EPA/AA/CTAB/PA/81-2
               An Approach for  Determining  Levels of  Concern for
                          Unregulated  Toxic Compounds
                             from Mobile Sources
                                      by


                                Robert  J. Garbe

                                  July, 1981
Technical  Reports  do  not  necessarily  represent   final  EPA   decisions   or
positions.  They  are intended  to  present technical  analysis  of issues  using
data  which are  currently  available.   The  purpose  in  the  release  of  such
reports is to  facilitate  the exchange of  technical  information and  to  inform
the public of  technical developments which may  form  the basis for a  final  EPA
decision, position or regulatory action.
           Control  Technology Assessment and Characterization Branch
                     Emission Control Technology Division
                 Office  of  Mobile  Source Air Pollution Control
                      Office of Air, Noise and Radiation
                     U.S. Environmental Protection Agency
                              2565 Plymouth Road
                             Ann Arbor, Mi. 48105

-------
                                                        EPA/AA/CTAB/PA/81-2
               An Approach  for  Determining Levels of Concern for
                         Unregulated  Toxic Compounds
                             from Mobile Sources
                                      by


                                Robert J. Garbe

                                  July,  1981
Technical  Reports  do  not  necessarily  represent  final  EPA   decisions   or
positions.  They  are intended  to  present technical analysis  of issues  using
data  which are  currently  available.   The  purpose in the  release  of  such
reports is to  facilitate  the exchange of technical  information and to  inform
the public of  technical developments which may  form  the basis for a final  EPA
decision, position or regulatory action.
           Control  Technology Assessment and Characterization Branch
                     Emission Control Technology Division
                 Office  of  Mobile  Source Air Pollution Control
                      Office of Air, Noise and Radiation
                     U.S. Environmental Protection Agency
                              2565 Plymouth Road
                             Ann Arbor, Mi. 48105

-------
Introduction

This report presents an approach for determining  ranges  of  concern for vehicle
emissions of  toxic unregulated pollutants. The  information used   in  deriving
this approach  was generated either in-house  at  EPA or  under EPA  contract  to
Midwest Research  Institute in Kansas City,  Missouri; and/or Southwest  Research
Institute in San Antonio, Texas.

This work represents  an  EPA technical effort designed as  one input that  may
bear on EPA policy with respect to implementation of section 202(a)(4)  of the
Clean Air Act  amendments of 1977.  As such,  this  report  does not represent
EPA policy  at  this time.   This report will  also be of interest  to parties
outside  EPA,   such  as the  automobile manufacturers,  who  are  involved  with
deciding  whether  unregulated  pollutants  from  motor  vehicles   constitute   a
public health  hazard. This  approach,  outlined  in  the  subsequent  sections  of
this report, has  four separate parts, the  last of  which  is a summary  of the
previous  three.   An  example of how  the approach  works   is  presented for  a
sample mobile source pollutant;  sulfuric acid
The range of compounds which are expected to be assessed by  this  approach  are
those  compounds  which   have   non-zero  "thresholds"  (e.g.  not   genotoxic)
associated with  them.    This approach  is  not intended for  the evaluation  of
mutagenic, teratogenic, and/or carcinogenic effects of substances emitted from
mobile sources,  but  may  be used to  evaluate  the non-genotoxic health  aspects
of a substance which may have other harmful effects also.

It  is  important  to  point  out  that  this  approach  is not  intended to  be  a
rigorous  examination  of  all  the   issues  and   variables   surrounding   any
hazardous  pollutant  question.   Such an  effort  would require  an  extensive
program  similar  to   the  procedures  used to  determine and  support  the   NAAQS
process.  The intent of this approach is to identify  and prioritize   hazardous
pollutants that  are  emitted from mobile sources  so  that  research  by EPA  and
the  automobile   manufacturers  can   focus   on  those  emission  products   of
significant concern.  If  one pollutant,  for  example,  is shown to fall  within
a potential problem  range,  then additional work can be initiated  to  determine
in more detail the hazards of that pollutant.

The first  step  in determining  the range of concern  for  a pollutant  involves
compiling a list  of  emission factors £f_ the various motor vehicle categories.
These  categories are  assigned  to  as many  discrete  subsets  of  the  mobile
source population as are necessary  to  characterize  the  current  and future
mobile source fleet.

The second  step  is  to  review the  relevent  health effects  literature  on  the
pollutant of interest.   The basic  review would involve selecting valid health
effects  studies   on  a  particular  unregulated pollutant  which fall  within  a
range of  exposure concentrations  suspected  to be  of  concern to human health.
The lowest value of  this range  would represent the lowest concentration at

-------
                                      -3-


which  negative physiological  effects*  can  be  detected.   The  highest  value
would  represent that  level above which the hazards are  so well defined  as  to
encourage regulation.  A simplified means of approaching the  upper  level  would
be to  use  a Threshold Limit Value  (TLV) as the  basis  from which  to  estimate
the range  of concern.  The TLV  can be adjusted  for  exposure time and  safety
margin and subsequently can be used to estimate the range of concern.

The third step is  to  use  an appropriate set of pollutant  dispersion  scenarios
to convert  emission  factors to concentrations that people may be  exposed  to.
Both worst case and average cases for a number of situations are evaluated.

The fourth  step will combine the  results of_ the previous  three  steps and
determine ranges of concern for unregulated pollutants from mobile  sources.

Background

The Clean  Air Act  (CAA)  was  amended  in August  1977 to  include   section  202
(a)(4) and  206 (a)(3)  dealing  with the emissions of  hazardous  pollutants from
vehicles produced after 1978.  The specific language of the statute  is:

202 (a)(4)

    "(4)(A)    Effective   with   respect  to   vehicles  and   engines
    manufactured after model year  1978, no emission  control  device,
    system,  or element  of  design shall  be  used in  a  new  motor
    vehicle  or new motor  vehicle  engine for  purposes of  complying
    with standards prescribed  under  this subsection  if such  device,
    system,  or element  of design  will cause or  contribute to an
    unreasonable risk to public health, welfare, or safety in its
    operation or function.

    (B)  In determining whether  an unreasonable  risk exists  under
    subparagraph (A),  the  Administrator shall consider, among  other
    factors,  (i) whether  and  to what  extent  the  use  of any  device,
    system,  or element  of  design causes,  increases,  reduces, or
    eliminates  emissions   of  any   unregulated  pollutants;   (ii)
    available methods for reducing or  eliminating any risk to public
    health,  welfare,  or  safety whch may be  associated with  the use
    of  such devices,  systems,  or elements  of design  which may be
    used  to conform  to  standards prescribed  under  this  subsection
    without  causing or contributing  to such unreasonable  risk.  The
    Administrator  shall  include  in  the consideration  required by
    this paragraph all  relevant  information  developed  pursuant to
    section 214."
*It is recognized  that  controversy  exists as to what a negative  physiological
effect is  in  relation to regulatory action  pursuant  to protecting the  public
health.   Since  this  approach  attempts  to  separate   clear  health  problem
emissions from non health  problem emissions,  it is prudent to  be  conservative
in selecting the lowest level of the range.

-------
                                      -4-

206 (a) (3)

    "(3) (A)   A certificate of conformity  may be issued under  this
    section   only  if   the   Administrator   determines   that   the
    manufacturer (or in the  case  of a vehicle or engine  for  import,
    any  person)   has   established   to   the  satisfaction   of   the
    Adminstrator  that  any  emission  control  device,   system,   or
    element of design installed on, or incorporated in, such  vehicle
    or  engine   conforms   to  applicable   requirements   of   section
    202(a)(4).

    (B)  The  Administrator may conduct  such  tests  and may  require
    the manufacturer (or any such person) to  conduct such tests  and
    provide   such   information   as   is   necessary   to  carry   out
    subparagraph  (A)  of  this  paragraph.   Such  requirements  shall
    include a  requirement  for  prompt reporting of  the emission  of
    any unregulated  pollutant from a system, device  or  element  of
    design  if  such  pollutant was  not  emitted,  or was  emitted  in
    significantly lesser amounts, from the  vehicle or engine  without
    the use of the system,  device, or element of design."

In response to these 1977 CAA requirements  EPA began a  program  to  modify  the
existing  procedures  dealing  with  hazardous  compounds  or  conditions.  The
existing sections pertaining to the hazardous  pollutant areas were contained
in 40CFR 86.078 subsection 5(b) as follows:

    "Any system  installed  on or incorporated in a new  motor  vehicle
    (or new motor vehicle engine) to  enable such vehicle  (or  engine)
    to conform to standards imposed by this  subpart:

         (i)   Shall  not  in  its  operation  or  function  cause  the
         emission  into  the  ambient  air of  any  noxious  or  toxic
         substance that  would not  be emitted in the  operation  of
         such  vehicle  (or  engine)  without  such system,  except  as
         specifically permitted by regulation; and

         (ii)  Shall not in its operation,   function, or  malfunction
         result  in  any  unsafe  condition   endangering  the  motor
         vehicle,  its  occupants,  or  persons, or  property  in  close
         proximity to the vehicle.

         (2)  Every manufacturer of new motor  vehicles  (or new  motor
         vehicle engines) subject to any  of  the standards imposed  by
         this  subpart  shall,  prior  to   taking  any  of  the  action
         specified in section  203 (a)(l) of the  Act,  test or  cause
         to be  tested motor vehicles (or motor  vehicle engines)  in
         accordance with good engineering practice to ascertain  that
         such  test  vehicles   (or   test  engines)   will  meet   the
         requirements of  this section for   the  useful life  of  the
         vehicle (or engine)."

-------
                                      -5-

Furthermore  40  CFR  86.078-23(d),  which  deals  with   data   required  as  a
precondition for certification, states the following:

    d)   "A  statement   that  the   vehicles   (engines)   for   which
         certification is  requested conform  to  the  requirements  in
         86.078-5(b) and that  the  descriptions of  tests  performed  to
         ascertain   compliance  with   the   general   standards   in
         86.078-5(b)  and  the  data  derived  from  such  tests  are
         available to the Adminstrator upon request."

In past  years,  the manufacturers  simply submitted the required statement  but
not the data on which it was based.

As a first attempt at bringing attention to and developing procedures  for  the
implementation  of  section  202(a)(4),  a  working  draft  Advisory Circular  was
sent to  the  manufacturers  for constructive comment on October 18,  1978. As  a
result  of  the  input received from  the manufacturers   and  other  interested
parties, EPA has proceeded  to  refine  the approach  to implementation  of  section
202(a)(4) of the Clean Air  Act.   EPA had issued Advisory Circular (AC)  76  (on
June 28, 1978)  requiring the  manufacturers  to  continue   to provide  statements
to the  effect  that the  new model  vehicles  certified to  be in  compliance with
the vehicle emission standards would not contribute to an unreasonable  risk to
public health.  On November 30,  1978, EPA  issued  AC  76-1 which continued  the
procedures set forth in AC 76 for the 1980 and later model years.

In  addition  to  the aforementioned  activities,  EPA  has been developing  and
documenting measurement  methodologies for  the  automobile industry  to  use  in
investigating  unregulated  pollutants  from  mobile  sources.  Two  reports   on
unregulated  pollutant  measurement  methods  have  been   produced  and   widely
distributed  among  auto  manufacturers   (1,2)*.   Also,   EPA  has  produced   an
initial data base  on unregulated  pollutants under  a variety of conditions from
vehicles utilizing different emission control systems  (3,4,5).

Methodology

The following sections will discuss  the  methodology developed  for   determining
a range  of  concern for  toxic  unregulated pollutants  from mobile sources.   It
should  be  reemphasized  that this  is  an approach  to this problem but  not  the
only approach.   It is felt  that  this methodology provides  a  valid  procedure
that is  neither  too complex nor  too simple  for  the  problem.   It is also felt
that this methodology could be used for a wide range of  unregulated  compounds
and yield  acceptable results.**  Figure  1  is provided  to  illustrate  the  way
this approach is structured.
*   Numbers  in  parentheses refer  to references
    which are listed at the end of this report.
**  We  invite  comments from  interested parties
    on  this approach  to  implementing  one  part
    of section 202(a)(4) of the Clean Air Act.

-------
                                       -6-

                                    Figure 1


              Flow Diagram  - Toxic Pollutant  Range of  Concern
                             Preliminary
                              Estimated
                                Range
                                  of
                                Concern
  Substance
     of
  Concern.  .
  Identifled
                                   II
                            Hcalch Effects
                             . Literature
                                 Search
                               Mobile  Source
                             Emission  ['actors
                              (determined or
                                estimated)
                           III
                       Dispersion
                      Models Re-lacing
                      Emission Factors
                      To Health. Effects
 "No Problem"
Implies Low Level
    of Effort
  Monitoring
    "Concern"
Implies  Voluntary
    Action  by
     Industry
                    IV
                Range of
                  Concern
              Emission Level
                "Determined
                                                                          Emission Control
                                                                           Systems of
                                                                              Concern
                                                                            Identified
                                                                          . (If
 "Danger"
  Implies
Re&ulat Lon

-------
                                      -7-

    Emission Factors
An essential part  of  the information for this methodology  is  accurate vehicle
emission  factors.   An  emission factor may  be determined  from the  available
literature and is the mass per unit distance or unit  time of  a pollutant which
is emitted  from a particular  vehicle  type over  a  certain  driving  cycle  or
mode.  A  complete  listing of  vehicle emission  factors for  the pollutant  of
interest  is needed for  the full range of  vehicle types and  driving  schedules
that contribute  to the  fleet  emissions  of the pollutant.   Since  most  of  the
exposure  situations (to  be discussed  later)  involve highway situations or city
street canyons, emission factors are  needed  on the vehicle types which  may  be
present in  these  locations.   Therefore,  emission factor  information will  be
needed for  the categories listed in Table I.  This list is not  intended to  be
all-inclusive  and  the  groupings  may change  from specific pollutant case  to
specific  pollutant  case  depending  on  the  information  available  and  the
information needed.
                                    Table I
                          Emission Factor Categories

  I.     Heavy Duty Vehicles (HDV)

    A.   Gasoline-Fueled Heavy Duty Trucks  (HDT-G)

    B.   Diesel-Fueled Heavy Duty Trucks (HDT-D)

 II.     Light Duty Vehicles (LDV)

    A.   Gasoline-Fueled Light Duty Vehicles (LDV-G)

    B.   Diesel -Fueled Light Duty Vehicles (LDV-D)

III.     Light Duty Trucks (LOT)

    A.   Gasoline-Fueled Light Duty Trucks  (LDT-G)

    B.   Diesel-Fueled Light Duty Trucks (LDT-D)

 IV.     Motorcycles

  V.     Others

In general, under  each classification, an  entry for  the   emission factor   of
the  unregulated pollutant  under  study is  needed,  if it is  appropriate.
Sometimes certain emission factors are not  needed.  For example,  if  the

-------
                                      -8-

unregulated pollutant had  to do with tetraethyl lead  or motor mix and/or  its
combustion  products,  an emission  factor for  those vehicles  that use  Diesel
fuel would not be necessary,since Diesel fuel does not contain lead.

On the other hand,  subclassifications  more  detailed than those shown  in Table
1, may be  needed,  if  the  emissions   of  the  unregulated  pollutant  are  not
adequately described by that  level  of  detail.  For example, for  sulfuric acid
emissions the presence or  absence of an oxidation catalyst and the details  of
the  air  injection   system  are  known  to  influence   the   emission  factor
significantly, so some more detailed subclassification is necessary.

The "Other" category is for  gas  turbine-powered vehicles or electric or  hybrid
vehicles,   etc,   a   catch-all   category   of   unusual   and/or   potentially
problematical  power-plants  or  vehicles which  might  emit   the  unregulated
pollutant of interest.

Unregulated  pollutant  emission  factors  are  frequently  a  function  of  the
vehicle  and engine  type,  emission control system  design,   vehicle  driving
schedule,  vehicle condition  (e.g. malfunction),  vehicle  mileage  and  other
variables.  The  combination  of  these  factors provide a  fleet  average  emission
factor.  However, simplifying assumptions can be made  for a given pollutant  so
that the combination of these factors can be reduced to a workable calculation.

The fleet  average emission  factor  for the  unregulated pollutant of  interest
should be calculated  for a variety of  vehicle fleet situations.   Two primary
sources of  information used  to  determine fleet  average emission factors  are
the Pedco report  (6)  and  the EPA publication "Mobile  Source Emission  Factors:
For Low Altitude Areas  Only" (7).  This information can be used  to determine
fleet emission factors  for use  in the ambient concentration models.   Examples
of  these procedures  can  be found  in  the  sulfuric   acid  example  in  later
sections of this report.

The information  contained  in the  Pedco report  allows  the  determination  of
weighting factors for  the  different major categories  of vehicles in  Table  1.
The information  contained  in the EPA  mobile source  emission factors  report
allows for a further demarkation of the weighting  factors to include different
emission control  system designs.   Both of these references aid in calculating
a fleet average emission factor.

When one  vehicle category  is of  particular  interest,  such as  the 3-way  plus
oxidation catalyst light duty vehicle  in the sulfuric acid example, the  fleet
emission factors  should be calculated  for various percentages of  this vehicle
category. Using  values of 25%,  50%,  75%  and  100%  of  the  vehicle  category
comprising  the  entire vehicle  fleet will allow  the effects  of  the  specific
vehicle/control system type to be estimated.

In most cases it will  be desirable  to  compare the emissions and  corresponding
air quality  of  a specific  vehicle  category  to  the ranges of  concern for  an
unregulated pollutant.  Thus, the above method of determining fleet  emission
factors will  probably be  most desirable.  However,  other suitable methods  of
making  this  determination  are  also possible  using different  references  and
averaging  schemes.   This   method has  been  picked  because  of  the   author's
familiarity with the cited  references  and the simplicity of  application.

-------
                                      -9-

II    Health Effects

A  health effects  literature review  is  necessary  to  determine a  range  of
concern  for a  particular pollutant.  Health effects information at  low levels
of  exposure  are  quite  important  for  this  assessment,  especially  since  the
health effects  at  these levels  are not  always  as easily detected,  or  readily
apparent.  Also,  chronic exposure  at low levels,  such as exposure  to a  low
concentration  of  pollutant  for  1-2   hours  per  day,  5  days   a  week  for  a
significant portion of  the subject's  life,  are  very important due to the  fact
that this  is  the  type  of  exposure which most  people  will be encountering  in
their  day  to  day activities.   Of particular  interest,  of  course,  is  that
health effects  information involving  human subjects.   Complete and appropriate
information  on  these  lower  levels  is  a  very  desirable   element  in  the
determination of a range of concern  for mobile source emissions.   The  lowest
level  of the  range of  concern  can be  defined  as the lowest concentration  of
the  pollutant  at  which adverse  physiological  effects  are  detectable.   This
definition  provides a  useful  and workable concept  for  later  calculations
involving  vehicle  emissions,   but does not  adequately  address  the   actual
causative factors  relating  to hazardous pollutants.  Other concepts which  are
of critical importance  in the ultimate  determination of  a pollutant  hazard  are
exposure  and  dose  or  effective   dose.  Exposure  can  be  defined as  that
parameter resulting from the calculation of concentration multiplied by time.
Dose is  defined as some  fraction of  the exposure which  is  presented   to  the
critical organ  or  subpart  of the organism of interest.   However,  in order  to
provide  a reasonable workable methodology, only concentration and  occasionally
exposure information  will  be  used.   It is recognized  that  if  a  potential
concern  is  determined  by this methodology then a more detailed  assessment
involving dose  determination may need  to be done.  In  general,  this type  of
terminology is used in  conjunction  with experiments involving various doses  of
the chemical of interest.  Below the  lowest  level of the  range there should  be
no  evidence,  from the  available  literature, which would lead  to the  belief
that there is  a hazard  to human  health.  The highest  value  of  the range,  on
the other hand,  can be defined  as  the  concentration  above which studies  show
hazards  so  well  defined as  to encourage regulation.   Both the  highest  and
lowest levels of the range are  not  static quantities and  may vary  depending  on
the  interpretation of  the  health  effects  data  and  results  from  new  tests.
Each,  however,  provide important information on  the  estimation  of  the  levels
which may be hazardous to humans.

The literature  search,  that  will  result in  a set of relevant citations  on  the
pollutant of interest,  comprises  the  first  step in the health effects  review.
A thorough computer assisted search using the  computer data bases TOXLINE  and
TOXBACK  as well as a manual journal  search  is  usually appropriate  to  uncover
most of  the health effects data on a  given pollutant.  An  example of this  type
of search is contained in Appendix II.

Prior  to the evaluation  of  the literature  search  information,  a preliminary
range  of pollutant level  concern  with  respect  to public  exposure  should  be
established  that  will  bracket   the  range  of  concern.    The  selection  of  a
preliminary range  of concern allows  a  closer focus on the literature  review,
eliminating  from  primary  consideration  those  studies  which  provide   health
effects  data   due  to   very  high  concentration  exposures.  To  simplify  this

-------
                                      -10-


calculation, the upper  value  of the range will  be the Threshold Limit  Value
(TLV) as listed  by the ACGIH, if available  and  appropriate.  The lower  value
of  the  range  should represent  the lowest  concentration  suspected  to  cause
health  effects.   This  lower  value  can  be  determined  from  the   available
literature.  In the  event that  such a  level cannot readily  be  determined,  the
lower  level  will   be  estimated  according  to  the  guidelines  provided  for
hazardous pollutants from multimedia sources (MEGs)(8).

The  studies obtained during  the literature search  and  document acquisition
should be reviewed and  rated  as to  their scientific merit and  suitability  for
further use.  The  studies which are rated as being of high quality  and  which
are  suitable  for  use  in range  of concern  determinations  should be used  to
develop a concise  set of  tables listing the results of all  related  studies.*
These tables are ultimately used to sort out the data  prior  to a decision on a
range of concern.

The whole process  of the health effect  literature review can be considered  as
an  effort  to   narrow,   or  validate,   the  preliminary  range  of   exposure
concentration suspected to be of  concern which was determined at the  start  of
the process.  The  detailed  information  may  allow  the  higher and lower  values
of this range to be changed such that the differences  between the high and  low
become smaller.

Once  the  appropriate information has  been  tabulated,  a  large  table will  be
prepared compiling,  separately,  all the information  for  the animal  and  human
studies.  These  studies will be  arranged  in numerical order  from highest  to
lowest concentration, noting  especially the exposure  time,  and will later  be
used  in  conjunction with the ambient  air  scenarios  to graphically  represent
the  conversion from an  ambient  air  concentration  range  of  concern  to   an
emission factor range of concern.

In the  event  that  health effects information  is not  readily  available   for  a
particular pollutant at low concentrations,  but a TLV  has  been established  for
that same pollutant, an equation using  the  TLV can be used  to compute ambient
air  concentrations  for  various   exposure   times  (8).    This  equation   is
represented below.
    Rating of  the scientific  studies should  be performed  by a  competent,
    professional  in  the  particular  field  of  study  (i.e.  epidemiologist,
    toxicologist, pharmacologist).  The  rating process may vary but  should
    include  the  evaluation of  several aspects  of  each  study  such  as  the
    number of  animal subjects  in  each test  group,  the  suitability  of  the
    control   groups   used,   comparison   with   historical   controls   and
    experimental  results,  suitability of  analytical/pathological methods,
    etc. Peer  reviews and  judgements  of  independent experts are expected  to
    be helpful when available.

-------
                         TLV
          -11-
         TLV       4£
int      100   x   e
   where TLVint  = Threshold Limit Value based intercept

         TLV     = Threshold Limit Value
         100     = a safety factor (this value may vary
                   but for the purpose of this paper is
                   chosen to be 100)
          40     = hours worked per week (assuming 8 hours
                   per day, 5 days per week)
          ex     = exposure time per week (in hours) corresponding
                   to scenario X
It is desirable to have a full and complete literature search, but  this  is  not
always possible.  The  TLV-based  approach is one that is  not  far from being  a
last  resort.   The  TLV-based approach  is also  somewhat  conservative  as  the
following table indicates, as far as carbon monoxide (CO)  goes.


                                   Table II

                       Example of the TLV  - Based Method
CO TLV                                      50 ppm (10)

CO value based on daily
8 hr exposure (50/100) x (40/40)              0.5 ppm

8 hr NAAQS for CO                            9.0 ppm

CO value based on
daily 1 hr exposure (50/100) x (40/5)         4.0 ppm

1 hr NAAQS for CO                           35 ppm

It can  be  seen  that the TLV-based -method predicts  lower  values than those  of
the NAAQS, at least for CO.

-------
                                      -12-

III    Concentrations To Which People Might Be Exposed


In  order to  relate  the  health  effects  information  gathered  in  the  health
effects  literature  review to vehicle emission rates, a  set of models must  be
chosen   to  represent   situations  of  appreciable  mobile   source   pollutant
concentrations   with   significant   public   exposure.    This   involves   the
application  of  pollutant  dispersion  modeling  techniques  to  estimate  the
concentrations  of  mobile  source  pollutants.   Since  toxic  (non  carcinogenic)
pollutants are  of  primary interest in  this  report,  special emphasis has been
given to those models and  situations  which may  reflect  short-term scenarios  to
which the public is exposed.

In relating the  public  exposure of a mobile  source  pollutant to known  health
effects information, the most important parameter is  the  dose  which  the  public
receives.  However, a pollutant dose  determination for the  general public is a
very difficult parameter  to  measure,  and  there is expected  to  be considerable
variation  in  the  parameter  throughout  the  population.    Therefore,  for  the
purposes of this approach,  the exposure  time of the general populace  to  the
pollutant of interest in most cases will be assumed to be of approximately 1-2
hours per  day,  probably  on  each day  of  a  5 day  work  week.   This overall
exposure time may  be spread among one or more  of  the scenarios  listed  below.
This  assumption is made  to  keep  this approach simple with  the  intent that  if
the resulting  analysis  shows that a  significant  potential  problem may  exist,
then a more detailed analysis of both  the health effects and the exposure  or
dose may be desired.

Public exposure  to  mobile source pollutants  occur in a variety  of  situations
from  potential  short-term,  high  concentration  events  involving    personal
garages,   parking   garages  or  other  enclosed  spaces,   to   long-term  low
concentration events  in an area wide scenario.  For non  genotoxic pollutants,
the most relevant  situations  to be  concerned about  with  respect  to  public
health would  appear  to  be  high  concentration short-term situations  (acute
exposures).  This is not to  imply that  chronic  effects of these  pollutants  are
not of  concern, but  rather  that these  chronic situations  require  data  from
long term health studies,  of which very few  have been done.  This report will
be  concerned  with  short-term  high  concentration   exposures.    These   acute
exposures are probably repeated often, perhaps five or more  times in  each week
so that  a  chronic  exposure to a  repeated acute dose  may be the most  relevant
exposure regimen.  However,  it is expected that little, if  any,  health effects
data will be available  for these  conditions  and that the most  closely related
health data  will  be  used as  a  comparison  to the  mobile  source  case.   The
scenarios which should be used in this assessment are  listed in Table  III.

-------
                                      -13-
                        Table III - Exposure Scenarios

I.  Enclosed Space

         A. Single Family Garage

         B. Parking Garage

         C. Roadway Tunnel

II. Street Canyons

III.     Expressway

         A. On Expressway

         B. Beside Expressway

IV. Localized Area Sources

         A. Parking Lots, Airports

These  scenarios  may  vary  in  the  maximum concentrations  of  mobile  source
pollutants  but  each  scenario  has  the  potential  for elevated  concentrations
under certain conditions which may lead to a public health hazard. For  each  of
the scenarios  listed  above,  it  is possible to  select an  infinite  number  of
input variables to determine  the  results.   Among the variables which may  have
strong effects  on the results are meteorological variables  such as  windspeed
and atmospheric stability, model parameters such as the mixing  factor,  as  well
as  physical  parameters  such as  number  of  lanes  per  highway,   number  of
vehicles,  fleet  composition,  emission  factors,  and building  height.  Within
this  myriad of  choices we  selected two  collections  of variables  for  each
scenario  to  represent an "average"  and  a  "severe"  situation  with respect  to
the particular  scenario.   Each average  and severe  situation was  chosen  to
represent  a  "real  world"  situation.   A  more  detailed  discussion  of   the
situations listed above, along with  the  the models and conditions chosen,  can
be found in Appendix I.

Enclosed Space Models

The enclosed  space scenarios  which  we  have  chosen  to  model  include  private
residential  garages,  parking  garages,  and  roadway  tunnels.   Each  situation
will be estimated using some form of Turk's equation (see Appendix I).

Residential Garage

The example conducted for  the residential garage involves two  situations,  one
with  an  operator  entering the garage,  opening the  garage  door, starting  the
car, warming  the  car  up for  30  seconds,  and driving out of the garage.  This
situation results, for  example, in a  garage  concentration of 58 ppm CO  in the
case of an emission rate of 7 grams/min. CO. The  second  situation involves the
previous  condition  followed by a  5 minute  period of idle,  with the  garage
door  open,  before the car is driven out.   In  this case,  the CO has  increased
from 58 ppm to  280 ppm. The  equations and charts in Appendix I can be  used  to
perform similar calculations for other pollutants.

-------
                                      -14-.

Parking Garage

The  parking  garage scenario  is being estimated  by the  use of  two  separate
parking  structures.  An  underground  parking  structure   (Los  Angeles  Music
Center) is being used to  simulate  a severe  case while an above ground parking
structure (San Antonio,  Texas)  is  being used  to simulate  an average or typical
case for this scenario.

The  ambient  air   concentrations   for  the  parking  garage   scenarios   were
calculated using the following equations;

    1.   For contribution of initial concentration in garage

          G! = C0 exp (-% FV2 t/V)

    2.   For contribution of pollutant concentration in incoming
         ventilation air

              C2 - C± [  1 - exp (-R-J. FV2 t/V)]

    3.   For contribution of vehicles within the garage
                   ng
              C3 = RiFv3 [1 - exp  (-Ri FV3 t/V)]

    4.   The total ambient concentration is:

              C = GI + C2 + 03

where:

    C    = nominal pollutant concentration,  ug/m^

    Co   = initial pollutant concentration at time, t = o, ug/rn^

    GI   = contribution of initial pollutant concentration, ug/m^

    C2   = contribution of ventilation air,  ug/m^

    63   = contribution of pollutant source, ug/m^

    Ci   = concentration of pollutant in ventilation air,  ug/m^

    Fv   = effective ventilation factor,  dimensionless

    n    - number of vehicles or emission sources, dimensionless

    q    = pollutant source rate of emission, g/min

    R^   = ventilation rate, ft^/min

    t    = time, minutes

    V    = volume,

-------
                                     -15-

In the severe case, the exposed population is assumed to  be  located  at  level 5
of the parking structure and  is  leaving  the  structure after a major  event,  20
minutes after the  garage started  to  empty.   Then, for an emission factor (EF)
of 1  g/min of pollutant,  and an initial concentration assumed  to be  1  mg/m3
the following factors pertain to this severe  case.

                    Severe  case  parking garage  (1  g/min EF)

     Level 5                9,600 ug/m3  for Ig/min EF (ramp  intake)
                           46,100 ug/m3  for Ig/min EF

     Ramp 5 to 4           14,900 ug/m3  for Ig/min EF

     Ramp 4 to 3           12,300 ug/m3  for Ig/min EF

     Ramp 3 to street      10,100 ug/m3  for Ig/min EF

As in all these situations, the emission factor (EF)  is a direct  multiplier  so
that  pollutant  concentrations  can  be  directly  factored   up  from  the  base
level.  To  illustrate  this severe case,  CO can be  used.   If an  emission rate
of 7  g/min  is used for  both  ramp and level areas  then the following  numbers
are generated.

               Level 5       9,600 x 7 g/min  =   67,200 ug/m3  (Ramp air)
                            46,100 x 7 g/min  =  322,700 ug/m3

                            Total Level  5    =  389,900 ug/m3

                                                or 339 ppm

               Ramp 5 to 4              14900 x 7  = 90 ppm

               Ramp 4 to 3              12300 x 7  = 75 ppm

               Ramp 3 to street         10100 x 7  = 61 ppm

This example illustrates the  substantial build up of  mobile  source   pollutants
which may  occur in an underground,  mechanically-ventilated   garage during  a
high vehicle use period.

The typical  case provides  factors for an  above  ground, naturally ventilated
parking structure  in  San Antonio, Texas.  For a  1  g/min emission rate  these
factors are:

              Typical Parking Garage Scenario

                   Parking level      3,900 ug/m3  for 1 g/min EF
                   Ramp Room         13,750 ug/m3  for 1 g/min EF

To illustrate this example for  CO,  assuming  a  0  ppm  inlet air concentration,
the concentrations corresponding to a 7  g/min emission rate  are:

              Parking level    7 x  3900   =  24 ppm
              Ramp Room        7 x 13750   =  84 ppm

-------
                                     -16-

Roadway Tunnel

The average and severe roadway  tunnel  situations  have  been simulated by use of
two  separate  roadway tunnels.   The  severe  case  has  been selected  as  the
Baltimore Harbor tunnel, which  is  a long, heavily used  tunnel  with  a somewhat
aged  ventilation design.   The  typical  case  is  a  Minnesota  highway  tunnel.
Details on these two  tunnels and on the modelling methods used for  the tunnel
scenario can  be found in  Appendix I.  The   multiplicative factors  for  these
situations are:

              Tunnel  Scenarios  (1 g/mile emission rate)

                   Typical   -    1123 ug/m3 for 1 g/mile EF
                   Severe    -    2856 ug/m3 for 1 g/mile EF
Street Canyon

The  street  canyon  situation  will  be  simulated  by  a  form  of  an  equation
developed  by researchers  at  SRI,  Int'l.  and  is  represented  by the  general
equation below.
         Cv
                   7 x 106 QL
                    (U + 0.5) (S + 2)
where:        Cy        = ambient concentration resulting from the
                          vehicles in the street canyon, ug/m

              QL        = vehicle emission rate, g/sec m

              U         = rooftop wind speed m/sec

              S         = slant distance from exhaust to receptor, m

              7         = empirical correlation factor

              10^       = units conversion, grams to ug

              0.5       = empirical wind correction, m/sec

              2         - empirical slant distance correction, m


The general form of  this  equation has been indirectly verified  by  wind  tunnel
tests.   Several  situations  have  been  worked  out  for  the  street  canyon
scenario.  The  example situations model a street  canyon in San  Antonio  and
Houston, Texas with  specific  physical characteristics.  This example, using  a
unity  emission  factor  of  1  g/mile  results  in  the  pollutant  concentrations
listed in Table IV.

-------
                                      -17-

                                   Table IV

                    Street Canyon Ambient  Air Concentration

                                       Concentration from Vehicle ug/m^

         Average Case        Sidewalk             31.3
                             In Vehicle          114.1

         Severe Case         Sidewalk             97.5
                             In Vehicle          334.8

Since  the  emission  factor  in  this  equation  is  a  direct  multiplier,  the
concentration of  any pollutant can  be  scaled up  by multiplying  the  emission
factor by a specific entry in Table iy.


Expressways

Expressway  related   exposure   to  mobile  source  pollutants  can  occur   in
essentially two  ways:   either by  being  in  close  proximity to the  expressway
(living and working close) or by being on  the  expressway  as a  commuter.   These
two situations call  for different  approaches to  modeling.  The close proximity
case  can  be estimated  by using one  of  the  available  line source  dispersion
models.  This report will make use of the G.M. model developed by Chock.  The
close  proximity  expressway example  uses  an   existing  Houston, Texas  freeway
and obtains an ambient  concentration  vs  distance from the  roadway for a  unity
emission  factor.  Details  on  this  can be  found  in Appendix I.   The  commuter
case  will  be  handled  by  use of  the EPA  Point,  Area and Line model  ("PAL")
which  can  input  both  point  and line  source.   The approach used  can best  be
described  as  a  relative  motion  procedure  viewing  each  vehicle  as  a  point
source upwind of the commuter.

The close proximity expressway situation can be  considered to  be a severe case
simulation.  A  typical  case is  not  provided here  because  the  ambient  air
concentrations resulting  from  this  situation  are very  low  compared  to  the
other scenarios and thus only the  severe case  assumptions provide  data  that  is
useful in  making up the  various  pollutant concentration  profiles.  The  short
term  factors  are calculated  for  a rush  hour time period (one hour  average)
with  an   average  traffic  density  of 3780  vehicles/hour.  The  factors  that
result from this expressway model treatment are listed below.

-------
                                      -18-

                                    Table V
              Close Proximity Expressway Situation (1 g/mile EF)

    Distance downwind
    from road edge,                         Ambient Concentration
         meters	              Short Term    ug/m^	Long Term*

           1                           397.0                   61.0
           5                           368.0                   55.0
          10                           334.0                   48.0
          25                           248.0                   35.0
          50                           171.0                   27.0
         100                           105.0                   14.0
         500                            26.3                    4.0
        1000                            13.6                    1.6
*   The long  term case has  been  developed using
    a  long  term  average  of observed  values  for
    CO  along  the  Houston  freeway.    The  values
    were   corrected   for   emission   factor   in
    deriving the table above.


The on-expressway  (commuter)  scenario  is simulated using a  typical  and severe
case.   This  scenario  used  a  specifically designed  computer program,  called
ONEX, to calculate ambient  concentrations  of  pollutants  in  the vicinity of the
vehicles on  the  expressway.   The typical exposure  case  used  a San  Antonio
expressway  running north  and south  during  the  peak  rush  hour.   A  4  lane
segment of this  freeway with an  average daily  traffic  of 30,000 vehicles per
day was used in the simulation.

Table  VI  presents  the  multiplicative  factors   for  the  typical  expressway
situation using  a range  of wind directions  and  speeds.  The range of  these
values  (61 to 124 mg/nP) is fairly close  considering  the large range  of  wind
speeds used (1.0 to 6.0 mph).

-------
                                     -19-
           TABLE VI AMBIENT CONCENTRATION FOR RECEPTOR ON EXPRESSWAY
                 TYPICAL EXPRESSWAY EXPOSURE SITUATION
                                  1  g/mile EF

                                               Outside downwind lane
   Wind Direction            Wind Speed        Ambient Concentration
  Degrees Relative           m/sec (mph)
         357.5                1.0 (2.2)                120
         355.0                1.0 (2.2)                120
         340.0                1.0 (2.2)                122
         315.0                1.0 (2.2)                124
         270.0                1.0 (2.2)                122

         357.5                2.0 (4.5)                114
         355.0                2.0 (4.5)                113
         340.0                2.0 (4.5)                109
         315.0                2.0 (4.5)                103
         270.0                2.0 (4.5)                 95

         357.5                3.0 (6.7)                109
         355.0                3.0 (6.7)                107
         340.0                3.0 (6.7)                 99
         315.0                3.0 (6.7)                 85
         270.0                3.0 (6.7)                 77

         357.5                6.0 (13.4)                96
         355,5                6.0 (13.4)                92
         340.0                6.0 (13.4)                75
         315.0                6.0 (13.4)                61
         270.0                6.0 (13.4)                72

The severe exposure case used  the  Santa Monica freeway in Los Angeles.    A 10
lane portion of this highway with  a 200,000 vehicle/day  traffic  load  was used
in  the  simulation.   Table VII  presents  the  multiplicative  factors  for  the
severe situation.

-------
                                     -20-
         TABLE VII,
AMBIENT CONCENTRATION FOR RECEPTOR ON EXPRESSWAY
SEVERE EXPRESSWAY EXPOSURE SITUATION
             1  g/mile  EF
   Wind Direction
  Degrees Relative

    357.5
    355.0
    340.0
    315.0
    270.0

    357.5
    355.0
    340.0
    315.0
    270.0

    357.5
    355.0
    340.0
    315.0
    270.0

    357.5
    355.0
    340.0
    315.0
    270.0

Area Wide Sources
    Wind Speed
    m/sec (mph)

    1.0 (2.2)
    1.0 (2.2)
    1.0 (2.2)
    1.0 (2.2)
    1.0 (2.2)

    2.0 (4.5)
    2.0 (4.5)
    2.0 (4.5)
    2.0 (4.5)
    2.0 (4.5)

    3.0 (6.7)
    3.0 (6.7)
    3.0 (6.7)
    3.0 (6.7)
    3.0 (6.7)

    6.0 (13.4)
    6.0 (13.4)
    6.0 (13.4)
    6.0 (13.4)
    6.0 (13.4)
Outside downwind lane
Ambient Concentration
	ug/m3	

      454
      467
      494
      506
      495

      453
      458  .
      428
      400
      369

      444
      435
      375
      327
      285

      399
      366
      275
      200
      148
Parking lots are an exposure  scenario which may  also  involve  large populations
of  people  in close  proximity to mobile  sources.  However,  EPA  studies  (see
Appendix  I)  have  not  shown  appreciable  levels  in parking  lots,  although
several measurements have  shown  levels  up to 50  ppm  CO  where the  parking  lot
enters  the  street.  Normally,  the levels  were   20  ppm  or  less  for  CO  even
during peak use of a sports  stadium  parking lot  where the intermittent use  is
important.  Thus, it appears  that parking lots do not produce as  high a  level
of automotive pollutants as do the other  scenarios.   Because  of  the difficulty
which was encountered  in  locating suitable models  and applying them, no  case
was developed for this scenario.

IV   Range of Concern Determination

Using the compiled information  from the previous  sections  of this  report,  it
is possible to convert  the health effects data to  corresponding mobile source
emission factors for the various  ambient air scenarios.

-------
                                     -21-

Figure 2_ represents  graphically  the  output of the mathematical models  used to
simulate the  different  scenarios used in  this  report.   An inspection  of  this
plot  (Fig.  2)  reveals  that,  for  any  specific  ambient  air  concentration  of
pollutant, a  large number  of vehicle emission factors, one for each scenario,
can  be  selected.  However,  if  an  ambient  air  concentration is  intended  to
represent a lower or upper level of  health effects concern, then an  additional
variable is important, and that  variable is exposure  time.  The health  effects
data used to  project or estimate a range of concern in the ambient  air always
has  a  specific exposure  time associated  with  it,  which relates  back  to  the
actual  dose  the  organism  receives  when   the  adverse health  parameters  are
measured.   When  comparing   these   health  data  to  mobile   source emission
scenarios it  is important  to remain consistent  regarding  the exposure  time.
However,  in  practice  it  has been  determined  that  very  few if  any  health
effects data  have exposure  times and exposure schedules similar to  the actual
human  exposure  to  mobile  source  pollutants.   Another  confounding   factor
related  to  this exposure  time issue is that the exposure  time  and exposure
schedule of a specific individual in the  population would be spread  out  among
a number of the mobile source scenarios in any  given day.  As an  example  from
Appendix I,  consider the  hypothetical exposure pattern of a  typical suburban
commuter.

                       Suburban  Commuter Exposure (Weekday)

1)   Starts car in garage               5 minutes       -personal garage  scenario
2)   Expressway to work                20 minutes       -Expressway  scenario
3)   Tunnel on trip                     5 minutes       -Tunnel scenario
4)   Central  Business district on trip 10 minutes       -Street canyon scenario
5)   Evening  reversal of above

It can be seen from  this example that this one  individual experiences several
types of automotive pollutant exposure in his/her  work day.

In order to provide, in this  methodology,  a workable  approach  to estimate  the
levels of unregulated  pollutants that may constitute a  human  hazard  it  has
proven necessary to assume a  standard exposure  time and schedule and then  try
to fit  the  available health data into  this regimen.   Thus, for the  purposes
of this methodology the exposure time of the general  population is assumed to
be 1-2  hours/day,  5-7 day/week  throughout  the year.  It is  felt that  this
type of exposure is similar  to what  most populations  are  affected  by in their
normal  activities.   In  most  cases,  health  effects  data   resulting  from
exposures  of  this  type  will not   be  available,  necessitating  the use  of
judgement in  arriving at  the ranges  of concern for mobile  source  pollutants.
It should be  emphasized at  this point  that  this methodology  is  intended  to
provide an  initial  conservative  assessment of the potential hazardous  levels
of unregulated pollutants  from  mobile sources.   If  a  significant  hazard
appears  possible,  then it may be prudent to perform  additional  health  and
exposure studies to  more  concretely  identify the  potential hazard to exposed
humans.

In addition  to exposure  time, other issues may be  important  to  the overall
evaluation  of  a  mobile  source  pollutant.   One  such issue  is   atmospheric
reactivity of a pollutant after  it is emitted, which  may  decrease  or increase
the exposure  of the  population to mobile source pollutants.  When  information

-------
                                                  • Figure 2




                               Pollutant Concentrations vs Emission Factors



«->
0
1

0
•rl
jQ
0
0
E
0
L
tn
o
L
0
•rl
J
a:
<
H
Z
UJ
m
<



1000
950 .
900 .
850 .
800 .

750 .
700 .

650 .
600 .


550 .


500 .

450 .
400 .
350 .

300 .

250 .
200 .
150 .
100 .
50 .
*
£
/

ENEYER/tA

-------
                                     -23-

on  the  reactivity of a pollutant is available,  it can be used to  modify the
emission  factors  by an  appropriate  amount.   However,  in most  cases  this
information  is  not  expected  to  be  readily  available  and  should  not  be
considered necessary for the use of this methodology.

After the  ambient air range of  concern is  identified using  the  appropriate
health  effects data  and  utilizing  the  assumptions on  exposure time  stated
previously, then  this  range of concern  can  be converted to vehicle  emission
factors for the various mobile source  exposure scenarios.   The result of this
exercise should be a table (illustrated fully in the example  provided  later
in the report as Table XII) compiling, in effect,  a  range of concern  for each
scenario and  situation.   In certain cases,  the use  of both the high  and low
levels  of  the range of health effects may be inappropriate,  such as when  a
severe situation  (e.g. Severe Expressway) is compared  to the low  level of the
range of  concern,  if  the  low level  were  derived  using  long term  exposure
data,   and   the   severe  situation  were   derived  using  parameters   (e.g.
meteorology) that would only occur infrequently.

An examination of this  table of ambient concentrations will reveal  that one
scenario will  be  dominant  in  that  it will have  the  lowest emission  factor
range of concern.   However, if this scenario  is an inappropriate  measure  of
the  potential  levels of  human health hazard  for the specific pollutant  of
interest, or if other scenarios are  of specific interest, then other  portions
of the exposure table will also be important.

Initially, the  derivation  of  the  emission range  of  concern  for a  specific
scenario will apply to the  average emissions of  that pollutant  for  the entire
vehicle fleet, except for the  residential garage  situation.  However,  in most
cases,  it  will  be desirable  to  compare specific vehicle/ emission  control
technology  combinations  to  this   range.  Obviously,  if  a   high  emitting
technology, with  respect to  a particular  pollutant  (compared  to the  other
technologies)  has a  maximum emission rate below the  lowest level  of  the
range,  then  that  technology poses  no  concern relative to  that  pollutant.
However, if  a specific  emission control  technology  has an emission  factor
which falls within the range  of  concern for  a  specific  scenario, then  some
question will  exist  with  respect  to that  technology and it will,  therefore,
be  subject  to  closer  scrutiny.    The  overall  evaluation  will  take  into
consideration additional factors which include the fleet average emissions  of
the  pollutant  and  the  maximum percentage  of  that  technology which can  be
expected in the  future.  Any  technology which has  an emission factor  above
the  highest level of the range  will  be  considered  as potentially hazardous
with respect to that pollutant.

Summary and Conclusions

This methodology has been designed to  be a valid  and workable  approach toward
estimating whether emissions  of  an unregulated  pollutant  are a  potential
hazard to public health.  The effort of  trying to  maintain  a   simple  approach
has  resulted  in  the  necessity of  using several  simplifying  assumptions  to
avoid complexities inappropriate  to the nature of  this   methodology.   If  a
potential  problem  is  uncovered  by  this   methodology  for  an   unregulated
pollutant, then a more  detailed investigation of  that specific pollutant can
be undertaken  to  provide a  more  complete evaluation of the hazard.   The  more

-------
                                   -24-


important  simplifying  assumptions  are  that   pollutants  are  non-reactive
within  the  time frames  considered,  short  term  exposures are most  relevant
and  health  effects  data can  be  used to   estimate  the effects  of  mobile
sources on public health.

To illustrate the way that  this  methodology works,  the last section  of  this
report provides an example of one pollutant, sulfuric acid.

-------
                                     -25-

                         Sulfuric Acid Example

This  section  is  presented  as  an  illustration  of  the  way  in  which  major
portions of this approach are  intended to work.   Sulfuric acid has  been chosen
as the  example pollutant  for the approach because  there is  a large amount  of
existing data  on  it, both  in  the health effects  and  emission factors  area.
In spite of  this  apparent depth of information and data,  no range of  concern
for mobile  source emissions has  ever been estimated  or  established for  this
compound.

Sulfuric Acid Emission Factors

Emission factors  for sulfuric  acid were collected  from a  number of available
sources  and are  listed  in Table  VIII.   These  emission  factors  have  been
compiled at  this  time only for  the  Congested Freeway Driving Schedule.   This
particular  driving   schedule  is most applicable  to  the  expressway exposure
situation,  but  may  still have  utility  for the street canyon situation.   The
emission  factors  for  the  enclosed  space conditions,  which  would best  be
derived from an idle or  slow  speed schedule, will  be the  subject of  further
work to identify concrete emission factors  for these situations.

While this example  is concerned with sulfuric acid only, the emission  factors
reported here  generally are measurements of aqueous soluble  sulfates.   In most
cases,  the  predominant soluble sulfate species  in mobile  source  exhaust  is
sulfuric acid  (10).   However,  other sulfates  such as  ammonium sulfate could  be
present.  For  the  purposes  of  this  example, it  is  assumed  that  the  mobile
source emission factors represent 100 percent  sulfuric acid.

These emission factors can  be  combined  to  calculate  fleet average emission
factors for the vehicle fleet  by using available information on vehicle  miles
traveled  (VMT) for  the  different  vehicle classes.   For  simplicity  in  this
example the VMT fractions will  be derived from information in the Pedco report
(6)  for calendar  year  1980  and  Mobile  Source  Emission   Factors: For  Low
Altitude Areas Only  (7).  In future assessments, other  references may  be  used
to perform these  calculations.   Table IX provides  a breakdown of  the  vehicle
class VMT's and the fleet average emission   factor for sulfuric acid.

Obviously this  particular set  of calculations does not represent any specific
fleet emission  factor.  Depending  on the make up of  the vehicle fleet  at  any
point in place or  time  that is  of  interest, the  fleet  emission factor  will
differ.  The  most  severe  case could be considered to be  the  scenario  where
high  sulfuric   acid  emitting   technology  is   the   predominant  member  of  the
vehicle  fleet.   To  address  this  possibility,   and  the   possible  presence

-------
                                     -26-

                                  TABLE VIII
                        Sulfuric Acid Emission Factors*
        Vehicle Category
Light Duty Diesel Vehicles
   w catalyst
   w trap Oxidizers
Light Duty Diesel Trucks
Heavy Duty Diesel Trucks
Light Duty Gasoline Vehicles
    Non Catalyst; no air pump
    Non Catalyst; air pump
    Oxidation Catalyst; no air pump
    Oxidation Catalyst; air-pump
    3-way Catalyst
    3-way Plus Oxidation Catalyst; air pump
Light Duty Truck
    Non Catalyst
    Catalyst, no air pump
Heavy Duty Gasoline Vehicles
Sulfuric Acid (mg/mile)**
           Avg

             9
           100
           100
            16
           100

            0.2
            1.0
           10
           20
            4
           30

            1.0
           20
            4
         Based   on  congested   Freeway   Driving
         Schedule  and  0.030  wt  %  Sulfur  for
         gasoline  and  0.2  wt %  for  Diesel fuel.
         These emission  factors may  change  with
         other   fuel  sulfur   levels  or   test
         cycles.
    **   Reference 9-14.

-------
                                   -27-
                                   TABLE IX

                Fleet  Average  Emission  Factors  -  Sulfurlc Acid*
Vehicle Class
Light Duty Diesel Automobiles
Light Duty Diesel Trucks
Heavy Duty Diesel Trucks
Light Duty Gasoline Vehicles
Non Cat.; no air pump
Non Cat . ; air pump
Ox . Cat . ; no air pump
Ox. Cat.; air pump
3-way Cat . ; no air pump
3-way Plus Ox. Cat.; air
Light Duty Gasoline Trucks
Non Catalyst
Catalyst
Heavy Duty Gasoline Trucks
Total Fleet Average
Fraction
VMT
0.015
0.002
0.027

0.147
0.098
0.289
0.261
0.012
pump 0.008
0.096
0.010
0.035
Sulfuric Acid
Emission Factor
(mg/mile)
9.0
16.0
100.0

0.2
1.0
10.0
20.0
4.0
30.0
1.0
20.0
4.0
11.8 mg/mile

EFxVMT
Fraction
0.135
0.032
2.700

0.029
0.098
2.890
5.220
0.048
0.240
0.096
0.200
0.140

*These calculations  were based  on available information  from the  reference
listed above  (7,8)  and  in Table 1.  Buses,  which  may be a  significant  source
of sulfuric acid emissions under certain conditions,  are not included  in these
fleet averages.

-------
                                     -28-

of future  technology,  the fleet  average emission  factor  can be modified  to
reflect  different  proportions  of these  higher  emitting  technologies.   As  a
worst case, a vehicle  fleet  consisting of 100 percent of  the  highest emitting
technology could be calculated.

Table  VIII  also  presents  emission  factors  for  vehicle/emission  control
categories which are  expected to be  the highest  emitters of sulfuric  acid
under a  variety of  conditions.  Obviously,   it  is these  technologies,  on  an
individual basis, that might  be expected to  constitute the most  likely  source
of an unreasonable  risk to  public  health.   Since  most  of these  technologies
are not yet in  common  use  except  on  an experimental basis, the potential  risk
can  be   considered   to  be  a  future  concern.   To  establish  bounds  on  the
potential risk  from sulfuric acid that  these technologies present,  they  will
be considered in a number  of  hypothetical calculations to comprise  25,  50,  75
and 100 percent of the total vehicle  miles traveled.

By using the  fleet  average emission factors  in  Table  IX and  the  hypothetical
calculations listed  above,  a list of  emission factors can be  calculated  to
use in subsequent steps.  This list is presented in Table   X.

Sulfuric Acid Health Effects
A literature review on the health effects of sulfuric acid was  performed  as  an
input to  the  determination of a range of  concern for mobile source  emissions
of  this   compound.   At  the  present  time  only  a  preliminary  draft  of  this
literature  is  available and  further  information may  modify the  conclusions.
The preliminary draft  of the literature search is  included  as  Appendix  II  to
this report.

As  indicated  in  the   methodology,   in  order  to  focus  the  health  effects
literature review, a preliminary  range  of  ambient levels has been  selected  to
bracket  the  region of  uncertainty  with  respect  to  sulfuric  acid health
effects.   This  range  has  been determined  to  be  10 ug/m^  -  1000 ug/m-^  for
sulfuric  acid.  The  lower  end of this range has  been selected to  approximate
the lowest level  at  which adverse physiological effects can be detected.  The
preponderance of  the evidence has shown  little  or no health effects  at levels
of  sulfuric  acid  below  this,  although  there  are  some   indications   that
sensitive  subgroups  of asthmatics may show  some reaction to  these levels  of
sulfuric acid.  To as great an extent as possible,  this  lower level also  takes
into  account   the   interactions   of   various   pollutants  such  as   S02  and
H2S04.

The upper  level of the range  is  chosen  to  be  the TLV recommended by  NIOSH and
the ACGIH as 1000 mg/m^ (9).  Above  this level several   studies  have  shown
an adverse  reaction  in healthy subjects which may  be  harmful under  repeated
exposures.

-------
                                     -29-
                                    Table  X
                   Sulfuric Acid Emission  Factors  -  Compiled

Fleet Category                                          mg/mile@

Fleet Average (FA)                                         12
FA + 25% 3W +OC*                                           17
FA + 50% 3W+OC                                             22
FA  + 75% 3W+OC                                            26
100% 3W-OC                                                 30

FA + 25% D+C**                                             34
FA + 50% D+C                                               56
FA + 75% D+C                                               75
100% Diesel Cat.                                          100

FA + 25% D+To***                                           34
FA + 50% D+To                                              56
FA + 75% D+To                                              75
100% D+To                                                 100
@Normalized to the Congested Freeway Driving Schedule
*   Light Duty Gasoline Vehicle - Three way + Oxidation Catalyst with air pump
**  Light Duty Diesel Vehicle - Catalyst equipped
*** Light Duty Diesel Vehicle - Trap-oxidizer equipped

-------
                                     -30-

Sulfuric Acid Ambient Air Concentrations

By  using  modeling  techniques  in  conjunction  with  sulfuric  acid  emission
factors,  ambient  concentrations  can  be determined  that  should  bracket  the
range  of  possible sulfuric acid emission concentrations  from  mobile  sources.
This  matrix  of   ambient   concentrations   (Table   XI)   is  composed  of  five
scenarios,  out  of  thousands  possible,  and was  chosen  at this  time as  one
example  of the exposure  ranges.   More  work  is  currently  being  done  to  more
concretely  specify  the  most  appropriate  scenarios  to  use  and  thus  the
scenarios may change in future  efforts  of  this kind.   The information in Table
XI is depicted graphically in Figure 2.

As  indicated  above,  the  range of  exposures to   the  general public can  be
estimated  by  considering   a   limited   number  of   specific  scenarios.    The
scenarios  selected are  all expected  to  be  dominantly  influenced by  mobile
source emissions.   Personal  garages, parking  garages, roadway  tunnels,  street
canyons,  and  urban expressways have  been  selected  to  bracket  the  range  of
sulfuric  acid  concentrations  from mobile  sources  that  influence  short  term
health effects in  the  exposed  population.   Each scenario  is developed as  both
severe  and average exposure  situations calculated  by  the  use  of  existing
ambient air modeling  techniques.  No  attempt  has  been  made to determine  the
cumulative effects  of  these  situations on general public  health.   Appendix I
contains  a detailed explanation  of the rationale for  choosing   the  specific
situations and parameters which lead to the numerical results presented here.

Table  XI  presents the ambient  air concentration of  sulfuric  acid  for  eleven
ambient  situations as  a  function of  vehicle  emission  rates.  Two  personal-
garage situations  are  presented, one  (average  or  typical)  using  a 30  second
vehicle warmup time and the  other  (severe)  using a five  minute vehicle  warmup
time.  These  two  situations  are  intended  to  simulate  summer  and  winter
conditions, respectively.

The two parking garage situations  simulate  average  and  severe  conditions,  with
an above ground, naturally ventilated garage for the  former and an underground
garage for the latter.  The  average  parking garage  case  is calculated  assuming
an exit  time  in which the vehicle spends equal  time  on the parking level  and
the ramp  level.  The  severe parking  garage  is calculated assuming  that  the
exposure takes place 20 minutes after  a major  sporting  event finishes, wherein
the exposed population is at  parking  level 5.  The  initial  concentration  of
sulfuric acid in this garage is assumed to be  low (1 mg/m^).

The roadway tunnel  situations  used two different specific  tunnels  to  estimate
an average and  severe  condition.   A new design,  two  lane  roadway  tunnel  with
moderate  traffic  flow is  used for  average conditions,  while  an old  design,
heavily used roadway tunnel is used for severe conditions.

-------
                                     -31-


The two street  canyon  situations are simulated by examining  the  parameters of
two street canyons.  The most  sensitive parameter in this model  appears  to be
the number of traffic lanes in  the  canyon.   The  severe  condition  is calculated
for a six lane  street  canyon with a 2400 vehicle/hr traffic  load and  with the
exposed population  located inside  of  the vehicle.   The typical condition is
calculated  for  a  two  lane canyon with  800  vehicles/hr of  traffic  and  a
sidewalk location of the exposed population.

The  expressway  situations  require  three  specific  estimations  to cover  the
range of possible  concentrations.  One highway condition tends to  estimate an
exposure involving a close proximity to  the  highway  such as would be gotten by
living  or  working  close   to   a heavily  travelled  freeway.   This  case  is
calculated on  a short  term basis for  a  distance  of  50 meters downwind of the
roadway.  The  other  two expressway situations simulate  a  commuter  (located in
the vehicle)  exposure, with the average case  using a  four  lane,  medium use
1400 vehicle/hour) and  a westerly wind at 1.0  meters/sec  and the  severe  case
using a ten  lane,  heavily travelled  (3600   vehicles/hr)  freeway  with  a  1.0
meter/sec westerly wind.

Determination of the Range of Concern

The  range  of   concern  for   sulfuric  acid  emissions from  automobiles  is
determined using the outputs  from the previous three areas,  emission  factors,
health  effects  and  exposure   estimation (the  emission factors  and  exposure
estimates have  already been combined in  Table XI). Using the  preliminary  range
(10 ug/m3  -  1000  ug/m3)  as  a  stepping  stone  for  this  effort,  along  with
the guidelines  explained earlier  in  the  methodology  section of  this report, an
upper and lower value can be determined for the final range of concern.

The literature  search reveals a human  study  which shows that  an acute  exposure
concentration   as  low  as  66.0  ug/m3  caused  significant   differences  in
lung  function  parameters  in  3  out of  18  subjects  tested  (Gardner  et.  al.
1976).  The   evidence   provided   in  the   literature   also   shows   that   no
physiological  effects  were detected for exposure  concentrations  below  66.0
ug/m3.   Since,   at  this time,  there  is no  available   information  definitely
concluding that there  are  adverse  physiological  effects at  concentrations  of
sulfuric acid  below  66.0 ug/m3, this value will  be  chosen as the  lower  value
in the range of concern.

The upper  value of  the range   will  remain at 1000  ug/m3  as was set for  the
preliminary range of concern.   This  TLV  for  sulfuric acid  is  the  time-weighted
average  concentration  for  a  normal  8-hour   workday  or 40-hour  workweek,  to
which nearly  all workers  may  be repeatedly exposed,   day  after  day, without
adverse  effects.  The  evidence  of  adverse   health  effects  above  this  level
would be sufficient to support  regulatory action.

-------
                                     -32-

Between  the chosen  limits of  this range,  there  are  scattered  data  points
providing  evidence of  adverse physiological  effects  caused  by  exposure  to
various  concentrations  of  sulfuric acid.   Therefore,  this  region  has  been
termed the  "range  of  concern"  for sulfuric acid concentrations  in  the  ambient
air.  This  range can  now  be used  in conjunction with  the  emission  factor data
to  graphically  present  the conversion of sulfuric  acid emissions to  ambient
air concentrations.

Once the  literature  search was completed  and  the appropriate  information was
tabulated  for  sulfuric acid,  a  large table was  prepared  compiling all  the
information for the animal and human studies (see  Appendix 111).  These tables
list  the  studies  according to  the exposure  concentration  of sulfuric  acid
(highest  to lowest  concentration).   Using  this  health  effects  information
along  with  the  emission factor  data  presented   in  Table  XI,  graphs  were
composed  representing  the  relationship  between  ambient  air  concentrations,
emissions  factors,  and  the various types of  public exposure  situations  (see
Figure 2-7).

According  to the  methodology described  earlier  in  the  report, the  lower and
upper levels which comprise  the health  effects  range  of  concern are compared
to  the  mobile source situations  to  calculate  the  emission  factor  range  of
concern.  The chief element of comparability between the  health effects range
and  the  ambient  situations is  exposure  time.   Most   of  the mobile  source
situations  simulate  short term exposures  (durations  of  an  hour  or less per
day) perhaps repeated several  times per week over   an  extended period.   The
average  exposure  situations appear more likely   to be  repeated often,  while
the  severe  exposure   conditions   would  likely   only  occur  on   infrequent
occasions.

With the above information,  the mobile  source  range of concern for   sulfuric
acid can  be estimated  for  the different mobile  source  situations.  Table  XI
lists the  vehicle  emission factors which  correspond to  the high  (1000  ug/m3)
and  low  (66  ug/m3)  portions  of   the  range  of  concern  for  sulfuric  acid.
Inspection  of  this  table shows  that  the  scenarios  result  in  a  variety  of
ambient concentrations  corresponding to  the health effects range of concern of
66 ug/m3 to 1000 ug/m3.

-------
                                     -34-
      959..
      758..
 C.
 o
 -p
 0>
 E
 o

 Q
 07
 o
 ••-•

«w^

cc:
559..
359..
S    259..
                                                         FIGURE 4

                                                     PARKING GARAGE
                                        *
                                        •
                                          9



                                        I
                                          I
                                    t
                                    •
                                    I
                                 I
                               I
I
                            I
                          g
                    i
                   I
      150..

      IBB..   /


       »-/
                9
        00   K5
  C.DDSTBI/-*
                              EMISSION FACTOR (milligrame/milo)

-------
                                  -35- -
                            FIGURE 5
                         ROADWAY TUNNEL
CCSftStfe*
                           EMISSION FACTOR (milli9ram8/milo)

-------
                                    -36-
I   553
 B
 O
-8
 u
520
453
 a
 D7
 O
 b

     250.
                                          FIGURE 6
                                      STREET CANYON
                                                                  «	1	j	«
                            EMISSION FACTOR

-------
                                      -37-
       653 .
 £.
 O
 -P
 0>
 B

 O
 •t-*
_Q

 O

 (0
 O

 O
453.
                                           FIGURE 7

                                          EXPRESSWAY
                              EMISSION FACTOR (milli3ram9/mile)

-------
                                      -38-

                                   TABLE XI
                    Emission Factors Required to Result in
                 Two Different  Ambient  Sulfuric  Acid  Levels***

Ambient Situation*                Emission Factor     Emission Factor
                                     mg/mile for        rag/mile for
                                     66 ug/m3            1000 ug/m3
                                     exposure	       exposure	

Street Canyon - Typical                1540                 23077
Expressway - Typical                    619                  9375
Expressway - Close Proximity            388                  5882
Street Canyon - Severe                  165                  2500
Expressway - Severe                     132                  2000
Personal Garage - Typical**
Parking Garage - Typical**
Roadway Tunnel - Typical                 59                   822
Roadway Tunnel - Severe                  22                   350
Parking Garage - Severe**
Personal Garage - Severe**
*   In order of increasing ug/m3 concentration for 1 g/mile
    (or 1 g/min) emission rate.

**  These  situations  were  not  evaluated  for  sulfuric  acid  because  of  an
    inadequate  data  base-  for  emission  factors  under  idle  and  low  speed
    conditions.

*** If the severe roadway tunnel situation is  of primary  interest  then a fleet
    emission factor of 22 mg/mile over an appropriate  driving  schedule will be
    enough  to  put  the  ambient concentration within  the range  of  concern.
    However,  if expressway operation  is of  primary  interest,  then  emission
    factors of  up to  132 mg/mile would yield  ambient  concentrations  below the
    range of concern even for severe conditions.

-------
                                      -39-


Summary and Conclusions - Sulfurlc Acid

1)  The  range of  concern  for sulfuric  acid  emissions from  motor  vehicles
    varies from 22-350 mg/mile to  1540-23077 mg/mile depending  on  the  scenario
    and situation of interest.

2)  The lowest level  of this  range (22 mg/mile)  is based  on  an ambient  air
    concentration of 66 ug/m^ for a severe roadway tunnel situation.

3)  No vehicle emission  factors  from the garage scenarios were  considered   in
    calculating  the  range  of  concern.   More  data  is  probably  needed   on
    emissions of sulfuric acid from light duty  vehicles under  idle   and  slow
    speed conditions to  evaluate  the effects  of the garage  situations on  the
    sulfuric acid range of concern.

4)  The roadway  tunnel  scenario  appears  to be  a controlling  factor  in  this
    methodology for the  sulfuric  acid  case.   There is some  doubt whether  this
    scenario  identifies  a  potential  mobile   source  problem  or  a potential
    roadway   tunnel   ventilation   problem.   If   a   potential   problem  were
    identified  resulting  from  roadway  tunnel  exposures   to   mobile source
    pollutants, then  it  is  possible that the most appropriate   solution  would
    be  to increase  tunnel  ventilation rather  than  to  reduce  the   vehicle
    emissions.

5)  This  preliminary  example of  sulfuric  acid has not  considered a  specific
    margin of safety.   It is possible  that the inclusion of  the  roadway tunnel
    scenario  as  the  controlling factor in the range  of concern consitutes  a
    margin of safety in view of conclusion 4 above, but  no specific  factor  has
    been calculated.

6)  The  current  vehicle fleet   emission  factor  for  sulfuric  acid   is  12
    mg/mile,   which  is well  below  the lowest  of the  ranges  of  concern  for
    sulfuric acid.

7)  With  respect to  specific vehicle  emission  control  designs, and referring
    to  Table  IV,   it  appears  that  the  emission  control design/   vehicle
    categories that  have emission  factors  most  often   appearing  within  the
    ranges of concern are Heavy  Duty  Trucks  and Light  Duty Diesel  vehicles
    with trap-oxidizers  (100 mg/mile).

-------
                                      -40-

                                  References
(1)   Analytical   Procedures   for   Characterizing   Unregulated   Pollutants
      Emissions from Motor Vehicles,  EPA Report No.  600/2-79-017.

(2)   Analytical  Procedures  for  Characterizing Unregulated  Emissions   from
      Vehicles Using Middle Distillate Fuels,  EPA Report  No.  600/2-80-068.

(3)   Regulated  and Unregulated  Exhaust  Emissions  from  Malfunctioning   Non
      Catalyst  and  Oxidation  Catalyst Gasoline  Automobiles, EPA  Report   No.
      460/3-80-003.

(4)   Regulated   and   Unregulated   Exhaust   Emissions   from  Malfunctioning
      Three-way Catalyst Gasoline Automobiles, EPA Report No. 460/3-80-004.

(5)   Regulated  and  Unregulated  Exhaust  Emissions  from   a  Malfunctioning
      Three-way Catalyst Gasoline Automobile,  EPA Report  No.  460-/3-80-005.

(6)   Air  Quality  Assessment  of  Particulate  Emissions  from  Diesel-Powered
      Vehicles, Pedco Environmental,  Inc.,  March 1978.

(7)   Mobile  Source Emission  Factors:   For  Low Altitude  Areas  Only, March
      1978, EPA Report No. 400/9-78-006

(8)   Multimedia  Environmental  Goals  for  Environmental   Assessment  Volume  1,
      EPA Report No. 700/7-77-136a.

(9)   Emission  of Sulfur-Bearing  Compounds  from Motor  Vehicle  and Aircraft
      Engines, EPA Report No. 600/9-78-028.

(10)  Regulated and Unregulated Emissions from Malfunctioning Automobiles,  SAE
      Paper 790606.

(11)  Emissions   from   Light  and   Heavy   Duty  Engines,    EPA   Report    No.
      460/3-79-007.

(12)  Exhaust Emissions from Malfunctioning Three-Way Catalyst-Equipped
      Automobiles, SAE Paper 800511.

(13)  Light Duty Diesel Catalysts,  EPA Report  No.  460/3-80-002.

(14)  Automobile Sulfate Emissions -  A Baseline Study,  SAE Paper 770166.

-------