 9:30 a.m.
   to
11:00 a.m.
                  Evaluation and
                  Sampling
                  Lexan Chamber
                  Clean Air
Lexan Chamber
Clean Air
11:30 a.m.
                  Post-exposure
                  Evaluation  and
                  Sampling
Same
Lexan Chamber
H2SOi,. Aerosol
100 yg/m3,
0.05-0.5 yg
particle size
Same
 Intervention Sequency  II:   Control,  Exposure,  Control  (CEC)
                   Control  (C)            Exposure  (E)         Control  (C)
                   Same daily procedure  as  for  CCE.
                                     187

-------
                                TABLE 6

        Time Course for Each of the Three Experimental  Sessions
Baseline Evaluation


Enter Lexan Chamber

Sitting Quietly
Exercise Bicycle Ergometer
for 15 Minutes

Sitting
Exercise Bicycle Ergometer

Leave Lexan Chamber


Post-exposure Evaluation
 8:30 a.m.
 9:30 a.m.

 9:35 a.m.
 9:40 a.m.
 9:45 a.m.

10:05 a.m.
10:30 a.m.


11:05 a.m.

11:45 a.m.
Baseline Pulmonary
Function Studies (PFS)
                                                     Resting VT After 5-, 10-,
                                                     15-Minute Exposure
Resting PFS After 1 Hour
of Exposure
2nd Post-exercise PFS After
2 Hours of Exposure
Subject will complete a short symptom
questionnaire before and on leaving the
chamber during each experimental  session.
                                       188

-------
               After each experimental session, the subjects will complete
a subject symptom inventory questionnaire to determine what specific
symptoms, if any, they experienced in conjunction with the experimental
session.

               Approximately 12 adult healthy young non-smoking male
volunteers are to be studied.  It is anticipated that most wtll be student
volunteers and other young men from the Chapel Hill community.  Each
potential subject will be given a history and physical examination before
being selected as a normal healthy volunteer.  In addition, a hematocrit,
white blood cell count, and differential will be obtained.  Subjects are
to be carefully questioned to exclude those taking any medications, those
with any recent pulmonary infections, etc.  Temperature, blood pressure,
pulse, and respiration are to be recorded just prior to starting the
experiment.  An informed consent form describing the study will be signed
by each subject prior to hfs participation.

               It is anticipated that there will be very little risk to
the healthy male subjects with a 2-hour exposure to these low levels of
sulfuric acid aerosol.  It is also anticipated that it will be difficult
to measure any detrimental health effects in any of these parameters at
this low level; the industrial human exposure level or standard is set
tenfold higher at 1 mg/m .  The only benefit to the subjects is a medical
examination and routine pulmonary function testing.

               The LEXAN Chamber facility at present will permit the
evaluation of 2 subjects per week.  The evaluation of at least 12 subjects
is anticipated.  Although a double-blind protocol is considered to be
highly desirable in terms of the statistical analysis, a single-blind
protocol control-exposure-control (CEC) was selected.  Since the two
groups of investigators (monitoring group and physical testing group)
work rather independently, the double-blind design is considered unnecessary
for this pilot study.  From a statistical point of view, a CEC protocol
will permit the population sample for this pilot study to be kept at a
minimum.  A three-way analysis of variance was considered the most suitable
test for a comprehensive analysis of the data.  In this way, variance
between individual days and/or hours as well as subjects can be tested
more effectively.  Other statistical tests may be employed in analyzing
the data to test alternate hypotheses.  In addition to on-line computer
analysis, all records will be analyzed manually by one of the investigators
to assure quality of the data.  All data will be reduced and statistically
analyzed by computer using standard statistical packages, e.g., SSPS.
                                        189

-------
     C.   Mobile Source Emissions

          In cooperation with the New York State Department of Environ-
mental Conservation at the Automotive Emissions Laboratory (AEL) in
Albany, New York, the EPA is conducting a research project to measure
the sulfate, particulate, and regulated emissions for a group of 49 in-
use catalyst vehicles obtained from the fleets of several cooperating
organizations and from private vehicle owners in the Albany area.  The
collection of data began in December 1975, and is projected to continue
throughout 1977.  This report is a summary document.  More detailed
results are available in monthly reports to the Project Officer, including
the metals emissions indicated in Table 7.

          1.   Test Procedures

               Each vehicle is tested periodically at intervals of approxi-
mately 5,000 miles to observe the effects of aging on emissions and on
fuel economy.  A particular sequence of test driving modes is followed:

                    •    a 1975 Federal Test Procedure (FTP);

                    •    a 1-hour, 50 mph steady cruise;

                    «    five successive EPA Congested  Freeway
                         Driving Schedule cycles
                                                                          i
The 1-hour, 50 mph cruise and the five CFS driving cycles are each followed
by 5-minute idle periods with the vehicle transmission in neutral.  This
combination of test-driving modes results in an experimental  sequence of
approximately 125 miles which lasts over 4 hours per vehicle  test.  Table
7 shows the measurements that are made during each test segment.  The
AEL test bay is equipped with 30 tons of air conditioning (cooling, heating,
dehumidification, humidification) providing 15,000 cfm of conditioned air
through two parallel sets of multiple overhead registers directed down
along both sides of the vehicle positioned on the dynamometer.  With this
system, room temperatures are typically held between 19°C and 24°C for all
testing modes.  A large fan is directed at the front of the vehicle and
a squirrel cage blower is directed diagonally from the side of the vehicle
ahead of the rear tires.  This blower is located behind the converter to
provide rear tire cooling in order to prevent tire failure.

               Flow through the constant volume sampler (CVS) is approximately
400 cfm, and the in-line dilution tunnel closely follows the  design of
several similar tunnels in use by EPA laboratories.  A clayton CTE-50 direct
drive variable-inertia, automatic road-load missions dynamometer is used in
all tests.  Laboratory standard gases are cross-calibrated by EPA Ann Arbor.
Sulfur dioxide is measured in dilute-filtered exhaust by a Lear-Siegler
SM-100-second-derivative UV spectrometer.  Sulfur dioxide mass emissions are
                                     190

-------
                                TABLE 7

                   Data List for Each Emissions Test
          Test Segment
            Measurements
Vehicle Check-In before 12-20
hour soak
Idle HC, CO; Engine Analyzer Check-in
(ignition, timing, power, balance, EGR,
etc.)  Visual emission system integrity.
Tank fuel sample:  Sulfur, Lead, Phos-
phorous, Reid Vapor Pressure, API
Gravity, Density, Distillation.
Odometer, Engine hour meter, Totalizing
fuel meter.
1975 FTP
Gaseous emissions:  HC, CO, NOX, C02,
S02 one 47 mm fluoropore filter for
entire FTP:  total particulate mass,
soluble sulfate by Barium Chioranilate
(BCA) S02 trace in dilute exhaust and
mass by continuous integration.
02 continous trace in raw exhaust.
1-Hour, 50 mph Cruise
Same as for FTP with 1 additional
47 mm fluoropore filter for Particu-
late Elemental Analysis by X-ray
Fluorescence:  Aluminum, Phosphorous,
Sulfur, Calcium, Iron, Copper, Bro-
mine, Zinc, Platinum and Lead.
5-minute Idle Periods
S02 trace, mass by continuous integra-
tion.  Soluble sulfate mass by BCA
from 47 mm filters collected:  A) one
filter for first idle after 50 mph cruise
and B) last 3 idle periods averaged
on single filter.
Five CFS Cycles Followed by
5-minute Idle Periods
Same as for FTP.
                                           191

-------
determined by continuous electronic integration of the output signal of the
spectrometer.  The optical chamber is thermostatically controlled at 50°C
to prevent condensation.  Dilute exhaust sulfur dioxide concentration is
also recorded continuously on a strip chart recorder/disc integrator in
order to observe the relationship of sulfur dioxide emissions to driving
mode.  Mass balance sulfur dioxide injection tests similar to propane
tests for HC have verified the ability of this system to quantitatively
detect sulfur dioxide, even In the presence of an auto exhaust background.

               The collected particulate is analyzed for soluble sulfate
by the semi-automated ban"um-chloranilate (BCA) technique.  The apparatus
is calibrated with prepared standards on a daily basis using a quadractic
fit to low-range calibration points.   Peak areas are integrated using a
Columbia Scientific Supergrater II.

          2.   Characteristics of Vehicles Tested

               Four cooperating institutional fleet operators provide
vehicles.  Privately owned vehicles are solicited by mailings to registered
owners of 1975 and 1976 model vehicles in the Department parking lot.  One
vehicle (1975 Chevrolet Wagon, 454 V-8), is fueled and maintained by
AEL staff to serve both as a control  vehicle for testing and as an over-
night loan car when privately owned vehicles are tested.

               Vehicles are not screened before entry to the test group.
Low-mileage vehicles are recruited when available, but no criteria of
age or tune-up condition (idle carbon monoxide, etc.) are used to select
or reject any vehicle.  Vehicles are tested in "as received" condition
after a thorough analysis of vehicle operating parameters.  Tank fuel is
used for testing whenever possible; commercial lead-free fuel is added at
AEL prior to the test if the vehicle as delivered has insufficient fuel
for the entire tests.

               Tests are scheduled at 5,000-mile intervals when possible;
however, scheduling constraints frequently dictate longer intervals,
especially for high-mileage fleet vehicles.  To date, 127 tests span
approximately one million total miles for the 49 test vehicles.  Over 90
percent of the tests are on cars with less than 28,000 miles.  Six of these
vehicles have had five or more tests, and five have had only one test.
The remaining 38 vehicles have completed two to four tests at this time.
The greater availability of larger vehicles for the project is reflected
by the fact that only six vehicles in the test group have inertia weights
less than 3,000 pounds.  Table 8 gives a breakdown of the 49-car group
and of the 127 FTP test data by manufacturer and by fleet vs. private
owner.
                                      192

-------
             TABLE 8
Classification of Cars and Tests


49 Cars:



1975
1976
127



24 Fleet Cars
24 Personal Cars
1 Control Car
Model Cars
Model Cars
Emission Tests:
76 Fleet Tests
46 Private Tests
5 Control Tests
General Motors
18
( 4)
(13)
( 1)
8
10
50
(19)
(26)
( 5)
Ford
9
(2)
(7)
(0)
4
5
18
( 5)
(13)
( 0)
Chrysler
22
(18)
( 4)
( 0)
17
5
59
(52)
( 7)
( 0)
                      193

-------
          3.   Average Emission Results

               Table 9 shows average emission results for all tests for
various subclassifications, using the same subclassifications.  Tables
10-12 show emission results for each manufacturer.  CFS emissions values
for individual tests are the average of the last three CFS cycles of each
test.  Test-by-test review of the sulfate emissions for the five CFS cycles
in each test has shown that stabilized emissions values are achieved in
most cases by the second CFS cycles and that the average of the last three
cycles is an accurate measure of stabilized emissions.

               In general, these results show low sulfate emissions and high
carbon monoxide emissions.  Idle carbon monoxide concentration and carbon
monoxide mass emission data exhibit a direct correspondence.  Few test results
meet all three regulated emission standards, with the carbon monoxide standard
being exceeded most often.

               Emissions certification standards for 49 states for both 1975
and 1976 vehicles are as follows:  1.5 gm/mi HC, 15.00 gm/mi CO, and 3.1
gm/mi NO.  Of the total 127 tests using 49 vehicles, 30 tests performed on
15 vehicles are at or below the standards for all 3 regulated pollutants.

               a.   Carbon Monoxide

                    Average carbon monoxide (CO) emissions for the 127 tests
were 25.1 gm/mi, significantly above the 15 gm/mi standard (see Table 9).
The average idle carbon monoxide concentration for the tests is 2.4 percent,
compared to 0.4 percent for the 53 tests in which FTP carbon monoxide
emissions are less than the 15 gm/mi standard.  This correspondence of high
FTP carbon monoxide emissions and high idle carbon monoxide is further indica-
ted by grouping FTP carbon monoxide data according to a dividing point of 1
percent idle carbon monoxide.  Average FTP carbon monoxide for vehicles
characterized by idle carbon monoxide below 1 percent is 10.3 gm/mi; whereas
the average for those above 1 percent idle carbon monoxide is 36.0 gm/mi.

                    Of 73 tests with more than 1 percent idle CO, only 9
(12 percent) were less than 15 gm/mi FTP CO.  Of 54 tests with 1 percent
or less idle CO, only 9 (17 percent) were greater than 15 gm/mi FTP CO, in
spite of factors such as choke operation that are not related to warm engine
idle CO, but can affect CO mass emissions.

                    Idle CO values obtained in this study are in close
agreement with data collected by the Department on a statewide basis.  Tests
on over 400 1975 catalyst cars at various locations showed that the percentage
of Chrysler, General Motors, and Ford vehicles with less than 1 percent idle
CO were 20 percent, 60 percent, and 80 percent respectively.  The data
collected in this study gives corresponding values of 27 percent, 50 percent,
and 72 percent respectively, suggesting that the results obtained on this
small group of cars is a fair indication of what catalyst cars are doing
in general.
                                        194

-------
vo
Ui
                                                                                      TABLE 9


                                                               Average Emission Results; All Tests, All Manufacturers
Classification
of Tests
All
FTP CO < 15 gm/mi
FTP CO < 15 gm/mi, HC < 1.5 gm/mi
FTP CO < 15 gm/mi, HC < 1.5 gm/mi,
NOX < 3.1 gm/mi
Idle CO £ 1%
Idle CO >H
FTP CO > 45 gm/mi
Stall in Bag I of FTP
Fleet vehicles
Private Vehicles
No.
Tests
127
53
49
30
54
73
24
28
76
46
No.
Cars
49
23
22
15
26
34
11
21
25
24
Idle
CO?
2.4
0.4
0.3
0.4
0.2
4.1
5.5
2.2
2.7
2.3


HC
1.7
0.9
0.8
0.7
0.9
2.3
3.4
1.4
2.0
1.4
FTP
gm/mi
CO
25.1
7.6
7.1
7.5
10.3
36.0
62.0
20.1
36.1
17.1
1 Hr 50 mph Cruise

NOX
2.9
3.2
3.2
2.5
3.0
2.7
2.7
2.5
2.6
3.3

HC
0.1
0.1
<0.1
0.1
0.1
0.1
0.2
<0.1
0.1
0.1
gm/mi
CO
0.8
0.3
0.1
<0.1
0.7
0.8
1.0
0.5
0.8
0.7

NOX
2.4
2.5
2.6
1.9
2.5
2.2
2.6
2.1
2.1
2.9
mg/mi
SO,,
4.1
5.1
5.4
5.7
5.1
3.3
3.3
6.3
4.0
3.9

HC
0.5
0.2
0.2
0.1
0.2
0.7
1.2
0.3
0.5
0.4
:CFS
gm/mi
CO
6.2
1.6
1.4
1.3
1.8
9.5
16.5
5.0
7.2
5.3


NOX
2.8
3.1
3.1
2.5
3.0
2.7
3.1
2.5
2.6
3.2

mg/mi
SO*
2.7
4.9
5.2
5.8
5.1
0.9
0.5
3.7
2.0
2.9

-------
O\
                                                                                      TABLE 10



                                                             Average Emission  Results; General Motors Vehicle Tests Only
Classification
of Tests
All General Motors Vehicles
FTP CO < 15 gm/mi
FTP CO < 15 gm/mi, HC < 1.5 gm/mi
FTP CO < 15 gm/mi, HC < 1.5 gm/mi,
NOX < 3.1 gm/mi
Idle CO £ 1%
Idle CO > U
FTP CO > 45 gm/mi
SET- 7 SO,, > 10 mg/mi
Fleet Vehicles
Private Vehicles

No.
Tests
50
29
28
20
25
25
3
2
19
26

No.
Cars
18
11
11
8
9
11
2
1
4
13

Idle
1.8
0.3
0.3
0.4
0.1
3.6
3.8
0.0
1.1
2.8


HC
1.3
0.9
0.7
0.7
0.9
1.8
4.1
0.6
1.3
1.5
FTP
gm/mi
CO
16.4
7.1
7.0
7.4
6.3
26.6
58.7
5.7
14.1
20.2
1 Hr 50
mph Cruise
gm/mi
N0x HC
3.0 <0.1
3.1 <0.1
3.1 <0.1
2.5 <0.1
3.2 <0.1
2.8 0.1
3.1 0.2
2.2 <0.1
2.7 0.2
3.4 <0.1
CO NO
0.9 2.5
0.1 2.5
<0.1 2.5
<0.1 1.9
0.1 2.7
1.7 2.4
5.0 2.6
<0.1 1.3
1.5 2.2
0.6 2.9

mg/mi
SO*
1.4
2.0
2.0
2.2
2.3
0.5
0.2
7.7
0.9
0.9

CFS

gm/mi
HC
0.4
0.2
0.1
0.1
0.2
0.5
1.3
0.1
0.4
0.4
CO
5.7
2.0
0.2
1.8
1.7
9.8
27.8
0.2
6.1
6.5
HOX
2.9
3.0
3.0
2.5
3.1
2.7
2.9
2.0
2.8
3.1

mg/mi
SO,
1.6
2.5
2.5
3.3
2.8
0.3
0.3
15.9
0.7
0.4

-------
                      TABLE 11



Average Emission Results; Ford Vehicle Tests Only
Classification
of Tests
All Ford Vehicles
FTP CO < 15 gm/mi
FTP CO < 15 gm/mi, HC < 1.5 gm/mi
FTP CO < 15 gro/mi, HC < 1.5 gm/mi,
N0y < 3.1 gm/mi
A
Idle CO < 1*
Idle CO > 1%
FTP CO > 45 gm/mi
SET-7 S0» > 10 mg/mi
Fleet Vehicles
Private Vehicles

No.
Tests
18
12
11
3
13
5
0
6
5
13

No.
Cars
9
6
6
2
6
3
0
4
2
7

Idle
COS
0.7
0.2
0.2
0.0
0.2
2.1


0.0
0.3
0.8


HC
1.0
0.9
0.8
0.6
0.9
1.3


0.6
0.8
1.0
FTP
gm/mi
CO
11.5
5.3
4.4
2.4
9.2
17.5


2.8
16.5
9.5
1 Hr 50 mph Cruise

NO
3.2
3.8
3.7
2.7
3.3
2.9


3.6
2.4
3.5

HC
0.3
0.3
0.2
0.2
0.2
0.4


0.2
0.1
0.3
gm/mi
CO
2.2
0.9
0.3
0.1
2.5
1.6

0.0
4.3
1.4

N0x
2.8
3.4
3.4
1.7
3.0
2.4
—
3.2
1.4
3.3
mg/mi
soT
8.7
10.3
11.1
15.7
9.4
7.1


14.5
7.1
9.4
CFS

gm/mi
HC
0.4
0.4
0.3
0.2
0.3
0.5


0.2
0.2
0.4
CO
3.9
1.2
0.8
0.1
3.3
5.5


0.3
6.2
3.0
NOX
3.3
3.9
3.9
2.8
3.5
2.4


3.9
2.2
3.7

mg/mi
wr~
8.5
11.7
12.5
22.1
10.2
4.1


17.2
7.5
8.9

-------
                      TABLE 12
Average Emission Results;  Chrysler Vehicle Tests  Only
Classification
of Tests
All Chrysler Vehicles
FTP CO < 15 gm/mi
FTP CO < 15 gm/mi, HC < 1.5 gm/mi
FTP CO < 15 gm/mi, HC < 1.5 gm/mi,
NOX < 3.1 gm/mi
Idle CO < H
Idle CO > \%
FTP CO > 45 gm/mi
SET-7 SO* > 10 mg/mi
Fleet Vehicles
Private Vehicles
Police Cruisers
Fleet -- No Police Cars
Fleet and Private Vehicles —
No Police Cars

No.
Tests
59
12
10
7
16
43
21
4
52
7
14
38
45

No.
Cars
22
6
5
5
11
20
9
4
18
4
2
16
20

Idle
3.5
0.9
0.5
0.6
0.3
4.7
5.7
0.1
3.6
3.1
4.4
3.2
3.2


HC
2.2
1.0
0.9
0.9
1.1
2.6
3.3
1.2
Z.3
1.4
3.0
2.1
2.5
FTP
gm/mi
CO
36.5
11.1
10.5
10.3
17.6
43.6
62.5
16.1
38.8
19.8
53.8
33.2
31.1
1 Hr 50

NO HC
A
2.6 0.1
2.8 <0.1
2.8 <0.1
2.5 <0.1
2.6 <0.1
2.7 0.1
2.7 0.2
2.3 0.1
2.6 0.1
2.8 0.1
3.0 0.0
2.4 0.1
2.5 0.1
mph Cruise
gm/mi
CO NO
0.2 2.1
<0.1 1.7
<0.1 1.8
<0.1 1.9
0.1 2.0
0.2 2.1
0.4 2.6
<0.1 2.0
0.2 2.1
<0.1 1.7
0.0 3.2
0.3 T.7
0.2 1.7

mg/mi
SO,,
4.9
7.5
8.7
11.3
6.0
4.4
3.7
11.4
4.9
5.0
2.0
5.9
5.8
CFS

qm/mi
HC CO
0.6 7.3
0.2 1.2
0.1 0.4
<0.1 0.5
0.1 0.8
0.8 9.8
1.2 14.9
0.1 0.2
0.6 7.7
0.4 4.4
0.6 13.7
0.6 5.5
0.6 5.4
N0x
2.6
2.3
2.4
2.4
2.5
2.7
3.1
2.2
2.6
2.5
3.8
2.2
2.2

mg/mi
SO,,
1.9
4.2
4.9
6.3
4.7
0.8
0.6
13.0
2.0
0.9
0.8
2.4
2.2

-------
                    Vehicles stalled at the beginning of the FTP in
28 tests.  In spite of the restart  (by Federal Register procedures),
average regulated emissions for these tests are lower than overall
average idle carbon monoxide is 2.2 percent for these tests, also
less than overall average value.

                    Tests on fleet-operated vehicles show considerably
higher FTP carbon monoxide emissions than do tests on privately-owned
vehicles.  Any comparison of fleet and private vehicles must be made with
care, however, because the manufacturer-mix differs for the two groups
(Table 8) and because the fleet vehicles have much higher mileage accumu-
lations.

               b.   Sulfate Emissions

                    Sulfate emissions for 127 tests average 2.7 mg sulfate
(SO.) per mile.  For tests with less than 15 gm/mi FTP carbon monoxide,
cease to 5.8 mg/mi in tests where HC, carbon monoxide, and NO  standards
are met.  Grouping sulfate emissions data according to vehicles' idle carbon
monoxide value also reveals a correlation.  Fifty-four tests on vehicles
having less than 1 percent idle carbon monoxide yeild an average of 5.1
mg/mi SO,, and 73 tests with greater than 1 percent idle carbon monoxide
give an average of 0.9 mg/mi SO,.   In 12 tests, sulfate emissions are over
10 mg/mi; in each case the idle CO  is 0.0 percent.

                    Significant variation in results are observed when the
data are grouped by manufacturer  (Tables 10-12).  The effect of air pumps
on Ford vehicles is seen in the low carbon monoxide and relatively higher
sulfate emissions.  Frequent rich carburetion encountered in Chrysler
products corresponds to the high  carbon monoxide and low sulfate emissions
observed.

                    In order to determine the sulfate emissions at idle,
two filters are collected during  idle periods of the test sequence
(see Table 7); one is collected during the 5-minute idle separating the
50 mph cruise from the first CFS, and a second is a composite sample from
the idle periods after CFS cycles three, four, and five.  These filters
are analyzed for soluble sulfate  on the BCA apparatus.  These data show
idle sulfate emissions to be very low.  Combining the driving cycle and
the following idle period, less than 2 percent of the sulfate collected
and 0.2 percent of the fuel sulfur  recovery is from the idle period.
                                     199

-------
               c.   Sulfur Dioxide Emissions

                    Two distinct types of sulfur dioxide (SOJ emissions
have been quantified.  One type is typified by high idle caroon monoxide
values; the other, by low idle carbon monoxide emissions.  Figure 7 shows
both types of sulfur dioxide emissions for the same car in two successive
tests.

                    The two types are observed when a car is decelerated
to idle.  The first type, which shows insignificant sulfur dioxide emissions
in any of the 5-minute idle periods, occurs on vehicles with low idle
carbon monoxide values typical of proper carburetion.  The second sulfur
dioxide emission pattern, which is characterized by a short burst of sulfur
dioxide upon deceleration, either at the end of the 50 mph cruise or at the
end of the CFS cycles, occurs in vehicles which have high carbon monoxide
values.  A strong sulfur odor is readily noticed in this case.

                    In order to quantify this spike phenomenon, the sulfur
dioxide concentration is integrated separately during each driving cycle and
idle period, yielding a sulfur dioxide mass value for each.   For comparison
purposes, a ratio of these masses is defined as follows:


                                 Spike S02 mass
          S02 purge fraction  =  	
                                 Cycle S02 mass + spike S02 mass

A purge fraction of 1.0 thus corresponds to a situation where the total  S02
emitted from a particular driving cycle (i.e., 1-hr, 50 mph cruise or
CFS cycle) is attributable only to the idle segment.  Table 13 gives average
purge fraction values for the four catalyst configurations for both the
post 50 mph cruise idle and for the post CFS idle.   CFS values reported are
the average of purge fractions for cycles 3, 4, and 5.   For all system
configuration, it is found that, within a given test, there is reasonable
consistency of purge fractions for all five CFS cycles.  S02 purge fraction
for the 50 mph cruise, classified by idle CO emissions, is shown in Figure
8.  The greater S02 purge fraction from tests with 1 percent or less idle
CO than for those with greater than 1 percent idle CO is readily apparent.

               4-   Effect of Mileage Accumulation

                    Table 14 shows average FTP regulated emissions for
various mileage ranges by manufacturer.  Data from tests on vehicles with
over 30,000 miles are predominantly from fleet vehicles with high mileage
accumulation rates.
                                      200

-------
10-
          1975 PLYMOUTH
             440 V8/A3
            12,753 miles
           IDLE CO=0.2%
              TEST 41
                    50 mph CRUISE
                    5-min IDLES
          1975 PLYMOUTH
             440V8/A3
            20JOOO miles
           IDLE CO=7.0%
              TEST 57
                                      90         120         150
                                            TIME, minutes
180
210
240
                 Figure 7. Two distinct patterns of sulfur dioxide emissions, high and low idle.

-------
                0.60
NJ
O
                0.40
u
<
ce
u.
Ul
O
1%
                                   MONOLITH
                                    WITHOUT
                                    AIR PUMP
                                                  MONOLITH;
                                                    WITH   i
                                                   AIRPUMP
PELLETED
WITHOUT
AIR PUMP
PELLETED!
  WITH   j
AIR PUMP i
                                                    Figure 8. Sulfur dioxide purge fractions, post 50 mph cruise idle.>

-------
                               TABLE 13
                    Sulfur Dioxide Purge Fractions

	Configuration	      50 MPH Cruise         CFS
(1)  Monolith without Air Pump (M)                0.46             0.25
(2)  Monolith with Air Pump (MA)                  0.12             0.10
(3)  Beaded with Air Pump (B)                     0.45             0.32
(4)  Beaded with Air Pump (BA)                    0.06             0.16
                                        203

-------
TABLE 14
Mileage
Interval
(Thousands) All Tests

0-5
5-10
10-15
15-20
20-30
30-40

No. No.
Tests Cars HC CO NOX
20 16 1.1 16.6 2.8
22 20 1.5 24.8 2.7
26 25 1.5 20.6 2.7
15 15 2.2 29.3 2.8
31 24 1.9 28.9 2.9
10 7 2.1 34..3 3.9
i ? ? i 11 <; i *
No.
Tests
7
7
9
8
12
5
7
(gin/mi )
GM Tests
No.
Cars HC CO NOX
6 0.9 12.0 2.5
7 1.2 20.2 2.4
8 1.0 12.0 2.7
8 1.7 22.7 3.0
9 2.0 19.9 3.4
4 0.8 10.1 4.3
? n Q Q n 1 1

No.
Tests
7
4
3
1
2
1

Ford Tests
No.
Cars HC CO NOX
5 1.1 8.0 3.6
4 0.8 5.0 3.6
3 0.7 7.2 4.0
1 1.2 21.1 2.1
2 1.1 28.9 1.6
1 1.1 30.4 1.3

Chrysler Tests
No.
Tests
6
11
14
6
17
4
1
No.
Cars
6
10
14
6
13
2
1
HC CO NOX
1.4 32.0 2.3
2.1 34.9 2.6
2.0 28.9 2.4
3.1 39.5 2.6
1.9 35.3 2.6
3.9 65.4 4.0
4 fi 7fi A H n

-------
                    Considering all tests, HC emissions rise from 1.1
gm/mi in the 0-5,000 mile interval to a fairly constant value of approxi-
mately 2.0 gm/mi after 20,000 miles.  The manufacturer breakdowns show
generally higher HC emissions from Chrysler products than from Ford or
General Motors.  Carbon monoxide emissions exhibit a slightly irregular
but definitely increasing trend with mileage.  In the 0-5,000 mile
interval, the carbon monoxide emissions for Chrsyler products are already
higher than for either General Motors or Ford vehicles, and the subsequent
increase with mileage is readily apparent.  The apparent reduction in
General Motors carbon monoxide emissions after 20,000 miles reflects the
preponderance of high mileage fleet vehicle tests in the data for over
20,000 miles.  Among GM products tested, fleet cars have lower FTP emissions
than private vehicles.  More Ford tests in all intervals are needed to
determine the significance of the indicated trend of increasing carbon
monoxide emissions with mileage.  In contrast to GM tests, Chrysler fleet
vehicles exhibit higher emissions than do the few Chrysler private vehicles
(Table 12).

                    Sulfur dioxide purge behavior has not been found to be
a function of vehicle mileage.  Detailed emissions results for two vehicles
over extended mileage periods are shown in Tables 15-16.  In Table 15,
the AEL control car consistently shows both low regulated emission and low
sulfur dioxide purge fraction for five tests spanning nearly 14,000 miles.
Through 3 tests (12,753 miles), a police cruiser exhibits similar behavior
(Table 16).  At 12,753 miles, idle carbon monoxide, FTP carbon monoxide,
and the 50 mph idle purge fraction are 0.2 percent, 15.89 gm/mi, and 0.05
respectively.  At the 20,200 mile test, however, idle carbon monoxide,
FTP carbon monoxide, and 50 mph purge fraction have all greatly increase to
7.0 percent, 80.72 gm/mi, and 0.4 respectively.  The change in S02 emission
pattern between these two tests is graphically illustrated in Figure 8.

                    Although one object of this study is to determine the
effect of vehicle mileage on sulfate emissions, no overall correlation is
apparent at this time.  High idle carbon monoxide, deactivated catalysts,
low fuel sulfur levels, and the absence of an air pump combine to yeild
low sulfate values that do not correlate with vehicle miles.

               5.   Conclusions

                    Testing of in-use catalyst vehicles has shown:

                    a.   Low sulfate emissions of 1.3 mg/mi for cars without
air pumps and 4.9 mg/mi for air pump cars as tested by stabilized CFS driving.
These values were obtained from fuel with an average of 0.017 weight percent
sulfur.

                    b.   High CO emissions by FTP testing and idle
measurements.

                    c.   Purge of large amounts of S02 from catalysts upon
deceleration to idle when idle CO is high.
                                   205

-------
                               TABLE 15

                   Summary of Test Results for Car 1
Test No.
Mileage
Idle Data
C0(%)
HC(ppm)
Regulated Emissions
(g/mi )
HC
CO
NOX
SOi, Production
(mg/mi )
FTP
50 mph
SET-7, Cycle 1
SET-7, Cycle 2
SET-7, Cycle 3
SET-7, Cycle 4
SET-7, Cycle 5
S02 Purge
Fraction
50 mph
SET-7
Sulfur Balance
(% Recovery)
FTP
50 mph
SET-7
17
2,715

0.0
90.


0.59
6.19
2.35

0.06
3.33
6.57
7-46
9.96
9.92
9.31


0.04
0.06


4.7
18.0
51.6
31
4,982

0.0
70.


0.57
5.21
2.16

1.64
6.53
13.37
13.38
16.79
18.08
16.40


0.02
0.07


17.0
39.7
61.1
56
8,319

0.0
90.


0.56
6.27
2.21

0.72
8.24
12.40
12.23
15.52
13.93
14.88


0.01
0.04


1.0
17.2
37.7
87
10,317

0.0
70.


0.70
4.96
2.62

0.71
7.06
5.15
3.66
4.55
7.66
5.86


0.05
0.03


0.7
11.7
36.6
124
13,732

0.0
40.


0.63
7.85
2.36

0.70
5.26
7.43
7.37
5.22
6.36
7.56


0.02
0.03


1.6
13-7
36.5
Fuel S(wt.  %)
                          0.017     0.019     0.021       0.018     0.018
                                          206

-------
                               TABLE 16
                   Summary of Test Results for Car 3
Test No.
Mileage
Idle Data
CO (%)
HC (ppm)
Regulated Emissions
(g/mi )
HC
CO
NOX
SOk Production
(mg/mi )
FTP
50 mph
SET-7, Cycle 1
SET-7, Cycle 2
SET-7, Cycle 3
SET-7, Cycle 4
SET-7, Cycle 5
S02 Purge
Fraction
50 mph
SET-7
Sulfur Balance
(% Recovery)
FTP
50 mph
SET-7
6
17

0.1
50.


0.99
14.32
2.98

0.59
3.33
1.97
1.64
2.40
2.41
2.96


0.17
0.06


5.2
41.2
91.0
30
9,298

0.0
125.


0.76
7.74
2.35

1.24
2.92
3.62
3.73
3.83
3.83
4.31


0.07
0.08


10.3
42.8
89.2
41
12,753

0.2
400.


1.03
15.89
2.11

0.44
1.19
0.78
0.87
0.81
0.85
0.76


0.05
0.03


9.5
29.4
75.3
57
20,200

7.0
460.


3.18
80.72
2.14

0.23
1.12
0.10
0.18
0.13
0.22
0.13


0.40
0.23


2.7
22.8
39.6
94
30,288

7.0
460.


3.40
90.69
2.38

1.66
3.44
1.00
0.49
0.56
0.52
0.47


0.33
0.24


2.5
22.6
36.7
110
35,226

5.8
400.


3.61
66.98
3.95

0.55
3.20
0.49
0.42
0.47
0.49
0.49


0.30
0.22


2.8
27.4
57.1
Fuel S (wt.%)
                     0.016
0.012
0.010
0.012
0.012
0.012
                                      207

-------
                    d.   Little particulate sulfate in idle periods
for either high or low CO conditions.

                    e.   Marked variation in emission pattern by vehicle
manufacturer.

                    f.   Frequent enrichment of idle mixture settings.


     D.   Vegetation Studies

          Investigators at the Department of Plant Pathology at the
University of Minnesota, St. Paul, Minnesota, are studying the effects
of sulfuric acid aerosol on vegetation under a grant from EPA.

          A greenhouse facility has been developed to expose vegetation
to submicrometer sulfuric acid aerosol.  The exposure system consists of
a bank of six clear acrylic tubes (30.5 cm diameter x 1.2 m length) lined
with clear teflon film (0.01 mm thick).  Two of the six tubes provide the
sulfuric acid aerosol and the remaining four serve as controls.   All the
tubes are provided with teflon inlet ports through which parts or whole
plant shoot systems can be inserted for the exposure.  The design facilitates
simultaneous exposure of comparable parts of the same plant to both the acid
aerosol and the control environments, thus providing an efficient system
for studying acute effects.

          The sulfuric acid aerosol is generated by nebulizing concentrated
sulfuric acid, followed by inertial removal of particles larger than 1 ym
and charge neutralization of the submicrometer particles.  Mass flow
calculations during the initial experiments indicate a sulfuric acid
aerosol concentration of 25-30 mg/rrH.  Preliminary cascade impactor studies
show that more than 50 percent of the particulate mass is less than 0.5 ym
in diameter and that 30 percent of the particles are below the limits of
inertia! impaction at atmospheric pressure.
                                                                           3
          Hybrid poplar, pinto bean, and soybean when exposed to 25-30 mg/m
sulfuric acid aerosol for 4 hours showed marginal and tip necrosis.  This
symptom is very similar to fluoride- and chloride-induced injury on broad-
leafed plants.

          Additional work is planned to further characterize the aerosol,
to screen a variety of plant species for their response to the aerosol
under different environmental conditions, and to determine the aerosol
neutralization characteristics of NH3 in relation to plant effects.
                                     208

-------
IV.  PROBLEM AREAS

     A-   Ultrafine Sulfuric Acid Studies

          Automobiles equipped with current catalytic converters emit
ultrafine sulfuric acid (H2S04) mist.  Animal exposure studies using
particles larger than ultrafine range (< 0.1 ym mass median dfameter)
show that as the particle size diminishes toward ultrafine, the pulmonary
resistance increases in exposed lungs.  The EPA needs data on the effects
of ultrafine H2S04 mist on the lung to compare with the large particulate
data.  There have been problems in the generation and calibration of the
ultrafine aerosol and in the development of analytic techniques for animal
exposures.  These problems are now near solution, and exposure studies
will begin soon.

     B.   Nitrogen Dioxide

          The effects of nitrogen dioxide (N02) on human health have become
of increasing importance at a rate which has overtaken the capability of
HERL facilities to study them.  Because of the new priority, the inadequate
N0« investigation facilities at HERL must be expanded to allow more sophis-
ticated human exposure studies.

     C.   Manganese Exhaust Products

          There has been some question regarding what compounds result from
combustion of automotive fuel containing manganese.  Until these questions
are reached, experiments in health effects of inhalation of manganese
compounds cannot be evaluated in terms of relevance to exposure to automotive
exhaust containing manganese combustion products.

     D.   Diesel Exhaust

          As the number of diesel-equipped vehicles grows, so does the concern
for the resulting increased diesel emissions and the associated risks to
health.  EPA is confronted with serious problems encompassing the generation
of appropriate diesel-emission samples, calibration of the equipment, and
isolation of the components of the extremely complex exhaust to determine
the effects of the single pollutants.  The vast number of components com-
prising the total complex diesel exhaust must each be isolated to identify
it with effects observed in exposure studies.  Potentiation from interaction
of these components can then be separated from the cause-effect relationship
of the single pollutants.  This problem will require close coordination with
the engineering effort and advances in state-of-the-art analytical technology.
                                        209

-------
V.    PLANS FOR FUTURE RESEARCH

     A.   Animal Studies

          1.    Sulfates

               The objective of a study initiated in the fall of 1976 is
to determine how metallic sulfates, sulfur dioxide, and sulfuric acid can
alter respiratory defense mechanisms, particularly the alveolar macrophage.
This work is being conducted under EPA grant to the School of Public
Health, Harvard University.  Since the grant was not initiated until Octo-
ber 1976, only preliminary experiments have been conducted.

               Immunosuppressive effects of metal sulfates will be the
focus of new studies undertaken by HERL at Research Triangle Park, N.C.
These planned experiments will be in-house.

               The catalytic converters currently used in vehicles have
been shown to generate or allow sulfuric acid (H2$04) emissions in the form
of ultrafine particulate.  It has been shown that HgSCty aerosols in larger
range particulate size produce harmful effects on animals.  It has been
difficult, however, to measure the health effects of H2S04 in the ultrafine
(< .1 ym mass median diameter) range.  Animal exposure experiments have been
designed to determine how the health effects of ultrafine aerosol differ
from the effects of aged aerosols.  It has been difficult to measure levels
of ultrafine H2S04 generated under laboratory conditions during attempts to
simulate "real-world" conditions, but recent designs have produced equipment
to overcome this barrier.

          2.    Metals

               No new platinum studies are planned, and most extramural
projects have been completed or are nearing completion.  Research for the
ongoing studies will continue with final reports due in 1977 or 1978.
Studies (discussed in Section III) for which final reports are shortly
expected are the following:

               a.   An investigation of whether a hypersensitive state
can be induced by exposure to platinum or to palladium metal.

               b.   An investigation of the use of leucocyte metabolism as
an indicator of health effects.

               c.   A study of the biological activity of platinum in the
induction of pulmonary neoplasia by benzo(a)pyrene and of the role of platinum
in changes of transpulmonary absorption, pulmonary clearance, and pulmonary
response to intrapulmonary particulates.
                                     210

-------
               d.   An evaluation of the effects of platinum sulfate
in major metabolic pathways in the central nervous system.

               e.   Chemical characterization of one MePt compound
prepared from K/tClg and methylcobalamin (MeB19) and isolation and
characterization MePt products from the reaction of Pt(S04)2 with MeB12.

               f.   Studies of the cellular toxicity and mutagenicity of
at least one MePt compound as well as several related Pt-salt complexes.

               g.   Evaluation of analytical methods for measuring trace
elements in biological matrices (especially mammalian tissue).   A series
of reports will document the various limits of detection achieved;
concentration, separation, and extraction techniques; sources of contami-
nation; and other considerations in order to recommend the most promising
of the methods developed applied, and reported to date.  It is  anticipated
that this series of reports on methods for individual metals (platinum,
palladium, nickel, manganese, chromium, cerium, and vanadium) will be
valuable to researchers in the area of trace metals, especially where
measurements of extremely small amounts of metal in tissue samples are
necessary for interpretation of experimental results.

               h.   The HERL at Research Triangle Park plans to do further
work on determining the effects of manganese oxides alone and in combina-
tion with sulfuric acid on animal tissues.
               In addition to these reports of separate studies, a
followup report is planned for publication in late 1978 to complement
the report of the 1975 review conference on platinum research which was
published in 1976 as "Proceedings - Catalyst Research Program Platinum
Research Review Conference."  Since platinum had received little attention
in research prior to the establishment of the CRP, this summary report
should become an extremely useful reference for future researchers in the
field.

     B.   Human Studies

          A project scheduled to begin in early 1977 will investigate the
potential health effects of low=level sulfuric acid aerosols in a concen-
tration range of 50 to 200 yg/m  with a mass median diameter (HMD) particle
size in the range of 0.05 to 0.5 micrometers.  It is hoped that this range
will lead to information valuable in light of the increased use of the
catalytic converter which produces a sulfuric acid aerosol.  Previous
relevant studies and the methodology adopted for this project receive
detailed discussion in Section IIIB.
                                     211

-------
          If there is found to be an observable effect from sulfuric
acid mist on the health parameters designated for the study, further
studies will be initiated to aid in the development of a regulatory
standard.  If there is no observable effect, studies in higher risk
groups, i.e., those with chronic lung disease, may need to be performed.

          Power plant emissions and automobile exhaust contain nitrogen
oxides (NO ) which are in themselves irritants or decomponse and react
with other components to form aerosol pollutants.  Research Triangle Park
has plans for studies of short-term human exposures to the oxides of
nitrogen in 1977.

     C.   Mobile Source Emissions

          The research project engaged in the measurement of sulfate,
particulate, and regulated emissions for a group of 49 in-use catalyst
vehicles will continue through 1977.  See Section IIIC for a report on
the current status of the project.

          HERL at Research Triangle Park, N.C., is initiating studies
using surrogate diesel particulate emissions to investigate effects on
health.  Bioassays will include in vitro screening tests for mutagenicity
resulting from exposure to diesel whole exhaust and to its isolated
components.

          The final HERL report on in-use catalyst-equipped vehicles is
to be completed in FY 78.

     D.   Vegetation Studies

          Additional work is planned by the Department of Plant Pathology
University of Minnesota, to further characterize sulfuric acid aerosol,
to screen a variety of plant species for thefr response to the aerosol
under different environmental conditions, and to determine the aerosol
neutralization characteristics of NHL in relation to plant effects.


VI.  CONCLUSIONS

     Although many projects performed by the Health Effects Research
Laboratory at Research Triangle Park, N.C., are continuing studies for
which final results are not yet available, several conclusions are indi-
cated at this time.
                                                   0
     Laboratory animals exposed to sulfuric acid aerosol for short periods
of time experienced a decrease in hematocrit, a greater increase in the
number and percent of polymorphonuclear leucocytes (PMN) than air-exposed
                                     212

-------
animals, and a greater decrease in the number and percent of lymphocytes
than controls.  Exposure to ozone immediately before exposure to sulfuric
acid resulted in an increase in respiratory infections as compared to
animals exposed only to one pollutant or to both pollutants in reverse
order.  Plants exposed to sulfuric acid showed marginal and tip necrosis
similar to fluoride- and chloride-induced injury on broadleafed plants.

     Results of metals studies indicated that manganese inhalation enhances
respiratory infections; that platinum does not induce allergy, but that
palladium, when complexed with albumin, does induce a delayed allergy; and
that platinum sulfate is a severe eye irritant, may have some cytogenetic
activity, and appears to have a mildly depressive effect on behavior.  It
has also been shown that platinum, as platinum sulfate or as potassium
hexachloroplatinate, can be methylated in the same way that mercury can
be methylated.

     Several conclusions can be drawn from the studies on mobile source
emissions.  Tests on 49 in-use catalyst vehicles show, in general, low
sulfate emissions and high carbon monoxide values.  Few cars tested have
met all 3 emissions standards (15.00 gm/mi CO, 3.1 gm/mi NO, and 1.5 gm/mi
HC); the carbon monoxide regulation was exceeded most often.  A direct
correspondence exists between high idle carbon monoxide and carbon monoxide
mass emissions, and carbon monoxide emissions are inversely related to
sulfate emissions.  In addition, the use of an air pump leads to low
carbon monoxide but to higher sulfate emissions.  It has also been found
that large amounts of sulfur dioxide are purged from catalysts upon deceler-
ation to idle when idle carbon monoxide is high.  The emissions were found
to vary significantly according to manufacturer.  Many cars revealed
enrichment of the idle mixture settings.  In relation to mileage, the tests
indicate that HC emissions rise with mileage to a fairly constant value
of 2.0 gm/mi after 20,000 miles and that carbon monoxide emissions show a
slightly irregular but definitely increasing trend with mileage.  Sulfur
dioxide purge behavior has not been found to be a function of mileage,
however, and no correlation between mileage and sulfate emission is
apparent at this time.
                                    213

-------
                             APPENDIX D-l

                  AEROSOL GENERATION AND MONITORING,
                            HUMAN EXPOSURE


     The basic design of the 2.4 x 2.4 x 2.2 meter acrylic plastic controlled
exposure laboratory which was designed for use with pollutant gases has been
modified for use with FLSO* aerosol.

     The air inlet system has been modified to include particulate and gas
(CBR) filter elements.  Preconditioned building air will be drawn through
these filters to insure that the test subject is exposed only to the pollu-
tants defined by the test protocol.  The ^$04 aerosol is added to the
clean air inlet after the filters and upstream from the mixing and chamber
inlet distribution elements.  These elements are designed for minimal effects
on aerosol particle size and concentration while insuring a uniform distri-
bution of aerosol concentration in the human exposure zone of the chambers.

     A block diagram of the present exposure chamber system with the CBR
filter units added to the supply and exhaust air subsystem is provided by
Figure 9.  The input filter unit occupies a volume of approximately 24
cubic feet near the entrance of the chamber air input supply duct.  The
chamber output filter is located outside the building near the chamber
exhaust blower.  The electrical output of the aerosol monitoring instruments
is used to control the rate of aerosol generation.  The aerosol generator,
exposure chamber, aerosol monitors, and a pollutant flow controller form a
closed loop control system similar to that presently used in the approved
ozone exposure study.  The response time of the aerosol monitors is less
than the 90-second air exchange rate of the chamber.

     The existing chamber safety system consisting of the alarm system and
pollutant flow inhibitor blocks illustrated in Figure 10 will be expanded
to include the aerosol generating system.  Any abnormal conditions detected
by the safety system will result in immediate shut down of the generators
and an automatic purge of the pollutant in the chambers.

     Sulfuric acid aerosol will be generated using the fuming sulfuric
acid technique described by Scaringelli and Rehme (13).  The particle
diameter of these aerosols in 50 percent Rh air was found to be approxi-
mately 0.5 urn.  In another animal exposure study described by Hinners,
ert al_. (14) using this H£S04 generation technique, the aerosol particle
diameter was found to be approximately 1 vim.  By adjustments to generator
air flow rates and other factors affecting ozone nuclei growth patterns
(dilution rates, dilutor design, generator temperature control, etc.), it
is expected that the H2S04 aerosol particle diameter can be controlled to
approximately 0.05 to 0.5 ^m with relative ease.  The mass concentration
will  be controlled at 100 ug/m3 ± 25 (measured by EAA) by adjustment of the
dry air flow over the surface area of fuming sulfuric acid in the generator.
Exposure atmospheres will be exhausted through high efficiency particulate
filters to the outside of the building.
                                       214

-------
Isi   :
t->
Ul
                        ORIFICE PLATE DIFFERENTIAL PRESSURE
               PREFILTER  '
             CARBON FILTER!
              PARTICULATE
              FILTER (HEPA)
                               POLLUTANT
                                 SUPPLY
                                   1
                                                                   ALARM
                                                               ANNUNCIATOR
              CONTROL SIGNAL
                               POLLUTANT
                                 FLOW
                               INHIBITOR
                                  ALARM SYSTEM:
                                 POLLUTANT LEAK,
                                  FIRE, SMOKE, &  :
                                 COMBUSTIBLE GAS
                                    MONITORS
                               j
                                                                                       CONTROL SIGNAL
                                                                            CONCENTRATION LIMIT SIGNAL!
POLLUTANT]
  FLOW   ;
 CONTROL :'
LINII i aiuNALt

*
FLOW
•nMTRfii
A SAMPLE'
POLLUTANT;
MONITORS i
A
i/ENT



DATA
RECORDING
SYSTEM
i ' "
                                              SIGNAL :
                                    C=TJ
                                         II  p
                                    LF
             POLLUTANT FLOW!
     ORIFICE!
      PLATE'
      MIXER'
                                                                 SAMPLE FLOW!
                                              TEMPERATURE)
                                               & DEW POINT
                                                 SENSORS   !
EXPOSURE)
CHAMBER '•
u>
PARTICULATE
   (HEPA)
   FILTER
b
   CONSTANT
 SPEED RADIAL
WHEEL EXHAUST
      FAN
                                 MAIN AIR FLOW1
ROOF
VENT
                                           Figure 9. Exposure chamber system with added filter units.!

-------
AIR
COMPRESSOR


PRESSURE
REGULATOR
— ™«

MOISTURE
TRAP
^^••^w
"1
                          450cm3/min
                                                                                  AEROSOL
                                                                                 GENERATOR
                                                                                                OUTSIDE
                                                                                                 VENT
                                                                                                     NO
                                         MICRO-
                                        METERING
                                         VALVE
                                                                     ROTAMETER
MEATLESS
   AIR
 DRYER
              ALARM
              BELLS &
              HORNS
5VDC
LIMIT
SIGNAL

POLLUTANT
FLOW
CONTROL
SYSTEM

                                           H2SO4 GAS
                                          TO CHAMBER
                                          AIR SUPPLY
                             ALARM SYSTEM:
                            POLLUTANT LEAK,
                             FIRE, SMOKE, &
                            COMBUSTIBLE GAS
                               MONITORS
                                                       LIMIT
                                   3-WAY  !
                                  SOLENOID
                                   VALVE  !
                                                                                                     CHAMBER
                                                                                                   TEMPERATURE
                                                                                                    & DEW POINT
                                                                                                     SENSORS
SIGNAL
                             DATA
                          RECORDING
                          STRIP CHART
                           RECORDER
                         J
                                                      CONDENSATION
                                                         NUCLEI
                                                        MONITOR
                         ELECTRICAL
                          AEROSOL
                            SIZE
                         ANALYZER
                             Figure 10. Method for control of the aerosol generator input air supply, j
                                                                                                            SAMPLE I
                                                                                                             FROM S
                                                                                                           'CHAMBER!

-------
     The methods for control of  the aerosol generator input supply
air will be similar to that presently used for the ozone generator.
As shown in Figure 10, dry compressed air will flow through a system
of pressure regulators, flow restrictor valves, a flow-controlling
servo valve, and a rotameter to  the input of the aerosol generator.
The electrical outputs of the  particle counter and aerosol analyzer will
be used to operate a servo controlled valve or mass flow controller to
maintain and adjust the volume of dry air through the aerosol generator
for the desired concentration  in the exposure chamber.  A normally closed
solenoid valve on the output line from the air supply will shut off the
air supply to the aerosol generator whenever it becomes necessary to stop
the flow of aerosol particulates.  The maximum flow of pollutant into the
chamber will be restricted by  the micrometering valve and a limit position
of the servo valve.

     Aerosol particle size and mass concentration monitoring will be accom-
plished using several measurement instruments.  Two instruments giving
real-time instantaneous response will be used to monitor changes in aerosol
size and number concentration.   One of these instruments (condensation
nuclei counter) measures the number of particles in the size range 0.001 pm
to 1.0 urn diameter.  The instrument (light-scatter particle counter) measures
the size and number of aerosol particles individually, one-by-one in the
size range of 0.3 pm to 6 ym diameter.  These two instruments will be used
by the control operator to adjust the aerosol generators to a steady output
rate within preselected size and concentration limits.  An additional instru-
ment, the electric mobility aerosol analyzer (EAA), is the primary instrument
to obtain average aerosol mass concentration and MMD exposure data over a
size range of 0.0032 to 1.0 pm.  This is a continuous, real-time instrument
that counts particles according  to size.

     Because these proposed studies will be conducted in controlled environ-
ments where the exposure aerosol is generated from pure HUSO* into a clean,
particle-free atmosphere, monitoring instruments need not be specific for
the chemical composition of the  aerosol.  However, the continuous monitoring
instruments must accurately measure the exposure aerosol concentration and
size.  To insure this accuracy,  standard manual calibration techniques will
be employed.

     Particle size calibration measurements will be periodically made
microscopically from samples collected by thermal precipitation, inertial
impact, and by filtration.  Also, the light-scatter particle counter will
be calibrated using aerosols generated from slurries of known particle
sizes of polystyrene latex spheres.  Mass concentration calibration measure-
ments will be made by collecting samples on high efficiency filters and
determining the mass of H^S04  collected using standard chemical analysis
methods (i.e., colorimetric and  conductimetric).
                                       217

-------
                              REFERENCES

 1.   Strong,  D.M.   In Vitro Stimulation of Murine Spleen Cells using a
     Microculture  System and a Multiple Automated Sample Harvester.
     J..  Immunol. Methods   2.:279-291  (1973).
 2.   Graham,  J.A., et al_.   Effect of Nickel Chloride on Primary Antibody
     Production  in the Spleen.  Environ. Health Persp. 12:109-113 (1975).
 3.   Gardner, D.E., F.J. Miller, J.W. Illing, and J.M. Kirtz.  Increased
     Infectivity with Exposure to Ozone and Sulfuric Acid.  Toxicol.
     Lett.  1:59-64 (1977).
 4.   Gardner, D.E., e£ al.   Alterations in Bacterial Defense Mechanisms
     of  the Lung InducecTby Inhalation of Cadmium.   Bull. Eur. Physiopathol.
     Resp.   13,: 157-174 (1977).
 5.   Eisenbud, M.    Arch.  Ind. Hyg.  Occ. Med.  6_:214 (1952).
 6.   Battigelli, M.D. and J.F. Gamble.  Air Quality Monograph #75-25.
     American Petroleum Institute (1975).
 7.   Lewis, R.,  et al.  U.S. Environmental Protection Agency, Res. Tri.
     Park,  N.C.  TJuTy 1972).
 8.   Airey, M.A.,  et. a\_.  Publication No. TR-1T900-3, Greenfield, Attaway,
     and Tyler,  Inc., San Rafael, CA  147 pp.  (1976).
 9.   Dorsch,  R.   Unpublished Dissertation, Wurzberg, Germany (1913).
10.   Toyama,  T.  and K. Nakamura.  Ind. Health  2_:34-35 (1964).
11.   Amdur, M.O.,  et a]_.  AMA Arch.  Ind. Hyg. Occ.  Med.  6_:305-313 (1952).
12.   Hazucha, M. and D.V.  Bates.  Nature 257/5521:50-51 (1975).
13.   Scaringelli,  F.P. and K.A. Rehme.  Anal. Chem.  4J_:707-13 (1969).
14.   Hinners, R.G., J.K. Burkhardt,  and C.L.  Purte.  Arch. Environ. Health
     16:194-206  (1968).
                                     218

-------
                  APPENDIX E
THIRD ANNUAL CATALYST RESEARCH PROGRAM REPORT
                 Prepared By

Office of Mobile Source Air Pollution Control
          Washington, D.C.  20460
    U.S.  ENVIRONMENTAL PROTECTION  AGENCY
     OFFICE  OF  RESEARCH AND DEVELOPMENT
     HEALTH  EFFECTS  RESEARCH LABORATORY
     RESEARCH TRIANGLE PARK, N.C.  27711

-------
                           TABLE OF CONTENTS
List of Tables	

List of Figures	

  I.      Introduction.,

 II.      Sulfuric Acid
          A.   Introduction	             224
          B.   Mechanisms and Laboratory Studies	             225
          C.   Sulfuric Acid Emission Data	             227
          D.   Decrease in Sulfuric Acid Emissions with
               Extended Mileage	             232
          E.   Sulfuric Acid Emissions from Future Prototypes             235
          F.   Sulfuric Acid Traps	             240
          6.   Sul fate Measurement Methods	             242
          H.   Correlation of Sulfuric Acid Tests Among
               Di fferent Laboratories	             243
          I.   Roadside Sulfuric Acid Levels	             244
          J.   Conclusions	             244

III.      Hydrogen Cyanide

          A.   Introduction and EPA Work	             245
          B.   GM Work	             249
          C.   Ford Work	             251
          D.   Chrysler Work	             254
          E.   Concl usi ons	             254

 IV.      Ruthenium Emissions

          A.   Introduction	             257
          B.   Chrysler	             257
          C.   Conclusions	             258

  V.      Ammonia

          A.   Introduction and EPA Work	             258
          B.   Ford Work	             259
          C.   GM Work	             260
          D.   Conclusions	•	             260
                                220

-------
                                                                         Page

 VI.       Polynuclear Aromatic Compounds

          A.   Introduction	            260
          B.   VW Work	            262
          C.   Conclusions	            268

VII.       Diesel Particulate Emissions

          A.   Introduction and Background	            268
          B.   Composition of Diesel Particulates	            269
          C.   Quantity of Diesel Particulates	            270
          D.   Conclusions	            272
                             221

-------
                            LIST OF TABLES
                                                                          Page
 1    Sulfuric Acid Emissions from Dispersion Experiment
     Cars	             228
 2    Sulfuric Acid Emissions from 6M Production Cars	             229
 3    Sulfuric Acid Emissions of Different Catalysts	             231
 4    GM Sulfuric Acid Tests of Transportation Pool Cars	             234
 5    Ford Sulfuric Acid Test Results on 1976-1977 Cars	             236
 6    Sulfuric Acid Tests on Prototype Vehicles	             237
 7    Ford Sulfuric Acid Tests on Low NO  Prototypes	             238
                                       A
 8    Chrysler Sulfuric Acid Test Results	             241
 9    EPA-ORD Test Results on Volvo (Rich Malfunction)	             247
10    Exxon Results on Volvo	             248
11    EPA-ORD Test Results on VW Certification Cars	             250
12    HCN Tests Results of 3-Way Catalyst Vega	             252
13    Chrysler HCN Laboratory Studies	             255
14    Chrysler HCN Laboratory Studies	             256
15    SM Ammonia Data	             261
16    Particulate Emission Rates of Light-Duty Diesel
     Vehicl es	             271
                                 222

-------
                            LIST OF FIGURES
                                                                         Page
1     PNA Percent Distribution in Exhaust Emissions	             264
2    Total PNA Emission Gasoline Versus Propane Operation	             266
3    PNA Emissions as Function of CO Emission, in Europe
     Test	             267
                                    223

-------
I.   INTRODUCTION

     As hydrocarbon, carbon monoxide, and nitrogen oxide emissions are
controlled, various unregulated emissions may be altered.  Introduction
of a new emission control system can increase or decrease the level of a
particular unregulated pollutant or, in extreme cases, result in formation
of entirely new pollutants.  It is important that these pollutants be
characterized to be certain that control of regulated emissions creates
no problems in other areas.

     The unregulated pollutants discussed in this section are sulfuric acid,
hydrogen cyanide, ammonia, ruthenium, polynuclear aromatic hydrocarbons,
and diesel particulate.

II.  SULFURIC ACID

     A.   Introduction

          The use of oxidation catalysts results in oxidation of both
hydrocarbons and carbon monoxide.  In addition, oxidation catalysts oxidize
a portion of the sulfur dioxide in the exhaust to sulfuric acid.  The sulfur
dioxide in the exhaust is formed by combustion of trace quantities of
organic sulfur compounds in gasoline to sulfur dioxide.  Gasoline
contains an average sulfur level of 0.03% which is very low compared to
other fuels.  As a result, sulfur dioxide emissions from gasoline-fueled
vehicles are less than 1% of the total sulfur dioxide emissions.  Most
sulfur dioxide emissions come from stationary sources.

          The sulfur dioxide in the atmosphere is photochemically oxidized to
sulfuric acid.  Even if the oxidation catalyst converted all  of the
sulfur dioxide to sulfuric acid, the resultant sulfuric acid contribution
from these vehicles would be less than 1% of the total sulfuric acid
in the atmosphere.

          However, it is possible to have high localized levels of sulfuric
acid along heavily traveled roads.  EPA has determined that the condition
of most concern is a heavily traveled freeway with maximum traffic flow.
EPA developed a driving cycle called the Congested Freeway Driving
Schedule (the CFDS) with an average speed of 35 mph to represent this
condition.  During the past year, various companies have obtained extensive
emission data using this cycle.  Laboratory studies have also been done
to determine how different parameters (e.g. catalyst composition) affect
sulfuric acid formation.  Work has also been done to measure sulfuric
acid levels along roadways and to determine vehicle sulfuric acid emissions
at higher mileages.
                                    224

-------
     B.   Mechanisms and Laboratory Studies

          Experiments were conducted by GM, Ford, and Chrysler using
laboratory apparatus or, in some cases, engine dynamometers or vehicles, to
determine the effect of various parameters on sulfuric acid formation.  The
laboratory work performed was less extensive than that conducted In
previous years when little was known about the formation of sulfuric acid
over automotive catalysts.

          GM studies showed that rhodium-containing catalysts produce less
sulfuric acid than those containing only platinum or palladium.  This confirms
vehicle work done under an EPA contract with Exxon which showed that a
rhodium-containing catalyst had unusually low sulfuric acid emissions.

          Ford studies on the effect of noble metal composition on sulfuric
acid formation show that it is possible to produce a palladium catalyst
that forms less sulfuric acid than either a platinum-palladium catalyst
or a platinum-rhodium catalyst.  Additional Ford work shows that a
rhodium-containing catalyst also has very low sulfuric acid emissions.

          Ford conducted extensive engine dynamometer tests on 24 catalysts
divided into four series (6 catalysts per series) to evaluate their
sulfuric acid emissions.  The catalysts were aged 100 hours for pre-
conditioning prior to sulfuric acid testing.  Catalyst light-off
temperature and HC/CO conversion efficiency was measured at 800°F.
Sulfuric acid measurements were made under conditions approximating 55 mph
steady state operation of vehicle.  Catalysts tested included Pt/Pd, Pt/Rh
(with varying ratios of Pt/Rh), and platinum.  The major finding of
these tests is that rhodium-containing catalysts emit less sulfuric acid
than catalysts without rhodium.  There is a marked change with catalysts
containing more than 10% rhodium, which emit far less sulfuric acid
(about 10% conversion) than catalysts with less than 10% rhodium
(about 40% conversion).  Ford indicated that there is a loss in HC control
with the increased rhodium content.  However, it appears that the change
in HC emissions is less than that for sulfuric acid.

          Ford also conducted some laboratory studies on catalysts that have
been installed on vehicles and operated for 50,000 miles.  Formation of
sulfuric acid was compared between the 50,000 mile Engelhard II B
catalysts and fresh catalysts.  It was found that the aged catalysts
had much lower sulfuric acid formation depending upon factors such as
presence of reducing gases (certain HC compounds and CO), and space
                                    225

-------
velocity.  Even though the aged catalysts had high activity for propane
and CO conversion, they had a two to six-fold decrease in sulfurlc acid
formation versus the fresh catalysts.  These results are 1n accord with
vehicle test data discussed later which show very low sulfurlc acid
emissions from catalyst vehicles with extended mileage.

          GM conducted further experiments on the effect of temperature,
space velocity, 0? levels, and CO levels on sulfuric acid formation.  It
was found that low oxygen levels, higher CO levels, high catalyst temperature,
and possibly low space velocity all lead to low sulfuric acid formation.
These results are in complete agreement with previous tests run by GM,
as well as past studies by Ford, Chrysler, other automobile manufacturers,
and EPA contractors.

          GM and Ford both reported results of laboratory studies on
sulfuric acid storage.  Storage of sulfuric acid occurs on fresh catalysts
under most operating conditions, resulting in low sulfurlc acid emission
rates.

          This effect is overcome on a vehicle, generally within the
first 1000 miles.  After that, high temperatures result in release of
sulfuric acid (frequently as S02) with subsequent low temperature opera-
tion permitting storage once again.  Since much of the stored sulfurlc
acid can be released as S02 at the higher temperatures, sulfurlc acid
emissions can be lower for systems which store significant quantities of
sulfuric acid (e.g., pelleted catalyst systems).  Again, this recent
work agrees well with earlier work done by the automotive companies and
EPA.

          Ford reported some laboratory test results on both fresh and
aged 3-way catalysts.  The results on fresh catalysts show that all
three 3-way catalysts tested (M-152C2-3, M-253, and M-257) have comparatively
low sulfuric acid formation (e.g., 40% formation) compared with a fresh
Engelhard II B oxidation catalyst under the same conditions (e.g., 80%
formation of sulfuric acid).  These data indicate that some 3-way catalyst
formulations have inherently lower sulfuric acid formation than conventional
oxidation catalysts independent of other conditions (i.e., 02 level).
Dynamometer-aged 3-way catalysts were also evaluated in the laboratory
reactor apparatus.  Three-way catalyst samples aged for 470 hours showed
about 10% sulfuric acid formation in a laboratory reactor compared with
about 50% sulfuric acid formation for fresh catalyst samples.

          Chrysler did some limited studies using a vehicle (Car 685) to
investigate the effect of CO and 02 levels on sulfurlc add formation.
Chrysler modified the exhaust system of this car to allow injection of
known quantities of CO and 02>  Sulfuric acid emissions were measured at
40 mph steady state conditions.  The results of this work agree with
earlier work showing that oxygen levels of "]% or less result in low
sulfuric acid emissions (e.g., 1 mg/mi) while higher oxygen levels of 2%
or more result in increased sulfuric acid emissions (e.g., 40 mg/mi or
more).
                                       226

-------
          Ford also did further work on how exhaust oxygen levels affect
sulfuric acid emissions.  A Pinto was equipped with a 3-way plus oxidation
catalyst system.  A managed air injection system (with a flow control
valve in the secondary air flow upstream of the oxidation catalyst) was
used to give variable amounts of oxygen (0.5%, 1%, 2%, and 5%).  Complete
test results are not available yet.  Preliminary results suggest that
lower oxygen levels result in lower sulfuric acid levels, but perhaps at
the expense of some CO control.

          Nissan has also done some engine dynamometer work on sulfuric
acid using a 4-cylinder Datsun engine with either a pelleted or monolithic
catalyst.  This work shows a sharp increase in sulfuric acid emissions
when the air-fuel ratio increases from 14.5 to 15.5 but only a moderate
increase in sulfuric acid when the air-fuel ratio is increased further.
This work also shows that monoliths and pellets emit the same quantity
of sulfuric acid at temperatures over 600°C. Below this temperature, the
monoliths tested by Nissan emitted more sulfuric acid presumably due to
lower storage capacity.  The Nissan work also shows lower sulfuric acid
emissions from rhodium or barium (a stabilizer used in the substrate)
containing catalysts.  Additional work by Nissan shows sulfuric acid
emissions to be greater for catalysts with more noble metal dispersed
near the surface.

     C.   Sulfuric Acid Emission Data

          GM tested seven of the 1975-76 vehicles used in their Sulfuric
Acid Dispersion Experiment.  These vehicles were tested on a driving
schedule similar to the driving experienced around the GM Proving Ground
track during the Sulfuric Acid Dispersion Experiment.  This test cycle
took two hours to complete with sulfuric acid measurements being taken
every half hour (i.e., four measurements over two hours).  The fuels
used contained 0.032 wt. % sulfur.  The vehicles were tested at both the
GM Proving Ground and Research Labs.  The last sulfuric acid measurement
for each vehicle is listed in Table 1.

          These/tests showed an overall sulfuric acid emission rate of
51 mg/mi emission rate calculated for the entire 352 car fleet used for
the Sulfuric Acid Dispersion Experiment.  There is reasonably good agree-
ment between the two laboratories.

          GM also measured sulfuric acid emissions from 12 other 1975-76
production vehicles, using the Congested Freeway Driving Schedule (CFDS).
It is assumed that 0.032 wt. % sulfur fuel was used.  These results are
shown in Table 2.
                                      227

-------
                                TABLE 1

                     Sulfuric Acid Emissions  from
                      Dispersion Experiment Cars

Car                                Lab*            Sulfuric  Add  (mg/mi)

'75 302 Ford (Federal)            GM PG                 31.8
                                  GM RL                 41.1
'76 302 Ford (Federal)            GM PG                 73.8
                                  GM RL                 61.0
'75 Chevrolet (Federal)            GM PG                 14.4
                                  GM RL                 14.0
'75 Chevrolet (California)         GM PG                 69.6
                                  GM RL                 94.0
'75 Chevrolet (California)         GM PG                 73.9
                                  GM RL                 98.8
'75 318 Chrysler (California)     GM PG                 84.6
                                  GM RL                 78.2
'76 Pontiac                       GM PG                 32.7
                                  GM RL                 55.1

*Proving Ground = PG
 Research Labs   = RL
                                    228

-------
                                TABLE 2
                     Sulfuric Acid Emissions from
                          GM Production Cars
Car
Emission Control System
Sulfuric Actd (nig/ml)
'75 Cadillac
'75 Oldsmobile
'75 Nova
'75 Vega
'75 Vega
'76 302 Ford
'76 Cutlass
'76 Cutlass
'76 Cutlass
'76 Cutlass
'76 Nova
 '76 Nova
California AIR
California no AIR
California AIR
California AIR
Federal pulsed AIR
Federal AIR
California no AIR*
Federal no AIR
California no AIR
California AIR
Federal no AIR
Federal no AIR
          79
          17
         136
          18
           1
          58
          17
           2
           2
          44
          0.3
          0.3
 *3.6% exhaust 02 for this car; 02 not measured for other cars
                                       229

-------
          These test results show that cars with air injection and
presumably high exhaust oxygen levels have high sulfuric acid emissions
while cars without air pumps and presumably low exhaust oxygen levels
have much lower sulfuric acid emissions.  Unfortunately, GM did not
measure oxygen levels from these cars which would be useful in Interpreting
these results.  Also, GM did not report the mileage of these vehicles.

          Ford has measured sulfuric acid emissions from two 1976 Ford
LTDs, a California version and a Federal version, both with the 400 CID
engine. Ford not only obtained emission data on these cars but also
investigated the effects of spark timing, EGR, oxygen concentration, and
catalyst composition on sulfuric acid formation.

          With the Federal car, Ford found high sulfuric acid emissions
(about 50-60 mg/mi) with oxygen levels from 1% to 5%.  Under 1% oxygen,
the sulfuric acid emission levels decreased dramatically to under 20
mg/mi (as low as about 5 mg/mi in one test with about 0.3% oxygen).
However, sulfuric acid emissions were under 10 mg/mi on this car over
the CFDS with oxygen levels anywhere from 1.5% to 6.5%.  It is unusual
to see significantly higher sulfuric acid emissions at a steady state
speed than over the CFDS.  Other work conducted by EPA, GM, Exxon,
Southwest Research Institute, and Chrysler did not indicate this phenomenon.
Perhaps the higher sulfuric acid emissions are due to lower "spikes" of
reducing gases such as HC and CO over a steady state mode versus a
transient driving cycle.  It could be that the particular Ford tested
was more sensitive to such a phenomenon than other vehicles.  It could
also be that the immediate preconditioning before the CFDS resulted  in
purging of stored sulfuric acid with subsequent storage over the CFDS.
It was found that changing the spark timing (from 6* to 16° BTDC) and
plugging the EGR did not affect sulfuric acid over the CFDS.

          Ford investigated the effect of using Pb/Pd, Pt/Rh, or Pd
catalysts all with the same substrate on both the California and Federal
cars. Results of the tests on these catalysts are shown in Table 3.
These results show that Pt/Rh catalysts do not always give lower sulfuric
acid emissions than other catalysts.

          Ford Also ran tests comparing different versions of the CFDS
denoted as SET 7 and SET 7D.  The SET 7 is extremely difficult if not
impossible for the driver to follow all the time while the SET 7D has
the "noise" modified by removing those parts difficult for the driver to
follow.  Furthermore, the present requirements outlining allowable driver
tolerances for the FTP (+ 2 mph within 1 sec) resulted in the earlier
                                     230

-------
                                TABLE 3

            Sulfuric Acid Emissions of Different Catalysts
Vehicle
Catalyst
Driving Cycle
Sulfuric Acid (mg/ini)
Ford - Federal
  Pt/Pd
  Pt/Rh
  Pd
                       Pt/Pd
                       Pt/Rh
                       Pd
  45 mph
  45 mph
  45 mph
                      CFDS
                      CFDS
                      CFDS
          70
          28
          10
                              10
                              10
                               8
Ford  - California
  Pt/Pd
  Pt/Rh
  Pd
  CFDS
  CFDS
  CFDS
       28-40
          25
          42
                                      231

-------
SET (the one with high "noise") having different drivers following the
CFDS in different ways.  That is, a driver could either follow or not
follow the "noise" and still have a valid CFDS.  Since the "noise" was
difficult to follow, many drivers would not follow it.  The modified
CFDS (SET 7D) is easy enough for the average driver to follow the "noise"
so that most cycles would be driven the same.  Ford ran a total of about
6 SET 7 and 70s on a 1976 California Ford.  Both cycles gave about 50
mg/mi sulfuric acid emissions with the SET 7D being about 18% lower than
the SET 7.  These results are in general agreement with those obtained by
EPA even though the Ford results show greater differences.  Extensive
EPA tests (both in-house and at Southwest Research) on about 20 repeti-
tions of the CFDS (both without any noise whatsoever, a precursor of
SET 7, and SET 7D) show the SET 7D to have about 15% lower sulfuric
acid emissions.

          Chrysler ran sulfuric acid tests on two 1976 production vehicles,
The first car (Car 176, a Dodge Coronet equipped with a 318 CID engine,
a 90 cu. in. catalyst, and an air pump) showed very high sulfuric acid
emission (82 mg/mi) over the CFDS.  The second vehicle (Car 155, a Dodge
Dart equipped with a 225 CID engine, a 90 cu. in. catalyst, and an air
pump) showed much lower sulfuric acid emissions (9.4 mg/mi) over the
CFDS.

          Most of the test results on prototype cars are somewhat higher
than those found in the EPA baseline study.  This study involved sulfuric
acid testing of 78 current and future prototype cars.  These tests
showed catalyst cars without air injection to emit about 10 mg/mi sulfuric
acid while catalyst cars with air injection emitted about 20 mg/mi
sulfuric acid.

     D.   Decrease in Sulfuric Acid Emissions with Extended Mileage

          Very extensive sulfuric acid emission data have been taken on
oxidation catalyst vehicles at low mileages.  The bulk of these data were
taken in 1975 and summarized in the EPA Status Report on Sulfate Emission
Control Technology as well as the EPA Report on Sulfate Emission Control
Technology and the EPA Report on the Baseline Study.  In the Baseline
Study, about 60 vehicles were tested for sulfuric acid emissions. However,
almost no sulfuric acid data have been taken on vehicles at extended
mileage.  EPA funded a major contract with Southwest Research Institute
to measure sulfuric acid emissions on four catalyst vehicles as they
accumulated 50,000 miles.  Exxon Research and Engineering funded an In-
house study to accumulate 50,000 miles on 20 cars and measure sulfuric
acid emissions at various intervals.  Both studies showed a significant
decrease in sulfuric acid emissions occurring at roughly 20,000 miles.
While sulfuric acid emissions can be high initially, most of the vehicles
had emissions below 10 mg/mi after the drop occurred.
                                     232

-------
          As a  result  of  this work,  EPA has  asked the  following companies
to measure sulfunc acid  emissions about every  5,000 miles on certain
vehicles as they  accumulate  AMA  type mileage.

     Manufacturer            Number of Vehicles to be Tested

       GM                                    12 cars

       Ford                                   8

       Chrysler                              6

       AMC                                    2

       Toyota                                 2

       Datsun                                 2

       VW                                     2
These cars would probably  come from existing fleets accumulating mileage
(e.g., precertification fleet) to  keep the cost of such a program at a
lower level.  The manufacturers  have indicated their willingness to
participate  in  this  program and  to have these tests completed by the end
of 1977.

          Aside from this  program, both GM and Ford have reported current
data on sulfuric acid emissions  of cars at higher mileages.

          GM is testing six vehicles (1975 Impalas) in a GM transportation
pool as they accumulate mileage.   Four of these cars were designed to
meet the 1976 California standards while the other two cars are Federal
cars.  The cars currently  have from 5,000 to 15,000 miles.  The results
to date are  given in Table 4 for the CFDS and 60 MPH cruise.  These
results show a marked degradation  in sulfuric acid emissions over the
CFDS but very little degradation during 60 mph test conditions. Additional
emission testing will be done as these cars accumulate mileage.  GM did
not report gaseous emissions (HC,  CO, NO ) from these cars.
                                        /\

          Ford ran a series of sulfuric acid emission tests on some of
their 50,000 mile 1977 certification cars.  The following test sequence
was used with 0.03%  sulfur fuel:
                                        233

-------
                                TABLE 4
                       GM Sulfuric Acid Tests of
Car
R5947
(Federal)

R5949
(California)
R5951
(California)

R5948
(Federal)
R5950
(California)
R5952
leage

700
5000
600
10000
15000
1800
5000
1700
5000
11000
15000
600
5000
10000
800
10000
15000
Transportation Pool
Fuel Sulfur

0.004%

0.004%


0.004%

0.036%



0.036%


0.036%


Cars
Sulfuric Acid
CFDS
0.3
0.4
6
1
0.8
2
0.8
4
3
2
1
108
55
15
64
36
14
Emissions (mg/nri)
60 mph
7
9
16
14
12
14
15
45
73
76
58
134
111
132
112
121
106
                                        234

-------
                222 miles  AMA preconditioning
                LA-4  prep  followed by cold soak
                FTP
                505 sec  of LA-4
                2 minutes  idle
                CFDS
                10 minutes idle
                CFDS
                10 minutes idle
                CFDS
                10 minutes idle
                CFDS
                10 minutes idle
                HFET                                                 s
                                                                    )
                                                                    )
Each test was repeated  on two additional  days.  Ford also tested three
1976 California Capri vehicles at 4,000 miles.  The test results are
given in Table  5.  All  of the vehicles had Pt/Pd catalysts and are
50,000 mile vehicles unless  noted otherwise.

          These vehicles  frequently,  but  not always, show low sulfuric acid
emissions at 50,000  miles.   However,  three of the cars had significantly
higher sulfuric acid emissions (18-108 mg/mi) than the others.  Ford
points out that two  of  these cars also had low CO emissions, showing an
inverse relationship between CO and  sulfuric acid for this system.
Ford is also measuring  sulfuric acid  on some of the 1978 certification
cars.

     E.   Sulfuric Acid Emissions from Future Prototypes

          GM tested  12  future prototype vehicles for sulfuric acid emissions.
These vehicles  included lean burn and 3-way catalyst vehicles.  The lean
burn vehicles had higher  sulfuric acid emissions while the 3-way catalyst
vehicles had low sulfuric acid emissions.  The results of these tests
for the CFDS are given  in Table 6.  GM did not supply gaseous emission
results or any  other details on the control systems other than what is
listed in the table.

          Ford  measured sulfuric  acid emissions from a variety of oxidation
catalyst, 3-way and  oxidation catalyst, and dual catalyst cars designed
to meet low NO  standards.   Even  though two of the oxidation catalysts
had 50,000 miles on  them,  the sulfuric acid emissions were still signi-
ficantly high.  The  sulfuric acid emissions from the 3-way catalyst/
oxidation catalyst prototypes were also high, because of SO, oxidation
occurring over  the oxidation catalyst with air injection before the
oxidation catalyst.  The  conventional 3-way catalyst would, of course,
be expected to  have  very  low sulfuric acid emissions.  Finally, Ford
found high sulfuric  acid  emissions (69 mg/m1) from the Gould dual catalyst
system.  Independent EPA  tests  confirm that the Gould dual catalyst
system does have high sulfuric  acid emissions.  The results of the Ford
tests are given in Table  7 for  the CFDS.
                                    235

-------
                               TABLE 5
                   Ford Sulfuric Add Test Results
on 1976-77 Cars
Vehicle
4A
5A
11A
10A
10A
10A
21B
8A
14A
16A
17B
14B
14B
21A
5C13*
5C15*
5C16*
Engine
Disp.
2.3 litre
2.3
2.3
2.3
2.3
2.3
2.3
2.8
250
302
302
400
400
400
2.3
2.3
2.3
litre
litre
litre
litre
litre
litre
litre
CID
CID
CID
CID
CID
CID
litre
litre
1 i tre
Catalyst
El
MB
UOP
MB
MB
MB
El
El
UOP
UOP
UOP
El
El
El
MB
MB
MB
Sulfuric Acid FTP Emissions
Emissions - (gm/mi)
CFDS (mgpm) HC CO NOx
0
0
6
0
0
0
0
1
1
45
107
1
1
17
21
31
9
.51
.61 ,
.07
.61
.72
.47
.64
.67
.98
.65
.80
.34
.79
.60
.57
.62
.39
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
20
15
10
11
05
05
10
12
16
10
07
23
15
15
08
07
07
2.
5.
2.
4.
1.
1.
2.
2.
5.
0.
0.
9.
7.
3.
0.
0.
0.
86
20
62
47
53
72
92
16
25
66
12
57
88
13
12
76
44
1
1
1
1
1
1
1
0
0
1
0
1
1
2
1
1
0
.29
.03
.65
.75
.54
.50
.31
.96
.93
.13
.94
.06
.02
.33
.00
.10
.64
*1976 California 4K Vehicle
                                    236

-------
                                     TABLE  6



                    Sulfuric  Acid  Tests  on  Prototype  Vehicles
Car
74 Chevelle
74 Chevelle
74 Chevrolet
74 Chevrolet
74 Chevrolet
74 Oldsmoblle
74 Oldsmobile
75 Chevelle
74 Chevelle
75 non-GM
76 Chevelle
Control System
lean burn engine, production
catalyst
lean burn engine, production
catalyst
closed loop fuel injection,
3-way catalyst
closed loop fuel injection,
3-way catalyst
closed loop carburetor,
3-way catalyst
production catalyst, no AIR
production catalyst, no AIR
production catalyst, AIR
production catalyst,
warm-up catalyst, AIR
stratified charge engine, fuel
injection, monolith catalyst
production catalyst, AIR
Fuel Sulfur
Level
0.06%
0.031%
0.06%
0.06%
0.031%
01031%
0.031%
0.031%
0.031%
0.012%
0.031%
Exhaust Oxygen
Level
2.1%
2.1%
0.7%
0.3%
0.7%
1.5%
1.6%
1.5%
1.5%
0.5%
1.5%
Sulfuric Acid (mg/mi)
99
42
0.2
0.1
0.2
44
76
41
15
6
4
switching

-------
                                     TABLE 7

                            Ford Sulfuric Acid Tests
Vehicle

8D1-302-5P
(50K oxidation
  catalyst)
8D1-302-5P
(oxidation catalyst
 from 8D1-302-6P,
 50K)

8D1-302-6P
(50K oxidation
  catalyst)
8D1-302-1P
(3-way and
 oxidation catalysts)
on Low NOx
HC
gm/mi
0.06
„„
0.06
0.07
0.07
..
0.01
0.05
0.06
_
0.06
0.06
0.05
-
0.05
0.05
0.03
0.03
_
0.04
0.09
_
0.07
0.09
_
0.08
0.08
0.15
-
0.16
0.15
0.16
-
0.18
0.18
Prototypes
CO
gm/mi
0.01

0.03
0.00
0.05
_
0.00
0.01
0.02
_
0.05
0.03
0.02
_
0.04
0.00
0.02
0.02
.
0.01
0.68
_
0.66
1.06
-
0.97
0.83
0.16
_
0.14
0.13
0.17
-
0.14
0.18

NOx
gm/mi
0.57
_
0.58
0.61
0.64
_
0.62
0.62
0.57
_
0.56
0.56
0.65
_
0.62
0.62
0.65
0.60
_
0.65
0.46
_
0.47
0.42
_
0.42
0.43
0.60
_
0.70
0.67
0.55
_
0.55
0.56

H2S04
mg/mi
139.4
103.2
107.1
110.9
162.1
106.9
114.5
115.4
88.5
86.8
95.0
87.8
103.6
95.6
92.3
104.6
102.5
117.9
119.7
106.8
12.6
19.9
17.4
13.7
12.8
14.4
15.3
56.2
51.2
61.4
62.3
69.8
53.6
71.4
61.2
                                              238

-------
8D1-302-7P
(3-way and
 oxidation catalyst)
Gould catalyst
(1975 400 CID LTD)
*FTP data
                                   (TABLE 7 cont'd)
0.14
-
0.13
0.14
0.13
-
0.13
0.14
0.33*
0.28
0.40
-
0.43
0.29
0.28
-
0.20
0.26
0.46*
0.39
0.49
_
0.50
0.49
0.61
_
0.67
0.64
0.25*
0.17
27.9
33.3
25.8
34.5
37.4
43.0
45.7
40.1

69.0
                                        239

-------
          Chrysler has run sulfuric add emission tests on advanced
prototype vehicles including the lean burn (with and without oxidation
catalyst), start catalyst, and dual catalyst (Gould reduction catalyst
followed by an oxidation catalyst) vehicles.  The lean burn vehicles
(cars 209 and 271) without catalyst showed very low sulfuric acid emissions
while the lean burn vehicle (Car 4028) with oxidation catalyst showed
higher sulfuric acid emission (41 mg/mi).  Chrysler tested three start
catalyst systems (Cars 454, 332, and 270) which were 1977 California
prototypes. All three cars had 22 cu. in. start catalysts and 90 cu. in.
main oxidation catalysts with air pumps.  These three cars produced 53,
25, and 62 mg/mi sulfuric acid emissions over the CFDS which is in the
same range as that expected from conventional oxidation catalysts without
start catalysts.

          Chrysler tested three dual catalyst type vehicles.  The first
of these, Car 178, had a Gould reduction catalyst followed by a 152 cu.
in. monolithic catalyst with an air pump.  This car emitted a very low
5.4 mg/mi sulfuric acid over the CFDS which is similar to values Chrysler
reported in 1975. The second car, Vehicle 232, has a monolithic Chrysler
reduction catalyst followed by a 90 cu. in. oxidation catalyst and air
pump.  This vehicle also produced very little sulfuric acid (13-25 mg/mi
over the CFDS).  The third vehicle (Car 623) also has a monolithic
Chrysler reduction catalyst followed by a 90 cu. in. oxidation catalyst.
This car has an air modulation system which injects about 7% oxygen
upstream of the reduction catalyst for the first 90 seconds after a cold
start.  After this, sufficient air is injected into the oxidation catalyst
to result in a 2% oxygen level (a diverter valve is used to achieve
this).  With the 2-3% oxygen level, 11.1 mg/mi of sulfuric acid is
emitted over the CFDS.  However, if the diverter valve is inoperative
and all of the air is injected Into the oxidation catalyst (7% oxygen),
sulfuric acid emission levels of 46.3 mg/mi are observed over the CFDS.

     The sulfuric acid emission results for the start catalyst and dual
catalyst systems over the CFDS are given in Table 8.

     F.   Sulfuric Acid Traps

          GM is the only company reporting information on sulfuric acid
traps.  EPA has also done extensive work on sulfuric acid traps through
a contract at Exxon Research and Engineering.  The sulfuric acid trap is
simply a chemical placed in a container (a muffler or catalyst-like
container is satisfactory) on the vehicle after the oxidation catalyst.
All of the exhaust passes through this container allowing the sulfuric
acid to react with the sorbent.

          The EPA work at Exxon showed several promising sorbents, the
most promising of which was 85% CaO, 10% AUG., and 5% Na20.  This
sorbent can be fashioned into either pellets or rings.  Tests on the
sorbent show extremely high efficiency (about 90%) in trapping sulfuric
                                    240

-------
              TABLE 8

Chrysler SuIfuric Acid Test Results
Vehicle

454 (start
  catalyst)
332 (start
  catalyst)
  Right Exhaust
  Left Exhaust
  Left Exhaust
  Dummy Main

  Left Exhaust
  Dummy Start

 270  (start
  catalyst)

 178  (dual
  catalyst)

 232  (dual
  catalyst)
 623  (dual
   catalyst)
Sulfuric Acid
(mg/mi )
51.4
59.4
49.7
13.6
17.1
14.5
18.5
12.9
14.2
7.6
14.8
10.0
13.2
2.9
6.8
16.9
61.8
5.4
13.1
15.6
25.1
11.1
46.3
Oxygen
%
6.4
4.9
4.8
3.2
3.2
3.0
4.0
4.0
4.0
3.0
4.0
3.7
4.0
4.0
3.3
3.7
4.8
2.2
5.3
6.1
6.1
3.3
7.3
                                               CO
                                             0.099
                                             0.466
                                             0.649
                                             0.850
                                             0.822

                                             0.220
                                             0.270
                                             0.304
                                             0.308

                                             0.060
                                             0.043
                                             0.078
                                             0.068

                                             0.090
                                             0.358
                                             0.268

                                             0.134
                                             3.257
                                             2.949
                                             2.2J7


                                             2.66
                                             1-29
                    241

-------
acid even after extended mileage (25,000 miles),  However, the sorbent
swells causing an.unacceptably high pressure drop causing the vehicle to
lose power and increase its fuel consumption.  EPA and Exxon spent a
second year in a contract extension to try to remedy these problems and
were unsuccessful.  The second year Exxon program did identify a number
of sorbents that could be considered promising.  The overall conclusion
of the EPA-Exxon work is that, if vehicle sulfuric acid reductions are
needed, the sulfuric acid trap is a potentially promising candidate but
will need a very large development effort.  Even with a large development
effort, it is not certain that a viable sulfuric acid trap would result.

          In their work, GM screened many different sulfuric acid sorbents
and found two promising candidates, CaO and a Na20-A1203 sorbent.  The
latter sorbent is especially attractive in that ft is sufficiently
selective that sulfuric acid alone with no S02 is trapped.  A vehicle
test on the CaO material showed 95% efficiency when fresh but only 70%
efficiency (over the CFDS) or 40% efficiency (at 40 mph) after 32,000 km.

          The Na20-Al203 trap has been tested only 1600 km and shows no loss
in efficiency.  Even though the GM program is less extensive than the
EPA-Exxon program, the GM results seem to be in agreement with those of
EPA-Exxon.

     G-   Sulfate Measurement Methods

          Both GM and Ford reported work on measurement methods for sulfuric
acid.  This work involves both the collection methods used for automotive
sampling and the analytical methods used on the collected samples.

          GM has investigated CVS flow rate and concludes that either a 300 or
600 CFM flow rate is adequate for a large (18-inch diameter - 20-foot
length) dilution tunnel.  However, GM feels that a lower flow rate of 335
CFM will not work with a smaller dilution tunnel (8.4-inch diameter -
10-foot length) due to high sulfuric acid losses.  The 335 CFM flow rate
is used on almost all current CVS units.  GM also stated that sample zone
temperatures of over 200°F will result in lower sulfuric acid values
(possibly by Toss of sulfuric acid from the filter or incomplete condensation
of sulfuric acid).  GM also found that tunnel inlet temperatures of less
than 200°F (caused by long connecting tubing between the vehicle and tunnel)
cause low sulfuric acid readings.  GM stated that smaller variations in
paired samples were noted with critical flow venturi sampling versus CVS
sampling.  All of these items are being investigated by an EPA contract
currently in progress at Southwest Research.
                                      242

-------
          GM also stated that neither 5300 nor 25,000 CFM cooling fans
result in vehicle temperatures during dynamometer operation that are
typical of on-road operation.  EPA generally uses a combination of
several 5300 CFM fans to assure adequate cooling; EPA has also purchased
a 25,000 CFM fan.  EPA has just started a contract with Olson Labora-
tories to investigate the configuration of cooling fans that will result
in vehicle cooling approximating that occurring during road operation.

          Finally, GM notes that their work shows the perchlorate titration
system to be equivalent to the barium chloranilate system for analysis
of filter samples containing sulfuric acid.  Ford has done similar work.
EPA work also shows both systems are equivalent.

          Ford has extensive work (both in-house and some EPA contract
work) in progress investigating analytical methods for sulfuric acid and
SOp. Ford is currently working on an instrument for continuous S02
measurement of automotive exhaust.  Furthermore, Ford is doing some work
on analysis of sulfuric acid by mass spectrometry.  In particular,
electron impact mass spectral analysis offers the potential of distinguishing
between sulfuric acid itself and salts of sulfuric acid (ammonium sulfate).
Such instrumentation would be very useful in ambient air monitoring,
where sulfuric acid particles could be distinguished from its salts.
Such distinction is valuable since the two compounds probably have
different health effects.

     H.   Correlation of Sulfuric Acid Tests Among Different Laboratories

          During the past year, several vehicles tested by GM and Ford have
also been tested by EPA.  Vehicles tested by the GM Proving Ground were also
tested by EPA in the Sulfuric Acid Baseline Program.  These tests showed
generally good agreement between the laboratories.

          Ford recently completed construction of dilution tunnels at
their Scientific Research Laboratory and another Dearborn Laboratory.
Ford tested a 1976 vehicle equipped with a 400 CID engine at both of
their facilities and sent the car to EPA for correlation testing.  The
test results show reasonably good (but not complete) agreement among the
three laboratories in the sulfuric acid emission values obtained.

          As part of the sulfuric acid deterioration factor program to
be conducted by the manufacturers and EPA during the coming year, sulfuric
acid correlation testing will be needed among the different laboratories.
A round robin car will be used for this purpose.
                                      243

-------
     !•   Roadside Sulfuric Acid Levels

          Both the GM and Ford submissions discussed work on measuring
roadside levels of sulfuric acid.  In particular, GM completed a major
experiment this past year titled the Sulfate Dispersion Experiment.  This
work involved running 352 cars, equipped with catalysts and afr pumps
(designed for the 1976 California standards), around a loop at the GM
Proving Grounds.  The cars were preconditioned with 0.032% sulfur fuel and
operated at about 55 mph.  The average sulfuric acid vehicle emission rate
during these tests was about 0.037 mg/mi.  Ambient air levels of sulfuric
acid were monitored.  Also, a tracer gas (SFg) was emitted from eight pickup
trucks participating in the experiment.  Measuring the level of SFg and
sulfuric acid under various meteorological conditions allowed verification
of an EPA predictive air quality model and also gave an independent
measurement of roadside sulfuric acid levels.

          It was found the normal levels of sulfuric acid in the air
samples due to the vehicles were 3-5 yg/m .   Under adverse meteorology,
increments of 15 yg/m  were found.  The increased level of sulfuric acid
inside the vehicles was also measured and found to be about 4 yg/m .
There was very good agreement between results obtained with the SFg
tracer and the sulfuric acid measurements.

          These results showed that the EPA Highway Model overpredicts
sulfuric acid emissions at the pedestrian level downwind from the roadway.
The overprediction occurs for stable meteorological conditions and gets
worse as the wind speed decreases (when roadside sulfuric acid levels
would be highest).  The EPA Highway Model is more or less satisfactory
under unstable meteorological conditions (when sulfuric acid levels
along roadsides would be much lower).  Apparently, "mechanical mixing"
due to traffic flow is an important parameter and not accounted for in
the EPA Highway Model.  Various government agencies (EPA, DOT, and DHEW)
as well as other automobile companies helped participate in this work.

          In addition to the massive GM experiment, Ford has been conducting
monitoring experiments in the Allegheny Tunnel on the Pennsylvania
Turnpike.  Initially, Ford monitored rubber tire particulate there and
then diesel and sulfuric acid particulate.  The cars driving through  the
tunnel would be those designed for the Federal standards and therefore
have lower sulfuric acid emission rates than the Califoria cars used  in
the GM experiment.  However, to date, no sulfuric acid increment has
been noted from the introduction of catalyst cars.

     J.   Conclusions

          Work during the past year confirmed earlier studies showing
that exhaust oxygen levels are the most important single parameter
affecting sulfuric acid emissions.  Low exhaust oxygen levels (such as
                                   244

-------
those found with 3-way catalyst cars) can result in sulfuric acid emission
levels from catalyst cars of under 5 nig/mi which is equivalent to levels
found with non-catalyst cars.

          Sulfuric acid tests on current catalyst cars run this past year
agree well with previous tests.  Oxidation catalyst cars without air
injection usually emit under 10 mg/mi of sulfuric acid.  Current oxidation
catalyst cars with air injection emit greater quantities of sulfuric
acid, generally 20 mg/mi or higher.  Tests of advanced oxidation catalyst
vehicle prototypes (which generally have air injection) show similar
sulfuric acid emission levels.  Prototype 3-way catalyst cars with a
down-stream oxidation catalyst and air injection show sulfuric acid
emission levels similar to conventional oxidation catalyst cars equipped
with air injection.

          One major area needing additional investigation is the effect of
mileage on sulfuric acid emissions.  Limited test results so far show
high mileage oxidation catalyst vehicles have lower sulfuric acid emissions,
Sulfuric acid emissions are reduced by a factor of two and, in some
cases, even more.  Some results have even approached the level of non-
catalyst cars (1 mg/mi).  Significant data are still needed to determine
whether this reduction with extended mileage is such that sulfuric acid
emissions are not a problem.  The car companies are running tests on
catalyst vehicles as they accumulate mileage to determine the extent of
this reduction.

          Finally, during the past year, GM completed a major test track
experiment measuring roadside sulfuric acid levels.  This work showed
roadside sulfuric acid levels far below levels predicted by EPA air
quality models. These lower levels, combined with a possible reduction
in sulfuric acid emissions at higher mileages, may result in sulfuric
acid emissions being considered much less of an air quality problem than
was previously the case.

III. HYDROGEN CYANIDE

     A.   Introduction and EPA Work

          In 1974, it was reported by Bell Laboratories that automotive
catalysts could, under very unusual conditions in a laboratory reactor,
form hydrogen cyanide (HCN) from laboratory gases.  EPA found in early
1976 that vehicles using a 3-way catalyst could produce HCN under rich
malfunction modes.  Since Volvo and Saab were certifying 3-way catalyst
systems for use in California in 1977, EPA had to make a rapid determina-
tion on the levels of HCN that would be acceptable from a health stand-
point.  EPA was concerned primarily about the worst case situation, I.e.,
                                      245

-------
localized levels of HCN that could occur in indoor parking garages and
along heavily traveled roadways.  The highest HCN emissions were observed
under rich malfunction modes when CO emissions were also highest
(approaching levels found in some uncontrolled pre-1968 vehicles).

          EPA considered two types of worst case situations in determining
acceptable levels for HCN exposure in localized areas.  The first was
a closed environment situation (i.e., garage-type exposure) while the
second was a highway-type exposure situation.  In the closed environment
situation, an upper bound of 5 ppm HCN could result from a maximum raw
exhaust level of 10 ppm.  The corresponding CO level under these conditions
would be 6500 ppm.  The adverse health effects of CO would overshadow
possible adverse effects of HCN by more than two orders of magnitude.
For the highway type exposure, a 150 mg/mi HCN emission rate would
result in a 1.1 ppm HCN level with 17,000 vehicles/hour along a heavily
traveled freeway, under the worst meteorological situation, with each
car emitting the maximum amount of HCN.  EPA-ORD states that 1 ppm HCN
exposure would not have unacceptable health effects.

          It was also thought that a primary factor affecting HCN formation
was the rhodium content of the catalyst.  The Engelhard catalysts planned
for use by Volvo contained about 16% rhodium.  Most 3-way catalysts
contain rhodium, sometimes at lower levels (e.g., 7%).  Many oxidation
catalysts also contain small amounts of rhodium (i.e., 7% or less).  Some
1975 Ford oxidation catalysts (those made by Matthey Bishop) and oxidation
catalysts used by many of the foreign manufacturers contain rhodium.

          EPA therefore required that all 1977 certification cars equipped with
rhodium catalysts be tested by EPA-ORD for HCN emissions.   All such cars
tested to date have been below the HCN levels that had been established
as safe with a large margin of safety on a preliminary basis by EPA.

          Typical data found for a Volvo equipped with a 3-way catalyst by
EPA-ORD are given in Table 9.

          Another Volvo 3-way catalyst system (Car 7712) was tested by
Exxon Research and Engineering under both normal and malfunction conditions.
The rich malfunction mode was induced by disconnecting the oxygen sensor.
Instead of the pyridine-pyrazolone wet chemistry method used by EPA for
HCN analysis, Exxon used a selective ion electrode method.  The results of
the Exxon tests are shown in Table 10.
                                    246

-------
                               TABLE  9
                    EPA-ORD Test  Results on Volvo
Test
Idle*
FTP
CFDS
HFET
30 mph
40 mph
50 mph
HC
gm/mi
0.32
1.66
1.13
0.88
0.84
0.80
0.94
CO
gm/mi
4.42
40.5
28.2
22.4
24.5
20.3
23.9
(Rich Malfunction;
NOx
gm/mi
0.062
0.503
0.314
0.383
0.032
0.07
0.21
HCN
mg/mi
1.70
66.3
110.6
78.8
16.4
47.1
113.8
NH3
mg/mi
7.9
493
627
735
110
176
270
H?S
mg/mt
0.47
6.88
10.1
7.59
NA
NA
NA
COS
mg/mi
0.03
1.89
2.11
1.72
1.32
0.71
1.16
*g/minute
                                       247

-------
                                    TABLE 10
                          Exxon  Tests  Results  on Volvo
HC
Test gm/mi
NORMAL OPERATING CONDITIONS:
Idle* 0
FTP 0.21
CFDS 0.08
FET 0.05
30 mph 0.06
40 mph 0.03
50 mph 0.06
RICH MALFUNCTION MODE:
Idle* 2.97
FTP 1.08
CFDS 0.77
FET 0.68
30 mph 0.76
40 mph 0.90
50 mph 0.85
CO
gin/mi
0
2.70
0.55
0.31
0.88
0.08
0.45

29.12
24.97
17.19
14.36
19.27
15.97
24.55
NOx
gm/mi
0.45
0.32
0.11
0.08
0.02
0.16
0.08

0.43
0.74
0.63
1.14
0.14
0.50
0.68
                                                             HCN
                                                            mg/mt
                                                            0.4
                                                            2.3
                                                            1.9
                                                            2.6
                                                            1.0
                                                            0.6
                                                            0.6
                                                            0.4
                                                           22.
                                                           21
                                                           19.
                                                           29.0
                                                           62.0
                                                          146.5
.5
,7
,3
              NH3
             mg/mt
              5.8
             36
             29
             28
             10
             10
    ,5
    .1
    ,3
    ,6
    ,6
             43.8
  9.4
327.8
493.9
425.4
187.
481
               ,1
               .4
            749.0
*g/minute
                                             248

-------
          It can be seen from  these data that HCN emissions are negligible
under normal operating conditions.  However, HCN emissions definitely
increase under rich malfunction  conditions to levels as high as 110
mg/mi.

          Rhodium-containing oxidation catalyst also has the potential
of producing HCN under rich malfunction conditions.  Since oxidation
catalysts generally contain less rhodium than some 3-way catalysts, they
would be expected to produce far less HCN.  EPA-ORD tests on three VW
1977 certification cars show this to be true.  The test results for one
of these cars (an Audi 100) are given in Table 11.

          Test results for two other VW certification cars (an Audi 80,
Car 8389 and a VW Rabbit, Car  6846) were similar.  The highest HCN
emissions observed for these three cars was 7 mg/mi for the VW Rabbit
over the FTP under the rich malfunction mode.

     B.   GM Work

          In the 1975 Status Report, GM discussed tests they ran on non-
catalyst cars and 1975 vehicles  for HCN emissions.  The non-catalyst
cars emitted 11-14 mg/mi which should be considered the baseline with
which all of the catalyst cars should be compared.  The 1975 oxidation
catalyst production cars emitted about 2 mg/mi.

          In the 1975-76 work  discussed in the most recent report, GM
tested several additional cars for HCN emissions using a modified colorimetric
method.  The first cars tested were two 1975 GM production cars (Cars
R5454 and R5452) designed to meet California standards with the standard
Pt-Pd pelleted catalyst.  The  first car was tested with the air pump
disconnected to enrich the exhaust.  CO emissions were lower (5.1 gm/mi)
with the air pump disconnected than when it was connected (8.9 gm/mi).
Under either condition, HCN emissions were very low at 2 mg/mi.  The
second car was run with both the standard and a specially designed rich
operating carburetor.  HCN emissions were very low under both conditions
being 2 and 0 mg/mi respectively, while CO was 3.1 and 96.1 gm/mi.
While GM does not specify test conditions, it is assumed these cars were
tested on the FTP.

          GM also tested two experimental cars for HCN emissions (Cars
CH42216 and ES66344).   The first vehicle contained an experimental 3-
way catalyst (HN-2217) and a closed-loop carburetor with an oxygen
sensor.  This car had FTP emissions of 3.1 HC, 85.5 CO, 0.7 NO .
However, HCN emissions were only 4 mg/mi.  Presumably, this veRicle was
operating under a malfunction  condition giving the high CO emissions.
                                      249

-------
                               TABLE 11

             EPA-ORD Test Results on VW Certification Cars

Vehicle
Audi 100
(Car 5202,
certifica-
tion tune)



Audi 100
(Car 5202,
rich mal-
function,
1.6% CO)

Test
FTP
CFDS
FET
Idle
(925 rpm)*
Idle
(2000 rpm)*
FTP
CFDS
FET
Idle
(925 rpm)*
HC
gm/mi
0.18
0.13
0.11
0.01

0.12

0.36
0.22
0.12
0.11

CO
gm/mi
1.94
0.72
0.62
0.12

0.03

9.92
3.83
1.44
4.28

NOx
gm/mi
1.24
1.40
1.40
0.02

0.05

1.15
1.05
1.16
0.02

HCN
mg/mi
3.03
1.40
0.97
0

0

5.65
3.37
1.09
1.08

''g/minute
               Idle           0.02      0.11      0.05      0.08
                (2000 rpm)1
                                      250

-------
          The second car was  a  Vega with electronic fuel  injection, an
oxygen sensor, and Degussa experimental 3-way catalyst  (HN-3032).  This
vehicle was tested under both normal  operating conditions and rich
malfunction conditions.  The  rich malfunction was  induced by enriching
the car electronically in the open loop mode.  The highest HCN value
found in these tests was 18 mg/mi which is only slightly higher than HCN
values found with non-catalyst  cars during previous tests.  The car was
then retested in the closed-loop mode using an HN-3032  catalyst that had
been aged 50,000 miles.  HCN  emissions were 9 mg/mi which is slightly
higher than the fresh catalyst.  The  car was not tested in a malfunction
mode with the aged catalyst.

          The test results for  this car are given  in Table 12 for the
FTP.

          GM did not specify  the rhodium content of the 3-way catalysts
tested. However, HCN emissions  from these catalysts are much lower than
those found for the Volvo 3-way catalyst systems.

     C.   Ford Work

          Ford has done more  extensive work on HCN emissions than any
other automobile company to date.

          The initial Ford work involved lengthy validation of both the
pyridine-pyrazolone and the ion selective electrode method for HCN.
Both of these methods can be  used to  analyze HCN trapped in a potassium
hydroxide solution contained  in a bubbler.  Ford bubbles a small sample
from their dilution tunnel into the bubbler.  The first test done by
Ford showed the injection of  6.74 ppm HCN, approximately 90 mg/mi,
resulted in 93.8% HCN recovery.  The  tests were repeated with lower
levels of HCN injected.  Quantities of HCN corresponding to 3.14 ppm or
approximately 45 mg/mi resulted in poor recovery, the reasons for which
are unknown at the present time.  Ford then added known quantities of
HCN to the dilution tunnel stream when a vehicle was being tested at a
45 mph cruise condition.  These tests showed an HCN recovery rate of
only 50% indicating an interference problem with both methods.  Ford
work identified the problem to  probably be a low concentration of potassium
hydroxide in the bubblers (caused by  reaction of acidic exhaust components)
which allowed HCN to escape.  Modifying the method by using more hydroxide
resulted in a recovery of 94% of the  HCN.  Ford work shows the limit of
detection of both methods is  about 0.06 mg/mi of HCN.  While more work is
needed to refine the method,  the method as it stands now can be used for
reasonable approximation of automotive HCN emissions.

          Ford vehicle tests  were conducted on a Granada equipped with
a 302 CID engine and a 3-way  catalyst followed by an oxidation catalyst
(Vehicle 9P), a Pinto equipped  with a 2.3 litre engine with a similar
system, and a 1978 prototype  Pinto with a 3-way catalyst.
                                     251

-------
                               TABLE 12


                HCN Test Results of 3-Hay Catalyst Vega
                          HC        CO        NOx        HCN        NH3
Test Condition           gm/mi     gm/mi     gin/ml     mg/mi      nig/ml

Fresh Catalyst --
     closed loop         0.29       2.8      0.23        2          2
open loop -
6% rich
12% rich
18% rich

0.30
0.83
1.75

3.4
19.7
60.0

0.86
0.13
0.11

2
16
11

3
211
259
50,000-Mile Catalyst —
     closed loop         0.63       5.36     0.60        9          8
                                    252

-------
          Tests on the Granada were conducted with the air injection
functioning normally and with the air injection after the 3-way catalyst
disconnected simulating a rich malfunction condition.  This car was tested
over the CFDS, idle, and 45 mph steady state conditions.  With the air
pump disconnected, HCN emissions were 12.3 mg/mi compared to about 0.15
mg/mi with the air pump functioning.  CO emissions over the CFDS were
about 16 gm/mi with the air pump disconnected and only 2 gm/mi with the
air pump functioning.

          The Pinto (Vehicle T791) with the 3-way plus oxidation
catalyst system was tested under the same conditions as Vehicle 9P.  The
HCN emissions were under 5 mg/mi for all tests.  CO emissions were
extremely low both with the air pump disconnected (under 0.1 gm/mi) and
with it functioning correctly (also under 0.1 gm/mi).  It is possible
that the low CO on Vehicles 9P and T791 are partly responsible for the
low HCN formation.

          The 1978 3-way catalyst prototype (Vehicle 61E28) was tested
with a line to the carburetor being plugged to deactivate the oxygen
sensor.  This resulted in a 12:1 air-fuel ratio.  The vehicle was also
tested in a 13:1 air-fuel ratio configuration.  The car was tested under
under idle (slow and fast), 50 mph, the CFDS, a hot start FTP, and the FET.
HCN emissions were in all cases under 2 mg/mi.  However, CO emissions
were considerably higher than usual (96, 113, and 88 gm/mi over the CFDS,
hot start FTP, and FET) giving ample opportunity for HCN formation.

          Finally, Ford has done some laboratory studies with catalyst
samples in their pulse flame reactor and also with simulated exhaust
gas samples blended from gas cylinders.

          The simulated exhaust gas experiments showed that higher NO
and CO concentrations caused higher HCN formation as expected.  Maximum
HCN from a Pt-Pd PTX catalyst and a pure platinum catalyst is formed
at 550°C.  The temperature for maximum HCN formation is slightly lower
for a rhodium-containing catalyst.  Space velocity (from 20,000 to
200,000 hr  ) did not influence HCN formation at 550°C.  It should be
noted that the simulated exhaust contained NO, CO, and hydrogen but no
water vapor (a component normally found in vehicle exhaust).

          However, the pulse flame reactor tests show that roughly
the same amount of HCN is formed from 550° to 800°C for all of the
catalysts tested (except one containing iridium).  Additional tests
showed that lower space velocities decrease HCN formation.  The pulse
flame apparatus results may be different from the laboratory test
apparatus results because the gas used in the former experiments contained
water vapor.
                                      253

-------
     D.   Chrysler Work

          The initial Chrysler work on HCN consisted of running engine
dynamometer tests with a variety of catalysts.  Catalysts tested included
an Engelhard Pt/Rh and three Chrysler catalysts (Pt, ruthenium-perovskite,
and Pt/Rh).  These tests were run with 1.5% to 5% CO (mostly 5%).  Even
though the CO levels were high, less than 1 ppm of HCN was formed.
Chrysler even ran an engine dynamometer test with the catalyst at 1300°F
where they thought HCN formation would be a maximum.  Again, almost no
HCN was formed.

          Chrysler then decided to run some basic laboratory tests
with simulated exhaust gases.  Using a gas containing 3% CO, 1% hydrogen,
and 0.2% NO in nitrogen, the following results were obtained with
different catalysts.

          Further tests with the platinum-rhodium catalyst shows that
doubling the CO input also doubles the HCN emissions.  Doubling the NO
input also increases the HCN emissions but by less than a factor of two.

          Chrysler did some further tube furnace laboratory studies
involving the platinum-rhodium catalyst in one furnace and a platinum
catalysts in a second furnace (in series).  Such an arrangement approximates
a dual catalyst system (a reduction catalyst followed by an oxidation
catalyst).  On a dual catalyst system, an air pump provides air injection
after the reduction catalyst but before the oxidation catalyst.  A vehicle
air pump failure can be easily simulated in the tube furnace by not adding
air between the two furnaces.  Chrysler did this and varied the temperature
of the oxidizing catalyst.  The results shown in Table 14 were obtained
using two different feedgases.  These results show the greatest HCN emissions
at about 1300°F.

          Finally, Chrysler reports some work done on analytical methods.
Chrysler uses the Liebig titration, the pyridine-pyrazolone method, and
the selective ion electrode method.  Chrysler states that heavy metals
(nickel, copper, iron, and zinc) and the sulfide anion interfere with the
selective ion electrode.  However,  Chrysler finds no evidence of any of
these interfering species in the liquid impinger sample they collected.

     E.   Conclusions

          Early studies by EPA showed HCN emissions from vehicles with
3-way catalysts operated under malfunction conditions.  These emissions
were always under 150 mg/mi for any driving mode and under 5 ppm for idle.
Almost no HCN emissions were found from vehicles with 3-way catalysts
under normal operating conditions.  EPA has concluded that these levels
of HCN do not represent a health concern.
                                      254

-------
                     TABLE 13

          Chrysler HCN Laboratory Studies

Catalyst                           HCN (ppm)
platinum-rhodium                       72
platinum                              0.9
ruthenium-perovskite                  0.5
rhodium                               160
                          255

-------
                               TABLE 14
                    Chrysler HCN Laboratory Studies
          Feedgas A


Temperature         HCN (pptn)
                                 Feedgas B'
   600° F


   800° F


  1000°F


  1300°F
(oxidizing catalyst

  removed)
 67


 65


 67


114




 65
Temperature
750°F
900° F
1050°F
1200°F
1350°F
1500°F
HCN (ppm)
1
3
9
49
92
88
1
 Feedgas A:  1.9% CO, 0.8% hydrogen, 0.19% NO, rest nitrogen
 >
 •Feedgas B:  0.7% ammonia, 1.9% CO, 0.8% hydrogen, rest nitrogen
                                     256

-------
          The manufacturers  have done extensive work developing analytical
methods to measure HCN emissions.  Also, considerable work was done
measuring HCN on a number of non-catalyst and catalyst cars.  Non-catalyst
cars have been found to emit from  11-14 mg/mi of HCN.  Oxidation
catalyst cars emit far less  HCN than non-catalyst cars even under malfunc-
tion conditions.

          Therefore, currently available data indicate HCN emissions
from catalyst cars is not a  problem.  However, HCN measurements should,
of course, be done on future emission control systems as they are
developed.

IV.  RUTHENIUM EMISSIONS

     A.   Introduction

          Ruthenium is being considered by several companies for use
in catalysts.  Chrysler has  considered it for reduction catalysts, 3-way
catalysts, and start catalysts.  Ford has conducted extensive work with
ruthenium for use in reduction catalysts.  In addition, both Exxon and
Gulf had done work in the past on  ruthenium-containing reduction catalysts
but are not now conducting any further work.  The basic reason ruthenium
is being considered is its low cost and attractiveness for nitrogen
oxide control.  Ford has considered use of ruthenium in the past for
NO  reduction catalysts and  is currently considering ruthenium as a
possible component of 3-way  catalysts.

          The higher oxides  of ruthenium (RuOJ are volatile and easily
formed under oxidizing conditions.  Various companies in previous years
(Ford and Exxon) have tried  to stabilize ruthenium so that it is not lost
from NO  reduction catalysts.  These attempts were unsuccessful due to
high loss of catalytic activity in the process or the loss of ruthenium.
Recently, DuPont has worked  on stabilizing ruthenium in a rare earth
oxide perovskite structure.   This work is still in progress.  Since
ruthenium oxides are very toxic, it seems that use of ruthenium catalysts
depends on no significant emission of their oxides.  EPA is not currently
in a position to quantify what, if any, levels of ruthenium emissions
would be considered significant.   EPA health effects studies are investi-
gating the toxicity of ruthenium and ruthenium oxides.

     B.   Chrysler

          Chrysler is the only company to have reported studies on
ruthenium emissions.  Chrysler ran dynamometer and vehicle tests.  Chrysler
also sent the vehicle they tested  to Exxon for independent tests.
The dynamometer tests showed from  14 to 36 mg/km of ruthenium being
emitted.  At this rate, all  of the ruthenium would be lost in 32,000 miles.
                                       257

-------
          Chrysler has also conducted ruthenium tests on Research Car
232.  Car 232 contains a small (58 cu. in.) ruthenium perovskite monolith
catalyst close to the exhaust manifold.  This catalyst can function as a
reduction catalyst.  A conventional Pt oxidation catalyst follows the
reduction catalyst.  Particulate samples were collected from the dilution
tunnel using a Fluoropore filter which was analyzed for ruthenium by
neutron activation.  This car was run on AMA durability for 25,000 miles.
The Chrysler tests indicated significant ruthenium loss.  After 25,000
miles, the car was sent to Exxon for particulate tests.  These tests
showed again that ruthenium was being emitted at a substantial rate
(about 150 mg/mi).  The Exxon results were about three times higher
than the Chrysler results.  The Chrysler results indicated that all
of the ruthenium on the catalyst would be lost in about 30,000 miles
(if the emission rate were constant).

     C.   Conclusions

          Tests by Chrysler show that ruthenium is in fact lost from
ruthenium-containing catalysts.  While ruthenium is not being used in
any production catalysts currently, it is being considered for future
catalysts so that the health impact of these emissions can be determined.

V.   AMMONIA

     A.   Introduction and EPA Work

          Ammonia emissions have been known to occur for some time now
with dual catalyst systems under rich operating conditions.   With a dual
catalyst system, ammonia formed over the reduction catalyst is frequently
reoxidized to NO over the oxidation catalyst.  Ammonia formation in a
dual catalyst system thus generally results in loss of NO control.  It
is possible to have incomplete oxidation of the ammonia over the oxida-
tion catalyst which could result in significant ammonia emissions.  Also,
an oxidation catalyst system malfunction (e.g., air pump failure) can
result in significant ammonia emissions.

          In addition to possible ammonia formation with dual catalysts,
ammonia can be formed with 3-way catalyst systems.  Early EPA data
show that the Volvo 3-way catalyst can have significant ammonia emissions
under rich malfunction conditions.  Data obtained by EPA-ORD and Exxon
on ammonia emissions was given in detail in some of the tables in the
HCN section.  Essentially, these data showed 20-30 mg/mi ammonia under
cyclic operation conditions with the vehicle correctly tuned and as
high as 300-500 mg/mi under rich operating conditions.
                                       258

-------
          Ford points out  in  their  status  report that ammonia will be
emitted from their 3-way catalyst systems.  Ford has also asked EPA
from time to time on what  level  of  ammonia emissions would be acceptable
from a health standpoint.   EPA  is currently investigating possible health
effects from automotive ammonia emissions  including both entire air
quality regions and localized situations (e.g., indoor parking garages and
heavily traveled freeways).   Currently, there is inadequate information to
permit determination of the level of  automotive ammonia emissions that
would be acceptable.

     B.   Ford Work

          Ford routinely measures ammonia emissions in all of their
3-way catalyst screening work.   A parametric study performed by Ford
on catalysts identical in  all aspects but Pt/Rh ratio shows a linear
decrease of net NO  conversion  (which is inversely related to ammonia
formation) with increasing Pt/Rh ratio.

          Laboratory studies  of various catalysts show increasing ammonia
formation as the A/F ratio becomes  "richer" in the laboratory pulsator
apparatus.  For example, catalyst M268A shows an increase in ammonia
formation from 0% of total  NO  to 20% as the A/F ratio changes from
14.38 to 14.08.  Catalyst  M268B shows a change in ammonia formation of
only 0% to 2% under these  conditions.  Catalyst M275D2 produces from
about 0% to 40% ammonia as the  A/F  ratio is changed from 14.48 to 14.08.
These tests were generally run  at 550°C.

          Ford has also done  some ammonia measurements of 3-way catalysts
being durability tested.   Testing was done with isooctane containing
0.007 gm/gal lead, 0.0008  gm/gal phosphorus, and 0.02% sulfur.  As the
catalyst ages, the A/F ratio  units  for optimum NO, CO, and HC conversion
generally shifts about 0.1  A/F  units  to the rich side.  Ammonia formation
was noted to decrease significantly in some of the catalysts (M-270 and
M-273), thereby aiding net NOV  conversion.
                              X
          Ford also reported  some engine dynamometer work on ammonia
formation with catalysts containing differing amounts of rhodium.  The
results were those obtained in  the  laboratory apparatus mentioned
earlier.  These tests showed  the ammonia formation was found to be inversely
proportional to the rhodium content of the catalyst.  Perhaps sulfur
poisoning of the aged catalyst  results in different ammonia formation
characteristics than a fresh  catalyst.
                                     259

-------
          Finally, Ford reported ammonia formation characteristics
of an experimental 3-way catalyst containing 3% TiCL (titanium oxide).
The Ti02 apparently affects oxygen storage capacity and the rates of
oxygen transfer.  The ammonia formation from this catalyst was very high,
26% at a point 2% rich of stoichiometric.

     C.   GM Work

          Last year GM reported that ammonia emissions from four catalyst
prototype vehicles are 1-3 mg/mi over the FTP, as shown in Table 15.

          GM work on 1975 production cars without air injection under
rich malfunction conditions show ammonia emissions of 43-165 mg/mi.   In
this year's status report, GM reports ammonia emissions on a 1975 Vega
with a 3-way catalyst.  Since this car was also tested for HCN emissions,
the ammonia test results were given previously in the HCN section.  This
car showed 2-8 mg/mi of ammonia under normal operating conditions, 3 mg/mi
when operating 6% rich, and 211-259 mg/mi when operating 12-18% rich.

          GM also tested four 1975 catalyst cars and one 1974
non-catalyst car for organic amines over the FTP.  Amine compounds
are very similar to ammonia.  No amine emissions were found using
the wet chemistry iodiometric analysis method which has a limit of
detection of 2 mg/mi.

     D.   Conclusions

          Ammonia emissions may be significant from vehicles with 3-way
catalysts but only under malfunction conditions.  Ammonia emissions
should be measured from vehicles with 3-way catalysts as they are being
developed.

          Ammonia emissions from vehicles equipped with oxidation
catalysts do not appear to be significant.

VI.  POLYNUCLEAR AROMATIC COMPOUNDS (PNA)

     A.   Introduction

          Polynuclear aromatic compounds (PNAs) are of interest due  to
their cardnogenicity.  The PNAs are a group of multi-ring aromatic
hydrocarbon compounds with very low volatility and high molecular
weight.  About 90% of total PNAs in the atmosphere come from stationary
sources (in particular from coal burning).  Only about 10% of the total
PNAs come from mobile sources.  Even though little PNA work has been done
on diesels, it is estimated that about half of this 10% comes from diesels.
The other half of the 10% comes from other motor vehicles.
                                    260

-------
                               TABLE 15
                            6M Ammonia Data
Car
Catalyst
Ammonia (mg/nri)
0-39680
0-38608
E564346

C4361
HN 2236
HN 2236, AIR
HN 2221 and 2364 (dual catalyst,
  reduction and oxidation)
HN 2217 and 3059 (3-way and
  oxidation catalyst, closed
  loop fuel injection)
     3
     2
     1
                                    261

-------
          Sometimes benzo(a)pyrene (BaP) is measured by itself instead
of measuring total PNAs (which consist of about 20 compounds).  The
amount of BaP found can generally be correlated to total PNAs.  BaP
is relatively easy to measure and is more carcinogenic than most other
PNAs.

          PNA compounds in automotive exhaust have been studied for some
time.  Early contract work through CRC-APRAC from 1968-72 thoroughly
characterized PNAs.  This work was done at Exxon and involved testing
several cars.  It was found that post-1968 cars had about half the PNA
emissions of pre-1968 cars.  The control of HC and CO also resulted
in decreases of PNA.  Exxon did some work on prototypes equipped with
advanced catalysts and thermal reactors.  Both systems were found
to greatly reduce PNAs (generally by 95% or more compared to uncontrolled
vehicles.  Subsequent work by UOP showed that catalysts preferentially
oxidize PNA compounds resulting in very low PNA emissions from vehicles
with catalysts.

          Very little work has been done on measuring PNAs from light-
duty diesels.  However, work at Southwest shows heavy-duty diesels emit
much greater quantities of PNAs than do gasoline engines.  It is important
that PNA work be done on light-duty diesels.

     B.   VW Work

          The only PNA data reported in the submissions came from VW.

          Volkswagen (VW) presented an interesting summary and
discussion of their testing concerning the levels of PNA from different
types of emission control technologies, alternative engines, and
alternative fuels.  VW did not provide complete details concerning
the test procedures utilized in VW's testing for PNAs.   Also missing
from the VW discussion was any mention of PNA emissions from diesel
engines.

          VW reported that their studies showed extremely low levels
of PNA concentrations in automobile exhaust emissions.   These levels
are reported to be approximately 0.001 gm/km (0.0016 gm/mi) on the
European Emission Test Procedure.  This level is higher than found
by other investigations.

          In addition to utilizing the European Test Procedure, the
data were collected under the following conditions:


          (1)  Europe Fuel ERF-G1 was used in all procedures.  This
               type of fuel has a composition of 55% paraffins, 8%
               olefins, and 37% aromatics.  Its lead content is 0.71
               gm/litre (2.68 gm/gal).  Lead-free fuel  (VK 91) was
               used with noble metal catalyst-equipped engines.  The
               lead-free fuel contains 1% more aromatics.
                                   262

-------
          (2)  The motor oil was a standard brand heavy duty oil.
               The oil was changed before every PNA test to avoid
               PNA enrichment of engine oil.

          (3)  The test vehicles had accumulated between 2,000 and
               15,000 miles of city driving at the start of the test.


          Volkswagen presented the data in Figure 1 which illustrates
the distribution by percent of PNAs in automobile exhaust emissions
produced by a variety of engine design concepts.  The column labeled
"Carcinogenic properties" is VW's estimate of the relative severity or
danger of producing cancerous effects in humans.  This opinion does not
reflect EPA's position indicating which PNAs are more carcinogenic than
others.  It should also be noted that the PNA listed in Figure 1 are
less stable as they proceeded down the list and hence easier to break
down with exhaust aftertreatment.

          Volkswagen drew the following conclusions concerning PNA from
vehicles equipped with gasoline-fueled engines.


          (1)  There is a certain and insignificantly varying pattern
               of distribution of the measured ten individual PNA
               components from fluoranthene through coronene.  The
               lower four-ring PNAs represent approximately 80% of
               this pattern.  VW mentioned that among the ten measured
               PNA components there are three each which combine two
               or three individual PNA components.  There is one
               triplet each in the chrysene and in the benzofluoranthenes,
               and one twin in the idenopyrene.

          (2)  The present production vehicles produce the following
               scatter on European Test Procedure - PNA total emissions
               of between 2000 ygm/test and 10,000 ugm/test, and
               between 50 ygm/test and 215 vgm/test in the benzo(a)pyrene
               emissions.

          (3)  The following percent reduction of PNA emissions are
               obtained with other concepts when compared to gasoline-
               powered vehicles.
                    production vehicles           =    H

                    catalytic design concepts     =      5%

                    thermal  reactor concepts      =      15%

                    methanol vehicles             =      10%
                                    263

-------



PNA


FLUORANTHENE
PYRENE
CHRYSENE + M226 +
BENZO(A)ANTHRACENE
BENZOFLUORANTHENE
BENZO(E)PYRENE
BENZOIAIPYRENE
PERYLENE
INDENOPYRENE +
DIBENZIAH {ANTHRACENE
BENZO(G,H,I)PERYLENE
CORONENE
CARCINO-
GENIC
PROPERTIES

+ SLIGHT
++ MEDIUM
+++ STRONG
0
0
+
++
+
•M-+
0
+
+++
0
0


PRODUC-
TION
VEHICLES
VEHICLE

CATALYTIC
CONCEPTS

CONCEPTS

THERMAL
REACTOR
CONCEPTS


METHANOL
VEHICLES

% OF TOTAL PNA
24
36
20
3
2.5
2.5
0.5
2

-------
          (4)  The distribution pattern of vehicle concepts other
               than production vehicles show limited but characteris-
               tically shifted PNA patterns.  The four-ring PNAs are
               lower in catalytic concepts than in the thermal reactor
               concepts while the six and seven-ring benzo(ghi)perylene
               and coronene components show a distinct increase in the
               catalytic concepts.  This indicates that these after-
               burner concepts reduce the latter components less when
               compared with the lower PNAs.

               On the other hand, engines operated on methanol show an
               extremely strong presence of 80% of the lower four-ring
               fluoranthene and pyrene components.  This means that
               the absence of the PNAs and of the other aromatic
               components in methanol fuel not only produces a PNA
               emission that has been reduced by one order of magni-
               tude but that higher PNA synthesizing is impeded.

          (5)  The investigation shows the total of the more or less
               carcinogenic individual components (as far as such
               carcinogenic effects are known from various (sources)
               to amount to approximately 20% or 30% per volume of
               total PNA emission.

          VW produced another graph (Figure 2) which shows the results
obtained with an engine using propane and gasoline.  VW notes that propane
operation reduces the PNA content to 10% (similar to methanol operation)
when compared with gasoline operation.

          VW feels this fact permits an interesting conclusion in
regard to the influence of engine oil, i.e., that engine oil influence
on PNA emissions in gasoline operation is bound to be lower than 10%.
VW stated that it is of additional interest that identical hydrocarbon
emissions were measured in the European Test Procedure under both modes
of operation.  They feel this confirms the experience found elsewhere
in a great many areas that the level of hydrocarbon emissions can not be
used to draw conclusions about PNA emissions.

          Another intestesting graph (Figure 3) from VW presents the
PNA emissions compared to CO emissions.  This graph shows the results
obtained from different gasoline engine-equipped vehicles.  Each
point represents one test vehicle.  The graph allows for the following
VW findings:
                                     265

-------
ho
          TEST
               1400 —
               1200 —
               1000 —
              < 800

              a.
              _J

              5
              I- 600
                400-
                200-
                        914

                                  1328
                                  yys
                                  rSS*
         616
        ii
                                                         70
                                                                    55
                                                         GASOLINE OPERATION


                                                         PROPANE OPERATION





                                                         AIR-COOLED 1.6 LITER ENGINE
                                                                                                    240
                                                                                          64
                                                                                                               191
                          FLT
PYR        CHY


       + M 226
BFLT        BEP
PER        IND


      +DahATC
BghIP       COR
                                                                                                                           67
                                                                   BAP
                                             Figure 2.  Total PNA emission, gasoline versus propane operation.

-------
                                                 GASOLINE PROCESS, HYDROCARBON OPERATION
                               * WANKEL WITH THERMAL REACTOR
                               D THERMAL REACTOR CONCEPTS
                               A CATALYST CONCEPTS
                               O WATER-COOLED
ON
                           0     20]    40     60    80
                         **CATALYSTr«— THERMAL REACTOR
                            RANGE           RANGE        *•*
140    160    180    200   220
   PRODUCTION VEHICLE RANGE
	WITHOUT EMISSION CONTROL -
240
      260    280  CO
       9/TEST
                                       Figure 3.  PNA emission as function of carbon monoxide emission, European test.

-------
          •    All measuring points are located in a band that
               spreads with the increase in CO emissions, and the
               upper limit of which is set by the ascending slope
               of approximately 5000 pgm PNA/100 gm CO and the bottom
               limit by 1000 ygm PNA/100 gm CO.

          •    Water-cooled engine-equipped production vehicles produce
               a much higher PNA emission than do air-cooled engine-
               equipped production vehicles.

          t    The distinguishing characteristics between water-cooled
               and air-cooled systems are further affected by additional
               engine characteristics.  Nevertheless, ft does not appear
               to be unrealistic to expect the markedly cooler combustion
               chamber walls in water-cooled engines to tend to produce
               higher PNA emissions because these close-to-the-wall
               zones with their inherent incomplete combustion due to
               wall-quenching, must be considered highly responsible
               for the production of PNA emissions.  These quench zones
               also tend to be greater in water-cooled engines.

     C.   Conclusions

          The VW work confirms earlier work done by Exxon and UOP which
shows about 5% of the amount of PNA from non-catalyst cars is found in
catalyst cars.  VW also showed that use of methanol as an alternate fuel
reduces PNA emissions and causes different types of PNAs to be emitted.

          An important area not addressed by any of the manufacturers
is PNA emissions from light-duty diesels.  With the planned introduction
of the VW and Oldsmobile diesels, this area is important.

VII. DIESEL PARTICIPATE EMISSIONS

     A.   Introduction and Background

          Mobile sources contribute a small but still  significant part
of the total suspended particulates in ambient air.  For example, prior
to the introduction of unleaded fuel, automobiles using leaded fuels
contributed between 2 and 13% of the total suspended particulates in
urban areas.  With about half of the air quality control regions around
the country over the total suspended particulate air quality standard
of 75 yg/rtH, automotive particulate emissions are important to the attain-
ment of the total suspended particulate standard.

          With vehicles burning fuel with maximum lead concentrations
(containing about 2.5 gm lead/gal), the resultant particulate emissions
are 0.25 gm/mi.   On the other hand, a non-catalyst vehicle burning
                                       268

-------
unleaded fuel emits only 0.01 gm/mi of particulate.  Catalyst-equipped
vehicles emit much less particulate vehicles burning unleaded fuel.  An
approximate particulate emission rate for these catalyst cars is 0.02
gm/mi and is based on testing of over 100 cars.  Parttculates from non-
catalyst vehicles burning unleaded fuel consist mostly of high molecular
weight organic compounds.  Particulates from catalyst cars burning
unleaded fuel consist primarily of sulfuric acid and its water of
hydration.

          As the use of unleaded fuels in catalyst-equipped vehicles
displaces the use of leaded fuels (which emit 0.25 gm/mi particulates),
the automotive contribution to urban air total suspended particulates
will decrease from 2-13% to about 0.2-1.3%.

          The introduction of large numbers of light-duty diesels may
reverse this trend and increase the automotive contribution to total
suspended particulates.  Therefore, it is important to investigate
the quantity of diesel particulates emitted.

     B.   Composition of Diesel Particulates

          The composition of light-duty diesel particulates has  not been
completely determined.  It is known that a major constituent of  these
particulates is elemental carbon with the ability to adsorb other
compounds on it much like activated charcoal.  The types of compounds
that have been found adsorbed are S02, sulfate, high molecular weight
organic compounds, and polynuclear aromatic hydrocarbons.

          The quantity of hydrocarbons including polynuclear aromatics
(PNA) present in diesel particulates is currently being investigated
by EPA in contracts with Southwest Research Institute (SwRI).  A
problem impeding progress in this area is that the sampling and
collection procedures are in a rudimentary stage of development, parti-
cularly for PNA.  Some measurements are currently being conducted for
one PNA species, benzo(a)pyrene, in light-duty diesels under the EPA
contract work with SwRI.  Also, PNA emissions were investigated  to some
extent by a previous CRC contract with Gulf Research.

          The amount of sulfate in diesel particulate has been
determined accurately by EPA work (both in-house and contract work)
and found to represent only a small  fraction (about 1-2%) of the sulfur
in diesel fuel.  While this conversion to sulfate is about the same
for the diesel as for non-catalyst gasoline cars, diesel  sulfate emissions
are greater than those from non-catalyst gasoline cars (0.1 vs.  0.01
gm/mi) due to the higher sulfur content of diesel fuel  compared  to
gasoline (about 0.23% vs. 0.03%).  Still, sulfate emissions account for
a very small part of the total particulates from diesels.
                                       269

-------
          There is some overlap between diesel parti oil ate and hydro-
carbon measurements.  Some of the compounds measured as gaseous hydro-
carbons by the heated FID (at 375°) used in the Federal Test Procedure
are also collected as particulates since the dflutton tunnel filter is
about 120°F.  The amount of overlap has not been quantified accurately,
but some fraction of the particulate is being measured and subject to
the HC emissions standards.

          Diesel particulates are very small in size with most of the
diesel particulates below 1 micron in size.  Attempts to determine the
size distribution by impactors have not been successful since impactors
separate particulates from 1 to 10 microns in size.  This information
on particulate sizes is important to the determination of how far they
penetrate into the respiratory system.

          The health effects associated with diesel particulates have not
been determined.  EPA has a contract with SwRI to do some bloassay tests
on diesel particulates and is also doing some in-house animal tests in
Cincinnati.

     C.   Quantity of Diesel Particulates

          Almost all of the work done to obtain emission factors on light-
duty diesels has been done by EPA either through contract or in-house
testing.  Little data have been provided by the manufacturers.

          These programs have employed the standard dilution tunnel
approach to measurement.  Tests completed to date have included the
vehicles shown in Table 16.  The results are shown in the table.

          The Oldsmobile prototype tested by EPA is similar to but not
identical to the version being considered by GM for production in 1978.
EPA contract work at SwRI involved testing a number of diesel vehicles.

          Particulate emissions rates are given for several cycles including
the FTP, CFDS, and FET.  Of special interest are the FTP results, since
this driving schedule represents driving patterns in an air quality
control region as a whole.  Since gaseous emissions over the FTP are
used in air quality models and for emission factors in the EPA
booklet, AP-42, particulate emissions factors over the FTP can be used
likewise.  The CFDS represents congested freeway driving while the
FET represents rural-highway-type driving.
                                   270

-------
                                                             Table   16
                                            Partlculate Emission Rates of Light-Duty Diesel Vehicles



y

.
..
.
.
,
a.
*
a.
„
*
0.
1.
2.
3.
4.
5.
6.


Vehicle/Model

Mercedes 220D
Mercedes 240D
Mercedes 300D
Peugeot 204D
Perkins 6-247
VW Diesel*
VW Gasoline
Oldsmobile Diesel
Oldsmobile Gasoline*
Pinto Diesel
Postal Van Diesel
VW Rabbit Diesel
Chrysler Diesel
VW Rabbit Diesel
Mercedes 300D
Oldsmobile Diesel
Nissan Diesel
Peugeot 504
Engine
Disp.
(CID)

134
146
183
83
247
90

350
260

165
90
200
90
183
350
122
129
Inertia
wt
(Ibs)

3500
3500
4000
2500
4500
2250

4500
4500
2750
3000
2250
4500
2250

4500
3500
3000
FTP Gaseous Emissions
Test
Lab

SwRI
SwRI
SwRI
SwRI
SwRI
SwRI
SwRI
SwRI
SwRI
AA
AA
AA
AA
AA
AA
AA
RTP
RTF
HC
g/mtle

0.18
0.29
0.16
1.11
0.72
0.37
0.23
0.76
0.39
0.24
0.14
0.11
0.26
0.23
0,16
0.52
0.25
0.49
CO
g/mile

1.30
0.97
0.85
1.71
2.87
0.79
3.70
2.00
2.16
1.21
1.47
0.98
1.22
1.11
0.92
1.92
1.10
1.45
NOx
g/mile

1.05
1.27
1.72
0.68
1.50
0.87
1.01
1.13
1.37
0.76
2.54
1.22
1.82
0.93
2.17
1.47
1.37
2.30
Sulfates
mg/mile
SET

9.21*
14.12*
16.53,
8.21
18.26
7.02
1.65
16.62
20.90
8.53
10.14
9.34
12.24
5.47
7.85
8.52

7.0
*Gasoline car included for comparison to diesel counterpart
 Sulfate values for these tests are over the FTP.
 GM requested the diesel fuel economy for the Oldsmobile be kept confidential  at  this  time.
                                                                                                             Particulates g/mile
FTP
0.60
0.48
0.49
0.38
0.81
0.29
0.007
0.92
0.009
0.35
0.47
0.32
0.19
0.31
0.43

0.3
0.51
SET
0.43
0.36
0.37
0.24
0.49
0.26
0.002
0.58
0.016
0.38
0.27
0.23
0.17
0.20
0.27
0.42

.31
HFET
0.38
0.31
0.39
0.30
0.54
0.25
0.003
0.48
0.021
0.28
0.24
0.26
0.17
0.20
0.25
0.39
0.33
0.40
Fuel Eco.
FET
33.5
33.7
30.0
43.8
28.3
53.7
36.1
2
23.0
54.5
37.8
51.0
25.7
51.7
29.5
2
33.8
35.9
(MPG)
FTP
25.9
25.7
23.8
35.9
25.7
42.7
24.6
2
15.5
44.8
29.9
31.6
n /
40.6
23.8
2
26.2
26.4
   ro

-------
          The Mercedes 300D was tested by both SwRI and EPA.  The
participate results obtained by EPA (0.43, 0.27, 0.25 mg/mi over the
FTP, CFDS, and FET) agree well with those obtained by SwRI  (0.49, 0.37,
and 0.39 gm/mi over the FTP, CFDS, and FET).  The gaseous emission
numbers also agree well.  The same Oldsmobile diesel was also tested
at both labs but EPA did not obtain participate mass measurements during
the tests.  The particulate numbers on the VW diesel from SwRI (0.29,
0.26, and 0.25 gm/mi over the FTP, CFDS, and FET) agree well with those
obtained by EPA (0.32, 0.23, and 0.26 gm/mi over the FTP, CFDS, and FET.
The gaseous emission results between the two laboratories on the VW
agree reasonably well, but the HC at EPA (0.11 gm/mi) is lower than that
found at SwRI (0.37 gm/mi).

          The four tests on the three Mercedes vehicles tested gave an
average particulate mass of 0.50 gm/mi over the FTP.  The two Peugeot
vehicles tested gave 0.45 mg/mi of particulate over the FTP.  The Nissan
had particulate emissions of 0.30 gm/mi.  The VW Rabbit diesel also had
particulate emissions of 0.30 gm/mi.  This level of particulate emissions
is still over 40 times the level found from the VW gasoline prototype.
The Oldsmobile diesel emitted a relatively high level of particulates
(0.92 gm/mi over the FTP).  This level is 100 times the level found from
the gasoline counterpart (an Oldsmobile 350 gasoline vehicle).  However,
GM states that this Oldsmobile diesel  is a prototype and may be changed
before final production.

          The two gasoline counterparts (Oldsmobile and VW had somewhat
lower particulate emissions than found for vehicles equipped with catalysts
(generally 0.02 mg/mi).  The 0.02 gm/mi number is based on extensive tests
of over a hundred cars and should be considered a number for general
comparison of gasoline and diesel cars.  For individual models, it is
more appropriate to use numbers from diesel-gasoline counterparts if
available.

          The average particulate emissions from the four tests on the
three Mercedes, the two tests on the VW, the test on the Oldsmobile, the
to Peugeots, and the Nissan are about 0.5 gm/mi over the FTP.  The number
can be used as an approximate diesel particulate emission factor for air
quality models.  It can be updated as more data are available.  The
numbers from the experimental vehicles (Pinto, Postal Van, and Chrysler
diesels) were not included in this average but would not significantly
change it.

     D.   Conclusions

          Tests conducted to date show very high particulate emissions
from light-duty diesels (0.5 gm/mi) which is about 25 times comparable
values for gasoline vehicles equipped with catalysts (about 0.02 gm/mi).

          The particulates consist of elemental carbon as well as various
hydrocarbon compounds and small quantities of sulfate.  EPA does not know
yet what the health effects of these particulates are.
                                    272

-------
                                    TECHNICAL REPORT DATA
                            (Please read Instructions on the reverse before completing)
 . REPORT NO.
 EPA-600/3-78-012
                              2.
 4. TITLE AND SUBTITLE

  THIRD ANNUAL CATALYST RESEARCH PROGRAM REPORT
                                                            3. RECIPIENT'S ACCESSIOWNO.
             6. REPORT DATE

                January 1978
              . PERFORMrfjG ORG
                                                                        RGANIZATION CODE
 7. AUTHOR(S)

   Criteria and Special  Studies Office
             8. PERFORMING ORGANIZATION REPORT NO
 9. PERFORMING ORGANIZATION NAME AND ADDRESS
   Health Effects Research Laboratory
   Office of Research  and Development
   Research Triangle Park, N.C. 27711
             10. PROGRAM ELEMENT NO.

                 1AA601
             11. CONTRACT/GRANT NO.
 12. SPONSORING AGENCY NAME AND ADDRESS
  Health Effects Research Laboratory
  Office of Research and Development
  U.S, Environmental Protection Agency
  Research Triangle Park. N.C. 27711
RTP.NC
13. TYPE OF REPORT AND PERIOD COVERED
 Annual  1/76 -  12/76	
             14. SPONSORING AGENCY CODE
              EPA 600/11
 15. SUPPLEMENTARY NOTES
   Annual Report  to  Congress
 16. ABSTRACT
        This report  constitutes the Third Annual Report of the ORD Catalyst Research
   Program required  by the Administrator as  noted in his testimony before the Senate
   Public Works Committee on November 6, 1973.   It includes all research aspects
   of this broad multi-disciplinary program  including: emissions characterization,
   measurement method development, monitoring,  fuels analysis, toxicology,  biology,
   epidemiology, human studies, and unregulated emissions control options.
   Principal focus is upon catalyst-generated sulfuric acid and noble metal particulate
   emissions.
17.
                                KEY WORDS AND DOCUMENT ANALYSIS
                  DESCRIPTORS
                                              b.lDENTIFIERS/OPEN ENDED TERMS
                             COSATI Field/Group
   catalytic converters
   sulfuric acid
   desulfurization
   catalysts
   sulfates
   sulfur
   health
automotive  emissions
unregulated automotive
 emissions
health effects  (public)
                13 F, B
                06 T
                14 B
 8. DISTRIBUTION STATEMENT
   RELEASE TO PUBLIC
19. SECURITY CLASS (ThisReport)
 UNCLASSIFIED
             21. NO. OF PAGES
                279
                                               20. SECURITY CLASS (Thispage)
                                                UNCLASSIFIED
                                                                         22. PRICE
EPA Form 2220-1 (9-73)
                                             273

-------