EPA/AA/CTAB/PA/81-18
                        Summary of Current Status of
             EPA  Office  of Mobile Source  Air Pollution Control
                         Characterization Projects
                              Thomas M. Baines
                                August, 1981
                                   NOTICE

Technical  reports  do   not  necessarily  represent  final  EPA  decisions  or
positions.  They  are intended  to present technical analyses 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.
                    U.S. Environmental Protection Agency
                     Office of Air, Noise and Radiation
                Office  of  Mobile  Source Air Pollution Control
                    Emission Control Technology Division
          Control  Technology Assessment and Characterization Branch
                             2565 Plymouth Road
                         Ann Arbor, Michigan  48105

-------
                                      -2-

                                Table of Contents


                                                                      Page

I.  Overview	        3


II. Summary and Conclusions  	        6


III.Recommendations for Future Work  	       12


IV. Characterization Results 	       15

    A.   Heavy Duty Diesels	'     15

         1.   Normal Operation 	       15
         2.   Malfunction Conditions 	       21

    B.   Light Duty Diesels	       23

    C.   Light Duty Gasoline Vehicles   	       23

    D.   Fuels	       31

         1.   Fuels Variables	       31
         2.   Alternate Fuels	       36

              a.   Middle Distillates Derived From Alternate Sources   36
              b.   Gasoline Derived From Alternate Sources ...       40
              c.   Alcohol Fuels	       40

    E.   Nitrosamines	       43

    F.   Identification of Types of Compounds Responsible
         for Ames Test Activity in Diesel Particulates 	       45

    G.   Development of Method to Collect/Characterize
         Gas-phase Hydrocarbons  	       45

    H.   Ames Testing	       47


V.  References	       54


VI. Appendixes	       58

-------
                                     -3-
I.  Overview

The Characterization Program  of  ECTD has focused on .expanding  the knowledge
of pollutants  as  they are  emitted  from various  mobile sources.   This  work
has been  done  within  ECTD,  by in-house and  extramural programs, as  well as
monitoring  the characterization  efforts  performed by other  organizations
both within EPA and by industry and others.   Some  of  the  principal  guiding
objectives  of  the  ECTD  characterization  program can  be  summarized by  the
following points.

1.  The  characterization of  pollutants not  normally  tested  and  that  may
    represent a human health concern.

    Currently, and  in  the past,  a large amount  of  effort  has  been expended
    by industry and EPA  characterizing the hydrocarbon, carbon monoxide  and
    oxides  of  nitrogen  emissions  from a  variety  of  engines  and  vehiclesi
    However, there may be other  compounds  being  emitted by  vehicles that  may
    be hazardous.   Especially of interest would  be those compounds that  may
    have  a deleterious  effect  on  human  health.  Some  of  these  compounds
    would  be  emitted  in  varying   amounts   from uncontrolled as  well  as
    controlled  engines.    Other  compounds   (such  as  catalyst   attrition
    products)  could be  emitted  mostly   from vehicles  which  have emission
    control systems designed  to lower  HC,  CO and  NOx.  Various  systems to
    improve both  emissions and  fuel economy could  have a large  impact on
    unregulated emissions.   Since motor  vehicle technology  is changing so
    rapidly now  in  response  to  the need for  improved  emissions and  fuel
    economy,   it   is   critical  that  ECTD   characterize   new   systems   for
    unregulated pollutants.  Consequently, the ECTD program has been  focused
    on  characterizing  a  broad range  of  compounds from present and  future
    engine and vehicle "technologies.

2.  Testing for a variety of pollutants under malfunction conditions.

-------
                                      -4-
    Much,   if  not  most,  of   the   testing  performed  by  EPA  and   other
    laboratories has  been done with  the  engines  tuned  to  manufacturers'
    recommended  specifications.  However,  many vehicles  that  are  in  use
    today  operate  under   conditions  of   tune  that  do   not  meet   the
    manufacturers'   recommended  specifications.   This   could   result   in
    increased  emissions  of  a  variety of  both regulated  and   unregulated
    pollutants, some of which  could have  negative human health effects.   As
    a  consequence,   ECTD  has  tested  a   variety  of  engines/vehicles  for
    pollutants  of   concern under  malfunction  conditions  to estimate  the
    impact that such  vehicles/conditions would  have on  the environmental
    loading of pollutants.

3.  Fuel parameters.

    The fuel  situation in the  United  States is  currently very  dynamic,  in
    that we  are rapidly developing alternate sources of fuel to  supplement
    conventional petroleum sources.   These alternate  sources include  lower
    grade  petroleum crudes,  as  well  as  alternative  and  synthetic  fuels
    derived from coal and oil shale.  Some of the new fuels  will  be  alcohols
    including  methanol.   These newer  fuels may have a  dramatic effect  on
    emissions  and,  as  such,   ECTD  has performed  some  characterization  on
    these  emissions  as well  as remained  abreast  of  the  field  in  general.
    Also,  some  testing  has been done on  emissions  from Diesel vehicles,  as
    these  emissions may  be  impacted  by  fuel  parameters.   The  ECTD  fuels
    characterization  program  will   continue  in an  effort  to   more  fully
    characterize the  future  fuels.   This  work  is  of  critical importance  in
    that   it  helps   assure    that  alternate   fuels  are   environmentally
    acceptable.  One  can  probably tailor fuel  composition and processes  to
    obtain maximum  environmental  benefit  from  these  fuels if one does  this
    characterization before these 'fuels are widely produced.

-------
                                       -5-
4.  Characterization of  pollutants from  engines/vehicles  that are  involved
    in a transition environment.

    There  are  many engines/vehicles   that  are  involved  in  a  transition
    environment  created  by  various  regulatory  initiatives, .fuel  economy
    incentives,  etc.   ECTD   is   very   interested  in  characterizing   the
    emissions from  these  vehicles/engines to be able to evaluate  the impact
    that  this  transition may  have.    For  example,  the  heavy-duty  Diesel
    engine manufacturers  are  now  currently  changing many  of their  engines
    from the  traditional,  naturally aspirated type over to  the  turbocharged
    type.   In  an  effort  to  evaluate  this  change  and  how  it  affects
    pollutants,  some  testing  has  been  done on  equivalent  engines.   Also,
    there is  currently  a  trend towards Dieselization of both  the  light-duty
    fleet  as  well  as   the  mid-range  heavy-duty  fleet.    Comparative
    application engines for both of these  fleets have been tested  so  that an
    estimate  can be made  of how such  a change will  impact the environmental
    loading of  pollutants.  Also,  a variety of other technologies  have  been
    evaluated so that  their  influence  can also be estimated.   These  include
    the  influence  of  indirect injection  in  a heavy-duty Diesel engine,  the
    influence of high pressure injection in a Diesel  engine,  and others.

    With these  objectives in  mind, a variety  of  programs  and  projects  have
    been performed.   The more  recent  and more  important  of  these  projects
    are summarized in the following section.

-------
                                       -6-

II. Summary and Conclusions

EPA-OMSAPC   is   conducting  a   thorough   characterization  of   unregulated
emissions on a variety  of  current and prototype engines.  Extensive  work is
also  underway  to  see  how  various  fuel  parameters  affect  regulated  and
unregulated  emissions.   This  latter  work  includes  projects  on  alternate
fuels such as methanol as well as fuels derived from coal or oil shale.

Based on the recent work the following conclusions can be made.

1.  The  effects   of  a  number  of  engine   parameters   on  regulated  and
    unregulated  emissions  were  examined  for  several  heavy   duty  Diesel
    engines.  It was  found for  two Daimler-Benz engines  that  turbocharging
    generally decreased all emissions with the major  exception being NOx,
    for which a slight  increase was noted.

    It was found on a Caterpillar  3406 engine  that  advanced  injection timing
    increased   HC,   CO,   NOx,   sulfates,   and   aldehydes  while   reducing
    particulates and  extractable organics.  It  was found on a  Caterpillar
    3406 engine that changing  from a direct to  an  indirect  injection system
    increased  HC,  brake   specific  fuel  consumption,   and   sulfates  but
    decreased CO, NOx,  particulates,  organic extract, and  odor.   With a Mack
    ETAY(B)  673A  engine, water  injection (30%  volume)  was  found  to reduce
   . NOx slightly,  increase HC and  CO slightly, but  increase  particulates and
    extractable organics greatly.

    Different fuel pumps and  the effect  of EGR  on  regulated  and  unregulated
    emissions were also examined.   The use  of a high pressure pump  reduced
    particulates greatly but increased NOx.   Use  of EGR  decreased  NOx but
    increased particulates.

2.  The  soluble organic  emissions  from  four  different  heavy  duty  Diesel
    engines  were   compared  with  those from a  heavy  duty  gasoline  engine
    burning  leaded fuel.   The   Diesel  engines  showed  a  higher  percent  of
    organic extractables from the  particulate  than  the gasoline engine.  The
    percent extractables decreased  greatly  as  load increased for the Diesel
    engines.

-------
                                      -7-
    Also,  the  gasoline  engine  was  compared  to  a  Diesel used  in  similar
    applications.  The Diesel engine  had  much  lower HC and CO with  slightly
    higher  NOx  emissions.   The Diesel  emitted a  much  higher  quantity  of
    particulates  and  percent  extractables  but  had  somewhat:   lower   BaP
    emissions.  Sulfate  emissions  were greater but  aldehydes  were less  for
    the Diesel engine.

3.  Limited testing  has been done  on Diesel  bus  engines under  malfunction
    conditions which  would  simulate  emissions  under in-use conditions where
    engines are  frequently  not  tuned up  to manufacturers'  specifications.
    The one engine tested so far shows an increase  in HC,  CO,  brake  specific
    fuel  consumption,  and  a  large  increase  in  both particulates   and
    extractable organics.  A second engine is being tested.

4.  Work has been done  comparing  emissions  from an Oldsmobile and VW Diesel
    with their  gasoline counterparts.  These  tests showed generally  higher
    HC,  much  higher  particulates,  higher BaP,  higher  sulfates but lower  CO
    from  the  Diesel  versions.    There was  no  clear   trend  with  aldehydes
    although  the  aldehydes were usually  higher with  the Oldsmobile  Diesel
    versus gasoline engine.

5.  EPA  has conducted extensive tests on four oxidation and  four  three-way
    catalyst  equipped gasoline  vehicles  under  malfunction conditions.   The
    malfunction  conditions  tested   include   rich-best-idle,  12%  misfire,
    disabled  EGR,  higher oil consumption,  and faulty  air injection.   Both
    the  oxidation and three-way catalyst-equipped  vehicles generally  showed
    higher regulated and unregulated emissions.

    These  results were  compared  with  data  from  a  non-catalyst equipped
    vehicle.

    It appeared  as  though sulfate, ammonia, and  nitrous  oxides were  higher
    from  catalyst  equipped  cars  than  from  the  one   non-catalyst   car.
    Hydrocarbons  and  aldehydes  were  highest  from  the   non-catalyst  car.
    Ammonia and  cyanides  were highest from the three-way  catalyst cars.   It

-------
                                       -8-
    appeared  that  the   unregulated   exhaust  emissions  composition   from
    catalyst equipped cars operating under malfunctioning conditions  was,  in
    general, distinguishably different from that of the non-catalyst cars.

6.  EPA has also  conducted  a program to  examine  how Diesel  fuel  parameters
    affected emissions  from typical  light  duty Diesels  (a  VW and  Mercedes
    Benz 240D).  Five different Diesel fuels were  tested.  It  was  found that
    the minimum quality  fuel  frequently  adversely affected  emissions.   A
    follow-on project was conducted in which  very specific  fuel  parameters
    were varied  to determine  their  effect on  emissions.   This work showed
    that phenol emissions were strongly affected by the shale oil blend.

    Particulate rates were  reduced with  the fuel  containing the  light-end
    blend and  the  olefin blend,  by 14  percent  and 6  percent,  respectively.
    The two blends containing aromatics  were  associated  with increases  in
    particulate emissions of about 27 percent.  The  cetane-improved aromatic
    fuel blend  increased particulate  emissions about  68  percent over  those
    observed with  base  fuel.  Heavy  ends  fuel apparently did not cause  an
    increase in particulate emissions.   Also,  blends containing  aromatics
    resulted   in   slightly  less   organic  solubles   extracted  from   the
    particulate matter.   Olefins,  light ends,  and  heavy  ends of  fuels  were
    associated with increases in the amount of organic  solubles.

7.  EPA  has just started  a  program  to   investigate  how  alternate  fuels
    (including  fuels  derived  from coal  and  oil  shale)  affect  emissions.
    Four fuels  (a #2 Diesel fuel,  #2 Diesel fuel  marine  from oil shale,  a
    Paraho  JP-5 from  oil shale,  and a Coal Case  5A fuel  (a  fuel  containing
    solvent  refined  coal (SRC  II) and  petroleum materials  blended by  the
    U.S. Army)) are  being tested with a VW  Diesel.   There was virtually  no
    change  in HC, CO and  NOx emissions with the shale  marine  fuel.   This was
    also  true  of  the  particulate  emission  rate.   The aldehydes,  however,
    decreased  slightly.   The phenols  stayed much  the   same,  as  well as  the
    odor.   There  was  a  slight  increase in  the  fuel economy.   In looking  at
    the Paraho JP-5, a  slight  increase was  seen in HC,  CO,  NOx as well as  a
    slight  decrease in  fuel  economy.  The  particulate was unchanged,  as well
    as aldehydes  and  odor while phenols decreased  somewhat  with use of  the

-------
                                       -9-
    JP-5.  The coal  case  resulted  in a rather  dramatic  increase in  HC,  CO,
    NOx  and  particulate,  on  the order  of approximately  50 percent.   Fuel
    economy also increased,  and phenols and odor went up  slightly.

    A  solvent  refined coal  (SRC-2)  fuel  is  currently being tested.   While
    EPA has tested only Diesel  fuels  to  date,  EPA plans  to  test  a  Mobil  MGT
    process  gasoline  (derived  from  methanol  made  from  coal)  and  other
    gasolines as they  are available.   EPA will  also test more  Diesel  fuels
    as they become available.

8.  EPA  has tested a heavy  duty Volvo Diesel  bus  engine designed  to run on
    methanol  and  Diesel  fuel, each injected  through  its  own  injection
    system.  The emission results  from this engine were compared with  those
    from a conventional Volvo  Diesel  engine using regular Diesel fuel.   The
    use  of  the dual-fuel  resulted  in a 50%  reduction in total particulate
    emissions compared to the  single  fueled engine.   The particulate  emitted
    from  the  engine  using  methanol  had  70%   organic  extractables  from  the
    particulate, indicating  far less  elemental carbon type  substances  being
    formed compared  to the  conventional  Diesel fueled  engine,   where  about
    30%  organics were found.   Less sulfate was also  produced  due to  lower
    fuel  sulfur  levels.    It   was  also   found  that  the  use  of   methanol
    decreased  NOx  by  56%  for   the  13 mode test  and  35%  for the  transient
    cycle.

    The  use  of  methanol  resulted  in a hydrocarbon  emissions   increase,  as
    measured by the heated FID, of about 40% for the  13  mode and 70%  for  the
    transient  cycle  compared  to the Diesel fueled engine.   Large quantities
    of unburned alcohol (which  is only partially measured  by the heated FID)
    were  emitted.   In one  case,  the amount  of unburned  methanol  was  four
    times  the   amount of   hydrocarbons.    Aldehyde  emissions  were   also
    increased with the use of alcohol.  The use of methanol  resulted  in a CO
    increase  of about  60%.   It  was  found   that  the use  of  an  oxidation
    catayst reduced  both  hydrocarbons and CO  but some  increase in  NOx  was
    found.  The catalyst did not reduce particulate levels.

    EPA  is  currently  testing  a Ford Escort  with Anafuel  (gasoline  with  a
    methanol based additive)  and plans  to test a VW  designed . to  burn  100%
    methanol fuel.

-------
                                      -10-
9.  EPA has also conducted work measuring nitrosamines  (a  potent carcinogen)
    in both Diesel  crankcase  emissions and vehicle  interiors.   Nitrosamines
    were found  in Diesel  crankcase  emissions, possibly  resulting  from  oil
    additives.   Nitrosamines  have  also  been  found in  vehicle  interiors.
    Finally,  nitrosamines  may be present  in  Diesel exhaust, although  this
    possibility   is   still   unconfirmed.    Preliminary   data   show   that
                                            3
    nitrosamine  levels  may be high  (1 jig/m ) along roadways arid apparently
    these nitrosamines  are mobile source related.   It is not  known if  the
    Diesel  crankcase  and  vehicle nitrosamines  are  entirely  responsible  for
    the roadside nitrosamine levels reported.

10. Various research  projects,  including some  at EPA,  have  identified  some
    of  the  compounds  responsible  for  the  Ames   test activity  of  Diesel
    particulates.   Apparently,  polynuclear  aromatic hydrocarbons  themselves
    are  responsible  for  only a  small  portion of  the activity.   However,
    various oxygenated  polynuclear aromatic hydrocarbons  (including hydroxy,
    ketone,  carboxaldehyde,  quinone,  acid  anhydride,  and nitro  compounds)
    have  been identified  in  the "transition"  fraction  that  accounts  for
    significant Ames  test activity.

11. EPA is  currently  developing a method to collect the gas phase  HC in  the
    exhaust  for  bioassay  tests.   A  promising method being developed by  ORD
    appears  to  be a  cleaned-up XAD-2  resin.   The  hydrocarbons  collected  by
    this method  may not have much Ames  test  activity  compared with  that  of
    the  particulate.    However,  OMSAPC  has  run some  tests  with a  back-up
    filter  containing treated Diesel  particulates  which  shows  that the  HC
    collected does have Ames test activity that could be significant.

12. A  number  of Ames  bioassay  tests  have   been   run  on  various  samples
    collected  by  OMSAPC   recently.   Ames  test work   on  the  Diesel  fuels
    project (discussed  in Conclusion 6)  showed  that  an  increase  in  Ames  test
    activity  is  associated with  fuel  aromatics.   An  increase  in Ames  test
    response  was also  seen when  the  cetane  improver  was used.   Ames  test
    work  on  samples  collected  from  the  dual  fueled Volvo  Diesel engine
    (discussed in Conclusion 8) indicated that  the  particulate from the dual
    fueled engine had lower Ames  test activity than did the  baseline engine

-------
                                  -11-
(i.e. pure  Diesel fuel)  over the  transient cycle  but higher  activity
over the 13  mode  cycle.   This apparent  discrepancy  may be explained  by
the  fact  that  the  dual-fueled  engine  showed  much  higher  Ames  test
response for heavily versus  lightly  loaded modes.   The 13 mode  cycle  is
considered to  be  more heavily  loaded than  the  transient cycle.   There
are also some data  showing that the presence  of  a catalyst resulted  in
increased Ames test activity.

-------
                                      -12-

III. Recommendations for Future Work

The ECTD recommendations for future work are the following:

1.  Synfuels.  With  the  dramatic change in petroleum and  fuel sources about
    to  occur,  it  is  extremely  important  that EPA  remain  abreast  of  the
    developments in  this area.   This  is especially important wi'th respect to
    the  influence  that these fuels may have on exhaust  emissions products.
    There  is currently  a  large  level  of  effort  on-going  in  the area  of
    alternate  fuel  plant  source emissions.   However,  very  little work  is
    currently  planned   on  ultimate   use  or  in-use  emissions   from  the
    combustion of  these  fuels  in vehicles.   The fuels  that should be tested
    in vehicles include  coal liquids  such as refined SRC-2,  refined EDS and
    H  coal  liquids  (made from  direct  coal liquefaction).   The  shale  oil
    liquids  should include those  that may be available  through the Union Oil
    process.  Also it  will be important to analyze different types of shale
    products, such as  those  made by in-situ retorts and  other such methods.
    Other  fuels that will  be important  to  characterize include  coal derived
    methanol, as  well as  actual  methanol  fuels (blends)  from non-petroleum
    sources.    Such  alcohols    should   be   looked    at   as   well   as
    alternative-source  fuels.   Also,  gasolines should be  tested, such  as
    those  from the Mobil MGT process  as well as Naptha  fractions,  and others
    that   are  currently   envisioned   to be  produced.   By  analyzing  the
    emissions  from  these  fuels,  EPA  will then  be able  to estimate  the
    influence that they will have on ambient air quality.

2.  New  Technologies.   As new technologies  are  developed  it will be  very
    important  to  remain abreast  of .the pollutants  that  may be  emitted  by
    them.   Many  technologies  that have been  introduced  in  the  past  have
    resulted in lower  quantities of some pollutants.   However,  some of them
    have  also  resulted in  increasing  some  pollutants  that have   been  a
    concern.  It is  important  to remain abreast of these  new techologies  so
    that  if  any  pollutants are  emitted  that are  hazardous  to human health,
    an appropriate assessment of  their  impact  can  be made.  New technologies
    might  include  the Ethyl lean  burn  system, various Diesel systems,  such
    as  the turbocompound  and  ceramic adiabatic engine or  possibly various
    Diesel particulate  trap  technologies.   New systems will  be  developed  in
    the continued desire for improved fuel economy.

-------
                                      -13-

3.  Bioassay.  It  will be  very  important to  continue to  analyze  emissions
    from  a  variety  of  vehicles  with  bioassay  techniques.   For  example,
    emissions  from locomotives,  possibly  aircraft,   motorcycles  and  other
    such  sources  that  have never  been analyzed should be  given at  least  a
    preliminary bioassay  characterization.   Also,  there is  evidence showing
    that  certain  fuel  additives appear  to  increase biological  activity  and
    this should be further investigated.

    To  date, ECTD  has relied  heavily  on  the Ames   test  to more  or  less
    routinely  test  Diesel particulates  and  other  samples.   While  the  Ames
    test  has  been  a  useful   tool   in comparing  and  investigating  the
    bioactivity of various  systems, the  Ames  test does have some limitations
    (e.g.  it  may  not   respond   accurately   to  some  nitrogen  containing
    compounds).   Future  efforts should  involve helping  develop  and use  a
    simplified tier bioassay test  system so  that more  accurate bioassay test
    data are obtained.

4.  Malfunction conditions.  Additional  work should be done  especially with
                                                  •V           \
    Diesel engines under  conditions outside  of  thos,e  normally ^recommended by
    the  manufacturer.   By  analyzing  emissions  frjpm vehicles  under  these
    conditions,  a more  realistic  view  of  the pollutant  burden emitted  by
    such vehicles can be obtained.

    This is  especially important for  heavy  duty Diesel buses which operate
    in  urban environments  where  the  Diesel  emissions can  come in contact
    with large  numbers  of  people.   This is  also  important  for light  duty
    Diesels  since  very  little  work  investigating  their  emissions  under
    malfunction conditions has been done.

5.  Gas  phase  analyses.   EPA-ORD  is currently completing the  development  of
    a gas  phase sampling technique.   ECTD  should perform some  testing  with
    this  technique  to  more adequately  and  fully  characterize  the  gaseous
    emissions from  a  variety of different  mobile sources.   In  this manner,
    EPA  will  be  able  to   obtain  a  thorough  estimate  of  the   potential

-------
                                      -14-

    impact of emissions  from  these mobile sources.  This sampling  should be
    performed on both  light and heavy duty engines, including  both gasoline
    and Diesel engines.

6.  Specific pollutants.  In  the  past,  there has  been some testing done on
    specific  pollutants   that   are   potentially   very  hazardous.    Some
    additional  work  should   be  performed,  for  example,  in  the  area  of
    nitrosamines.  Nitrosamines should  be analyzed  in Diesel  exhaust  where
    there  is  a  tentative indication  that they  may be  found.   Also,  there
    should  be  some   additional   nitrosamine  monitoring  along  roadways  to
    confirm the preliminary evidence  of  their  existence there.  Also  in the
    past there have been specific  compounds, such as dioxins, that  had  to be
    analyzed in a  variety  of  exhaust sources.   As  these "pollutants of  the
    month" arise, EPA  should  be in a  position  to respond in a  rapid fashion
    and be able to analyze them.

-------
                                       -15-

IV. Characterization Results

A.  Heavy Duty Diesels

1.  Normal Operation

The  more  recent  work  with  heavy  duty  Diesels  has  been  done  under  EPA
contract Number  68-03-2417,  wherein a number  of different  engine  variables
were  analayzed  for  their   effect  on   engine   pollutant  emissions.    The
variables  which  were  investigated included  effects  of 1) turbpcharging,  2)
exhaust  gas  recirculation  (EGR),  3)  changes in  timing, 4) combustion system
(direct  or indirect injection),  5)  fuel  injection  pressures,  and  6)  water
injection.  The data from this study of  these  six variables  are presented in
Table 1* and are discussed in the following paragraphs (1, 2  and 3).**

The  engine which was used  to determine  the effects  of turbocharging  was  a
Daimler-Benz OM-352 and  the turbocharged version, the  Daimler-Benz  OM-352A.
Turbocharging, it is noted, decreased almost all of  the pollutants  for which
analyses  were  made.  The  major  exception  to  this was NOx,  which  showed  a
slight  increase.   Aldehydes  were almost unchanged,   and  the brake  specific
fuel consumption decreased.

The  effects  of  timing  were determined by  testing a  Caterpillar  3406 direct
injection  engine.   By  advancing the timing,  increases were observed  in  HC,
CO,  NOx,  brake  specific  fuel  consumption,  sulfate  and a  very  dramatic
increase in aldehydes.  The amount of particulate was reduced as  well as  the
percent  of  extrac table  organics.   By  retarding   the   timing.,  a  slight
reduction  in HC was observed, as  well   as  a very slight  increase  in  CO,  a
rather  dramatic  decrease in  NOx,  and a slight  increase  in brake  specific
fuel consumption.   The  particulate  rate  increased rather  markedly (i.e.,  by
a  factor  of  almost  three),  the  percent  of  extractable  organics  also
increased, as did the BaP  emission  rate  and the  sulfate  emission rate.  The
odor level decreased;  however, aldehydes.increased by almost  a factor of two.
* The data in this and other tables are usually based on single emission tests.
**Numbers in parentheses designate references given at the end of the paper.

-------
                                                                  Table 1
IIC, g/kW-hr
CO. g/kW-hr
NOx, g/kW-hr
BSFC. kg/kW-hr
Participate, g/kW-hr
Organlcs, X
BaP, ug/ku-hr
Sulfate, og/kU-hr
Odor, TIA
Aldehydes, mg/kW-hr
Effect of Varying Engine Parameters on Pollutant Emissions
Daimler-Benz Caterpillar 3406 DI Caterpillar 3406
OM-352
Naturally
Aspirated

3.20
7.00
9.98
0.294
1.33
34.2
1.43
19.0
2.2
375.0
OH-352A
Turbo-
chargcd

3.02
4.08
11.95
0.284
0.75
29.2
1.17
18.7
2.32
372.0

Standard
Timing
\
0.47
3.13
13.09
0.253
0.47
17.9
0.20
28.5
2.0
97.9
5X
Advanced
Timing

0.53
5.63
18.53
0.262
0.36
12.9
1.26
34.4
2.0
326.3
10Z
Retarded
Timing

0.43
3.16
7.02
0.273
1.36
19.2
0.69
37.2
1.8
165.1
Dl - Std.
Without
EGR

0.47 ••
3.13
13.09'
0.253
0.47
17.9
0.20
28.5
2.0
97.7
Timing
With
EGR

0.23
6.43
7.35
0.268
1.24
19.5
0.11
28.7
1.7
107.0
- 13 Mode Cycle U
Caterpillar
3406 DI
Direct
Injection

0.47
3.13
13'. 09 '•-.'.
0.253
0.47
17.9
0.20
28.5
2.0
97.9
3406 ID1
Indirect
Injection

, 0.16
1.68
: 6.88
0.272
0.37
11.1
0.14
41.0
1.4
106.9
                                                                                                                   MACK ETAY(B) 6/3A
                                                                                                                   Standard
                                                                                                                   R. Bosch
                                                                                                                   Pump(APE)*
APS R.
Bosch Pump
111 Pressure
0.64
2.13
8.87
0.243
0.821
16.3
0.23
A4.9
1.8
87.0
0.69
1.08
12.06
0.234
0.40
16.9
0.08
42.8
1.7
105.0
  HACK ETAY(B) 673A

Standard   Standard
Bosch      APE Pump
Pump(APE)* + 30^ H20
                0.60
                1.64
                8.52
                0.240
                0.71
               40.0
                0.18
               42.2

               49.0
            0.62
            1.
            6.
  .75
  .49
 0.239
 0.30
49.6
 0.22
39.3

53.0
   These were 2 separate experimental runs  using  different APE pumps.
   Thus, the data are not necessarily the same.

-------
                                      -17-
The  effect  of  EGR was  tested  with  the Caterpillar  3406 direct  injection
engine with  standard timing.  By  applying  EGR,  a slight  increase in  HC,  a
doubling of CO, a  50%  reduction  in NOx, a slight increase  in brake specific
fuel  consumption   and  a  2.5  fold  increase in  particulate  with  associated
increase in the percent  of  extractable  organics  were  observed.   The BaP went
down  almost  by factor of  two,  the sulfates remained  the same, there  was  a
slight increase in odor and aldehydes were virtually unchangedi

The   effect   of  direct   and  indirect  injection  were   studied   with  the
Caterpillar 3406  engine,  which  was tested  in  the direct  injection mode and
then  changed  to the indirect injection  mode.   With this  change  to indirect
injection,  a  dramatic decrease  was observed  in HC,  a  decrease  in CO,  and
again  a  decrease  in  NOx.    Brake   specific  fuel   consumption  increased
somewhat; however,  the particulate rate decreased as well  as the  percent of
extractable  organics.    The  BaP  rate  also  decreased  but  the sulfate rate
increased somewhat.   There was  a rather marked decrease  in odor; however,
aldehydes were virtually unchanged.

The  effect  of  the  pump  system  was  investigated with  a  Mack ETAY(B) 673A
engine, wherein the  engine  was tested with  a standard  Robert  Bosch fuel pump
and  then  with a very  high pressure pump.   The  results  showed a  slight in-
crease in HC,  a decrease in CO,  a slight increase  in NOx, and a  slight de-
crease in brake specific fuel consumption.  The  most  dramatic  influence was
the  approximately  50%  reduction  in the  particulate  emission rate  with almost
no  change  in  the  percent of extractable organics.   The  BaP rate with the
high  pressure  injection  was 1/3  that obtained with the  normal  pump.   Almost
no  change  in sulfate, a slight  decrease in odor,  and a  slight increase in
aldehydes were  seen.

The  same Mack  engine was used  to analyze the effect of water injection on  a
variety of  emissions.   This testing involved  the injection of  approximately
30%  water along with the fuel  (DF-2)  and here a  slight  increase was  seen in
HC  and  CO,  a  slight reduction  in NOx,  and  brake specific fuel  consumption

-------
                                       -18-
was  almost unchanged.  However,  there was  a  very dramatic  decrease  in  the
particulate  emission rate  with an  accompanying  increase  in  the percent  of
extractable  organics.  The BaP  rate also increased;   however,  it was  noted
that most  of this increase was  at idle conditions where presumably  such  an
increase  could be reduced  by modulating water  injection to eliminate  water
injection  at  idle   speeds.   The  aldehyde  emission  rate was  virtually  un-
changed.

Table  2  presents the percent organic  extractables  for particulate from four
Diesel  engines  and  one gasoline  fueled engine  burning  leaded  fuel.   This
showed a number  of things  including  the trend towards much higher percent  of
organic  extractables from  the  Diesel  relative  to the  gasoline  engine.   It
also  showed  that  there was  generally a  dramatic decrease  in  the  percent
extractables  as  the  load  on the  engine increased  for  the  Diesel  engines.
The  medium duty  fleet  of  trucks is  currently  undergoing  a  rather  marked
conversion  from  gasoline  to Diesel  engines.   In  order to  assess  how such  a
conversion  may influence  regulated  and unregulated  emissions,  a  number  of
comparsions  can  be  made  between  two  engines  of  comparable  applications  in
heavy duty trucks.   These data  are presented in Table 3 for two engines; one
a Caterpillar  3208 which  employed EGR  and which  was  designed for California
NOx  standards,  and   a Chevrolet  366  CID gasoline-fueled  engine  designed for
the  same  1977 California  standards.   Here,  much  lower hC and  CO emissions
were generally seen  with  the  Diesel engine along  with  slightly  higher NOx
emissions.   Much lower  brake  specific  fuel consumption  was  seen  with the
Diesel  engine.   However,   the  Diesel  engine  emitted  over   12  times  more
particulate than the corresponding gasoline  engine as well as  a  much Higher
percent of organics  extractables.  The  BaP emissions,  however,  were  somewhat
lower  from the  Diesel  engine than  from  the  gasoline engine.   Sulfate was
much higher,  due,  of course,  to  the  much  higher level of  sulfur  in the
Diesel fuel.  The aldehydes from the Diesel  engine were  lower  by  a factor of
approximately 10 compared to the gasoline engine.

-------
  Condition
Speed/Load.(%)

Inter /02

Inter /50

Inter/100

Idle

Inter/CT*

High/100

High/50

High/02

High/CT*
-19-
Table 2
Percent Organic Extractables
From Particulate For
Four Diesel And One Gasoline
Fueled Engines
MACK
ETAY(B)
'673A
34.8
9.3
3.5
15.5
NT@
1.0
6.3
38.5
NT
Cater-
pillar
3208 EGR
17.0
6.7
2.7
8.2
23.3
2.7
2.2
9.4
23.7
Cater-
pillar
3406IDI
12.9
4.5
5.1
9.9
NT
0.8
11.7
35.9
NT
Daimler
Benz OM-
352 N.A.
37.3
57.2
7.1
17.1
NT
7.4
51.6
59.1
NT
Chev-
rolet
366
2.4
4.0
2.0
6.1
5.6
3.9
2.6
1.7
2.9
@ NT = Not tested
* close throttled

-------
                                       -20-
                                   Table 3

                                Emissions  From
                       Two Engines Used in Comparative
                        Application Vehicles  - 23 Mode
                               Steady  State Test
 HC, g/kW-hr

 CO, g/kW-hr

 NOx, g/kW-hr

 BSFC, kg/kW-hr

 Particulate, g/kW-hr

 Organics, %

 BaP, ug/kW-hr

•\         \
1 \Sulfate, mg/kW-hr

 Odor, TIA

 Aldehydes, mg/kW-hr
                         Caterpillar 3208
                         with EGR - Diesel
Chevrolet 366 CID
California (1977) Stds,
       Gasoline
1.78
8.51
5.07
0.290
3.48
10.97
2.31
24.36
2.4
109.0
3.34
73.76
4.56
0.463
0.28
3.92
3.02
1.38
1-NA
1017.0

-------
                                       -21-
2.  Malfunction Conditions

Almost all of the emissions  data  that are available from  heavy  duty engines
have  been  gathered  from  engines   that  were  tuned  up  to  manufacturers'
specifications.   However,  many  of   the  engines  that  are  operating on  the
roads  today  are  not  operating under a  similar  set of  conditions, due  to
factors   such   as  normal   wear   patterns,   maintenance   procedures,   etc.
Consequently  our  perception  of   their  emissions  could  be  dramatically
different than may be the actual case.

In  order  to  test this  hypothesis,  two engines were  to  be  tested  under  EPA
Contract  No.  68-03-2706 with  Southwest  Research  Institute (4).   The  first
engine tested was a Cummins  VTB-903 engine  which is an engine normally used
in  bus applications.  The engine was tested  in  its baseline condition (manu-
facturers'  recommended  specifications)  and  then altered  to  a  malfunction
condition.  This  malfunction condition was  an increase in the fuel injector
lash and a bleed-off of a certain portion of the  turbocharger  boost.  Such a
condition was  one in  which  there would  be  virtually no  noticeable decrease
in  power  and  therefore  the  operator would probably not bring  in the vehicle
for maintenance but would probably  continue  to  operate it  on  the road.  The
data  from this  condition are presented in Table  4.   When  tested  under this
malfunction condition,  increases  in  HC,  CO, and  a  very   slight  decrease  in
NOx  were  seen along  with  an increase  in  the  brake   specific   fuel  con-
sumption.   The brake  specific particulate  also  increased by  a  factor  of
approximately  two,  along with  the   organics, which  went   up by  a  factor  of
three  on  a g/km basis.   Sulfates decreased  somewhat, odor changed slightly,
depending  on  whether  the data  were taken on the thirteen mode  test or  the
transient  cycle.   Aldehydes increased  by  a  factor of   approximately  five
under  the  13  mode  test and stayed the  same  under  the transient  test.
Phenols generally increased  quite a  bit under the 13 mode test condition but
stayed almost the same or decreased  somewhat under the transient test.

EPA is currently  testing a Detroit  Diesel 6V-71 N engine  under a set of mal-
function  conditions  that are  similar to  those  of the  Cummins in  that they
are  severe enough  to alter  the  emissions  yet  not severe enough  that  the
operator  would  return the vehicle  to the shop  for maintenance.   These data

-------
         -22-

          Table 4

  Emissions Results from
A 4-Strbke Cycle Heavy Duty
 Engine  (Cummins VTB-903)
Operated Under Baseline And
  Malfunction  Conditions

    13 Mode                  Transient
Baseline/Malfunction
HC, g/kW-hr
CO, g/kW-hr
NOx, g/kW-hr
BSFC, kg/kW-hr
Particulate, g/kw-hr
Organics , %
Organics, g/kw-hr
Sulfate, mg/kW-hr
Odor, TIA
Aldehydes, mg/kW-hr
Phenols, mg/kw-hr
0.96
2.20
9.29
0.254
0.32
38.4
0.123
39.5
1.58
33.4
5.7
3.94
3.55
9.25
0.269
0.57
55.4
0.316
33.3
1.74
108.12
66.0
Baseline/Malfunction
1.79
1.87
7.83
0.269
0.50
51.2
6.256
38.4
1.17
101.0
15.4
6.98
3.12
7.63
0.288
1.08
76.9
0.829
33.4
1.32
99.0
10.2

-------
                                       -23-
will be available within  the next several months.  Malfunction  data of this
type are very  important as they give a  realistic view of how  these engines
operate under  actual conditions.   Additional work should  be done  on heavy
duty engines  but an  more extensive  effort  should be  performed with light
duty Diesel engines for which no data of this type are available.

B.  Light Duty Diesels

Table 5 presents  data from the work performed under  EPA contract 68-03-2414
wherein  the emissions  from what  were,  at  the   time  of testing,  prototype
Diesels were  compared against  their  gasoline counterparts  (5 arid  6).  This
work showed that the hydrocarbon emissions  were generally higher  from the
Diesel than the  gasoline  fueled vehicles.   However, CO was lower as well as
oxides of nitrogen.   The  fuel  economy for the Diesel engine  was,  of course,
much better from the Diesel than  the  gasoline  fueled  engine.  The  Diesel
engine  also  had a  much  higher  particulate  rate  as  well  as  an  organic
emissions rate.   The  BaP  and sulfate emission rates  from  the  Diesel engine
were much  higher than  those  from the gasoline  engine.   Aldehydes  from the
Oldsmobile  vehicles  were much  higher   with the  Diesel than  the  gasoline
engine; however,  aldehydes were almost  equal from  the Volkswagen Diesel and
gasoline vehicles.

Continued work should  be done on a  spot monitoring  basis  to evaluate  the
emissions from light  duty Diesel engines.

C.  Light Duty Gasoline Vehicles

Much of the testing  of  light duty gasoline  vehicles has  been performed when
the  vehicles   have  been  tuned  to  manufacturers'  specifications.   However,
many vehicles  operate in  a condition that is outside those specified  by the
manufacturer.   In  order   to estimate  the  levels of  pollutants  that  are
emitted  by  vehicles  under  such  operating  conditions,  EPA  conducted  a
thorough  program (7,  8,  9,  10,  11, 12  13)  to  test a  number of  vehicles
operated  under several  malfunction conditions.   These   conditions  included
rich-best-idle,  12%  misfire,  disabled exhaust gas  recirculation, higher  oil
consumption  (obtained  by oil  injection),   and  rich-best-idle  with  no  air
injection.

-------
                                     -24-
                                  Table 5

                           Emissions Results From
                      Comparative Diesel and Gasoline
                        Fueled Vehicles - FTP Cycle
                   Oldsmobile Cutlass
Volkswagen Rabbit
Displacement in 3
HC, g/km
CO, g/km
NOx, g/km
Fuel
Economy, mpg
Particulate, g/km
Organics , %
BaP, ug/km
Sulfate, rag/km
Odor, TIA
Aldehydes, mg/km
Diesel
350**
0.47
1.24
0.70
21.7
0.57
13.3
4.54
9.96
1.2
80.7
Gasoline
260**
0.24
1.34
0.85
15.6
0.006
10. 7@
0.17@*
1.37
N.T.@@
14.0
Diesel
89.7
0.23
0.49
0.54
42.7
0.18
21.4
2.67
3.66
1.1
39.6
Gasoline
97.1
0.14
2.30
0.63
24.6
0.004
42.3*
0.17@*
0.04
N.T.
35.1
@  = FET cycle
*  = Hot FTP cycle
@* = BaP values are considered conservative and should be used
     with caution due to inadequacies of sampling method.
@@ = NT=not tested
** = These engines were selected to have equal performance (e.g.
     equivalent vehicle acceleration) rather than equal displacement,

-------
                                       -25-
The data  from  four  oxidation  catalyst  equipped vehicles  and  one  baseline
non-catalyst  equipped vehicle  are  presented  in  Table  6  which  gives  the
number of  vehicles  that  experienced  emissions increases  during malfunction
relative  to  the baseline  state.   Most  of  the malfunctions resulted  in  an
increase  in  hydrocarbon  emissions as  well  as carbon monoxide.   The misfire
and disabled EGR conditions  resulted in  increases  in  oxides  of  nitrogen,
whereas  the  remainder  of the  malfunctions  did not  have  any  influence • on
NOx.  Most  of the malfunctions resulted in increases in  most of  the  other
emission  (e.g. unregulated) with a few minor exceptions.

In  trying to estimate  the importance of  these data,  the  maximum  emission
rate  (MER)  that was  experienced by any  of  the vehicles  tested  under  any  of
the  conditions  tested  was related  to the  threshold limit  value  (TLV)  as
given in  Table  7.  The  threshold  limit values are  those  established by OSHA
for exposure in an eight-hour work  shift.   For the  compounds listed it would
appear  that  the higher  the MER  to TLV ratio the more  likely  the maximum
emission  rate  would  be a source  of concern.   However,  it  should  be  empha-
sized that  definite  conclusions are  currently inappropriate for any  of the
compounds  studied  with a possible  exception  of  those for  which  automotive
emissions  standards   have been established.   EPA  is currently looking  at
various  unregulated  emissions and  plans to  define an acceptable  level from
many  of  these  unregulated emissions to the extent  possible.   It should also
be  observed that  some  of  the  compounds which are  considered  by  OSHA and
other  agencies  as  being  non-toxic can have a  potential  indirect harmful
effect.   An example  of  this  would  be nitrous oxide, due to its  potential
negative  effect  on the  upper  atmospheric ozone layer.  In reviewing Tables 6
and  7 it is  observed  that the maximum  FTP  composite emission  rates  for  HC
and  NOx  greatly  exceeded the  standards for  1983.  The  maximum value for
total particulates, however,  was  less than one-half  of the  proposed 0.2 gpm
particulate  standard  for  1983 Diesel vehicles which  would  be expected since
gasoline  vehicles  using  unleaded  fuel typically have  very  low particulate
emissions.   The  maximum undiluted exhaust emission  rates  for carbon monoxide
and oxides  of  nitrogen exceed  the TLV by over 100.  Other  compounds  in the
undiluted  exhaust  which  exceed  an  established   TLV  include  benzene  and

-------
                                           -26-
                                     Table  6

          Number of Cars with Emissions Increase During Malfunction*
      -Baseline (non-catalyst)  and Oxidation Catalyst  Equipped Vehicles-

                               Number of  Cars  for Which Emissions Increased/
                              The Total Cars Evaluated in that Configuration
                          Rich Best    12%     Disabled   High Oil   R.B.I, and
                            Idle     Misfire     EGR      Consump.   No  Air Inj.
Regulated  Emissions
 Hydrocarbons                 4/5       5/5        a
 Carbon Monoxide              5/5       5/5        a.
 Oxides of Nitrogen           0/5       5/5       5/5

Particulates
 Total Particulates           2/4       4/5       3/5
 Sulfate                     1/5       2/5       0/5

Compound Group  Totals
 Aldehydes & Ketones          4/5       5/5       5/5
 Individuals Hydrocarbons**  4/4       2/2       1/2
 Organic Sulfides             2/3       2/4       0/3
 Organic Aminesb              5/5       1/4       2/5

Other Compounds
 Ammonia                     3/5       3/5   ^   2/5
 Cyanide & Cyanogen           4/5       3/5   "   5/5
 DMNAC                       	       	       	

 Hydrogen  Sulfide             4/5       1/5       1/5
 Nickel Carbonyld
 Nitrous Oxide                2/5       2/3       3/3

Other Elements0
                      3/3
                      3/3
                      0/3
                      3/3
                      3/3
                      1/3

                      1/3
                      1/3
                      1/3

                      0/1
                        2/2
                        2/2
                        0/2
                        2/2
                        1/2
                        1/2

                        1/2
                        1/2
                                  1/1
                                  1/1
                        2/2

                        1/2
  Changes were relatively minor and inconsistent

  Changes in organic amines were relatively minor
c Discussed separately

d No nickel carbonyl was ever detected
 * Vehicles were:
                                               Emission  Controls
     Year       Make and Model

     1977       AMC Pacer
     1978       Chev. Malibu
     1978       Chev. Malibu  (Cal.)
     1978       Ford Granda
     1978       Ford Mustang  II
Engine
CID/CYL

258/6
305/8
305/8
302/8
302/8
Oxidation
Catalyst

  None
Pelleted
Pelleted
Monolith
Monolith
                                                               Air
 **Individual HC consist of  lower  molecular weight hydrocarbons (methane,
   ethane,  ethylene, acetylene,  propane, propylene,  benzene,  and toluene)
   measured by gas chromatography.

-------
                                           -27-
                                      Table 7

                 Relative Importance of Malfunction Emission Rates
         Baseline  (non-catalyst) and Oxidation Catalyst Equipped Vehicles
Regulated Emissions
 Hydrocarbons
 Carbon Monoxide
 Oxides of Nitrogen
                      Maximum
                      Emission
                      Rate  (MER)
                      in mgAma
10800b
52900
 4200
                    Equivalent
          Emission  MER in
          Standard  Undiluted
          for 1983  Exhaust in
          in mg/km  mg/m3  (MEC)
 255
2113
 622C
 7000
34400
 2700
                      Threshold
                      Limit Value     Ratio
                       (TLV)in mg/m3   of MEC
                      OSHA       Re    to TLV
55
 9
625
300
Particulates
 Total Particulates
, Sulfates
   48*
   17
 124d
   31
   15
           1
           1
 31e(12)f
 15e
Compound Group Totals
 Aldehdyes & Ketones       228b
 Organic Sulfides            2
 Organic Amines             <1
                        228
                           2
                                 10
Other Compounds
 Ammonia
 Cyanide & Cyanogen
 Hydrogen Sulfide
 Nickel Carbonyl
 Nitrous Oxide
   57
    7b
    4
    0.00
   70
   37
    7
    4
    0
   60
                 00
                       35
 0.007
                                  5
                                 10
 Formaldehyde
 Benzene
  1511
  230t
             151
             182
            4
           33
                38(2)
   * FTP rate recorded for the regulated emissions
     Value for non-catalyst car - Highest HC  value  for  catalyst car was 2300
   c 255 mg/km in California
   d Proposed for diesel cars
   e Based on values for some similar or Related  compounds as set by OSHA(9)
     or Recommended by the 1968 American Conference  of  Governmental Industrial
     Hygientists(lO)
   f Value for catalyst cars given within the parentheses if the value for
     non-catalyst and catalyst equipped vehicles  differs substantially

-------
                                       -28-
formaldehyde.   Even  though   benzene  and  formaldehyde  are  both  suspected
carcinogens, they also  are  toxic and the TLVs  are based on  their  toxicity.
(The maximum value  for ammonia  was  approximately equal to  the TLV).  As  a
matter of possible  interest,  the value  for  carbon dioxide was  over  20 times
the  established  TLV.  For  the remaining compounds  and compound groups,  it
was  necessary  to estimate an  equivalent  TLV based  on best  available data.
Based on  these estimated values,  the  maximum levels  for  total particulates
and  sulfate were above the estimated values  for an equivalent TLV.

Emissions  from  a  total  of  four  3-way and  3-way  plus  oxidation  catalyst
vehicles  have  also  been  analyzed and  the  resultant  data  are  presented  in
Tables 8  and  9.   The same general trend  is  apparent with these  vehicles  as
with the  oxidation  catalyst  vehicles,  wherein the emissions  of  a variety  of
compounds tend to increase with the  malfunction conditions.   The  same type
of analysis was  made with regards to  the maximum emission rate  compared  to
the  OSHA  eight-hour TLV.   It is observed that  the maximum undiluted exhaust
emission  concentrations  for  carbon monoxide and oxides of  nitrogen exceeded
the  TLV   by  more than  the  factor of  100.    Other compounds  for which  the
maximum  emission rate  exceeded an  established  TLV  included  benzene  and
ammonia.   (The maximum values  for organic  sulfides,  hydrogen  sulfide,  and
formaldehyde  were  almost equal  to  the TLV).   As  a  matter  of   possible,
interest,  the   value  for  carbon  dioxide  was  again  over  20  times  the
established  TLV.  For the  remainder  of  the  compounds it  was  necessary  to
estimate  an equivalent TLV  based on  best  available  data.   Based   on these
estimated values,  the maximum  levels  for total particulate,  sulfates,  and
cyanide and cynanogen were above the estimated values for an equivalent TLV.

It appears  as  though the  following general  comparisons may be  made  for this
malfunction  work.    Sulfate,  ammonia,  and nitrous oxides  were higher  from
catalyst  equipped cars  than  from the  one  non-catalyst car.   Hydrocarbons  and
aldehydes were highest  from  the  non-catalyst  car.  Ammonia  and cyanides were
highest from  the three-way catalyst  cars.   It  appears that  the unregulated
exhaust  emissions  composition  from  catalyst equipped cars  operating under
malfunctioning conditions  is  in general distinguishably different:  from that
of non-catalyst  cars.

-------
                                            Table 8 - Selected FTP Regulated and Unregulated EmI salons
                                               From 3-Way Catalyst  Equipped  Gasoline  Fueled  Vehicles



Car 41, Pinto (3-Uay +
Hydrocarbons
Carbon Monoxide
Oxides of Nitrogen
Fuel Consumption
Total Partlculates
Sulfates
Ammonia
Cyanide & Cyanogen
Car 42, Sunblrd (3-way
Hydrocarbons
Carbon Monoxide
Oxides of Nitrogen
Fuel Consumption***
Total Partlculates
Sulfates
Ammonia
Cyanide & Cyanogen
Car 42, Saab 99 (3-way
Hydrocarbons
Carbon Monoxide
Oxides of Nitrogen
Fuel Consumption
Total Partlculstes
Sulfates
Ammonia
Cyanide & Cvannnen
Car 51, Marquis (3 way
Hydrocarbons
Carbon Monoxide
Oxides of Nitrogen
Fuel Ccr.suBpilon
Total Partlculates
Sulfatea
Ammonia
Cyanldu & Cyanogen


Unmodified
Oxld/Alr
0.1
1.3
0.5
11.1
15.3
3.13
6.0
0.1
only)
0.2
2.6
0.5
10.9
5.6
0.21
11.1
1.01
only)
0.1
2.1
0.1
10.8
4.9
0.19
ou.o
1.1
••• 0*ld./A«r>
0.1
1.6
0.8
15.8
2.0
0.85
5.1
0.31
Disabled
Oxygen
Sensor

0.1
1.9
0.3
12.0
2.9
1.03
10.2
- .

1.5*
57.5*
0.1
12.7**
11.1
0.32
68.1
4.13

0.7*
20.0*
0.2
11.6
8.9
0.40
4uJ.5
66.9

0.1
1.0
2.5*
15.5
5.8
1.30
3.4
0.13

12Z
Misfires

0.4
2.4
0.3
11.2
6.5
0.80
6.53
0.1

1.1*
4.7
0.4
10.9
7.5
0.16
36.7
4.09

2.2*
11.1*
0.1
13.0**
17.1
0.31
142.5
7.5

1.7*
8.9*
0.6
16.4
10.0
0.13
30.9
1.57

Disabled
ECRb

0.1
1.8
0.7
11.8
19.7
12.25
13.2
0.0

1.4*
47.6*
0.2
12.2**
28.7
0.12
93.9
6.97

_ _
—
—
—
	
—
—
—

1.5*
36.5*
0.6
17.1
5.8
0.17
172.4
69.96
High
Oil
Cons.

	
—
—
—
—
—
—
—

0.2
3.0
0.6
11.3
6.0
0.20
5.31
5.40

0.1
2.3
0.1
11.8
6.4
u.ll
63.3
5.5

__
—
—
—
__
—
—
—

Other
MalfC

1.2*
38.7*
0.1
11.6
7.5
0.05
63.3
10.1

	
—
—
—
—
—
—
—

0.8*
29.9*
0.1
11.4
12.3
0.18
318.1
36.7

1.8*
52.4*
0=4
18.5**
11.3
0.21
253.0
112.29
                                                                                                                                                             I
                                                                                                                                                            l-o
8Wlth air to bypaaa for Car 51.
^Oxygen sensor disabled with Cars  41  and 42 and air  to bypass with Car 51.
cDlsabled oxygen sensor and air  pump  with Car 41, disabled oxygen senosr and rich-
 bcst-ldle with Car 43, and disabled  cold temperature sensor with Car 51.
•Values exceeding 1978 California  standards (0.25, 5.6, 0.9 for HC, CO, NOx).
'•Greater than ten percent Increase in  fuel consumption relative to unmodified.
•••Liters / 100km

-------
                                        -30-
                                          Table 9

                        Relative Importance of Maximum Emission  Rates
                           From 3-way Catalyst Equipped Vehicles
Maximum
Emission
Rate (HER)
in mg/km
Emission
Standard.
for 1983
in mg/km
Equivalent
HER in
Undiluted
Exhaust in
mg/m3 (MEG)
Threshold
Limit Value
(TLV) in mg/m3
OSHA(8) Rc
Ratio of
MEQ to TLV
Regulated Emissions
  Hydrocarbons
  Carbon Monoxide
  Oxides of Nitrogen
 1,800
52,400^
 2,500=
  255
2,113
  6221
 1,100
31,400
 1,500
55
 9
571
167
Particulates
  Total Particulates
  Sulfate

Compound Group Totals
  Aldehydes & Ketones
  Organic Sulfides
  Organic Amines
    27
     3
     1
               124
               24
                2
                1
                    1
                    1
                    10
                     7
                    24
Other Compounds
  Ammonia
  Cyanide & Cyanogen
  Hydrogen Sulfide
  Nickel Carbonyl
  Nitrous Oxide
   253
   112
     3
     0
    88
              152
               67
                3
                0
               53
          35
           0.007
                     5
                    10
                     4
                    13
  Formaldehyde
  Benzene
     2
   125
                2
               75
           4
          33
a FTP rate recorded for the regulated emissions and total particulates
  255 mg/km in California
C Based on values for some similar or Related compounds as set by OSHA(8)
  or Recommended by the 1968 American Conference of Governmental Industrial
  Hygienists (9)
  Proposed for Diesel cars
  Values were 44 for the HFET

-------
                                      -31-
D.  Fuels

1.  Fuels Variables

Under EPA  contract 68-03-2440 with  Southwest  Research Institute,  the emis-
sions from two  different  light duty Diesel vehicles  (Mercedes  Benz 240D and
Volkswagen Rabbit)  were tested as  a function  of  five different  fuels that
were blended  for  use in this  program  (5  and  14).   These fuels  were   1) the
Emissions Test  Fuel that  has been used by  EPA for Diesel  testing  since the
late 1960's,   2)  a  fuel  that has the  properties  of  the  "national average"
Diesel  Fuel,    3)  Diesel  Fuel number  1,  4)   a  minimum  quality  fuel more
typical of the  direction  towards  which today's Diesel  fuels  are headed, and
5)  a premium  quality fuel.   The data from  this  testing  are  presented  in
Tables  10  and 11.   When  the vehicles were tested with those  fuels,   little
change  was seen in hydrocarbon emissions with the exception of the minimum
quality  fuel,  which  increased  the HC  emission rate.  The  same was  true  of
carbon  monoxide  emissions.   The  Volkswagen  was   generally  insensitive  to
fuels variables with regards to  NOx  emissions;  however,   the  NOx emissions
went up  in the  Mercedes with the  use of the minimum  quality  and the premium
quality fuel.   Fuel economy  generally  decreased with the minimum quality and
the  premium  quality  fuels.   The  most marked  change  came with particulate
with the minimum  quality  fuel resulting in  a  much  higher particulate rate in
both vehicles.   The  percent extractables also  was higher.  These results,
when combined together, showed  a  much  higher  organics emissions rate for the
minimum quality fuel  in the  Volkswagen; however,  in the Mercedes Benz  virtu-
ally no change  in the organics emissions  rates in  terms of g/km was seen as
a  function of the  five fuels.  BaP  was generally  unaffected by  four  of the
fuels;   however, the minimum  quality  fuel  showed a  marked  increase in the BaP
emission  rate.  With aldehydes, a marked increase  occurred with the minimum
quality fuel.

This work was  followed  by EPA contract 68-03-2707  wherein  the  very specific
fuel properties were varied with a  base  fuel  to note  their  effect on emis-

-------
HC, g/km

CO, g/km

NOx, g/km

Fuel Economy,
  mi/gal

Particulate, g/km

Organics, %

Organics, g/km
              ;\
BaP, ug/km    \\
               I1
Aldehydes, mg/'km

-32-
Table 10



Emissions Results
From a Volkswagen Rabbit Diesel
Operated on Five Different Fuels
- FTP Cycle -
2D
Emissions-
Test
Fuel
0.18
0.49
0.59
41.2
0.225
10.8
0.024
1.13
28.7
2D
National
Average
0.20
0.51
0.65
42.0
0.218
13.2
0.029
1.33
12.6
ID
"Jet A"
0.17
0.55
0.57
41.2
0.177
15.0
0.027
1.18
18.0
2D
Minimum
Quality
0.71
0.81
0.58
38.1
0.375
15.6
0.059
3.64
52.5
2D
Premium
Quality
0.20
0.52
0.63
39.9
0.194
13.6
0.026
1.05
13.4

-------
             -33-
            Table 11

        Emissions  Results
From a Mercedes Benz 240D Diesel
Operated on Five Different Fuels
          - FTP Cycle -
HC, g/km
CO, g/km
NOx, g/km
Fuel economy,
mi /gal
Particulate, g/km
Organics, %
Organics, g/km
BaP, ug/km \
Aldehydes, mg/km
2D
Emissions
Test
Fuel
0.12
0.57
0.78
27.7
0.329
11.2
0.037
0.39
16.9
2D
National
Average
0.19
0.64
0.79
27.3
0.314
9.8
0.031
0.46
15.9
ID
"Jet A"
0.09
0.57
0.73
27.6
0.235
12.0
0.028
0.32
16
2D
Minimum
Quality
0.20
0.71
0.88
25.1
. 0.380
7.6
0.029
0.62
23.9
2D
Premium
Quality
0.12
0.68
0.85
25.1
0.292
8.9
0.026
0.23
16.9

-------
                                      -34-
sions  (15).   The  test vehicle  was  a Mercedes  Benz  240D.   The  data  are
presented in Table  12.   The principal results  that  were observed  from this
work are the following.

    1.   Regulated gaseous  emissions  were  not strongly affected by  the fuel
         variables studied.  Hydrocarbon and  CO emissions  increased slightly
         with  the  high aromatic  and  cetane-improved  aromatic blends.   The
         cetane-improved  aromatic  blend also resulted  in  a  small NOx  in-
         crease.  The  cetane improver (DII-3)  is  an alkyl  nitrate  compound
         made by by Ethyl Corporation.

    2.   Aldehyde  emissions were  largely  unaffected  by fuel  changes,  with
         the   exception   of  a  fuel  designated   EM-405-F   (base   fuel   +
         isoquinoline  at  0.1%  N).    The   increase  in  aldehydes  was  sub-
         stantial, but not  completely understood.   The  light-end and heavy-
         end blends  also  resulted in  slight  aldehyde increases,  but  not  of
         the magnitude observed with EM-405-F.

    3.   Phenol emissions (which are  co-carcinogenic)  were  strongly affected
         by the shale  oil blend.   They increased to over 60  times  the level
         observed  with base  fuel.   The  only other  pronounced increase  in
         phenols occurred while  testing  EM-405-F,  base  +  isoquinoline (0.1%
         N).
    4.   Particulate  rates  were  reduced  with  the light-end  blend and  the
         olefin blend,  by 14  percent  and  6  percent,  respectively.   The  two
         blends containing  aromatics  were  associated with  increases  in par-
         ticulate  emissions of about  27  percent.   The  cetane-improved aro-
         matic  fuel  blend  increased  particulate  emissions about 68  percent
         over  those  observed with  base fuel.   Heavy  ends fuel apparently did
         not cause  an  increase in particulate emissions, as  might  have been
         expected.

    5.   Blends  containing  aromatics  resulted  in  slightly  less  organic
         solubles  extracted from particulate matter.   Olefins, light  ends,
         and  heavy  ends  were associated  with  increases   in  the  amount  of
         organic solubles.

-------
               Table 12

        Emissions Results From
A Mercedes 240D Diesel Vehicle Operated
   on a Variety of Fuels - FTP Cycle
Base Fuel

C, g/km
3, g/km
Ox, g/km
uel Economy, ml/gal
Articulate, mg/km
rganlcs, Z
rganlcs, mg/km
aP, ug/kra
dor, T1A
Idehydea, mg/km
lenols, mg/km
(H 320)
0.09
0.52
0.75
24.9
165
18.4
30.4
0.53
0
0.61
1.34
Base +
Isoquinollne
0.05 Z N
0.08
0.47
0.76
24.8
162
• 16.0
25.9
0.26
0
t
0.31
4.33
Base +
Isoquinollne
0.1Z N
0.08
0.46
0.89
24.5
162
16.0
25.9
0.35
0.98
15.22
29.40
Base +
Shale Oil Cut
0.05Z N
0.09
0.52
0.87
24.1
180
12.7
22.9
0.96
0
0.00
127.96
Base +
31Z Exxon
"HAN" heavy
aromatlcs
0.12
0.59
0.88
24.7
212
11.8
25.0
0.37
1.31
2.79
0.66
Base +
6% Chevron
alpha
olef Ins
0.09
0.48
0.85
23.9
155
13.9
21.5
0.21
1.45
0.00
0.14
Base +
light
ends
0.11
0.48
0.82
24.7
142
19.9
28.3
0.51
1.35
8.18
0.21
Base +
30% Pure
aroma tics
0.13
0.60
0.92
24.6
219
11.7
25.6
0.73
0.89
0.00
1.30
Base +
heavy
ends
0.08
0.47
0.86
24.4
159
17.5
27.8
0.43
1.09
8.83
0.55
Base +
31Z "HAN"
+ 0.7Z
DII-3
0.13
0.66
1.04
24.4
287
10.8
31.0
0.43
1.05
0.00
0.18









1
OJ
Ln
1



-------
                                       -36-
    6.   BaP  emissions  were  generally  low  for  the  fuel  blends  tested,
         ranging from 0.25 to 0.96 ug/km.  A  slight  reduction was  associated
         with the olefin blend.  A significant increase in  BaP,  3  times,  was
         exhibited by the shale oil blend EM-430-F when compared to the base
         fuel.

2.  Alternate Fuels

    a.   Middle Distillates Derived from Alternate Sources

Under  the  current  task  order contract  (EPA Contract  No.  68-03-2884)  with
Southwest  Research Institute,  some  alternate  fuels evaluations  are  being
conducted  with  a Volkswagen  Rabbit  Diesel  (15).   So far,  four fuels  have
been  tested  which include a  base  Diesel fuel  #2,  a Diesel  fuel   marine  #2
obtained from  shale oil,  a Paraho JP-5  as well  as  a  Coal Case 5A.   These
fuels  are   described  in  Table  13.    The  emissions  from  these  fuels  are
presented  in Table  14.  Here  it was  seen that there was  virtually  no  change
in HC, CO  and NOx emissions with the  shale  marine fuel.  This was  also true
of  the  particulate  emission  rate.    The   aldehydes,   however,,   decreased
slightly.  The phenols stayed much  the same,  as well as the odor.   There  was
a slight decrease in the  fuel  economy.  In  looking at  the  Paraho JP-5,  a
slight increase was seen in HC, CO,  and  NOx  as  well  as  a slight decrease  in
fuel economy.  The  particulate was unchanged, as  well as aldehydes  and odor
while  phenols decreased  somewhat  with  use  of  the JP-5.   The  coal  case
resulted in  a rather dramatic increase  in  HC,  CO,  NOx  and  particulate  (on
the  order  of approximately  50 percent),  fuel  economy also  increased,  and
phenols and odor went up slightly.

A combination  blend of  35% Solvent  Refined  Coal  2  (SRC2)  has been  tested,
which resulted in smoke  levels that were very very high during the  first  105
seconds of the LA4.  The smoke and particulate  levels were  so high  that  the
sampling system  failed  and  thus  testing was unable to  be  continued.   The
SRC2 will  be  blended  at  a lower level (approximately 25%)  in an attempt  to
have  the  engine  operate  without  exceeding  the  limits  of  the   sampling
system.  Other  fuels such as an Exxon  EDS  fuel  and a  naptha  cut will  be
tested in  the near future.   Other  such fuels  will  also be  tested as  they
become available  (16).

-------
TABLE 13  FUEL PROPERTIES AND COMPOSITION
Substance
Fuel Code (EM-
Cetane No. (D613)
Cetane Index (D976)
Gravity, °API @ 60°F
Density, g/ml @ 60°F
Carbon, wt. %
Hydrogen, wt. %
Nitrogen, ppm (oxid. pyrolysis)
Sulfur (lamp) , %
Calculated H/C, numeric
Carbon No. range (G.C.)
Aromatics, vol. %
Olefins, vol. % (D1319)
Paraffins, vol. %

Viscosity, cs @100 °F (D445)
Gum, mg/100 ml (D481)
Total solids, mg/ic.
Metals in fuel, x-ray
Boiling Range, °C (IBP-EP, D86)
10% point
20% point
30% point
40% point
50% point
60% point
70% point
80% point
90% point
95% point
Residue, wt. % (D86)
Base
DF-r2
329-F
50
52
37.5
0.837
85.8
13.0
48
0.24
1.81
B-r24
21.3
1.7
77.0

2.36
14.3
7.4
oa
191-340
219
231
242
251
260
269
278
290
307
323
1.3
Shale Diesel
Marine
453-F
49
56
37.9
0.835
86.3
13.4
5
<0.005
1.85
9-20
28.5
2.1
69.4

2.61
0.3
0.3
oa
207-317
236
246
252
259
266
272
278
286
295
302
1.0
Paraho
JP-5
473-F
45
46
43.6
0.808
85.9
13.7
<1
<0.005
1.90

22.
2.
76.

1.38
0.0


179-248
189
192
196
198
202
206
211
218
228
237
1.5
Coal Case
5A
474-F
42
41
31.1
0.870
86.5
12.4

0.10
1.71

34.9
1.4
63.7

3.08
42.4


192-366
234
244
253
259
267
276
277
292
330
353
1.5
SRC-II
Med. cetane
475-F






























Broadcut
Mid-Continent
476-F
35
57
44.1
0.806
86.1
13.2

0.17
1.83

16.2
0.0 i
83.8 ^
I
1.53



21-354
93
137
159
197
230
250
262
282
327
_
1.0

-------
                         TABLE  13  (Cont'd). FUEL PROPERTIES AND COMPOSITION

Substance
Fuel Code (EM-
Boiling Range, °C (IBP-EP, D2887)
10% point
20% point
30% point
40% point
50% point
60% point
70% point
80% point
90% point
95% point
Residue, wt. % (D2887)
Composition, Volume %
Kerosene
Petroleum
JP-5
JP-8
Diesel
Petroleum
Shale DFM
Coal SRC- I I
Light Cycle Oil
LSR Naptha
HSR Petroleum
Shale
Coal (Simulated)
N-Butane
Base
DF-2
329-F
104-387
197
220
239
256
268
280
292
307
330
347
0.0


0.0
0.0
0.0

100.0
0.0
0.0
0=0
0.0
0.0
0.0
0.0
0.0
Shale Diesel
Marine
453-F
118-341
216
237
254
265
274
285
297
307
319
325
0.0


0.0
0.0
0.0

0.0
100.0
0.0
0.0
0.0
0.0
0.0
- __^-^6-°
0.0
Paraho
JP-5
473-F














0.0
100.0
0.0

0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
Coal Case
5A
474-F














17.3
0.0
0.0

66.7
0.0
16.0
0.0
0.0
0.0
0.0
0.0
0.0
SRC-II
Med. Cetane
475-F














0.0
0.0
0.0

65.0
0.0
35.0
0.0
0.0
0.0
0.0
0.0
0.0
Broadcut
Mid-Continent
476-F












.

22.0
0.0
0.0

23.0
0.0
6.2
5.2
7.4
4.8
20.9
0.0
10.5
of Cr, Fe, Ni, Cu,  Zn,  and Mq;  <70 ppm Pb;  <100 ppm Al and Si
<10 ppm

-------
                                      -39-
                                    Table 14

                                    Emissions
                             From Volkswagen Rabbit
                          Diesel Operated on a  Baseline
                   DF2  Diesel Fuel  and Three "Synthetic"  Fuels
Fuel •— :
Emissions
HC, g/km
CO, g/km
NOx, g/km
Fuel Economy, mi/ gal
Particulate, g/km
Aldehydes, mg/km
Phenols, mg/km
> Base
DF-2
0.31
0.96
0.66
36.9
0.25
14.0
12.0
Shale
Diesel
Marine
#2
0.31
1.06
0.67
35.5
0.27
9.0
11.0
Paraho
JP-5
0.38
1.20
0.70
35.5
0.25
13.0
0.5
Coal
Case
5A
0.39
1.21
0.83
37.8
0.32
N.Y.A.*
14.0
Odor, TIA
2.3
2.2
2.4
2.8
* N.Y.A. = not yet available

-------
                                      -40-

b.  Gasolines Derived from Alternate Sources

In the  near future, EPA will  be testing a  gasoline made from  methanol via
the  Mobil   MGT  process   and  also   a   gasoline   that   was   blended  to
specifications.   This  latter  fuel  was  blended  by  the  Army  Fuels  and
Lubricants  Laboratory  at   Southwest  Research  Institute  and  will  have
properties  similar  to  a synthetic  gasoline  from oil  shale.   Also  future
possibilities   include   the  SASOL/Kentucky   coal  run  that   is  currently
underway, which should  produce some  gasoline  that may become  available for
EPA  tests.   Also,  Union Oil Company  is currently under  contract to  DOE  to
produce some gasoline from  oil shale  which may be tested  by  EPA.  These last
two sources of  alternate gasolines are  very  tentative and  would only be seen
in the  somewhat distant  future.   Efforts continue  to obtain  these samples
for ECTD testing (16).

To  date,   ECTD has  performed   no   characterization  on  alternate  source
gasolines due  to their  lack of availability.  In spite of the  fact  that EPA
had a contract since 1976  with  Southwest  Research Institute  to do  gasoline
synthetic  fuel  work,   EPA  has  been  unable  to  obtain  synthetic  gasoline
samples  (17).   Efforts were  made  to obtain  some from the  SASOL  plant  in
South Africa,  but they  were unwilling in  the end  to  provide fuel samples  to
EPA.

c.  Alcohol Fuels

Some testing  and characterization has  been  performed by ECTD/CTAB  on vehi-
cles  fueled  with  alcohols  with more  such  testing  planned  for   the  near
future.  One major project  that  has just  been  completed involved the testing
of an  alcohol-fuel/Diesel-pilot compression-ignition  engine  from Volvo,  as
well as its heavy  duty  Diesel counterpart (18,  19).   The  dual-fueled engine
uses Diesel fuel for  pilot injection, which  initiates combustion  and  then
uses  a  variable  amount   of  alcohol,  injected  subsequent  to  the  pilot
injection  for  the  principal  combustion.   This  engine  was tested .with  both
methanol and  ethanol  plus  water, as  well as a limited number of  tests  with
an  oxidation  catalyst  added  to  the  dual  fueled  engine.  The counterpart
engine, fueled  exclusively with Diesel fuel,  was also tested.

-------
                                       -41-
The results from this work are presented in Table  15.   It  was found that the
use of methanol. resulted in a 50%  reduction  in total  particulate emissions
compared  to the  single  fueled  engine.   The particulate  emitted  from the
engine using  methanol  had  70%  organic extractables  from  the  particulate,
indicating far less elemental carbon  type  substances  being formed as was the
case  with the conventional  Diesel  fueled engine,  where about  30% organics
were  found.  Less  sulfate is  also produced due to  lower  fuel sulfur levels.
It was also found  that the use  of  methanol  decreased NOx by  56%  for the 13
mode  test and 35% for the transient cycle.

The use of  methanol resulted  in  a hydrocarbon emissions increase as measured
by the heated  FID, of about  40% for  the  13  mode  and 70% for .the transient
cycle  compared  to the  Diesel fueled  engine.   Large quantities  of unburned
alcohol (which  is  only partially measured by the heated  Fill'), were emitted.
In one case,  the  amount  of  unburned methanol was  four times the  amount of
hydrocarbons.   Aldehyde  emissions  were  also  increased  with  the  use  of
alcohol.   The  use  of methanol resulted in a CO  increase of  about  60%.  It
was found that the use  of  an oxidation catalyst  reduced  both hydrocarbons
and CO but  some increase in NOx  was found. The catalyst did not reduce par-
ticulate  levels.  Ames test data were accumulated  on these fuels and will be
presented in the Ames test data  section.

Testing has also been performed  on  a  Ford Escort fueled  by  gasoline as well
as a  blend  of Anafuel,  which  is  an  alcohol-based  additive blended with gaso-
line  (15).   The  preliminary  results  from   this  work,  based  on  limited
testing,  showed  a  slight   decrease   in  exhaust   HC,  approximately  a  70%
decrease  in CO, a  25%  increase  in NOx,  no  change in  fuel economy,  and  a
decrease  in particulate when the Anafuel  case  was compared  to  the gasoline
case.  No ethanol  or methanol was emitted from  the engine  when  either fuel
was used.   With  the  base fuel,  there  were  negligible quantities of aldehydes
emitted and when the Anafuel was used,  there  was  a very  slight  increase in
this  level of  aldehyde.   The vehicle  appeared  to run  well on  both fuels.
Again, the  results with  the Anafuel are preliminary and may  not be confirmed
with  further  testing.  Some  additional  EPA  in-house testing  showed similar
results and also a large increase in evaporative HC emissions.

-------
                                      AND  Tllli VG1A/O  TIJ-iUUA
                                                                              PXL.UX/ /n.

Composite emission Kates
Fcdeial Tost Procedure (fTP)
liyd coca i bun, lie*1
g/kw-hr, (g/hp-hr)
Carbon Monoxide, CO
g/kw-hr, (g/hp-hr)
Oxides of Nitrogen, NOXC
q/kw-hr, (g/hp-hr)
lirake Specific Fuel Consump.d
kg fuel/kw-hr, (Ib fuel/hp-hr)
Test Cycle
Total Individual IIC
uig/kw-lir
Total llnbuined Alcohols
mc|/kw-hr
Total Aldehydes6
iwj/kw-lir
Total rhenols
nuj/kw-hc
Total {'articulate
g/kw-hr, 1'i/lip-hr)
Snl fate, SO.=
mg/kw-lir, (t of Part.)
Soluble Organic Fraction (SOF)
mg/kw-hr, (1 of Hart.)
Dal> in SOF
llg/kw-hr
Amos Response (TA9U)
revertant/mg SOF
— — Engine-Teat Configuration -
Uiescl
Volvo TO-100C
13-niode
1.05
(0.78)
3.1U
(2.37)
ll.Bflf
(8.86)
0.262
(0.431)
7 -mode
100
Does Not
Apply
19
Not
Kun
0.69
(0.52)
45
(6.5%)
2OO
(20't)


Transient
1.15
(0.85)
4.04
. (3-01)
11.19
(8.34)
0.288
(0.473)
Transient
130
Does Not
Apply
14
35
0.70
(0.52)
38
(5.4%)
220
(321)


Methanol
Volvo TD-100A
13-niode
1.45
(l.OQ)
9.55
(7.12)
5.26
(3.92)
0.486
(0.799)
7-mode
67
2200
88
17
0.30
(0.23)
14
(4.6%)
200
(66%)


Transient
1.95
(1.45)
10.29
(7.67)
7.31
(5.45)
0.531
(0.873)
Transient
180
4900
250
24
0.39
(0.30)
16
(4.1%)
280
(73%)


Hethanol-Catalyst
Volvo TO- 100 A
13-modeh
0.16
(0.12)
O.B3
(0.62)
6.79
(S.06)
0.482
(0.792)
7-mode
32
950
140
14
0.51
(0.38)
220
(43%)
70
(14%)


Transient^
0.22
(0.1C)
3.74
(2.79)
7.89
(5.89)
0.547
(0.900)
Transient
66
890
260
48
0.37
(0.27)
98
(27%)
60
(16%)


Ethanol
Volvo TD-100A
13 mode
1.65
(1.23)
10.53
(7.85)
6.88
(5.13)
0.396
(0.651)
7-mode
Not
Run








Transient
2.27
(1.69)
12.89
(9.61)
7.38
(5.50)
0.435
(0.715)
Transient
600
2300
240
44
0.35
(0.26)
14
(4.0%)
190
(53%)


Ethanol -Catalyst3
Volvo TD-IOOA
13 mode
0.60
(0.45)
3.10
(2.31)
7.98
(5.95)
0.400
(0.657)
7-mode
Not
Run








Transient
0.63
(0.47)
4.24
(3.16)
8.62
(6.43)
0.448
(0.737)
Transient
220
480
250
33
0.38
(0.28)
89
(23%)
40
(11%)


Ethanoli 301 Water
Volvo TD-IOOA
13-moclo
1.U9
(1.41)
9.99
(7.45)
4.46
(3.32)
0.495
(0.814)
7-mode
220
860
91
Not
Run
0.33
(0.243)
17
(5.2%)
190
<5B'l)


Transient
Not
Run



Transient
Not
Run








 (>Hun  with  backpressure  device  engaged during idle and motoring.
  IIC vulue  reported  here in  based on measurements by IIL'ID.  FID response is
 ( very low  (or  unbiirned  alcohols- and other species of unregulated emissions
  13-mode HOX correction factor for intake humidity was computed but not
  applied
  Computed  on the  basin  of measured diesel consumption and alcohol consumption
  combined
•e
 ^Uenzaldehyde  was not included In the total composite value
  HOX  value is  reduced to 10.89 g/kw-hr (8.12 g/hp-hr) when the intake
d
 humidity correction for NOX is applied.
 I'henols were determined from only selec
                                nly  selected modes
 "Without  backpressure device  at  idli
 Average  of data with ami without backpressure  device

-------
                    TAHLIi  .1.5.      SUMMAUY OL1 COMPOSITE EMliJblUN KAius r KVJI-I mii  vwi-,v
                                     AND  THE VOLVO TD-100A  DIESEL PILOT/ALCOHOL  ENGINE

Composite Emission Rates
Federal Tost Procedure (FTP)
hydrocarbon, MC^
g/kw-hr, (g/hp-hr)
Carbon Monoxide, CO
g/kw-hr, (g/hp-hr)
Oxides of Nitrogen, NOXC
g/kw-hr, (g/hp-hr)
Brake Specific Fuel Consump.d
kg fuel/kw-hr, (Ib fuel/hp-hr)
Test Cycle
Total Individual IIC
mg/kw-hr
Total Unburned Alcohols
mq/kw-hr
Total Aldehydes6
mg/kw-hr
Total Phenols
mg/kw-hr
Total Particulate
g/kw-hr, (<|/hp-hr)
SulCate, S04=
mg/kw-hr, (* of Part.)
Soluble Organic Fraction (SOF)
mq/kw-hr, (I of Part.)
DaP in SOF-
pg/kw-hr
Ames Response (TA98)
revertant/mg SOF
Engine Teat Configuration
Diesel
Volvo TD-100C
13-mode
1.05
(0.78)
3.18
(2.37)
11. 88^
(8.86)
0.262
(0.431)
7-mode
100
Does Not
Apply
19
Not
Run
0.69
(0.52)
45
(6.5t)
200
(2fH)


Transient
1.15
(0.85)
4.04
. (3.01)
11.19
(8.34)
0.288
(0.473)
Transient
130
Does Not
Apply
14
35
0.70
(0.52)
38
(5.4%)
220
(32%)


Methanol
Volvo TD-100A
13-mode
1.45
(1.08)
9.55
(7.12)
5.26
(3.92)
0.486
(0.799)
7-mode
67
2200
88
17
0.30
(0.23)
14
(4.6%)
200
(66%)


Transient
1.95
(1.45)
10.29
(7.67)
7.31
(5.45)
0.531
(0.873)
Transient
180
4900
250
24
0.39
(0.30)
16
(4.1%)
280
(73%)


Methanol-Catalyst
Volvo TD-100A
13-modeh
0.16
(0.12)
0.83
(0.62)
6.79
(5.06)
0.4B2
(0.792)
7-mode
32
950
140
14
0.51
(0.38)
220
(43%)
70
(14%)


Transient^
0.22
(0.16)
3.74
(2.79)
7.89
(5.89)
0.547
(0.900)
Transient
66
890
260
48
0.37
(0.27)
98
(27%)
60
(16%)


Ethanol
Volvo TD-100A
13 mode
1.65
(1.23)
10.53
(7.85)
6.88
(5.13)
0.396
(0.651)
7-mode
Not
Run








Transient
2.27
(1.69)
12.89
(9.61)
7.38
(5.50)
0.435
(0.715)
Transient
600
2300
240
44
0.35
(0.26)
14
(4.0%)
190
(53%)


Ethanol-Catalysta
Volvo TD-100A
13 mode
0.60
(0.45)
3.10
(2.31)
7.98
(5.95)
0.400
(0.657)
7-mode
Not
Run








Transient
0.63
(0.47)
4.24
(3.16)
8.62
(6.43)
0.448
(0.737)
Transient
220
480
250
33
0.38
(0.28)
89
(23%)
40
(11%)


ethanoli-301 Water
Volvo TD-100A
13-mode
1.89
(1.41)
9.99
(7.45)
4.46
(3.32)
0.495
(0.814)
7-mode
220
SCO
91
Not
Run
0.33
(0.243)
17
(5.2%)
190
(581)


Transient
Not
Run



Transient
Not
Run








^Kun with  backpressure device engaged during idle and  motoring.
 IIC value  reported here in based on measurements by  IIF1D.   FID response is
(.very low  for  unhurried alcohols and other species of unregulated emissions
 13-mode NOX correction factor for intake humidity was computed but not
 applied
 Computed  on tho basin of measured dier.el consumption  and alcohol consumption
 combined
e
^Bcnzaldchyde  was not included in the total composite  value
 NOX value is  reduced to 10.89 g/kw-hr (8.12 g/hp-hr)  when  the intake
 humidity  correction for NOX is applied.
 Phenols were  determined from only selected modes
.Without backpressure device at idle
 Average of data with and without backpressure device

-------
                                      -43-
Future  work  in the  alcohol area will  include the  testing of  a  Volkswagen
that  is  fueled exclusively by  methanol  and compared  to  the results  from a
Volkswagen fueled exclusively  by  gasoline.   It is also hoped  that an Escort
designed for use with  100%  methanol can be obtained such  that its emissions
data  can  be  compared  to  the  data discussed above.   The Volkswagen  will be
tested  and its  emissions  analyzed  by  mass  spectroscopy  so  that  a  broad
characterization, in a qualitative  sense, can be made  between the emissions
from  these gasoline-fueled and methanol-fueled engines.

E.  Nitrosamines

Nitrosamines  are  a  group  of compounds  that  take   the  general  form  of
R?N=NO.  Nitrosamines  have been  described  as  among the most  broadly acting
and  the most  potent  carcinogens.    More  than  100 nitrosamines   have  been
tested  for carcinogenic  activity and most of  them have  induced  tumors in
rats.  A smaller number  have  been tested in hamsters,  mice,  or guinea pigs,
and again  most of  those  tested have  been carcinogenic.  As a  result nitro-
samines  are  considered  to be very  hazardous  to  humans  because of  their
assumed carcinogenicity in humans.

Very  little  monitoring of  roadside nitrosamines has  been  performed.   How-
ever, what limited  work  has been  done to date has  shown  instances of nitro-
samines  up to  approximately one  microgram  per cubic  meter.   It  appears as
though  these  roadside  levels  are mobile source related.   However,  only two
sources  of  nitrosamines  have been confirmed,  as  well  as  one  possible
source.   The   confirmed  sources  included are   1)  crankcase emissions  from
heavy-duty Diesels,  and  2) automobile  interior  emissions (20, 21,  22,  23,
24).  The  possible  source  of  nitrosamines  is indicated  by one datum point
that  showed nitrosamines  may be  present  in heavy-duty  Diesel  exhaust.  This
latter datum point is, as of yet,  unconfirmed.
The  work on  nitrosamines  from heavy-duty  Diesel  crankcases  indicated  that
the most likely source for  these  nitrosamines was  the  oil additives that are
used  for purposes of improving  the  performance characteristics of  the  oil.
There  is  a  clear  association  between  the  nitrosamine  emissions from  an
engine using  new  oil and the  level  of  nitrosatable amines  that is measured

-------
                                      -44-
by a bench procedure.  There  was  also an influence due  to  engines  (probably
those factors influencing NOx emission  rates)  as well as a  probable  associ-
ation -with  the  amount of time that the oil is  used.  However,  this  latter
association  has  not  been  totally  confirmed.    There would  also be  nitro-
samines from light-duty  Diesel blowby.  A  number of different  sources  have
indicated that neither nitrosamines nor  nitrosamine  precursors  are  emitted
in  the exhaust  of  gasoline  vehicles.  -It  is   very likely  that  gasoline
engines would also have nitrosamines in  their blowby.  However, the  blowby
from gasoline vehicles is currently recirculated through the engine.

The significance of the roadside  concentrations  is  not  easy to estimate  very
accurately, due to the lack of models  specifically  tailored for that  purpose
and  the lack of  adequate  nitrosamine  roadside  monitoring data.   However,
some  very  rough  approximations  have  shown  that  the  number  of  cancer
incidences  per  year  may  not be   insignificant   for  the  exposure  levels
detected.

Also,  some  recent  EPA work has shown nitrosamines  to be present: in  vehicle
                        A
interiors (23).  In  the Vehicle interior work, 56  vehicles  were sampled for
nitrosamines.  Of  these \56  vehicles, detectable levels of  nitrosamines  were
found  in  47.  Concentrations  of  detectable  nitrosamines in  these  vehicles'
interiors  ranged  from 0.01  to 0.16 micrograms  per  cubic  meter.  The  test
results indicated  an  N-dimethyl nitrosamine  (NDMA)  dependence  on mileage and
time with  NDMA  levels decreasing slightly with  time  and mileage.  Also,  an
NDMA dependence on -temperature was  observed, with higher  temperatures  giving
higher  levels of NDMA.  The levels  of NDMA in vehicles  during operation was
observed  to  decrease  somewhat,  but   they  rebounded  quite  rapidly  after
operation.   Also   similar  levels  of  NDMA  in  vehicles  were  observed  after
either  an overnight  soak  or for  vehicles  closed for  only  a short  time.
Nitrosamines  were  found  in  passenger   cars,  stationwagons,  passenger  and
cargo  vans,  pick-up  trucks,  as well as in new  and used heavy-duty  trucks.
Nitrosamines were  not detected in  motor  homes.   A very rough  analysis  was
made and  it was observed  that on  the  average,   the  daily  intake of  nitro-
samines from vehicle  interiors would  be less than  the amount  contained  in a
can of  beer  (before nitrosamines  in beer were  regulated  to  lower levels) and
more than the level  found  in a strip of bacon (bacon is  currently  regulated
for nitrosamine content).

-------
                                      -45-
F.  Identification  of  Types of Compounds  Responsible  for Ames Test  Activity
    in Diesel Particulates

Many  investigators  including  those  in  EPA,   the  automotive  industry,  and
private  investigators  such as  those  at  the University  of  California  at
Berkeley,  are  conducting studies  to  identify  the  compounds  responsible  for
the  Ames  test  activity of  Diesel particulate extract.   These  researchers
have  spent effort  in  separating  the methylene  chloride  extract of  Diesel
particulates  by high  pressure  liquid chromatography  (HPLC).   Three  major
fractions  have  been  obtained  by  HPLC:   non-polar,  moderately  polar  (tran-
sition), and highly polar.

The non-polar fraction consists  of aliphatic and polynuclear aromatic  hydro-
carbons  (many of  them alkyl substituted)  which were responsible  for only a
portion of the Ames test activity.

The  transition  and  polar  fractions  accounted  for most   of  the  Ames  test
activity.  About  50%  by weight  of the material  in the  transition  fraction
consisted  of various oxygenated  polynuclear aromatic hydrocarbons  (including
hydroxy,  ketone,  carboxaldehyde,  quinone,  acid  anhydride,  and  nitro  com-
pounds).   The polar fraction  consisted of carboxylic  acid polynuclear  aro-
matic hydrocarbons which also have significant Ames test activity  (25).

Appendix 1 (26)  lists  the compounds which  have been identified so far in the
joint project on  Diesel particulate fractions  being conducted  by Ford Motor
Company and EPA.

Appendix 2 lists the compounds which  have been identified  by  Sklarew et al.
of Battelle Northwest.

G.  Development  of Method to Collect/Characterize Gas-phase Hydrocarbons

OMSAPC has been  concerned about the  potential hazard  of  automooile  exnaust
hydrocarbons  for several years-  As such.  OMSAPC has been  very interested in
the development of  a method to  collect  artifact-free  samples ot  gas  phase

-------
                                      -46-
hydrocarbons  in  motor vehicle  exhaust for  bioassay  testing.   EPA/ESRL  has
evaluated both a  filter  cartridge  and a condensate method and  ECTD has done
some limited gas phase sampling and analysis.

Preliminary tests  performed  by ESRL have  been  done with a  method involving
filtering the exhaust  particulates  and then  condensing  components in the gas
stream.   The  condensate appeared  to have low  Ames  test  activity.   If  the
filter upstream of  the condenser were removed,  the condensate contained some
Diesel particuate and had somewhat higher Ames test activity.

The  filter  cartridge  method  involved passing  a  gas stream sample  after  a
conventional  particulate filter through a cartridge or  bed  of  treated XAD-2
resin.  After the  gas  stream was passed through  the  XAD-2,  the hydrocarbons
adsorbed  onto the XAD-2 were  removed from  the  resin by methylene chloride
extraction.   The  lower molecular weight hydrocarbons (e.g.  below C-10) were
sufficiently  volatile  that they  were probably lost  during  the extraction.
However,  hydrocarbons  above  C-10  were retained and could then  be subject to
the  Ames  test.  Since  the  conventional particulate filter  generally retained
hydrocarbon  compounds  above  C-15,  the XAD-2  traps  could  provide  a  good
method to collect hydrocarbons  in the C-10 to C-15 range (25).

A very preliminary  result  of  Ames testing on the  gas phase  hydrocarbon col-
lected by this method  for a VW  Diesel  Rabbit showed  that the activity may be
low  compared  to that of the particulate.

ECTD/TAEB has done  some work wherein a light-duty Diesel vehicle was used to
load a large  filter with  particulate.   The volatile  organic  material  was
then driven off by  heating  the filter to 1000°F which is slightly below the
ignition  point  of  carbon.     These  "organic  free"  filters  were  then  put
downstream  of a  particulate  filter in  a  stream  of  Diesel  exhaust.   Thus,
presumably  the  "organic free"  filters would be  subjected  only to  the  gas
phase  hydrocarbons  of  the  exhaust stream.   The carbon  on the downstream
filter provided a media  for  collection of  the hydrocarbons.   The preliminary
results of  this work  indicated  1) collection of some  hydrocarbon material
on   the  filter,   and  2)  some  apparently  significant  Ames  test  activity
associated with the hydrocarbons thus collected (27).

-------
                                       -47-
H.   Ames Testing

ECTD  has  conducted a  number  of  experiments  wherein  the  emissions  were
sampled  and  subjected  to the Ames  bioassay test.   One such study  was  with
the Diesel fuels  project  (discussed in section  IV  D 1.) and these  data are
summarized in Tables 16  and 17.   These  data seemed  to show  little  effect
from  fuel  nitrogen levels when  the nitrogen was introduced  in the form of
isoquinoline.   There also  appeared  to  be  a trend  towards a decrease  in
activity  when olefins  were  introduced  to  the  base fuel.   The heavy  ends
indicated  a  possible trend  towards increasing  Ames  test activity  and  also
the same could possibly  be  said for the cetane improver as it appeared to
increase .the  Ames  test  activity.   The other trend appeared  to  be  towards an
increase in Ames test activity with the  use of aromatics.  This included the
heavy aromatic naptha (HAN) as well as the individual aromatics (15).

These findings  tend to  be corroborated  with the preliminary results reported
by Sklarew from Battelle  Northwest.  Sklarew et  al.  rather extensively char-
acterized Diesel  particulate extract from filters provided  by  ECTD  from the
Southwest .Research Institute fuels  project discussed  above.   These samples
were  taken when the vehicle was operated on  1) base fuel,  2) base  fuel  +
isoquinoline  (0.05% N),   3)  base fuel +  isoquinoline (0.10% N), and   4)  base
fuel  +  HAN +  cetane improver.   They concluded  that  there was a  slight  but
positive  correlation between higher aromatic content in- the  parent  fuel and
mutagenicity  of  the chemical fractions from combustion  particulates of  that
fuel.   No correlation  was found  by them between increased  mutagenicity and
the   nitrogen  content  (from isoquinoline)   of the  parent  Diesel   fuel.
However,  they also observed  that the  fuel  with isoquinoline added  prior to
combustion  appeared to  have lower mutagenic  activity  than  the  same  fuel
without  isoquinoline.    Their  preliminary  data  suggested  that  the  high
aromatic  Diesel  fuels  gave rise  to higher  concentrations of indirect  acting
mutagens  in  the  aromatic hydrocarbons,  moderately  polar  or  neutrals  and
highly polar  neutrals than do the low aromatic Diesel fuels (28).

The other set of  Ames  test  data  that  has  been recently analyzed  was  that
from  the Volvo  testing  wherein  the dual fueled  Volvo  (which employed  Diesel

-------
                                  -48-
TABLE 16...  SUMMARY OF AMES BIOASSAY ANALYSIS  OF  ORGANIC SOLUBLES FROM
               PARTICULATE MATTER COLLECTED DURING  FTP
' Fuel Code
EM-395-F
EM-404-F
EM-405-F
1 EM-430-F
; EM-434-F
EM-438-F
EM-448-F
i
EM-401-F
EM-460-F
EM-461-F
'EM-463-F
Description
base
base + isoquinoline
(0.05% N)
base + isoquinoline
(0.10% N)
base + shale
oil cut
base + "HAN"
base + olefins
base + light ends
JP-7
base + individual
aromatics
base + heavy ends
EM-434-F *
cetane improver
RLI-16
Activation
No
Yes
No
Yes
No
Yes
No
Yes
NO
Yes
No
Yes
No
Yes
No
Yes
No
Yes
No
Yes
No
,, Yes
Model Predicted
Mean Slope, revertants/ug extract
TA-1535
0.0
0.0
0.0
0.1
0.0
0.2
0.3
0.0
0.0
0.1
0.0
0.0
0.0
0.1
0.0
. 0.1
• 0.0
0.1
0.0
0.1
0.0
0.1
TA-1537
0.2
0.3
0.7
0.5
0.1
1.1
0.2
0.2
0.7
0.5
0.1
0.2
0.4
1.3-
0.2
0.3
0.7
0.4
0.3
0.5
0.7
0.6
TA-1538
0.7
0.4
1.7
1.0
0.5
0.9
0.6
0.7
1.7
1.1
0.3
0.5
0.9
0.7
1.4
0.9
1.7
0.9
1.2
1.0
1.7
0.9
TA-98
3.2
0.9
3.3
3.0
1.9
0.8
1.3
1.0
3.7
1.1
0.8
1.3
4.1
0.9
3.0
3.3
3.5
2.9
3.6
2.7
5.7
2^5
TA-100
3.9
1.9
5.1
1.9
5.5
1.3
4.9
1.9
6.9
2.2
3.0
1.2
3.3
1.4
1.4
8.7
2.6
6.6
2.3
14.0
2.4

-------
                                       -49-
          TABLE 17.  AMES BIOASSAY ANALYSIS RESULTS IN REVERTANTS PER
                             DISTANCE DURING FTP
Fuel Code
: EM-395-F

EM-404-F

' EM-405-F

EM-430-F

EM-434-F

i EM-438-F

EM-448-F

EM-401-F

EM-460-F

EM-461-F

1 EM-463-F

Description
base

base + isoquinoline
(0.05% N)
base + isoquinoline
(0.10% N)
base + shale
oil cut
base + "HAN"

base + olefins

base + light
ends
JP-7

base + individual
aromatics
base + heavy
ends
EM-434-F +
cetane improver
RLI-16
Activation
No
Yes
No
Yes
No
Yes
No
Yes
No
Yes
No
Yes
No
Yes
No
Yes
No
Yes
No
Yes
No
Yes
Model Predicted
Mean Slope, 103 revertants/kiria
TA-1535
0.0 .
0.0
0.0
2.
0.0
5.
7.
0.0
0.0
3.
0.0
0.0
0.0
3.
0.0
3.
0.0
3.
0.0
3.
0.0
3.
TA-1537
7.
11.
16.
12.
3.
29.
5.
5.
18.
13.
2.
4.
11.
37.
5. .
8.
18.
10.
8.
14.
22.
19.
TA-1538
25..
14..
40.
23.
13.
24.
14.
16.
43.
28.
6.
11.
25.
20.
. 36.
23.
44.
' 23.
33.
28.
53.
28.
TA-98
112.
32.
77.
70.
50.
21.
30.
23.
93.
28.
17.
28.
116.
25.
77.
85.
90.
74.
100.
75.
176.
77.
TA-100
137.
67.
119.
44.
145.
34.
112.
43.
173.
55.
65.
26;
93.
40.
^ ,_
36.
223.
67.
184.
64.
433.
74.
^calculation incorporates particulate mass rates based on 47 mm Pallflex filters,
 percent organic solubles extracted from Pallflex "20x20" filters,  and  data  in  Table 23.

-------
                                       -50-
fuel pilot  injection and  alcohol main  injection) was  tested and  compared
against a baseline pure Diesel Volvo engine.

Table  18 shows  that  the   change  between  the methanol  and  methanol  plus
catalyst case resulted in a 20%  decrease in particulate and an  82%  decrease
in organics with use of the catalyst.  However, the use of  the catalyst also
resulted in a very  dramatic increase  in  the revertants per plate  as  seen in
the first column  of data.   This increase was  between about 300  and  5,400%.
However, when  the accompanying reduction in  organics was  incorporated into
the data, the increase in revertants  per kilowatt  hour column  was not nearly
as  dramatic.   However,  there  was  still generally  an increase per  kilowatt
hour associated  with  the  methanol with catalyst  configuration.   Basically
the same trend was exhibited by the ethanol and ethanol catalyst comparison.

Table 19 shows the  data from  the pure Diesel  fueled baseline  engine  compared
against  the dual-fueled  engine  using  methanol.   Both  transient  and  the
steady  state  data are presented.   Some  preliminary conclusions can  be made
from these  new  data.   For  example, the  Ames  test  activity results from the
dual  fueled engine using  methanol  tested over  the transient  cycle  were
generally lower  than  the baseline/pure  Diesel  fueled engine.  However,  the
dual fueled engine using  methanol and  Diesel fuel  seemed  to  show  higher
activity  than  the  baseline  engine over the  13 mode cycle.   This  apparent
discrepancy may be  explained  by  looking  at  the modal  data  which  showed that
the heavier loaded  modes tend  to  result in higher Ames  test activity than
lightly  loaded  modes  (an  interesting observation  in-and-of   itself).   This
trend  was  much more  pronounced  for  the  dual  fueled engine   using  methanol
than for the pure Diesel fuel engine.  This, combined with  the fact  that the
transient cycle is  more lightly  loaded than the 13 mode test, could  account
for the  above mentioned  apparent conflict.   The steady state  runs  (13  mode)
with the dual fueled engine using  methanol  and  Diesel fuel  with  the  catalyst
tended  to  have slightly higher  specific activity  in revertants  per  micro-
grams  of  extract  but  lower revertants per  kilowatt hour responses than the
same engine but without the catalyst.  This  is  in  some  conflict with  the
earlier data as presented in  Table 18 wherein  the  transient cycle comparison
showed both the specific activity  and the revertants  per kilowatt  hour  to be

-------
                                                              Table 18
                                            Average Percent Change Between Two Variables
Dual Fueled Volvo TD100A - Transient Composite Test Cycle
Variable A
Variable B
Particulate g/kW-hr
METHANOL
METHANOL W/CAT
-20%
Organics g/kW-hr
Ames Data
TA 98
TA 98
TA 100
TA100
TA1535
TA1535
TA1537
TA1537
TA1538
TA1538








-82%
Rev/Plate-ug Rev/kW
No Act.
Act.
No Act.
Act.
No Act.
Act.
No Act.
Act
No Act.
Act.

A






1001%
457
285
745
None*
1050
4463
1473
5402
296

ce Percent B
Change

Where: A
Bi
A2
B2
99%
0
-30
53
None*
125
694
194
331
-29
B - A
/inns A J
*• ' A
Al

= Results from
= Results from
= Results from
= Results from
ETHANOL
ETHANOL W/CAT
+8%
-79
hr Rev/Plate-ug
899%
534
165
1041
100
25
7608
18699
708
2809
B_ - A,
1+2 2
A2
2
Variable A tested on
Variable B tested on
Variable A tested on
Variable B tested on

Rev/kW hr
115%
36
-43
144
100
-67
1573
3883
73
523




2/11/80
2/11/80
1/27/81
1/17/81
METHANOL
ETHANOL
-13%
-36%
Rev/Plate-ug
126%
125
169
163
None*
200
208
214
242
58
'







METHANOL
W/CAT
ETHANOL W/CAT
+17%
-25%
Rev/kW hr
43%
42
72
68
None*
88
85
105
114
0**








Rev/Plate-ug
110%
146
40
186
225
-61
470
1797
-50
861








Rev/kW hr
58%
86
4
116
100
-63
325
1312
61
621








                                                                                                                                                                   I
                                                                                                                                                                   1/1
*  "None" indicates no change in the two variables.


** "0" indicates no or negligible Ames test repsonse.

-------
                  Table 19
Average Percent Change Between Two Variables
Dual Fueled Volvo TD100A
Variable A -^
Diesel-transient Diesel 13-Mode
Variable B —> Methanol-transient Methanol 13-Mode
'articulate g/kW hr -44% -57%
jrganics g/kW hr
\raes Data
TA 98
TA 98
TA 100
TA 100
TA1535
TA1535
TA1537
TA1537
TA1538
TA1538

No Act.
ACT.
No Act.
Act.
No Act.
Act.
No Act.
Act.
No Act.
Act.
+27%
Rev/Plate/ug
-81%
-76%
-81%
-65%
0%
0%
-94%
-90%
-89%
-83%

Rev/kW-hr
-76%
-69%
-75%
-54%
0%
0%
-91%
-86%
-85%
-79%
-0%
Rev/Plate/ug
+156%
+367%
—
3%
0%
0%
0%
0%
0%
0%
Methanol 13-Mode
Diesel 13-Mode
Methanol + Catalyst 13-Mode Ethanol + H20 - 13-Mode
+70% -52%
-65%
Rev/kW-hr
+156%
+367%
—
+ 3%
0%
0%
+ 50%
- 17%
0%
0%

Rev/Plate/ug
- 43%
- 21%
+ 13%
+ 44%
0%
0%
0%
0%
- 33%
+188%

Rev/kW-hr
-80%
-73%
-60%
-50%
0%
0%
-33%
-60%
-75%
0%
- 5%
Rev/Plate/ug
- 56%
0%
__
+ 60%
0%
- 67%
-100%
- 83%
- 83%
- 38%

Rev/kW-hr
- 56%
- 08%
—
+ 61% ,
0% £
- 67% '
-100%
- 83%
- 83%
- 38%

-------
                                      -53-

higher with  the catalyst  than  without the  catalyst.   Therefore, one  would
conclude that the test cycle  is  very  important when analyzing the Ames  test
data from those engines (18).

-------
                                      -54-
V.                           References

1.  Springer, Karl J.,  "Characterization of Sulfates, Odor, Smoke,  POM,  and
    Particualates From  Light  and Heavy Duty Engines  - Part IX",  EPA-OMSAPC
    Contract    68-03-2417    with    Southwest    Research,    Report    No.
    EPA-46013-79-007, June 1979.

2.  Ullman,  T.L.,  Springer,  K.J.,  and  Baines,   T.M.,   "Effects   of  Six
    Variables on  Diesel Exhust  Particulate",  ASME Paper  80-DEP-42,  February
    1980.

3.  Baines, Thomas M.,  Somers,  Joseph H., and Harvey, Craig A.,  "Heavy Duty
    Diesel  Particulate  Emission  Factors",  Journal  of  the  Air  Pollution
    Control Association, 29, 616, 1979.

4.  EPA-OMSAPC Contract  No. 68-03-2706 with Southwest Research,  "Unregulated
    Emissions Characterization of Heavy-Duty Diesel  and Gasoline  Engines  and
    Vehicles", contract  in progress.

5.  Hare, Charles T.,  "Characterization  of  Gaseous and Particulate  Emissions
    from Light  Duty  Diesels Operated on Various Fuels",  EPA-OMSAPC  Contract
    No.   68-03-2440   with   Southwest    Research   Institute,    Report   No.
    EPA-460/3-79-008, July 1979.

6.  Springer, Karl J.  and  Baines,  Thomas M.,  "Emissions  from Diesel  Versions
    of Production Passenger Cars", SAE Paper 770818,  September  1977.

7.  Dietzmann, Harry  E., Smith,  Lawrence R.,  Parness, Mary Ann,  and Fanick,
    E.   Robert,   "Analytical   Procedures  for  Characterizing   Unregulated
    Pollutant   Emissions  from   Motor   Vehicles",   EPA-ORD  Contract   No.
    68-02-2497    with    Southwest   Research    Institute,    Report    No.
    EPA-600/2-79-017, February 1979.

-------
                                      -55-
8.  Smith,  Lawrence,  Parness,  Mary Ann,  Fanick,  E.  Robert,  and  Dietzmann,
    Harry E.,  "Analytical  Procedures for Characterizing  Unregulated Emissions
    from Vehicles Using  Middle  Distillate Fuels", EPA-ORD  Contract  68-02-2497
    with Southwest Research Institute,  Report No. EPA-60012-80-068,  April 1980.

9.  Urban,   Charles,   "Regulated  and   Unregulated   Exhaust  Emissions   from
    Malfunctioning Non-Catalyst  and Oxidation Catalyst Gasoline  Automobiles",
    EPA-OMSAPC  Contract No.  68-03-2499  with  Southwest  Research  Institute,
    Report No. EPA-460/3-80-003, January 1980.

10. Urban,   Charles,   "Regulated  and   Unregulated   Exhaust  Emissions   from
    Malfunctioning   Three-Way  Catalyst   Gasoline  Automobiles",   EPA-OMSAPC
    Contract  No. 68-03-2588  with  Southwest  Research  Institute,  Report  No.
    EPA-460/3-80-004, January 1980.

11. Urban,  Charles,  "Regulated  and  Unregulated  Exhaust  Emissions  from  a
    Malfunctioning   Three-Way   Catalyst  Gasoline   Automobile",   EPA-OMSAPC
    Contract  No. 68-03-2694  with  Southwest  Research  Institute,  Report  No.
    EPA-460/3-80-005, January 1980.

12. Urban, Charles M. and  Garbe,  Robert  J.,  "Regulated and Unregulated Exhaust
    Emissions from Malfunctioning Automobiles", SAE Paper 790696,  June 1979.

13. Urban,   Charles   M.   and  Garbe,   Robert   J.,  "Exhaust  Emissions   From
    Malfunctioning Three-Way Catalyst Equipped Automobiles",  SAE  Paper 800511,
    February 1980.

14. Hare, Charles  T. and  Baines,  Thomas  M., "Characterization of  Particulate
    and Gaseous  Emissions  from Two  Diesel Automobiles as  Functions  of Fuel and
    Driving Cycle", SAE Paper 790424, March 1979.

15. EPA-OMSAPC  Task  Order Contract  No. 68-03-2884  with  Southwest  Research
    "Basic  Characterization  Support  for   the   Emission  Control   Technology
    Division", contract in progress.

-------
                                      -56-
16. EPA-OMSAPC  Task Order  Contract  No.   68-03-3073  with  Southwest  Research
    Institute,   "Pollutant   Assessment  Support   for   the  Emission   Control
    Technology Division", contract in progress.

17. Newman, Frank M., Russell,  John A.,   Bowden, John N.,  and  Johnston,  Alan
    A.,  "Impact  of  Coal  and   Oil  Shale  Products  on  Gasoline  Composition
    1976-2000,  Task  1   Final   Report",  EPA-OMSAPC  Contract  68-03-2377  with
    Southwest Research Institute, December 1976.

18. Ullmann, Terry  L.,  and  Hare, Charles T., "Emission Characterization of an
    Alcohol Fueled/Diesel Pilot  Compression-Ignition Engine and  its Heavy-Duty
    Diesel  Counterpart", EPA-OMSAPC  Contract  No.  68-03-2884,  Task  6,  with
    Southwest Reearch, Final. Report in Press.

19. Holmer,   E.,   Berg,  P.S.,   and   B-I   Bertilsson,   "The  Utilization  of
    Alternative  Fuels  in a Diesel  Engine  Using Different Methods",  SAE Paper
    800544, February 1981.

20. Hare,  Charles  T.  and  Montalvo,  Daniel  A.,  "Diesel  Crankcase  Emissions
    Characterization",   EPA-OMSAPC  Contract  No.  68-03-2196,   Task   4,  with
    Southwest Research, May 1977.

21. Fine,  David H. and  Goff, Ulku, "Nitrosamine Analysis of  Diesel  Crankcase
    Emissions",  EPA-OMSAPC  Contract 68-03-2719  with  New  England  Institute of
    Life Sciences,  Report No. EPA-460/3-81-008,  March 1980.

22. Baines, Thomas  M. ,  "Nitrosamines  and Other  Hazardous  Emissions From Engine
    Crankcases", EPA-OMSAPC Report No. EPA/AA/CTAB/PA/81-15,  June 1981.

23. Smith,  Lawrence  R.,   "Nitrosamines   in  Vehicle   Interiors",  EPA-OMSAPC
    Contract No. 68-03-2588 with Southwest Research,  Final Report in Press.

24. Goff,  E.  Ulku, Coombs, James  R. ,  Fine, David H. ,  and Baines, Thomas M.,
    "Nitrosamine  Emissions  from Diesel Engine  Crankcases",  SAE  Paper 801374,
    October 1980.

-------
                                       -57-
25. Somers,  Joseph  H.,  "Brief  Synopsis  of  EPA  Office   of   Research  and
    Development and  Health  Effects  Institute  Mobile  Source Work",  EPA-OMSAPC
    Report No. EPA/AA/CTAB/PA/81-10, May 1981.

26. Schuetzle,  Dennis,   Lee,   Frank  S.C.,  and  Prater,   Thomas  J.,   "The
    Identification   of  Polynuclear   Aromatic  Hydrocarbon   Derivatives   in
    Mutagenic Fractions of Diesel Particulate  Extracts",  International  Journal
    of Environmental Analytical Chemistry, 9_,  93,  1981.                  _

27. Penninga,  Thomas J., "Evaluation  of Mutagenic  Characteristics of  Diesel
    Gaseous Hydrocarbons", EPA-OMSAPC Report No. EPA/AA/TEB/81-22, July  1981.

28. EPA-ORD Mobile  Sources/Health  Effects Project Status Report, 1981  -  Work
    in progress.

-------
                                      -58-
VI. Appendixes

                                  Appendix 1

             Compounds Identified in Diesel Particulate Emissions


This Appendix is reprinted directly from the following publication:

Dennis  Schuetzle,  Frank S.C. Lee,  and  Thomas J. Prater,  Ford Motor  Co.,  and
Silvester  B.  Tejada,  EPA, "The  Identification  of Polynuclear  Aromatic  Hydro-
carbon  Derivatives in  Mutagenic Fractions  of  Diesel Particulate  Extracts",
presented  at  the  10th  Annual Symposium  on  the  Analytical  Chemistry of  Pol-
lutants,  Dortmund,  Germany,  May  28-30,  1980   which was  published  in  the
"International Journal of Environmental Analytical Chemistry",  9, 93 (1981).

-------
                                             -59-
                                           TABLE  IV
                      RELATIVE A1TJMDANCZS  OF  PAH  SPECIES  IDENTIFIED 3Y
             GC/MS (25m DEXSIL 300 CAPILLARY COLUMN7)  AND H2MS FOR THE PAH!
                    HPLC  FRACTIONS  OF A DIESEL  PARTIOJLATE  EXTRACT (OL-1)
                       Compounds

Phenanthrene and Anthracene
Methyl  (Phenanthrenes and Anthracenes)
Flvioranthene
Pyfene
Dihydro  (Fluoranthenes and Pyrenes)
Diaethyl  (Phenanthrenes and Anthracenes)
Benzofluorenes
Methylpyrenes
Behzylnaphthalenes
Trimechyl  (Phenanthrenes or Anthracenes)
Cyclopentapyrene
Berizo(a) Anthracene(BaA); Chrysene»and Triphenylene
Diphenylbenzene
Methyl  (BaA, Chrysene and Triphenylene)
Benzofluoranthenes
Benzo(e) Pyrene(BeP), Benso(a) Pyrene(BaP) and Perylene
Methyl Benzofluoranthenes
Methyl  (BeP, BaP, Perylene)
Benzo(g,h,i) Perylene and Anthanthrene
Coronene
Dibenzopyrenes

Formula
C14H10
C15H12
C16H10
CI&HIO
C16H12
Cl6H14
C17H12
C17H12
C17H14
C17H16
CisH^o
C18H12
C18H14
C19H"4
^20^12
C20Hi2
^21^14
^21^14
C22H12
C2/Kj.2
C24H14
1
Mass
178.078
192.094
202.078
202.078
204.094
206.110
216.094
216.094
218.110
220.126
226.078
228.094 .
23o.no
242,110
252.094
252.094
266.nO
266.110
276.094
300.094
302.no.
Relative2
Abundance
0.27
0.54
0.41
• 0.41
0.09
1.00
0.67
0.55
0.18
0.37
0.47
0.16
0.11
0.42
0.33
0.48
0.13
0.16
0'.25
0.06
0.02
 "Measured to within + 10 ppm at 12K resolution.
 2  ,                                .
  Abundance relative to dimethyl (phenanthrenes and anthracenes)

-------
                                                            TABLE V




                        CHEMICAL COMPOSITION OF THE fl  -  IJPLC FRACTIONS OF A DIESEL PARTICULATE EXTRACT  (NI-1)
                COMPOUND




Anthaulhroiie or Isomera (C22"l2)




Methyl Am hanthrene  or Isomers (C23"l4)




Coronene  (C2/,l'l2)




I)il>ai»/o|>yrune  (C2/lUi/,) - 2 leoiners




Methyl Coronenu




llydroxy Coronene




Cyclopentacoronene or
MASS
Found
276.097
290.108
300.092
302 . 106
314.109
316.091
326.111
Theory 24 25
276.094 0.181 0.009
290.109 0.362 Q.,269
300.094 0.121 ' 0.159
302.109 0.481 0.097
314.109
316.089 0.034 0.042
326.110
RELATIVE ABUNDANCE
26 27
'
-.-!••
1.000 0.480
0.068 0.128
0.010
0.049 0.051
0.053
(FRACTION (1)
28 29
-
-
0.293 0.073
0.202 0.174
.0.028
0.032
i
0.162 0.109

30
-
-
-
-
'-
~ i
cr>
0.043 °

-------
          COMPOUND


 2-Naphthalene Carboxaldehyde


 .1-Naphthalene Carboxaldehyde


 Methyl Naphthalene Carboxaldehyde


i Naphthalene Acetaldehyde


 Naphthalene Acetaldehyde


 Dimethyl Naphtliaiene-Aarhoxaldehyde

i
 9-Fluorenone


'Naphthalene DiearboxarWehyde


 Phenanthrene or Anthracene


 Naphthalene Dicar^oxaidehyde
i

.Phenanthrone or-iAnchroTte


 Naphthalene Dicar-boxald«hyde


j9-Anthrone


 Trimethyl Naphthalene Carboxaldehyde


 Phenanthrone or Attthrooe


 Trlmethylnaphthalene Carboxaldehyde


 Fluorene Carfaoxaldehyde


Methyl (Anthracene or Phenanthrene)


Methyl (Anthracene or Phenanthrene)


 Fluorene Carboxaldehyde


 9-Xanthone


Triaethyl Naphthalene Carboxaldehyde


Fluorene Carboxaldehyde


Methyl (Anthracene or Phenanthrene)
-51-
TASLE VI



>r£CIES IDENTIFIED BY GC/MS (S? 2250 COLUMN) IK THE
. 0? A DIESEL PARTICULATE EXTRACT-OL-I
FORMULA
C11H8°
C1IH8°
C12H10°
C12H10°
C12H10°
C12H8°2
C13H8°
C12H8°2
C14H10
C12H8°2
C14H10°
C12H8°2
C14H10°"
C14H14°
C14H10°
C14*14°
C14H10°
C15H12
C15H12
C14H10°
C13H8°2
C14H14°
C14H10°
C15H12
MASS
156.061
156.061
170.073
170.073
170.073
184.052
180.058
184.052
178.078
184.052
194.073
184.052
194.073
198.104
194.073
198.104
194.073
192.094
192.094
. 194.073
196.052
198.104
194.073
192.094
SCANtf
660
673
779
789
795
868
881
894
895
925
941
943
947
950
961
965
989
991
998
1000
•1004
1010
1021
1022
RELATIVE
ABUNDANCE
0.035
0.059
0.062
0.073
0.015
0.098
1.000
0.1^0
0.260
0.047
0.110
0.028
0.040
0.061
0.197
0.062
0.176
0.128
0.168
0.176
0.039
0.031
0.134
0.136

-------
                                          -62-
         COMPOUXD
Trimechyi  Naphthalene Carboxaldehyde

2-Methyl  S-Anthror.e

Methyl (Anthrone or Phenanthrone)

Dimethyl  (Anthracene or Phenanthrene)

9,10 Phenanthreae Quinone

(Pheaanthrene or Anthracene) Quinone

Dimethyl  (Anthracene or Phenanchreae)

9,10 Anthracene Quinoae

(Phenanthrene or Anthracene) Quisone
              •
Pyrene

Methyl 9,10 Phenanthrene Quinone

Unknown

trimethyl (Anthracene or Phenanthrene)

Methyl 9,10 (Anthracene or Phenanthrene) Quinone

Trimethyl (Anthracene or Phenanthrene)

Fluoranthene

Thioxanthone

Trimethyl (Anthracene or Phenanthrene)

Methyl 9,10 Anthracene Quiaoae

Pyrone

Thioxaathone

(Phenanthrene or Anthracene) Carboxaldehyde

9-Thioxanthone

(Phenanthrene or Anthracene) Carfaoxaldehyde

Fluoranthone or Pyrone
 i
Fluoranthone or Pyrone

Methyl (Phenanthrene or Anthracene)  Carboxaldehyde
FORMULA
C14E14°
C15K12°
C15H12°
C16H14
C14H10°2
C14E10°2
C16E14
C14H10°2
C14H10°2
C16H10
C15E19°2
C15H8°
C17E16
C15H10°2
C17H16
C16H10 '
C13H8°S
C17E16
C15H10°2
C16E10°
C13H8OS
C15H10°
C13E8°S
C15E10°
C16E10°
C16H10°
C16H12°
MASS
198.104
208.089
208.089
206.109
208.052
208.052
206.109
208.052
208.052
202.078
222.068
204.057
220.125
222.068
220.125
202.078
212.030
220.125
222.068
218.073
212.030
206.073
212.030
206.073
218.073
218.073
220.089
SCAN:?
1040
1057
1077
1090
1094
1105
1111
1122 .
1142
1157
1171
1172
1175
1193
1199
1205
1208
1217
1231
1235
1237
1249
1251
1260
1270
1284
1316
RELATTvl
ASU^DAI-CE
0.030
0 . 241
0.144
0.121
0.063
0.070
0.057
0.231
0.044
0.237
0.180
1.041
0.020
0.092
0.095
0.190
0.057
0.046
0.095
0.087
0.061
0.354
0.269
0.254
0.041
0.043
0.085

-------
                                          -63-
          COM?OU!vD
 Unknown

'Methyl (Phenanthrene or Anthracene) Carboxaldehyde

 Methyl Pyrene
            •
 Pyrene Quinone

• iMethyl (Phenanthrene or Anthracene) Carfaoxaldehyde

 Methyl Fluoranthene

 Fluoranthene Quinone

 Methyl (Phenanthrene or Anthracene) Carboxaldehyde

 Dimethyl  (Phenanthrene or Anthracene)
 Carboxaldehyde

 Pyrene or Fluoranthene Carboxaldehyde

 Dimethyl  (Phenanthrene or Anthracene)
 Carboxaldehyde.

 Benzo(d,h)  Anthrone

 Dimethyl  (Phenanthrene or Anthracene)
 Carboxaldehyde

 (Benzo(a)anthracene, Chrysene or Triphenylene)
 Carboxaldehyde

 (Benzo(a)anthracene, Chrysene or Triphenylene)
 Carboxaldehyde

 (Benzo(a) anthracene, Chrysene or Triphenylene)
 Carboxaldehyde

 (Benzo(a)anthracene, Chrysene or Triphenylene)
 Carboxaldehyde •

 1-Nitro Pyrene
  ihe relative abundance of 9-Fluorenone is arbitrarily given  a value of 1.000.
 •>
 "Mass accuracy is poor due to low ion abundance,  therefore, exact  empirical formula is
  'difficult to establish.
FORMULA

C16H12°
C17H12
C16H8°2
C16H12°
C17H12
C16H8°2
C16H12°
C17H14°
C17H1Q0
1/14
1/10
C17H14°
C19H12° •
C19H12°
C19H12°
C19H12°
C16H9N°2-
MASS
2262
220.089
216.094
232.052
220.089
216.094
232.052
220.089
234.104
230.073
234.104
230.073
234.104
256.089
256.089
256.089
256.089
247.063
SCAN)?
1318
1338
1338
1338
1355
1357
1357
1369
1413
1415
1440
1450
1462
1499
1518
1523
1536
1624
RELATIV:
ABUNDA:-;
0.02^
0.188
0.024
0.024
0.126
0.221
0.023
0.094
0.057
0.409
0.051
0.073
0.027
0.032
0.028
0.028
0.032
0.027

-------
                                          -64-

                                        TASLE VII
                A COMPARISON OF THE RELATIVE ABUNDANCES   0?   PAH DERIVATIVES

                    IDENTIFIED BY DIRECT PROSE ERMS AND GC/MS FOR THE

           TRANSITION  (f   HPLC FRACTION OF A DIESEL PARTICULATE EXTRACT (OL-i)
                                                                 RELATIVE ABUNDANCE
       POSSIBLE COMPOUNDS                   FORMULA    MASS   KRKS   GC/TlS71   HRMS -GC/MS



 9-Fluorenone                               C. ,H00   180.058  1.00   1.00        0.00
                                             J.J 0



 Hydroxy Fluorene*                          C13H10°  182'073  °'15   °-°°        0.15*



 Naphthalene Dicarboxaldehyde       .        C12H8°2  18*'052  °'29   °'32      -0.03



 Methyl (Anthracene  or  Phenanthrene)        C15K12   192. 09 A  0.45   0.43        0.02
 (Anthrone  or  Phenanthrone                  C14H10°  I94-073  °-96   °-83       0.13*

 Fluorene Carboocaldehyde

 Methyl Fluorenone

 Hydroxy (Anthracene or Phenanthrene)*
 Fluorene Quinone


 9-Xanthone                                 C13H8°2  196'052  °'66   Q-^       0.62*
 Hydroxy Fluorenone*






 Hydroxy Xanthene*                          C13H10Q2 198'068 • °'30   °'00       °-30*


 Dihydroxy Fluorene*



 Trimethyl Naphthalene Carboxaldehyde       CiAHiA°  198-10A  °-29   0.19   •    0.10



 Pyrene or Fluoranthene                     C.,H,n   202.078  0.37   0.43      -0.05
                                             ia 1U


 Unknown                                    C. CH00   204.057  0.80   1.04      -0.24
                                             -L.J o


 Anthracene or Phenanthrene (Carboxaldehyde)C.-H..O  206.073  0.76   0.65       0.11



 (Phenanthrene or Anthracene)  Quinone     '  C..H-0.  208.052  0.39   0.4.1      -0.02



                                -v            C..H..O  208.089  0.35"  0.39      -0.04
'Methyl(Anthrone  or  PhenanthroneJ             15 12




 Thioxanthone'                               C.-H-OS  212.030  0.47   0.39       0.08
                                             J.J o


 Methyl  Naphthalene  Dicarboxylic AcidL       c -11-0,  212.047  0.39   0.00       0.39*

 Anhydride:                                   13 5 3

 Hydro-xy Xanthone

-------
                                          ~65~                   RELATIVE ABUNDANCE
      POSSIBLE  COMPOUNDS                    FORMULA    MASS   HRMS   GC/MS   KRMS-CC/MS



 Fluoranthone or Pyrone                     C .H  0  21S.073 <0.15   0.17      -0.02
                                             iO J. J


'Methyl  (Anthrone or  Phenanthrene)           C.,E..O  220.089  0.60   0.59       0.01
                                             lo lx


(Dihydropyrone  or Dihydrofluoranthene)



 Aceanthrone  or -Aoephenanthrone
Methyl  (Anthracene or Phenanthrene) Quinone c1eH1o°2  222t068   °*39    °'37       °'01



Unknown





Dimethyl   (Phenanthrone or Anthrone)         C^H^O  222.104   0.76     -         0.76*



Dimethyl Hydroxy  (Phenanthrene  or Anthracene)*

Trimethyl  Fluorenone



Dihydroxy  Methyl  (Anthracene  or            C15H12°2  224-084   °'4*     ~         0.44*

Phenanthrene)*

Hydroxy Nitro  Fluorene*                   .  C13H9N03  227.058   0.18     -         0.18*





(Pyrene or Fluoranthene)  Carboxaldehyde     C^-H  0   230.073   0.48    0.52      -0.04

Benzo(d,H)  Anthrone


                                                                                \


Dimethyl  (Anthracene  or Phenanthrene)       ci7Hi/°   234.104   0.49    0.14       0.35*

Carboxaldehyde;                              iy 1
i
Unknown*



Dihydrodihydroxy  (Pyrene  or Fluoranthene)*  C-,H.-0.  236.084   0.30     -         0.30*
                                             XO 1-jL L


Hydroxy Trimethyl  (Anthracene or            C17H16°   236-120   °'18     "         0.18*

Phenanthrene)*



Nitro Methyl  (Anthracene  or Phenanthrene)*  CL5H11N02237.079   0.14     -         0.14*
1 Those "species that were not eluted in the GC/MS procedure but may be present based upon

 HRMS relative abundance values are designated bv a star*

2 Error in relative abundance values are jr7%.

  GC/MS identifications for this fraction are  presented  in Table  VI.

-------
                                              -66-
                                             TAELE VIII
       RELATIVE ABUNDANCE  Or  CHEMICAL  SPECIES  IDENTIFIED  BY  GC/MS (DEXSIL 300) IN THE
       ANALYSIS OF A  TRANSITION  (Y2) K?LC  FRACTION  OF  A DIESEL PARTICULARS EXTRACT (OL-1)
                        Compound

 9-Fluorenone
Metnyj. Fluorenone
Methyl Fluorenone
Methyl Fluorenone
I Naphto  (c,d) Pyrone
Dimethyl Naphthalene Carboxaldehyde
An'throne or Phenanthrone
Lvsphto, 2-3-B-Furan-4,9-dione
Naphthalene Dicarboxaldehyde
Fluorene Quinone
Fluorene Quinone
9-Xanthone
Unknown
Flucrene Quinone
Unknown
Naphthalene Dicarboxylic Acid Anhydride
(Phenanthrene or Anthracene)'Quinone
iferhyl (Anthrone or Phenanthrone)
(Anthracene or Phenanthrene). Carboxaldehyde
rluorene Quinone
CPhenanthrene or Anthracene) Quinone
Methyl Naphthalene Dicarboxylic Acid Anhydride
-iethyl Fluorene Quinone
Methyl Fluorene Quinone
Methyl Fluorene Quinone
 •iethyl Naphthalene Dicarboxylic Acid Anhydride
Diaechyl Fluorene Quinone
•isthyl Naphthalene Dicarboxylic Acid Anhydride
}imethyl Fluorene Quinone
 )imethyl Naphthalene Dicarboxylic Acid Anhydride
 )iaethyl Fluorene Quinone
 Jiaethyl Naphthalene Dicarboxylic Acid Anhydride
                                 «
 .-iaethyl Naphthalene Dicarboxylic Acid Anhydride
 Flucranthene or Pyrene) Quinone
 .ethyl (Pyrene or Fluoranthene) Quincne
 Benzo(a)Anthracene,  Chrysene or Triphenylene) Quinone
 Benzo(a)Anthracene,  Chrysene  or Triphenylene)  Quinone

Foraula
C13HgO
ci4BioO
C14H1QO
C14H10°
C12H802
C13H120
Ci4H100
Cj^2Hft03
C12H802
C13HS02
C13H802
C13H802
^12^10^3
C13H802
Cl2Ki003
C12H603

C15E120

C13H802
^14^10^2
Ci3H803
^14^10^2
C14Hio02
C14H10°2
C13H803
C15Hi202
C13EB03
C15H1202
C]_4Hi n03
C15H1202
Cl4Hl(j03.

Ci4H1003
Cj_gE802
C17H10°2
^18^10^2
C1SK10°2

Mass
180.058
194.073
194.073
194.073
184.052
184.089
194.073
198.032
184.052
196.052
196.052
196.052
202.064
196.052
202.064
198.032
208.052
208.089
206.073
196.052
208.052
212.047
•210.063
210.068
210.068
212.047
224.084
212.047
224.084
226.064
224.084
226.064

226.064
232.052
246.063
258.068
253.068

Scar. T?
719
758
787
801
808
846
853
862
863
876
888
900
901
904
912
913
913
923
928
935
948
970
973
979
995
980
1010
1015
1030
1030
1055
1059

1093
1148
1210
1225
1235
• Relative
Abundance
0.157
0.069
0.041
0.046
0.105
0.040
0.175
0.099
0.054 .
0.033
0.040
0.073
0.037
0.133:
0.030
1.000
0.123 :
0.162
0.020
0.053.
0.064
0.123 •
0.102
0.091
0.005
0.111
0.043
0.108
0.035
0.036
0.021
0.120
(2 Isomers)
0.012
0.026
0.005
0.006
0.012

-------
                                            -67-
                                           TABLZ IX
    A COMPARISON OF THE RELATIVE ABUNDANCES OF PAH DERIVATIVES IDENTIFIED BY DIRECT PROBE
   AND CC/MS FOR TEE TRANSITION (^2) H?LC FRACTION OF A DIESEL PARTICULARS  EXTRACT  (OL-1)
                                                                         Relative  Abundance.___
                                                                                               r.
Possible Compounds
-Fluorenone
ydroxy Fluorene*
aphthalene Dicarboxaldehyde
iaethyl Naphthalene Carboxaldehyde
athrbne or Phenanthrone
Methyl Fluorenone
9-Xanthone; Fluorene Quinone
ethyl Hydroxy Fluorene*
aphthalene Dicarboxylic Acid Anhydride
/droxy Xanthene*
Dihydroxy Fluorene*
iknown
3henanthrene or Anthracene) Carboxaldehyde
Phenanthrene or Anthracene) Quinone
ithyl (Anthrone or Phenanthrone)
Diaethyl Fluorenone
Methyl Hydroxy (Anthracene or Phenanthrene)*
Jthyl Fluorene Quinone: Dihydroxy Anthracene*
Laethyl Hydroxy Fluorene*
ithyl Naphthalene Dicarboxylic Acid Anhydride
Lhydroxy Methyl Fluorene*
^droxy Naphthalene Dicarboxylic Acid Anhydride*
rihydroxy Fluorene*
;thyl (Anthracene or Phenanthrene) Quinone
^thracene or Phenanthrene) Dicarboxylic
Acid Anhydride
Lnethyl Fluorene Quinone
jnethyl Naphthalene Dicarboxylic Acid Anhydride
Lnydroxy Methyl Fluorene*
iknown*
:knowh*
'luoranthene or Pyrene) Quinone
.hydroxy Dinethyl Anthracene*
.known*
Those species that were not eluced in the GC/MS
Formula
C]jH80
C13Hi00
C12H802
C].3Hi20
Gi4Hi00

C]^H802
Cl4H120
C12H603
Cl3H8°2

C12H1003
C15H100
C1AK802
C15H120


Cl4H]_o02
^15^140
C13H803
^14^12^2
C12H604
C13H1003
C15H1002
C1AH803

C15H1202
Cj_4H]_Q03
Ci5H]_402
C^3H804
^14^12^3
C16H8°2
Cl6Hl4°2

procedure
Mass
180.058
182.072
184.052
184.089
194.073

196.053
196.089
198.031
198.067

202.064
206.073
208.052
208.089


210.068
210.104
212.047
212.084
214.026
214.063
222.068
224.047

224.084
226.064
226.099
228.042
228.078
232.052
238.100
242.094
but nay be
KRMS
0.23
0.30
0.11
0.11
0.30

0.29
0.13
1.00
0.38

0.05
0.03
0.20
0.23


0.27
0.07
0.36
0.20
0.32
0.06
0.26
0.21

0.17
0.16
0.08
0.06
0.03
0.04
O.OS
0.03
present
GC/MS
0.16
-
0.12
0.03
0.33

0.33
-
1.00
-

0.07
0.02
0.19
0.15


0.20
-
0.34
-
-
-
(?) .,
_ (.'

0.10
0.17
-
-
-
0.03
-
••
based
' KRMS-GC/:
0.07
0.30*
-0.01
0.08
-0.03

-0.03
0.13*
-
0.38*

-0.02
0.01
0.01
0.07*


0.07*
0.07*
0.02
0.20*
0.32*
0.06*
i\ ""


0.07
-0.01
0.08*
0.06*
0.06*
0.01
0.08*
0.03*
upon HRMS
relative 'abundance values arc designated by a star*
GC/MS identifications for this fraction are presented in Table VIII.
Presence of compound not confirmed by GC/MS

-------
                   -68-

                Appendix 2

Compounds in Diesel Particulates Found by
     Battelle  Northwest  (Reference  28)

       	Compound	

  Polycyclic Aromatic Hydrocarbons
  Alkylated fluorenones
  Alkylated phenols
  Aliphatic acids
  Methyl esters
  Alkanones
  Phenanthrenequinone
  Benzoic Acid
  Xanthen-9-one
  Isobenzofuranone
  N-aminopiperidine

-------