States
Environmental Protection
Agency
                             V **f-
Environmental Sciences Research   -^ j x />, C
Laboratory                      /• j|-  ^
Research Triangle Park NC 27711      ''
Research and Development
EPA-600/S3-83-087 Jan. 1984
Project  Summary
Characterization of
Emissions  and  Fuel Economy  of
In-Use  Diesel  Automobiles

Richard E. Gibbs, James D. Hyde, Robert A. Whitby, and Delip R. Choudhury
  Exhaust emissions  from twenty
1977-1980 in-use light-duty diesel ve-
hicles were measured to determine the
effects of driving cycle, mileage ac-
cumulation and test conditions. Hydro-
carbons, carbon monoxide, carbon
dioxide,  nitrogen oxides and particu-
lates were measured from the Federal
Test Procedure (FTP),  Highway Fuel
Economy Test (HFET), Congested Free-
way Driving Schedule (CFDS), and
New York City Cycle (NYCC), 50 mph
cruise (50C), and idle. Individual partic-
ulate samples were Soxhlet extracted
with dichloromethane to partition the
particulate into extract (soluble) and
residue (insoluble). The extracts were
tested for mutagenicity by the Ames
Salmonella typhimurium/microsome
method. Detailed chemical analysis and
subsequent bioassay was performed on
selected composite particulate samples.
  Emissions (g/mi) and fuel consump-
tion by driving cycle generally increased
in the order 50C< HFET < CFDS < FTP
<  NYCC.  Vehicles in the General
Motors  group generally had  higher
emissions than the Mercedes-Benz and
Volkswagen groups and were more
sensitive to  driving cycle. The extract
showed very little cycle dependence but
the residue was very cycle dependent.
NOx emissions decreased with mileage
accumulation, while other emissions
increased or were unaffected.
  Fuel economy was determined by the
carbon balance method, by fuel meters,
and by fueling records. Over-the-road
fuel economy was always lower than
carbon-balance fuel economy and was
best approximated by the FTP.
  A new method for real-time particulate
measurements is described  using a
Tapered Element Oscillating Microbal-
ance (TEOM). The TEOM mass agreed
to within 10% of the gravimetric mass
on average with a response time of 8-
15s.
  Short studies were performed on the
effects of driving cycle sequence, the
dilution tunnel, sub-FTP temperature
and mutagenic artifact formation.
  Bulk extract samples were fraction-
ated and analyzed by GC, GC/MS and
HPLC/UV. The highest specific activity
was in the acidic fraction, but most total
activity was in the neutral  fraction
which contained fluorenones and oxy-
PAH's.
  This Project Summary was developed
by EPA's Environmental Sciences
Research Laboratory. Research Triangle
Park, NC. to announce key findings of
the research project that is fully docu-
mented in a separate report of the same
title  (see Project  Report ordering
information at back).

Introduction
  The diesel  passenger car  has  only
recently become a significant contributor
to automobile pollution. From  1975 to
1980 diesels  increased from 0.05% to
0.64% of the light-duty fleet in New York
State and penetration has been predicted
to be  as high as 25% by the year 2000.
  This study grew out  of a need for
comprehensive emissions data from in-
use diesel automobiles  as opposed to
certification data. Particulate emissions
have  been of special concern  because
they are a visible pollutant with a high

-------
potential for adverse health effects. Few
data were available to assess the effects
of non-FTP driving,  vehicle aging and
real-world use/abuse of light-duty diesel
automobiles. The compounds responsible
for the mutagenicity in diesel particulate
extract have not been adequately charac-
terized and the  applicability of current
emissions test methods to such areas as
mutagenicity  testing required further
study.
  Several short experiments were con-
ducted to investigate special data inter-
pretation situations.  The topics investi-
gated were the effects of driving cycle
sequence on emissions of the dilution
tunnel on emissions and mutagenicity, of
filtered exhaust  gas on particulate and
extract mutagenicity, and of sub-FTP tem-
perature soaking on particulate emissions.
These experiments were conducted with
a  limited number of vehicles and,
therefore, the results may not be generally
applicable.
Experimental Approach
  Table 1  lists the test vehicles by their
car number designations which are used
in subsequent data presentation.  Cars 1
and 5 were purchased new to be used as
loan cars  for people whose private
vehicles were being tested. Cars 2 and 3
were purchased new for use by the New
York State Thruway Authority to serve as
high mileage accumulation vehicles. Car
10 was also operated by the Thruway
Authority. Except for Car 21, the other
vehicles were  privately owned and
maintained. Vehicles were  divided into
groups according to manufacturer  as
follows:
     GM: Cars 2, 3, 4, 5, 7, and 16
     VW: Cars 1, 6, 8, 9, and 11
     MB: Cars 12,  13, 14, and 19
     Other: Cars 10,15,17, 18, and 21.
     (Car 18 was excluded from the GM
     group  and Car 20  was excluded
     from all groups.)
  The  test protocol employed two  or
three replicate driving cycle sequences to
test  different fuel/lubricating  oil test
conditions.  Each sequence of driving
cycle was called a  "Phase." The three
phases were:

     Phase 1 -vehicle tested as received;
     Phase 2 - project control fuel, as re-
             ceived oil;
     Phase 3 - project control fuel, fresh
             oil of manufacturer spec-
             ifications.

Table 2 shows the phases and driving
cycle sequences used.
  Hydrocarbons, carbon monoxide, carbon
dioxide, nitrogen oxides, andparticulates
were measured from the Federal Test
Procedure (FTP), Highway Fuel Economy
Test  (HFET), Congested Freeway Driving
Schedule (CFDS), New York City Cycle
(NYCC), 50 mph cruise (50C), and idle.
Fuel economy  was measured by the
carbon balance method for dynamometer
tests and by fuel  meters  and fueling
records for over-the-road  tests.
  Particulate samples were collected  on
50-  x  50-cm Pallflex T60A20 Teflon-
coated fiberglass filters. The particulate
was  Soxhlet extracted with dichlorome-
thane for 24 hours. The resultant extract
was  dried, weighed and an aliquot tested
in dimethylsulfoxide for mutagenicity by
Table 2.    Vehicle Test Driving Cycle
          Sequences
                          B
Table  1.   Vehicle Specifications and Dynamometer Test Conditions
Car it
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
Year
79
79
79
79
79
80
79
78
79
78
77
77
78
79
79
79
79
80
79
78
78
Make
VW
Olds
Olds
Olds
Olds
VW
Cadillac
VW
VW
Dodge
VW
M-B
M-B
M-B
Audi
Olds
Peugeot
Olds
M-B
Olds
Dodge
Model
Rabbit
Cutlass Cruiser
Cutlass Cruiser
98 Regency
Cutlass Cruiser
Rabbit
Eldorado
Rabbit
Rabbit
D-JO Mitsubishi
Rabbit
240-D
300-CD
240-D
5000
Delta 88
504
Cutlass Cruiser
3OO-SD (Turbo)
Delta 88
Tradesman 200
Engine
Displacement
1-4
V-8
V-8
V-8
V-8
1-4
V-8
1-4
1-4
1-6
1-4
1-4
1-5
1-4
1-5
V-8
1-4
V-8
1-5
V-8
1-6
1.5L
5.7 L
5.7 L
5.7 L
5.7 L
1.5L
5.7 L
1.5 L
1.5L
4.0 L
J.5L
2.4 L
3.0 L
2.4 L
2.0 L
5.7 L
2.3 L
5.7 L
3.0 L
5.7 L
3.3 L
Trans-
mission
M4
A3
A3
A3
A3
MS
A3
M4
M4
A3
M4
M4
A4
M4
MS
A3
M4
A3
A4
A3
A3
Dynamometer
H.P.
7.3
12.5
12.5
12.8
12.5
6.8
10.6
7.3
7.3
14.4
7.3
12.3
13.2
12.6
11.8
13.3
10.7
12.6
13.0
13.3
12.0
I.W.
2250
4000
4000
4500
4000
2250
4500
2250
2250
5500 (+)
2250
3500
4000
3500
3000
4500
3500
4000
4000
4500
4000
 Vehicle Tests
    1-34
Vehicle Tests
    35-80
50C, 30 min*
50C, 30 min.
CFDS
HFET X 3
SOAK, overnight
FTP
CFDS
HFET X 3
IDLE, 30 min
Repeated for
each of 3 fuel/
oil combinations=
Phase 1, 2. 3
50C. 15 min*
HFET X 3
SOAK, overnight
FTP
CFDS
HFET
NYCC
50C. 15 min
IDLE, 15 min

Repeated for
each of 2 fuel/
oil comt>inations=
Phase 1, 3
"Pre-Test conditioning, no data taken.

the Ames Salmonella typhimurium/micro-
some method. Most work was performed
with tester strain TA98 without metabolic
activation. Duplicate plates were run at
extract doses of 0, 10, 20, 30,40, 50, 75,
100, and 200 /ug. Linear  correlation
coefficients and their significance levels
were computed for bioactivity and
extract/residue parameters.
  Real-time diesel particulate  mass
measurements were made for the first
time using a Tapered Element Oscillating
Microbalance (TEOM).  The TEOM is  a
hollow glass rod, fixed at a wide base,
with a  removable filter element attached
to the  narrow top, and oscillating in an
electric field (Figure 1). The TEOM  has
been shown to  behave as a harmonic
oscillator with  a frequency dependent
upon the mass collected by the filter
element.
  Physical, chemical and bioactive charac-
teristics of the particulate were measured
to assess the effect of the dilution tunnel
on the  particulate. The effects of sub-FTP
temperature soaks on FTP emission were
investigated  by  cold-soaking a vehicle
outdoors during the winter.
  Large samples  of  particulate were
collected to  provide large amounts of
extract (1 -27 g) for chemical fractionation
and  identification  by  GC, GC/MS  and
HPLC/UV. The Ames assay was used as a
biological monitor to identify important
chemical fractions. The extracts were
first fractionated into  acidic, basic,  and
neutral compounds by liquid-liquid
partitioning. The neutral  fraction (about
90-95% by  weight) was  further  divided
into seven  subfractions by silica gel
adsorption chromatography.  All subfrac-
tions were bioassayed with tester strains

-------
    Side View
                                                Top View
                     Filter Element


                         Field Plates
                  LED
                     Tapered
                     Element
                          Conductive
                          Path to Fiber
 I/////////////////
   Sample Flow
 TEOM Operation

 1.  Electric field is set up between Held plates.
 2.  Image of tapered element is projected on phototransistor.
 3.  Oscillation of element initiated electrically or mechanically produces an AC voltage output
    from phototransistor.
 4.  AC voltage is amplified and applied to conductive path on element which maintains the
    oscillation due to interaction with field set up in Step 1.
 5.  Frequency of oscillation and hence mass on filter element is determined by frequency counter.
 Figure 1.    Schematic representation of TEOM instrumentation.
TA98 and TA100 with and without S9
activation.

Results and  Discussion
Gaseous and Paniculate
Emissions
  Table 3 summarizes the mean values
of paniculate and gaseous emissions and
related parameters for the three groups
of vehicles for Phase 3. Paniculate refers
to the material collected  by  the  EPA
procedure while residue and extract are
respectively the insoluble  and soluble
fractions of the particulate  after extrac-
tion.
  Figure 2 illustrates the effect of driving
cycles on  particulate emissions.  In
general  the other  emissions  (g/mi)
followed the same general trends of 50C
< HFET < CFDS < FTP < NYCC and VW <
MB < GM. The most prominent aspects
for  all emissions were the  very  large
increase for the NYCC (about double the
FTP) and the large decrease for the IDLE
relative to the driven cycles. The General
Motors group's emissions were greater
and much more sensitive to driving  cycle
than  those of  the Volkswagen and
Mercedes-Benz groups. The cycle varia-
tion  of the particulate  emission  rates
were  principally due to variations in the
residue emission rate  for all  vehicle
groups. Residue  emission  rates were
very cycle dependent for  the General
Motors group  but  only  slightly cycle
Table 3.   Fuel Economy and Particulate and Gaseous Emissions Summary All Cycles - Phase 3
                                                                                   
-------
          o Car 02
          * Car 03
          + Car 04
          x Car 05
          » Car 07
          * Car 16
h
 a  ~J
.s>  o
   o
   o"
                                   90
Figure 3.
          15   30   45   60   75
             Test Mileage. 1000's
            Mileage accumulation effects for
            paniculate from the General
            Motors group.

Fuel Economy
  Pue\ economy was measured by the
carbon balance method and by under-
hood flow totalizing  meters. Figure 4
shows the variations of carbon-balance
fuel economy with driving cycle. For each
vehicle group, the fuel economy increases
as  the average  speed  of the cycle
increases. Fuel economy for a given cycle
was always highest for the Volkswagen
group, followed by the Mercedes-Benz
group. Mileage accumulation generally
had no effect on the FTP fuel economy.
  Over-the-road fuel economy was most
closely approximated by the FTP economy
even though over-the-road useage was
at a higher average speed. At comparable
average speeds,  over-the-road fuel
economy was about 15-20% less than
carbon-balance fuel economy. In general,
the fuel economy for the CFDS very
closely approximated the maximum over-
the-road fuel economy.

Bioassay Characterization
  The Ames activity of the extract itself
generally  increased in the  order GM<
MB < VW  for all test cycles  and  all
methods of expressing activity. Figure 5
shows the Ames  activity in terms of
revertants//ug  extract as  a  function of
cycle. The activity generally increased as
the average speed of the cycle decreased
except that the NYCC generally  showed
less activity than the other driven cycles.

                                   4
                                          8-
                                        f
                                          o.
                                                            • General Motors
                                                            A Volkswagen
                                                            + Mercedes-Benz
       NYCC FTP  CFDS HFET 50 C

Figure 4.    Cycle variations of fuel economy,
           miles/gallon, by vehicle group.

The  FTP or the  CFDS usually had the
highest activity. Some large differences
in activity were noted with fuel/lubricat-
ing oil changes and with mileage accum-
ulation, but no definite trends could be
established.
  Linear  correlations of activity parame-
ters and exhaust parameters were
generally very weak (r values typically
ranging from 0.2 to 0.4), although often
statistically significant. Some parameter
pairs showed a consistency in the sign of
the correlation coefficient which was
independent of vehicle type and driving
cycle. The  bioactivity parameter  which
most frequently yielded a statistically
significant correlation coefficient was
revertants  per /ug extract  and it was
usually correlated negatively with extract
and correlated positively with residue.

Real-Time Paniculate
Measurements
  The TEOM (Figure 1), as tested, could
respond to dilution tunnel concentrations
as low as 1 to 2 mg/m3 with a response
time on the order of 8 to 15 seconds. Cars
1 and 5 were driven over the FTP Bag  3
and  NYCC  using both standard 47-mm
filter collection and the  TEOM.  On
average  (29 tests) the TEOM mass was
within 10% of the gravimetric determina-
tions. Figure 6 shows a trace of paniculate
mass emissions rate vs.  time and  the
corresponding acceleration vs.   time
trace for the FTP Bag 3. Periods of high
particulate emissions corresponded to
periods of  rapid  acceleration while  the
particulate  emissions decreased very
rapidly during deceleration. Portions of
the  curve  which  seemingly  indicate
negative particulate emission rates  are
                                                     Revertants/UG SOF
                                                     o General Motors
                                                     * Volkswagen
                                                     + Mercedes-Benz
                                                     * Other Cars
                                                                                  o.
                                        s
                                        "
                                                                                I
        NYCC  FTP CFDS HFET 50 C IDLE
Figure 5.   Cycle variations of specific Ames
           activity.
thought to be caused by the desorption of
water held on the particulate and filter.

Chemical  Characterization of
Extracts
  The  acidic  fraction had  the highest
specific activity, but was a small fraction
of the total mass, and therefore, was not
the main contributor to the mutagenicity
of the extract. The neutral fraction (about
90 to 95% by weight) had a lower specific
activity; but due to its mass was the main
contributor to the mutagenicity of the
extract.  The  subfraction  containing
PAH's showed very little activity. Twenty-
one PAH's were identified by gas chroma-
tography. The highest specific mutagenic
activity was found in the fourth subfrac-
tion which contained about 2-4% of the
mass  of the  neutral  fraction,  but  con-
tained  42-52%  of the direct-acting
mutagenicity  of  the  neutral  fraction.
GC/MS and HPLC/UV showed this frac-
tion to  contain  alkylfluorenones with
benzo(a)fluorenone the major constituent.
The fifth subfraction comprised 4 to 5%
(predominantly oxy-PAH's  of  the mass
and 13-20% of the activity of the neutral
fraction.
  For  experiments conducted only  with
Car 5 the cycle driven prior to the FTP cold
soak had no effect on gaseous emissions
and little effect on particulate emissions
except when that cycle was an idle. Idling
prior to a driven cycle always increased
particulate, but did not affect gaseous
emissions. Previously driven cycles had
no effect on gaseous emissions. When a
cycle was repeated consecutively during

-------
   8
   05
I §

1 N
-S o
8 3
o *-

18
<0
       Mass Rate tig/sec

Acceleration mph/sec
     0.   33.   67.  100.  133. 167. 200.  233. 267. 300.  333. 367. 400. 433. 467. 500.
<
                                    Time (seconds)

Figure 6.    TEOM mass rate vs. acceleration for Car 5 and Phase 3 of the FTP.
a  day,  the fuel  economy  gradually
increased.
  The day-to-day variability of particulate
and  gaseous emission measurements
over a period  of six or seven test-days
usually did not exceed 0.04 g/mi (standard
deviation). Multiple repetition of a cycle in
a single day had  ranges in the order of
0.01 to 0.04 g/mi. The ranges for fuel
economy over  a six test-day period were
0.5 to 0.8 mpg for Car  5.
  The dilution tunnel collected particulate
of varying physical, chemical, and biologi-
cal character. Proceeding  from the
warmest to the coolest section, particulate
matter decreased in bulk density, increased
in soluble organic content and decreased
in Ames specific  activity.  The specific
activity of the particulate extract from the
tunnel walls was higher than that of most
of the particulate matter collected with
filters.
  No effect  on specific activity or total
mutagenic activity  of  the extract  was
observed when particulate or its extract
alone were re-exposed to filtered dilute
exhaust. The effect  of cold ambient
temperature  on  FTP emissions was
investigated by cold soaking a  vehicle
outdoors during the winter. Comparison
of FTP data for soaks at  0°C as opposed to
20°C showed an  increase in  particulate
of 18-75%, a decrease in percent extract-
able of 13-36%, little change in the mass
of the extract, and an increase in specific
bioactivity of 230-400%.
                                  Richard E. Gibbs. James D. Hyde, and Robert A. Whitby are with the New York State
                                    Department of Environmental  Conservation, Albany,  NY  12233; Delip R.
                                    Choudhury is with the New York State Department of Health, Albany, NY 12201.
                                  Peter Gabele is the EPA Project Officer (see below).
                                  The complete report, entitled "Characterization of Emissions and Fuel Economy of
                                    In-Use Diesel Automobiles," (Order No. PB 83-262-071; Cost: $ 17.50, subject to
                                    change) will be available only from:
                                         National Technical Information Service
                                         5285 Port Royal Road
                                         Springfield, VA 22161
                                          Telephone: 703-487-4650
                                  The EPA Project Officer can be contacted at:
                                         Environmental Sciences Research Laboratory
                                          U.S. Environmental Protection Agency
                                         Research Triangle Park, NC 27711

-------
United States
Environmental Protection
Agency
Official Business
Penalty for Private Use $300
Center for Environmental Research
Information
Cincinnati OH 45268
                                   111!
                                   HH
                                   111!
                                   ill!
         U&OFFjgAyjAIL

m   <3'4  ]  ^Mll±-li.
                                                                     ^-QHJ>-
       PS   0000329
       U S  ENVJR  PROTECTION AGENCY
       REGION  5 U8RA*Y
       Z30  S DEARBORN STREET
       CHICAGO IL 60604
                                                                     U.S. GOVERNMENT PRINTING OFFICE: 1984-759-102/830

-------