PB85-127470
Emission from In-Use
Heavy-Duty Gasoline Trucks
(U.S.) Environmental Sciences Research Lab.
Research Triangle Park, NC
Nov 84
Apartment of Commerce
1 Techrkai biformation Service
-------
PB85-127U70
EPA-600/D-84-281
November 1984
EMISSION FROM IN-USE HEAVY-DUTY GASOLINE TRUCKS
by
Frank Black
William Ray
Foy King
William Karches
Ronald Bradow
Environmental Sciences Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Ned Perry
John Duncan
William Crews
Northrop Services, Inc.
Research Triangle Park, NC 27711
EPA Project Officer
Frank Black
ENVIRONMENTAL SCIENCES RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
RESEARCH TRIANGLE PARK, NC 27711
-------
TECHNICAL REPORT DATA
(Pteate read Ivtrvctiont on I fie rtvene before completing]
1. REPORT NO.
EPA-600/D-84-281
2.
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
EMISSION FROM IN-USE HEAVY-DUTY GASOLINE TRUCKS
t. REPORT DATE
November 1984
6. PERFORMING ORGANIZATION CODE
AUTMOR(S)
P.M. Black, W.D. Ray, F.G. King, W.E. Karches,
R.L. Bradow, N.K. Perry, J.W. Duncan and W.S. Crews
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
10. PROGRAM ELEMENT NO.
Environmental Sciences Research Laboratory - RTP, NC
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park.,North Carolina 27711
C9YA1C/01-2076 (FY-84)
11. CONTRACT/GRANT NO.
12. SPONSORING AGENCY NAME AND ADDRESS
Environmental Sciences Research Laboratory - RTP, NC
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
13. TYPE OF REPORT AND PERIOD COVERED
14. SPONSORING AGENCY CODE
EPA/600/09
15. SUPPLEMENTARY NOTES
16. ABSTRACT
'Apportionment of air pollution to sources requires knowledge of source emission
strengths and/or chemical and physical characteristics. The literature is deficient
in data useful for this purpose for heavy-duty motor vehicles, which can be
important sources of air pollution in select microenvironments. Emission factors
are developed in this study for heavy-duty gasoline trucks using chassis dynamometer
simulations of urban driving conditions. The sensitivity of the emissions to such
considerations as the characteristics of the speed-time driving schedule, vehicle
payload, and chassis configuration are examined. Emissions characterization
includes total and individual hydrocarbons, aldehydes, carbon monoxide, oxides of
nitrogen, total particulate matter, particulate organics, lead, bromine, chlorine,
and the fraction of total particulate less than 2 microns. Preliminary comparisons
of emissions obtained using transient engine and transient chassis test procedures
are also reported.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
c. COSATi Field/Croup
18. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19. SECURITY CLASS (This Report I
UNCLASSIFIED
21. NO. OF PAGES
25
20. SECURITY CLASS (This paftl
UNCLASSIFIED
22. PRICE
EPA Perm 2270.1 (R»». 4-77) PHEVIOU* EDITION n OBSOLETE
-------
NOTICE
This document has been reviewed in accordance with
U.S. Environmental Protection Agency policy anci
approved for publication. Mention of trade names
or commercial products does not constitute endorse-
ment or recommendation for use.
11
-------
Emission from In-Use
Heavy-Duty Gasoline Trucks
Frank Black,
William Ray,
Foy King,
William Karches
and Ronald Bradow
US. Environmental Protection Agency
Research Triangle ParK. N.C.
Ned Perry,
John Duncan
and William Crews
Northrop Services, Inc.
Research Triangle Park, N.C.
ABSTRACT
Apportionment of air pollution to sources
requires knowledge of source emission strengths
and/or chemical and physical characteristics.
The literature is deficient in data useful for
this purpose for heavy-duty motor vehicles,
which can be Important sources of air pollution
in certain microenvironments. Emission factors
ate developed in this study for heavy-duty
gasoline trucks using chassis dynamometer
simulations of urban driving conditions. The
sensitivity of the emissions to such consider-
ations as the characteristics of the speed-time
.driving schedule, vehicle payload, and chassis
configuration nre examined. Emissions charac-
terization includes total and individual
hydrocarbons, aldehydes, carbon monoxide,
oxides of nitrogen, total particulate matter,
partlculate organics, lead, bromine, chlorine,
and the fraction of total particulate less than
2 urn. Preliminary comparisons of emissions
obtained using transient engine and transient
chassis test procedures are also reported.
AIR POLLUTION CONTROL AUTHORITIES arc often
tasked with assessment of the relative
contribution of various sources, both' mobile
and stationary, to the degradation of local,
regional, and national ambient air quality.
This knowledge is necessary to develop and
implement control strategies that will
achieve the legislatively mandated ambient
air quality standards.
Several procedures "have been developed
in recent years to aid in the apportionment
of observed air pollutant concentrations to
various sources. Dispersion models have.
been used with knowledge of the location of
sources and receptors, source emission
strengths, and the meteorological conditions
that transport and disperse the emitted
pollutants. Source-receptor models have
also been used to estimate the proportional
contribution of sources by associating the
detailed composition of the various source
emissions with the composition of the air at
the receptor (1-4).* In their simplest
form, H single substance known to be
essentially completely emitted by one kind
of source is measured in the ambient air.
From t'ne source emissions characteristics,
the mass ratios of this "tracer" compound to
the compounds of interest are determined,
and then, by simply multiplying the measured
ambient concentrations of the tracer by the
ratios, the source contribution to the
observed concentrations of the compounds of
interest is calculated. This technique has
been called the chemical element balance
method. Lead has often been used as a
tracer or surrogate for mobile source
emissions (5,6). Somewhat more complicated
approaches use internal ratios of chemical
components for various sources in a matrix
solution. Since the analysis often involves
more than just elemental composition, this
method is referred to as chemical mass
balance. Other receptor models used for
source apportionment include target
transformation factor analysis, ridge
regression, and multiple linear regression
analysis (7-10).
These methods all require knowledge of
source emission" rates and/or composition.
Motor vehicles are difficult sources to deal
with because the vehicle characteristics,
e.g., size, shape, engine, control devices,
fuel, age, and usage pattern are widely
variant, and their emissions characteristics
are sensitive to these considerations.
*Numbers in parenthesis designate references
at end of paper.
-------
Furthermore, the emissions vary with ambient
conditions, e.g., temperature, humidity, and
barometric pressure. The practice that has
developed to accommodate many of these
considerations involves determination of
brscline emission factors using standardized
test conditions with the major categories of
motor vehicles, followed by mathematical
manipulation to adjust for the
characteristics of the vehicle population and
operating conditions in the area where air
quality is being examined. The computer model
used for this purpose, MOBILE 2, permits
vehicle-miles-traveled (VMT) weighting of eight
primary vehicle categories (11). The
standardized test conditions most often used to
generate exhaust emission rates with motor
vehicles and/or engines are representative of
summer urban driving, conditions (12). As such,
these emission rates must be manipulated in
order to be useful for conditions not
aligned with the standardized conditions.
MOBILE 2 permits estimation of fleet average
emission factors by the vehicle mix (trucks,
cars, "diesel engines, gasoline engines,
vehicle age, etc.), with corrections for
speed, ambient temperature, and other
operating modes different from those
examined with the standardized test
conditions. presently the model can be used
for total hydrocarbon (THC), carbon monoxide
(CO), and oxides of nitrogen (NOx). Modules
for particulate emissions and the
unregulated emissions do not currently
exist.
Sources of the basic emission factor
data are many, but the primary source is the
Environmental Protection Agency (EPA)
administered exhaust emissions surveillance
program (13). In this program
consumer-owned motor vehicles are examined
with test procedures similar to those used
for new motor vehicle and/or motor vehicle
engine emissions certification. For
light-duty motor vehicles these are chassis
dynamometer procedures and the emissions are
reported in grnms per mile (g/mi) units.
With heavy-duty motor vehicles the
procedures involve engine dynamometer tests
and the emissions are reported in work
equivalent gram per brake-horsepower-hour
(g/bhphr) units. Since vehicle usage
patterns are normally defined in terms of
miles traveled, it is necessary to convert
the heavy-duty g/bhphr values to g/roi values
tor use in estimating the air quality impact
of the emissions from these vehicles.
The efforts reported in this paper were
directed at an alternative to conversion of
g/bhphr data to g/mi values for heavy-duty
motor vehicles. The uncertainty asso< ..aipd
with defining the necessary bhphr/mi
conversion factors is eliminated by using
chassif; dynamometer procedures similar . to
those used with light-duty motor vehicles to
develop g/mi emission factors. Regulated
THC, CO, and NO emission rates were
x " .
determined; and the unregulated individual
hydrocarbon, aldehyde, total particulate,
organic particulate, particulate less than 2
lim, lead, bromine, and chlorine emission
rates were also examined to 3uppbrt possible
source-receptor model applications of the
data. Due to facility limitations this program
examined only trucks in classes 2B (CVW
8500-10000 Ib) to 6 (GVW 19501-26000 Ib).
Larger trucks and buses are being examined in a
similar pri/gram sponsored by EPA at Southwest
Research Institute, San Antonio, Texas
(Contract No. 08-02-3722).
It has been estimated that in 1980
heavy-duty vehicles were responsible for
about 10 percent of the THC emissions, 15
percent of the CO emissions, and 34 percent
of the NOx emissions from highway motor
vehicles (M). It has also been estimated
that heavy-duty vehicles were responsible
for about 44 percent of the engine-related
particulate emissions from highway motor
vehicles in 1977 (15). These percentages
were based on national average emission
factors and VMT. For specific
microenvironments, these values could be
much higher or lower depending upon the
relative miles traveled by heavy-duty and
light-duty motor vehicles.
The e-nlssion factors reported in this
paper are for gasoline heavy-duty trucks
operated over urban transient driving
routes. Various chassis configurations,
transient driving cycles, and payload
conditions were examined to study the
sensitivity of the. emissions to these
considerations.
The program is continuing, and a later
report will publish data for additional
trucks, including dicsels.
EXPERIMENTAL PROCEDURES
Many of the historical emissions data for
heavy-duty motor vehicles- were based on
steady-state engine dynamometer test
procedures involving a series of 13
operational modes at constant engine speed
and load (16). Emissions under the
steady-state modes of operation have been
mathematically combined to obtain emission
rates in grams per brake horsepower hour,
grams per -kilowatt hour, and grams per
kilogram of fuel consumed (17,18). These
results are, however, difficult to relate to
the mass emitted per unit distance traveled
by the vehicles. Further, motor vehicle
operation in urban areas general.ly involves
transient conditions of enginu speed and
load. In 1979, EPA described hesvy-duty
transient engine test procedures (19) and
transient chassis test procedures (20). The
first application of the transient chassis
test procedures for heavy-duty truck
emission rate determinations was reported by
Dietzmann et al. in 1980 (21,22).
The transient chassis lest procedures
-------
used ID this study were also based primarily on
the Recommended Practice (RP) of France et al.
(20). Laboratory simulation of motor vehicle
roadway conditions; requires accurate knowledge
of engine load as a. function of vehicle speed,
and a representative transient driving
schedule. The RP provides a general equation
of the form:
RLP = 0.67 (H-0.75)W + 0.00125
ILVW - (N x DW)] (1)
where RLP = road-load power nt 50 mph
1 (horsepower)
H = vehicle overall height
(feet)
W = vehicle overall width (feet)
LVW = loaded veh'ile weight
(pounds)
N « number of dynamometer rolls
supporting a tire .
DW » vehicle weight supported by
dynamometer
for determination of the road-load
horsepower at 50 mph for dynamometer simula-
tions. The aerodynamic drag component of
the RLP is represented by 0.67 (H-0./5)W,
the frlctional component by 0.00125 LVW, and
the dynamometer frictional compensation by -
0.00125
-------
, Table 1 - Description of Test Vehicles/Engines
. Dynamometer
Manufacturer
International
Harvester
General
Motors
General
Motors
Ford
Ford
Fon'
Chassis
1973
Stake-Bed
1975
Stake-Bed
1980
Van (Ryder)
Gas Saver)
1979**
Van
1979**
Stake-Bed
1976
Van
Engine
V8-345
V8-350
V8-366
(1979)
V8-6.1L
(1983)
V8-6.1L
(1983)
V8-351W
Odometer,
mi
105,000
35,000
73,000*
51,000*
51,000*
91,000
GVW
Ib
20,000
25,000
23,000
19,200
19,200
9,900
Inertia,
Ib
9,819
15.047
11,150
16,378
10,514
15,798
9.920
14,560
9,920
14,560
6,380
8,215
Load,
HP @
50 mph
47.5
55.4
43.6
50.4
48.8
55.4
60.2
66.4
42.9
49.1
47.5
50.1
* The engines installed in these chassis have accumulated mileages
less than 10,000 mi
** These are actually the same vehicle. The differences lie only in the
dynamometer road load simulations which were established to permit
examination of the sensitivity of emissions to chassis configuration,
i.e. stake bed versus van
transient chassis cycles for the purpose of
studying the importance of cycle
characteristics to observed emissions in this
project. One involved empirical determination
of the acceleration-deceleration
characteristics typical of heavy-duty trucks
being operated in an urban driving situation.
Subsequently, the driving schedule of the RP
was "smoothed" by removing the unrealistic
speed changes. The details of this process are
presented in Appendix A. The second approach
involved developing driving schedules by
operating an instrumented test vehicle in local
urban traffic. Fig. 1 illustrates the speed
versus time 'plots of each cycle examined in the
project, and Table 2 presents cycle
specifications. Fig. 2 displays the first 150
seconds of the RP cycle cs defined and as
smoothed by filtering at 0.5 Hz, which was
determined to be the highest frequency
associated with meaningful speed changes for
the trucks .examined (see Appendix A). The
smoothed cycle is veil within the acceptable
speed tolerances on the original cycle ,->.s
defined In the RP for satisfactory emissions
tests. .
The RP cycle has distinct periods
associated with Los Angeles and New York
nonfreeway. Jriving and with Los Angeles
freeway driving. The locally derived cycles
also include periods of nonfreeway and freeway
driving, in about the same proportions as the
RP cycle. The locally derived cycles were
developed by operating a truck empty and with
about half of rated payload over a designated
local road route. The average speed for these
local cycles was about 25 mph, as compared to
about 19 mph for the RP cycle. The local route
was also driven in reverse order to permit
examination of the sensitivity of emissions to
the location in the cycle of the high-speed
freeway driving period.
Specifications lor the fuels used in
this program are given in Table 3. Tests
with the International 345 (1973) and Ford
6.1L (1983) were completed with fuel No. 1;
the remainder were completed with fuel No.
2. With regard to particulate lead
emissions, it is important to note that i'uel
No. 1 contains 1.05 .j/gal lead; fuel No. 2
contains 1.48 g/gal.
A Burke Porter model 1059 chassis
dynamometer was used for inertia and
road-load simulation. The system uses 9.5
In. diameter rolls, flywheels, and . a DC
electric motor with a Reliance digital
microprocessor controller for Inertia
simulation in 1-lb increments from 1,000 to
18,200 Ib, and for simulation of the
aerodynamic and frictional components .of
-------
60
50
40
30
20
10
THD TRANSIENT CYCLE
JL
DURHAM ROAD ROUTE. EMPTY, LATE
DURHAM ROAD ROUTE, EMPTY, EARLY
DURHAM ROAD ROUTE,HALF LOAD. LATE
1100
Fig. 1 - Heavy-duty transient cycles
road-load. A Horiba Constant Volume Sampling
(CVS) system with selectable flow rate from 200
to 5000 cubic feet per minute (CFM) was used to
dilute and sample exhaust for subsequent
analysis. This system was operated at 1200 CFM
to maintain the diluted exhaust temperatures
below I25°F as required for particulate
sampling (32). The CVS included an 8 in.
diameter dilution tunnel with approximately 25
ft. from the point of initial dilution air
mixing to the participate filtration probe
system.
Exhaust characterization included THC,
CO, NOx, individual hydrocarbon and aldehyde
compounds, total particulate, organic
particulate, particulate less than 2(jm,
lead, bromine, and chlorine. THC, CO, and
NOx emission rates were determined using
-------
. Table 2 - Transient Test Cycle Specifications
Name
HD Transient Cycle
New York Nonfrceway
Los Angeles Nonfreeway
Los Angeles Freeway
New York Npnfreeway
HD Transient Cycle - Smoothed
Durham Road Route, empty-late
Durham Road Route, empty-early
Durham Road Route, half-late
Time ,
s
1,060
254
285
267
, 254
1,060
910
853 !
927
Distance,
mi
5.55
0.53
1.15
3.33
0.53
5.55
6;39
5.91
6.37
Average
Speed,
mph
18.86
7.56
14.55
44.94
7.56
18.86
25.29
24.96
24.72
Table 3 - Test Fuel Specification
Fuel Type
Lead (g/gal)
Sulfur (wt.%)
RVP (PS1)
API gravity, 60°F
Distillation
% 65°C
2 118°C
Z WC
End point, "C
% Recovery
% Residue
Octane, (R+M)/2
Gas Chromatography
7, Saturates
% Olefins
% Aromatics
No.l
Leaded
Regular
Gasoline
1.05
0.034
10.7
60.2
- 29
63 .
94 .
211
97
1.0
89.6
64.5
12.0
23.5
. . No. 2
Leaded
Regular
Gasoline
1.48
0.034
8.3
60.5
'
26.8
65
93
214
97
.1.2
88.6
64.9
15.0
20.1
standard flame ' ionization, nondlspersive
infrared, and chemiluminescence analytical
procedures, respectively. Previously
described capillary column gas
chromatographic procedures were used to
determine individual hydrocarbon compound
emission rates (33). This analysis quantified
83 individual hydrocarbon species. The
emission rates of 19 individual aldehyde/
ketone compounds were determined using
2,4-dinitrophenylhydrazone derivative
high-pressure liquid chromatographic
procedures (34). Total particulate mass and
methylene chloride extractable
particulate-phase organic mass . were
determined -using previously described
filtration-gravimetric ' and . solvent
extraction procedures (33), and the
particulate mass less than 2. pm was
determined using the cyclone size-selective
sampling procedu; .s of John et al. (35).
Previously described x-vay fluorescence
analytical procedures were used for -the
determination of lead, bromine, and chlorine
omission rates (36).
The emission factors reported in this
paper were based on the mean values of a
minimum-of three repetitions of each test
-------
TIIU.IK
Fig. 2 - Comparison of original and smoothed
versions of HDTC (0-150 sec)
condition. Cold start and hot start tests were
completed for each cycle, with the composite
results based on 1/7 cold start and 6/7 hot
start relative weightings. Data quality
control included reference to National Bureau
of Standards, when possible, for all of
analytical procedures.
RESULTS AND DISCUSSION
The objectives of this program included
development of emission factors for heavy-duty
gasoline trucks using transient chassis
dynamometer test procedures with study of the
sensitivity of emissions to such considerations
as payload, chassis configuration, and driving
cycle characteristics used to simulate urban
driving conditions. The van and
flat-bed/stake-bed trucks studied were
tested empty and at about half of rated
payload. All of the trucks were tested with
the Heavy-Duty Transient Cycle (HDTC) of the
RP, and with the appropriate Durham Road
. Route (DRR) cycles for the payload being
examined. The DRR cycles were with "late"
freeway driving periods unless otherwise
indicated. Tables 4 and 5 present THC, CO,
and NOx emission rates and fuel economy for
each vehicle and test condition.
Hydrocarbon emissions observed with the DRR
driving _ cycle averaged about 55 percent of
those with the .HDTC; CO emissions averaged
about 74 percent. The NOx emission rates
were approximately equivalent for the two
cycles, with the DRR values averaging about
101 percent of HDTC values. The DRR fuel
economies averaged about 124 percent of the
HDTC values. It Is . suggested that the
higher THC and CO emission rates . and the
reduced fuel economy associated with the
HDTC driving schedule result from attempting
to follow the numerous rapid up and down
speed shifts characteristic of this cycle.
With carbureted gasoline engines, these driving
patterns result in excestIve accelerator-pump
actuation with associated tuel r-i^h rc
-------
Table 4- THC end CO Enisuion Races*
Truck
1973
IH
Stake
1975
CMC
Stake
1980
CMC
Van
1979
Ford
Van
1979
Ford
Stake
1976
Ford
Van
Cycle
HDTC
DRR
HDTC
DRR
HDTC
DRR
HDTC
DRR
HDTC
DRR
HUTC
DRR
Test
Inertia,
Ib
9,819
15.047
9,819
15,047
11,150
16.-J78
11,150
16,378
10,514
15,798
10,514
15,798
9,920
14,560
9,920
14,560
9,920
14,560
9,920
14,560
6,380
. 8,215
6,380
8,215
THC,
{/ml
12.1 1 1.4
13.9 i. i;2
6.7 i 0,3
7.4 ± 0.3
29.6 i 3,6
31.4 i 1.9
19.4 ± 3.3
16.9 i 1,2
14.1 t 1.2
26.3 + 2.9
9.8 ± 0;6
15.8 r 1.6
15.6 ± 1.0
22.7 ± 2.9
6.7 ± 0.7
9.1 ± 0.9
15.3 ± 2.2
20.4 ± 3.3
6.8 ± 0.'5
9.8 i 0.9
6.9 1 0.4
10.0 ± 1.1
4.7 ± 0.2
6.3 ± 0.8
CO,
g/wl
213.6 i 16.9
233.1 i 25.1
171.0 ± 13.0
173.4 ± 7.9
211.1 ± 22.5
237.4 ± 25.3
185.3 ± 35.0
203.2 » 9.2
91.4 * 3,8
113.7 ± 5.9
51.6 i 4.3
74.2 ± 8.9
115.5 ± 9.5
147.6 i 22.4
78.3 ± 4.9
104.8 ± 13.0
98.8 ± 10.1
142.4 ± 9.0
65.4 ± 9.6
95.7 ± 15.3
73.9 i 8.6
101.1 ± 11.8
57.5 ± 0.3
84.1 ± 15.8
* Values based on weighting 1/7 cold start, 6/7 hot start
different at approximately the 5% level. Kjan
values and associated standard deviations are
also given for each speed.
The THC and CO emission rates were clearly
speed sensitive, with the higher values
associated with lower average vehicle speeds.
NOx emissions speed sensitivity was not clenrly
indicated. The data in Table b -suggest that
the lower inertia van configurations had higher
N'Ox at higher average speeds, hut this trend
was not apparent with higher inertia and
smaller vehicle frontal aroan (aerodynamic
drag).
Heavy-duty motor vehicles arc generally
marketed to carry a payload from point A to
point B. To study the sensitivity . of the
emissions to .payload, program test vehicles
were examined empty and at about half of rated
payload. The data in Tables 4 and 5 suggest
that THC, CO, and NOx emissions are less when
vehicles are operated empty; the empty emission
rates were about PO percent of the half-load
rates (THC, 78%, CO, 79%, NCx, 82%).
The dais reported for the 1979 Ford
permit examination of the sensitivity of:
urban emissions to chassis configuration.
By manipulating the road-load function with
the dynamometer controller, this truck was
tested as a van and as a stake-bed. As
indicated in Table 1, the empty bO-mph
roaj-load horsepower values were 60.2 and
42.9 for the van and stake-bed configurations,
respectively, and the half-payload values were
66.4 and 49.1, respectively. The NOx and CO
-------
Table 5 - NOx Emission Rates and Fuel Economy*
Truck
1973
IH
Stake
1975
CMC
Stake
1980
CMC
Van
1979
Ford
Van
1979
Ford
Stake
1976
Ford
Van
Cycle
HDTC
DRR
HDTC
DRR
HDTC
'
DRR
HDTC
DRR
HDTC
DRR
HDTC
DRR
Test
Inertia,
Ib
9,819
' 15,047
9,819
15,047
11,150
16,378
11,150
16,378
10,514
15,798
10,514
15,798
9,920
14,560
9,920
14,560
9,920
14,560
9,920
14,560
6,380
8,215
6,380
8,215
NOx.
8/mi
7.2 ± 0.9
9.4 i 0.5
7.6 * 0.9
9.5 ± 0.4
8.3 ± 0.7
10.2 ± 1.3
6.5 t 0.3
11.3 * 1.3
7.0 i 0.6
8.3 ± 0.8
7.8 ± 0.3
8.3 ± 0.5
8.3 i 0.4
9.5 t 1.3
7.8 ± 0.5
9.5 ± 1.0
7.3 ± 0.9
9.4 ± 0.4
6.1 ± 1.2
8.0 ± 1.3
7.9 ± 0.5
8.7 ± 0.8
9.6 ± 0.3
10.3 ± 1.1
Fuel
Economy,
ml/gal
4.7 ± 0.9
4.4 ± 0.3
6.5 * 0.4
6.0 ± 0.3
5.2 ± 0.3
5.1 ± 0.2
5.8 ± 0.6
5.7 ± 0.3
5.1 * 0.1
4.8 ± 0.2
6.5 ±0.2
6.1 ± 0.1
5.4 ± 0.3
4.7 ± 0.2
6.2 ± 0.3
5.5 ± 0.1
5.3 t 0.6
4.6 i 0.1
6.6 ± 0.4
5.9 ± 0.4
7.8 + 0.5
6.9 ± 0.2
9.6 i 0.2
8.4 i 0.3
* Values based on weighting 1/7 cold start, 6/7 hot start
emission rates were somewhat, sensitive to the
chassis configuration, with the stake-bed
values averaging 87 and 89 percent,
respectively, of the van values; THC emissions
were relatively insensitive to the chassis
configuraticn, with the average stake-bed and
van values nearly equivalent. Average fuel
economy was about 3 percent higher with the
stake-bed simulation.
Considering all of the data collectively,
for the vehicles and test conditions examined,
THC emission rates ranged from a low value of.
4.7 g/mi for en empty class 2B (GVW 8500-
10,000 Ib) van with the DRR driving schedule to
a high value of 31.4 g/rai for a half-loaded
class 6 (CVW 19,501-26,000 Ib) stake-bed with .
..he HDTC driving schedule. As will be
discussed later, the particulate-phase organic
omissions from the "high" vehicle were also
elevated, si.ggesting a possible lubricant
consumption problem. The CO emission rates
ranged from a low value of 51.6 g/mi, again
with an empty DRR driving schedule simulation,
to a high value of 237.4 g/ml with a half-load
HDTC driving schedule simulation. The range of
observed NOx emission rates was smaller from a
low value of 6.1 fi/mi to a high value of
11.3 g/mi, with empty DRR and half-load DRR
simulations, respectively. The two gasoline
trucks examined by Dietzmann et al. using RP
transient chassis test procedures had THC
and CO emission rates intermediate to these
-------
10
Table 6 - Comparison of Emissions - HDTC and HDTC (Smoothed)
Transient Driving Schedules*
Cycle
HDTC
HDTC
(Smoothed)
THC.
8/mi
13.4 ± 0.7
11.2 ± 2.5
CO.
g/mi
104.4 ± 9.1
88.0 ± 9.0
NOx,
g/mi
8.4.1 0.4
7.2 ± 0.2
Economy!
mi/gal
5.5 ± 0.3
5.9 t 0.1
* Values based on hot start tests
Table 7 - Sensitivity of Emissions to Early and Late
Cycle Freeway Driving Periods
Truck
1979
Ford
Van
(1983
engine)
1975
CMC
Stake
Cycle
DRR, E
DRR, L
DRK, E
DRR, L
THC,
g/mi
6.8 ± 0.7
.6.7 ± 0.7
19.2 ± 1.3
19.4 i 3.3
CO,
g/mi
76.3 ± 4.1
78.3 1 4.9
174.9 + 11.6
185.3 ± 35
NOx,
B/mi
6.9 ±'1.3
7.8 ± 0.5
7.4 ± 0.4
' 6.5 ± 0.3
Fuel
Economy ,
mi /gal
6.6 + 0.2
6.2 ± 0.3
6.2 ± 0.6
5.8 ± 0.6
ranges, 'aut slightly higher NOx emission rates
at 13-14 g/mi with the trucks tested at
16,000 Ib inertia (22). Direct comparison with
current EPA mobile source emissions model
(MOBILE 2) values is difficult because the
model uses VMT-weighted class (2B through 6 for
gasoline engines) emission factors with
expected increases associated with mileage
accumulation (deterioration factors).
However, the values being considered for
MOBILE 3 heavy-duty gasoline trucks for the
model years and mileages associated with the
vehicles examined in this project range from
about 3 to 11 g/mi THC, 97 to 193 g/mi CO,
and 5 to 8 g/mi NOx (37).
The gaseous organic emissions were
characterized for individual hydrocarbon and
aldehyde composition. A total of 83
hydrocarbon compounds anH 19 aldehyde
compounds were determined. Hydrocarbon
percentage compositions for several of the
more abundant compounds and for the major
organic subtractions (paraffinic, olefinic,
aromatic, and acetylenlc) are presented in
Appendix B. A summary of the organic
subfraction data is given in Table 9. The
hydrocarbon compositions varied somewhat
from vehicle to vehicle, but were not highly
sensitive to the transient driving schedules
or vehicle payloads examined. Methane
emissions ranged from about 3 to 10 percent of
THC, benzene from about 2 to 4 percent of THC,
and total olefinic from 15 to 30 percent of
THC. The higher-mileage engines had larger
olefinic and acetylenic. subtractions. The
aldehyde emissions are presented in Table 10.
As with the hydrocarbon emissions, the
percentage compositions of the aldehydes were
not highly sensitive to the vehicle loads or
transient driving schedules examined.
Therefore, the percentage values reported are
based on all of the tests with each vehicle.
The lower percentage values observed with the
1975 CMC probably resulted from the elevated
contribution of lubricant (i.e. oil burning
vehicle) to the THC emissions from this
truck. The total aldehyde emission rates,
based on summation of the 19 compounds
-------
11
Table 8 - Emissions Sensitivity to Transient Driving
Pattern Average Speed
Truck
1973
III
Stake
1975
CMC
Stake
. .
1979
Ford
Stake
1979
Ford
Van
1980 -
CMC
Van
1976
Ford
Van
Load
empty
half
empty
half
empty
half
empty
half
empty
half
empty
half
Test
Phase*
LANF
LAF .
NYNF
LANF
LAF
NYNF
' LANF
LAF
NYNF
LANF
LAF
NYNF
LANF
. LAF
NYNF
LANF
LAF
NYNF
LANF
LAF
NYNF
LANF
LAF
NYNF
LANF
LAK
NYNF
LANF
LAF
NYNF
LANF
LAF
NYNF
LANF
LAF
NYNF
Average
CnAn*4
bpeed »
mph
14.6
44.9
7.6
14.6
44.9
7.6
14.6
44.9
7,6
14.6
44.9
7.6
14.6
44.9
7.6
14.6
44.9
7.6
14.6
44.9
7.6
14.6
33.9
7.6
14.6
44.9
7.6
14.6
44.9
7.6
14.6
44.9
7.6
14.6
44.9
7.6
THC
0.66
0.42
1.00
0.68
0.40
1.00
0.79
0.24
1.00
6.84
0.24
1.00
0.77
0.18
1.00.
0.91
0.19
1.00
0.80
0.20
i.oo
0.93
0.31
.. 1.00
0.81
0.21
1.00
0.90
0. 26
1.00
0.73
0.53
1.00
. 0.70
0.42
1.00
Normalized
CO
0.72
0.59
1.00
0.72
0.58
i.oo
0.71
0.5f
1.00
0.62
0.49,
1.00
1.00
0.51
0.99
0.86
0.56
1.00
0.89
0.55
. 1.00.
1.00
0.75
0.95
0.76
0.29
1.00
0.75
0.38
1.00
0.55
0.34
. 1.00
' 0.56
' 0.41
1.00
NOx
0.87.
1.00
0.96
0.88
0.97
1.00
1.00
0.99
0.90
0.97
1.00
0.95
0.87
0.76
1.00
0.91
0.67
1.00
0.76
1.00
0.70
0.90 .
1.00
0.89
0.84
1.00
0.92
0.88
.0.82
' 1.00
0.61
1.00
0.58 .
0.79
1.00
0.70
* Heavy-Duty Transient Cycle (see .Table 2).
-------
12
i.o
0.1
0.9
IM
U
SPtEO
7.6
14.8
44.1
MEAD
1.00
0.7t
0.30
S.D.
0.00
0.09
0.11
7.t
14.6
VEHICLE SPEED. «ieh
44.9
1.0
0.9
0.1
0.7
0.6
0.5
SPEED
l.«
14.6
44.«
MEM
O.M
0.16
0.11
S.O.
0.14
o.to
0.12
7.6
14.6
VEHICLE SPEED, mph
44.»
Fig. 3 - Relative hydrocarbon emissions as a
function of cycle average speed
1.0
0.1
; 0.6
0.2
SPEED
1.6
14.6
44.9
MEAN
1.00
0.76
O.SO
S.D.
O.Ot
0.15
0.11
7.6
14.6
VEHICLE SPEED, mph
443
Fig. 4 - Relative carbon monoxide emissions as
a function of cycle average speed
Fig. 5 - Relative oxides of nitrogen emissions
as a function of cycle average speed
analyzed, were equivalent to about 6 percent of
the THC emission rates. Generally,
formaldehyde accounted for about 50 percent
of the total aldehyde emissions observed.
Particulate emission rates and
compositional data are presented in
tables 11 and '12. Total particulate
emission rate, the percentage of the total
particulate mass extractable with methylene
chloride solvent, and the percentage of the
total particulate mays less than 2 urn
aerodynamic dia'U'iter were determined.
Elemental analyses for Pb, Br, and Cl were
also completed and are reported for the fine
(less than 2 urn) particulate. The data for
the 1979 Ford with the low-mileage 1983
engine were highly erratic, vith abnormally
large proportions of coarse particulate
matter, suggesting less than adequate
mileage accusulscicn (break-in) for
particulate emissions measurement.
Excluding the. 1979 Ford data, the total
particulate emission rates ranged from about
0.14 g/mi to 0.54 g/mi. The DRR emission
-------
13
Table 9 - Hydrocarbon Emissions Composition*
Weight Percentage of THC (ranges from Appendix B)
Truck
Paraffinic
Olefinic
Aromatic
Acetylenlc
1973, IH
1975, CMC
1976, Ford
1979, Ford
('83 eng.)
1980, CMC
('79 eng.)
38.7
48.7
39.6
46.1
47.9
- 42.2
- 52.9
- 44.1
- 49.6
- 52.1
25.8
15.0
25.3
14.9
17.4
- 29.7
- 19.1
- 26.8
- 16.9
- 18.6
22.0
24.0
22.7
30.6
25.5
- 23.9
- 30.0
- 24.2
- 33.7
- 29.5
8.5
4.5
8.0
3.7
4.0
- 9.0
- 5.5
- 9.4
- 4.7
- 5.0
* Values based on weighting 1/7 cold start, 6/7 hot start
Table 10 - Aldehyde Emissions
Weight Percentage of THC
Truck
Formaldehyde
Total Aldehyde
1973, IH
1975, CMC
1979, Ford ('83 eng.)
3.69 ± 0.51
1.17 ± 0.15
2.79 ± O.Z4
6.67 ± 0.95
2.82 ± 0.35
5.82 1 0.79
* Values based on weighting 1/7 cold start, 6/7 hot start.
rates averaged about 74 percent of the values
observed with the HDTC, and the empty-payload
rates averaged about 86 percent of the
half-payload values. Methylene chloride
extractablo organic mass ranged from about 15
to 30 percent of the total pnrticulate mass,
except for the 1975 CMC truck, which had values
from 35 to 45 percent. Although further
characterization of the participate organics
was not undertaken in this project, it has
been reported that particulate organic
emissions fron leaded gasoline engines are
active in the Ames bloassay, suggesting the
presence of mutagenic compounds, and that
polynuclear aromatic (PNA) compounds such as
benzo(a)pyrene and nitropyrcne are present
(38).
There were no apparent sensitivities of
particulate size distribution or Pb, Br, and
Cl emission fractions to the driving cycles
or vehicle payloads studied, and therefore, the
remainder of the particulate data will be
reported by truck only, representing averages
of all cycles and payloads tested.
Giie-selective examination of the
partii'.ulate emissions indicated a range of
from 12 to 90 percent with aerodynamic
diameters less than 2 urn The low value
was associated with a low nileage engine and
resulted from a relatively high emission
rate of coarce break-In or wear products;
the high value was associated with an engine
emitting excessive organic aerosol which is
generally characterized by fine liquid droplets
(39). A fine particulate (less than 2 pro)
emission rate ot between 35 and 60 percent of
the total particulate rate would appear normal
for the vehicles examined. It should be noted
that the total particulate rate was defined
using a laboratory dilution tunnel and that
very large particles (greater than about 300
urn) are lost to the walls of the tunnel cue to
gravitational settling and therefore could not
be "counted" (40). Table 12 presents Pb, Br,
and Cl data for the fine particulate emissions.
Excluding the 1975 CMC data, Pb constitutes
about 30 to 45 percent of the total less than 2
urn particulate emissions. Lead is emitted from
motor vehicle engines primarily as the halide
PbClBr, and secondarily as binary complex
compounds of PbClBr nnd NH.C1 (41). The mass
ratios of Pb to Br to Cl observed in this study
were consistent at about 59 to 33 to 8. The
theoretical ratios for PbClBr would be 64 to 25
to 11. The observed Br/Pb mass fraction of
typical of atmospheric aerosols (42-45).
However, a reduction of this fraction on
atmospheric aging has been indicated by Pierson
and Brachnczck, who observed lower ratios
outside of roadway tunnels than inside (e.g.,
0.51 inside the Detroit and Canada Tunnel and
0.30 outside 1C m above street level; 0.42
inside the Allegheny Tunnel and 0.15 a few
-------
15
hundred meters outside of the exit) (46).
Two of the program trucks were fitted
with engines provided by the manufacturers
to permit comparison of fuel-specific
emission rates obtained with engine and
chassis transient test procedures (the
engine tests performed by the manufacturer).
The results with a 6.1-L Ford engine are
reported in Table 13. The General Motors
results will be available at a later cate.
The greatest differences were observed in
THC emissions, where the HDTC chassis rates
were about a factor of 3 higher than the
engine rates. The CO and NOx rates compared
more favorably. It should be noted that
Ford used the MVMA version of the engine
transient schedule which is smoothed
relative to the EPA version. As previously
discussed, the high chassis THC rates are
related to the driver's attempt to follow the
rapid up and down speed shifts characteristic
of the HDTC driving schedule.
SUMMARY AND CONCLUSIONS
Efforts by air pollution control
authorities to apportion observed ambient air
pollutants to sources require knowledge .of
source emission strengths and/or compositional
characteristics. For specific
microenvlronments, motor vehicles can be
very important sources. Motor vehicle
emissions are complex and widely variant
depending upon chassis (car., truck, bus),
engine, fuel, usage pattern, and ambient
conditions, among other considerations.
Available emissions data for heavy-duty
trucks useful for estimation of their air
quality impact are very limited. This is due
in part to the certification practice for this
vehicle category which develops work-specific
g/bhphr emission rates rather than the g/mi
rates normally required by air quality models.
The effort reported in this paper
developed g/mi emission rates, and detailed
emissions characteristics for a small fleet
of leaded gasoline heavy-duty trucks. The
sensitivity of the emissions to such
considerations as transient driving
schedule, vehicle payload, and chassis
configuration was studied using laboratory
dynamometer simulations.
The following observations and
conclusions were made:
1. THC emissions ranged from 4.7 g/mi
to 31.4 g/mi, CO emissions ranged . from
51.6 g/mi to 237.4 g/mi, and NOx emissions
ranged from 6.1 g/mi to 11.3 g/mi.
2. THC and CO" emissions were
sensitive to the characteristics of the
transient driving schedule. The emission
rates were higher with the HDTC (taken from
the EPA RP) with its characteristic rapid up
and down speed shifts than with the locally
derived cycles with less speed dither, THC
and CD emissions also varied with cycle
average- speed, increasing .as the average
speed decreased. NOx emissions were not as
sensitive to the cycle characteristics
examined,
3. THC, CO, and NOx emissions were
elevated by increased vehicle paylead.
Empty rates were about 80 percent of
half-payload rates:.
Table 13 - Comparison of Emissions Using HDV Transient Chassis and
Engine Procedures with 1983 MY 6.1L-4V Ford Engine
8/kg fuel
Pay load
Chassis Cycle
Stake-Bed HDTC
DRR
,
Van HDTC
DRR
Transient
Engine
Results* (47)
(inertia)
Empty
Half
Empty
Half
Empty
Half
Empty
Half
(9920)
(14560)
(9920)
(14560)
(9920)
(14560)
(9920)
(14560)
28.
32.
15.
20.
29.
37.
14.
17.
11.
THC
3
8-
8
1
4
2
4
5
7
t 4.1
i 5.3
± 1.1
± 1.9
±1.9
i 4.8
± 1.5
± 1.7
± 0.7
182.
223.
' 151.
196.
21.7.
242.
169.
201.
254.
CO
8
6
6
1
7
1
4
1
4
i 18
i 14
± 22
± 31
± 17
± 36
± 10
± 24
t 2.
.7
.5
.3
.4
.9
.7
6.
.8
8
NOx
13.5
15.1
14.1
16.4
15.6
15.6
16.9
18.2
13.1
± 1.6
± 0.6
± 2.8
± 2.8
i 3.0
± 3.0
± 1.1
t 1.9
± 0.5
* These results are based on the smoothed MVMA version of the Federal
Heavy-Duty Engine Transient Cycle
-------
16
4. CO and NOx . emission rates were
somewhat sensitive to the vehicle chassis
configuration, with stake-bed rates about 90
percent of van rates.
S. Detailed hydrocarbon composition
varied from vehicle to vehicle, but was not
highly sensitive to the transient driving
schedules or vehicle payloads examined.
Methane constituted 3 to 10 percent of the
THC, benzene constituted 2 to 4 percent, and
olefinlc hydrocarbons constituted 15 to
30 percent. Higher oleflnic emissions were
associated with higher levels of mileage
accumulation.
6. Aldehyde emission rates were
equivalent to about 6 percent of the THC
emission rates. Formaldehyde constituted
about 50 percent of the total aldehyde
emissions.
7. Total particulate emission rates
varied from about 0.14 g/mi to about
0.54 g/rai. . The emission rates were higher
with the HDTC than with the DRR cycles, and
were higher with larger vehicle payloads.
8. The total particulate emissions '
were generally characterized by about 15 to
30 percent methylene chloride extractable
organic mass. Approximately 35 to 60 percent
of the particulate mass was less than 2 urn
aerodynamic diameter. This fine particulate
matter was about 30 to 45 percent Pb. The
observed Br/Pb ratio was 0.56.
ACKNOWLEDGEMENTS
The authors wish to thank Roberta Sloan,
ant) Phil Carter for data processing assistance,
James Braddock for particulate analysis
assistance, Silvestre Tejada for aldehyde
analyses, Bob Kellogg for elemental lead,
bromine, and chlorine analyses, Warren Daniels
for test vehicle preparation, Roy Carlson for
vehicle acquisitions, and Susan Bass for
manuscript preparation.
The content of this publication does
not necessarily reflect the views or
policies of the U.S. Environmental
Protection Agency, nor does the mention of
trade names, commercial products or
organizations imply endorsement by the U.S.
Government.
REFERENCES ..
1. S. K. Friedlander, "Chemical Element
Balances and Identification of Air
Pollution Sources." Environ. Sci.
Technol., 7, 235-241, 1973.
2. G. S. Kowalczyk, C. E. Choquette, and
G. F. Gordon, "Chemical Element
Balances and Identification of Air
Pollution Sources in Washington, D.C."
Atmos. Environ., 12*. 1143-1153, 1978.
3. M. T. Kleinman, B. S. Pasternack,
M. Eisenbud, and T. J. Kneip,
"Identifying and Estimating the
Relative Importance of Sources of
Airborne Particles." Environ. Scl.
Technol., 14, 62-65, 1980.
4. G. Gartrell, and S. K. Friedlander,
"Relating Particulate Pollution . to
Sources: The 1972 California Aerosol
Characterization Study." Atmos.
Environ., 9, 279-299. 1975.
5. J. J. Huntzicker, S. K. Friedlander,
and C. I. Davidson, "Material Balance
for Automobile-Emitted Lead in Los
Angeles Basin." Environ. Sci.
Technol., 9, 448-457, 1975.
6. G. R. Cass, "Lead as a Tracer for
Automobile Particulates: Projecting
the Sulfatc Air Quality Impact of
Oxidation Catalyst-Equipped Cars in Los
Angeles." Environmental Quality
Laboratory, California Institute of
Technology, Pasadena, CA, 1975.
7. P. K. Hopke, "The Application of Factor
Analysis to Urban Aerosol Source
Resolution," In "Atmospheric Aerosol:
Source/Air Quality Relationships,"
E. S. Macias, P. K. Hopke, editors, ACS
Symposium Series No. 167, 21-49,
American Chemical Society, Washington,
D.C-., 1981.
8. C. K. Liu, B. A. Roscoe, K. C. Severin,
and P. K. Hopke, "The Application of
Factor Analysis to Source Apportionment
of Aerosol Mass." Am. Ind. Hyg. Assoc.
J., 43, 314-317, 1982.
9. T. G. Dzubay, and R. K. Stevens,
"Interoomparison of Results of Several
Receptor Models for Apportioning
Houston Aerosol." Presented at
Conference on Receptor Models Applied
to Contemporary Pollution Problems,
Danvers, MA, October lyt52.
.JO. S. C. Davis, "Modified Factor Analysis
. of Selected RAPS Aerosol Data."
EPA-600/S2-80-195, U.S. Environmental
Protection Agency, Research Triangle Park,
-------
17
NC, April 1981.
11. "User's Guide to Mobile 2 (Mobile
Source Emissions Model)."
EPA-460/3-81-006, U.S. Environmental
Protection Agency, Ann Arbor, Ml,
February 1981.
12. "Control of Air Pollution from New
Motor Vehicles and New Motor Vehicle
Engines: Certification and Test
Procedures." 40CFR. Part 86, 240-749,
1983.
13. "Compilation of Air Pollutant Emission
Factors: Highway Mobile Sources." EPA
460/3-81-005, U.S. Environmental
Protection Agency, Washington, DC,
March 1981.
14. National Air Pollutant Emissions
Estimates, 1940-1980." EPA
450/4-82-001, U.S. Environment*!
Protection Agency, Research Triangle
Park, NC, January 1982.
15. F. M. Black, J. «. Braddock, R. L.
Bradow, C. T. Hare, M. N. Ingalls, and
J. M. Kawecki, "A Study of Particulate
Emissions from Motor Vehicles." Draft
Report, U.S. Environmental Protection
Agency, Research Triangle Park, NC,
April 1983.
16. "Heavy-Duty Engines for 1979 and Later
Model Years, Certification and Test
Procedures." 40 CFR, Part 86; 42(r?4),
Part III, September 8, 1977.
17. T. K. Baines, and J. H. Somcrs,
"Heavy-Duty Diesel Particulate Emission
Factors." APCA Paper No. 78-54.3
Houston, TX, June 1978.
18. T. M. Baines, J. H. Somers, and C. A.
Harvey, "Heavy Duty Diesel Particulate
Emission Factor." JAPCA, 29(6),
616-621, 1979.
19. "1983 and Later Model Year Heavy-Duty
Engines< Proposed Gaseous Emissions
Regulations." 49CFR, Part 86, 44(31),
Part II, 9494-9539, February 13, 1979.
20. C. J. France, W. Clemmoi.s, and
T. Wysor, "Recommended Practice for
Determining Exhaust Emissions from
Heavy-Duty Vehicles under Transient
Conditions." Technical Report SDSB
79-08, U.S. Environmental Protection
Agency, At.n Arbor, MI, February 1979.
21. H. E. Dietzmann, M. A. P.irnf ,ransicnt Procedure."
SAE 801371, Presented at Fuels & Lubri-
cants Meeting, Baltimore, MD, October
1980.
22. H. E. Dietzmann, M. A. Parness, and
R. L. Bradow, "Emissions from Gasoline
and Diesel Delivery Trucks by Chassis
Transient Cycle." ASME 81-DGP-6.
Presented at Energy-Sources Technology
Conference, Houston, TX, January I98I.
23. "Road Load Measurement and Dynamometer
Simulation Using Coastdoun Techniques."
SAE Handbook J1263, Society of
Automotive Engineers, Uarrendale, PA,
1980.
24. F. T. Buckley, C. H. Marks, and W. H.
Watson, "Analysis of Coast-down Data to
Assess Aerodynamic Drag Reduction on
Full-scale Tractor-trailer Trucks in
Windy Environments." SAE 760107,
Society of Automotive Engineers,
Warrendale, PA, 1976.
25. L. C. Montoya, and L. L. Steers,
"Aerodynamic Drag Reduction Tests on a
Fullscale Tractor-trailer Combination
with Several Add-on Devices." NASA
TMX-56028, NASA Flight Research Center,
Edward, CA, December. 1974.
26. E.J. Saltzman, R.R. Meyer, and D.P.
Lux, "Drag Reductions Obtained by
Modifying a Box-shaped Ground Vehicle."
NASA TM X-56027, NASA Flight Research
Center, Edward, CA, October 1974.
27. S. Cha, R. L. Sloan, and r'. M. Black,
"Study of Models for the Prediction of
Average Road Load of In-use Delivery
Trucks." SAE 831799, Society of Auto-
motive "Engineers, Warrendale, PA,
November 1983.
28. C. M. Urban, "Dynamometer Simulation of
Truck & Bus Road Horsepower for
Transient Emissions Evaluations.." SAE
840349, Society of Automotive Engi-
neers; Warrendale, PA, February 1984.
29. M. Smith, "Heavy-Duty Vehicle Cycle
Development." EPA-460/3-7S-008, U.S.
Environmental Protection Agency, Ann
Arbor, MI, July 1978.
30. C. J. France, "Transient Cycle
Arrangement for Heavy-Duty Engine and
Chassis Emission Testing." Technical
Report HDV 78-04, U.S. Environmental
Protection Agency, Ann Arbor, Ml,
August 1978.
31. Motor Vehicle Manufacturers Association,
"A Review of the Heavy-Duty Transient
Certification Test Cycle." Arthur D.
Little, Inc., Cambridge, MA, July 1981.
32. "Control of Air Pollution from New
-------
18
33.
34.
35.
36.
37.
38.
39.
40.
41.
Motor Vehicles and New Motor Vehicle
Engines; Fartlculatc Regulation for
Heavy-riuty Diesel Engines (Proposed Rule).
40 CFR, Part 86, 46(4) Part III: 1910-
1967. January 7, 1981.
F. Black, and L. High, "Methodology for
Determining Particulate and Gaseous
Diesel Hydrocarbon Emissions." SAE
790422, Presented at Congress and
Exposition, Detroit, Ml, February 1979.
S. B. Tejada, "Determination of Aldehydes
on 2,4-Dinltro-phenyl-hydrazine
Derivatives in Ambient Roadway Atmospheres
and in Actomobile Engine Exhaust by HPLC."
Preprint, U.S. Environmental Protection
Agency, Research Triangle Park, NC, March
1984.
W. John, and G. Rcischl, "A Cyclone for
Size-Selective Sampling of Ambient
Air." JAPCA, 30(8), 872-876, 1980.
J. Wagraan, R. L. Bennett, and K. T.
Knapp, "Simultaneous Multivavelength
Spectrometer for Rapid Elemental
Analysis of Particulate Pollutants." In
"X-ray Fluorescence Analysis 'of
Environmental Samples," 35-55, Ann
Arbor Science Publishers, Inc., Ann
Arbor, MI, 1977.
"MOBILE 3 Basic Emission Rates by
Vehicle Class." Draft report, Office of
Mobile Sources, U.S. Environmental
Protection Agency, Ann Arbor, MI,
February 1984.
J. M. Lang, L. Snow, R. Carlson, K. M;
Black, R. Zweidtngcr, and S. Tejada,
"Characterization of Paniculate
Emissions from In-use Gasoline-fueled
Motor Vehicles," SAE 811186, Society of
Automotlva Engineers, Warrendale, PA,
October 1981.
K. Carpenter, . and J. H. JoTinson,
"Analysis of the . Physical
Characteristics of Diesel Particulate
Matter Using Transmission Electron
Microscope Techniques." SAE 790815,
Society of Automotive Engineers,
Warrendale, PA, September 1979.
K. Habibi, "Characterization of
Particulate Lead in Vehicle Exhaust-
Experimental Technique." Presented at
American Chemical Society Symposium on
Air Quality and Lead, Minneapolis, . MN,
April 1969.
D. A. Hirschler, L. F. Gilbert, F. W.
Lamb, and L. M. Niebylski, "Particulate
Lead Compounds ir Automobile Exhaust
Gas." Ind. Eng. Chem. 49(7), 1131-1142,
July 1957.
42. T. G. Dzubay, "Chemical Element Balance
Method -Applied to Dichotomous Sampler
Data." Annals of the New York Academy
of Sciences, 338, 126-144, May 1980.
43. R.. H. Hararaerle, and W. R. Pierson,
"Sources and Elemental Composition of
Aerosol in Pasadena, California, by
Energy-Dispersive .X-ray Fluorescence."
Environ. Sci. Technol., 9, 1058-1068,
1975.
44. K. W. Boyer, and H. A. Laitinen, "Lead
Hall le Aerosols, Some Properties of
Environmental Significance." Environ.
Sci. Technol., 8, 1093-1096, 1974.
45. G. L. Ter ' Haar, and K. A. Bayard,
"Composition of . Airborne Lead
Particles." Nature, 232, 553^-554,
'1971.
.46. W. R. . Pierson, and W. W. Brachaczek,
"Particulate Matter Associated with
Vehicles on the Road." SAE 760039,
Society of Automotive Engineers,
Warrendale, PA, February 1976.
47. R. E. Blsaro, Private communication.
Environmental and Safety Engineering
Staff, Ford Motor Co., Dearborn, MI,
February 1983.
APPENDIX A: PROCEDURES FOR SMOOTHING
TRANSIENT MOTOR VEHICLE DRIVING CYCLES
OF
The . transient driving schedule
identified for heavy-duty motor vehicle
emissions, testing in the EPA Recommended
Practice. (RP) was developed using CAPE-21
in-use truck driving pattern data. These
data were processed to the recommended
chassis cycle using Markov modal
mathematical procedures. The Motor Vehicle
Manufacturers Association has suggested that
this model is inappropriate for fhe purpose.
of generating a test cycle representative of
heavy-duty motor vehicle driving patterns,
and that, the resultant cycle has excessive
fluctuation of acceleration rates with
random up and down speed shifts. Attempts
to follow this cycle with its excessive
speed dither -could significantly influence
emissions, particularly with motor vehicles
using carbureted gasoline engines.
A .procedure has been developed to
smooth the cycle, eliminating unrealistic
speed dither, and permitting examination of
the importance of this cycle characteristic
to emissions. Any signal-time display can
be examined In the frequency domain by using
available. Fourier Transform . techniques.
Actual speed versus time data were obtained
-------
19
by operating program rest vehicles in local
urban traffic. The frequency domain for
purposeful speed changes and operator speed
control system interactions was then
examined by establishing the Fourier
Transform of the speed-time data. The
amplitudes of the frequency components of
the recommended cycle not present in the
actual roadway driving data were edited to
zero in the Fourier Transform HE" '.ng;
subsequently, the inverse transform was
generated to define the smoothed cycle.
The Fourier Transform converts time series
data into frequency data by calculating the
discrete transform:
N-l
Xd(K)
for K
X(n)e
-J2nn K./N
(A-l)
The original speed tfata array then contains
D(l), a real number representing
the DC (direct current) value
of the speed signal; and
l)(2), a real number representing
the value of the discrete
Fourier Transfbra at the
NYQUIST frequency (1/2 the
sampling frequency).
The remainder of the array "D" contains the
real and imaginary components of the Fourier
coefficients. A selected starting point in
the array can be set to zero, eliminating or
"filtering" the higher-frequency data from
the array. The filtered data are then
converted back to time series speed data by
taking the inverse transform. The mileage
associated with the speed-time data array
can be readily adjusted to the original
mileage. However, experience has indicated
this adjustment to be less than 0.1 mi.
Figure A-l illustrates the product of
this process for a section (574 s - 768 s)
of the Los Angeles freeway mode of the RP
cycle smoothed by filtering at 0.15 Hz. The
character of the original speed-tine trace
is very "blocky", with several rapid up and
tiown speed shifts, nil of which were
eliminated in the .smoothing process.
Selection of the frequency appropriate for
filtration of the RP cycle Involved
examination of actual roadway speed-time
data in the frequency domain. Figure A-2
illustrates the frequency data for the
roadway driving and the RP cycle. The
roadway data reach a near constant,
low-amplitude "noise" level at about 0.5 Hz,
whereas the original cycle continues at
high-amplitude over the . plotted frequency
range. Thus, 0.5 Hz was selected as the
frequency for smoothing the RP cycle. As a
test of the appropriateness of this
S «
r rr r
t«*»CUIi.Cfc.l««#»t MODI
HOCVCIE FlitfftfOtO US HI
I I
XV
TUU.MC
Fig. Al - Transient cycle, smoothing by
frequency filtering
DURHAM ROAD ROUTE. IMMY.EARl*
0« 01
mountv.H,
Fig. A2 - Frequencies associated with speed
changes in the HDTC and DRR driving
schedules
-------
20
frequency, plots of acceleration rate versus
time were examined for both the filtered and
unfiltered versions of the RP and local
roadway driving cycles. Figure A-3
illustrates sections of these plots scaled
to permit ready visualization of changes.
With the local roadway data, very little
change was observed by filtering at 0.5 Hz;
however, significant differences were
observed with the RP data (note the
different time scales for the two cycles).
Filtering at 0.4 Hz caused excessive change
to the local road route data. Similar
observations were made when comparing the
speed-time data.
The described smoothing process permits
a cycle to be modified so that it can be
followed by a vehicle category of interest.
Average speed and trip length are
unmodified, and the resultant speed-time
trace is within the permitted speed tolerances
of the RP on the original cycle.
APPENDIX B:
COMPOSITION
DETAIL HYDROCARBON EMISSIONS
Tables Bl through B5 detail the
hydrocarbon emissions percentage compositions
for each truck and test condition examined. A
total of 83 compounds were analyzed and
appropriate summations were used to determine
the percentages in each major organic
classification. All values are based on 1/7
cold start, 6/7 hot start relative weightings.
HD TRANSIENT CYCLE, FILTERED
DURHAM ROAD ROUTE, EMPTY, EARLY
DURHAM ROAD ROUTE, EMPTY, EARLY, FILTERED
50
100
150 0
TIME, sec
100
200
300
Fig. A3 - Acceleration rates for HDTC and DRR driving schedules,
original and filtered at O.SIlz
-------
21
Table SI - 1973 IB Hydrocarbon Emissions Compositions
Weight Percentage of THC
Methane
Echylene
Propylene
N-Butane
Isopentane
N-Pentane
Isooctane
Benzene
Toluene
Paraffinic
Olefinic
Aromatic
Acetylenic
DRR-Empty
9.7 ± 0.34
16.4 ±0.38 .
4.4 ± 1.54
5.1 ± O.'.S
3.3 ± 0.21
1.9 ± 0.14
0.7 ± 0.04
3.9 ± 0.20
4.9 i 0.03
39.2 ±0.71
29.7 ± 1.10
22.0 ± 0.72
9.0 ± 0.64
DRR-Half
.9.4 ± 0.55
16.0 ± 0.26
4.8 * 1.00
4.7 ± C.52
. 3.0 ± 0.54
1.5 ± 0.21
0.8 ± 0.06
4.0 ± 0.11
5.0 ± 0.11
38.7 ± 1.34
29.5 ± 1.64
22.9 ± 0.32
9.0 ± 0.03
HDTC-Erapty
8.3 .
12.8
4.4
4.1
2.9 .
1.7
0.8 .
3.1
4.2 .
42.2 . .
25.8 .
23.5 .
8.5 .
HDTC-Half
8.6 ±0.35
14.3 ± 0.45.
3.6 ± 0.36
4.8 ± 0.16
3.3 ± 0.13
1.8 ±0.06
0.9 ± 0.07
3.5 ± 0.10
4.7 ± 0.16
40.6 t 0.59
26.5 ± 0.82
23.9 ± 0.71
9.0 ± 0.30
Table B2 - 1975'CMC Hydrocarbon Emissions Compositions
Methane
Ethylene
Propylene
N-Butane
Isopentane
N-Pentane
Isooctane
Benzene
Toluene
Paraffinic
Olefinic
Aromatic
Acetylenic
Weight Percent of THC
DRR-Empty
3.7
. 4.9
2.3
4.7
4.5
2.8
2.1
2.6
5.6
49.5
15.0
30.0
5.5
Table B3 - 1980
DRR-Half
4.1 0.73
6.4 0.42
2.2 0.33
1.6 0.36
2.6 0.94
1.6 0.79
1.1 0.37
2.3 0.89
3.6 1.33
52.9 6.41
18.7 1.84
24.0 7.20
4.5 1.04
CMC Hydrocarbon
HDTC-Empty
3,6 ± 0.21
5.9 ± 0.42
2.4 ± 0.55
3.0 ± 0.63
3.7 1 0.18
2.2 ± 0.12
1.4 ± 0.19
2.7 ±0.13 .
4.6 ± 0.33
48.7 ± 0.73
17.2 ± 0.76
.28.8 ± 0.87
5.3 ± 0.37
Pxisslons Compositions
HDTC-Half
4.1
6.5 .
2.3
2.3 .
2.9 .
.1.9 . '
0.7
2.4 ..
-3.9 .
51.0
19.1 .
24.5
5.4
Weight Percentage of THC
Methane
Ethylene
Propylene
N-Butane
Isopentane
N-Pentane
Isooctane
Benzene
Toluene
Paraffinic
Olefinic
Aror.dtic
Acetylenic
DRR-Empty
2.4 ± 0.26
4.0 ± 0.21
1.9 1 0.17
2.3 ± 0.09
2.6 ± 2.14
1.5 ± 1.21
1.2 ± 0.12
2.3 ± 0.16
3.9 ± 1.74
47.9 ± 6.40
18.6 ± 3.46
29.5 ± 3.47
4.0 ± 0.35
DRR-Half
*
. ,
*
, ,
*
HDTC-Empty HDTC-Half
3.3 .
J.8 .
1.6
2.9 : .'
5.0 -. .
2.9 .
1.0
2.2 .
5.7
52.1 .
17.4 .
25.5
5.0
* * .
.
*
, ,
-------
22
Table B4. 1979 Ford Hydrocarbon Emissions Compositions
Weight Percentage: of THC
Methane
Ethylene
Propylene
N-Butane
Isopentane
N-Pentane
Isooctane
Benzene
Toluene
.Paraffinic
Olefinic
Aromatic
Acetylenlc
DRR- Empty
2.7 ± 0.24
4.5 ± 1.12
2.1 ± 0.51
4.3 ± 0.67
3.6 ± 0.86
2.4 1 0.36
0.9 ± 0.11
2.2 * 0.22
4.9 + 0.12
49.6 ± 3.63
15.4 t 1.39
31.3 ± 3.11
3.7 ± 0.28
DRR-Half
2.8 .
4.8 .
2.1 .
2.9 .
-2.7
2.0 .
0.8 .
2.3 .
5.0 .
47.7
14.9 .
33.7 .
3.7
HDTC-Empty
3.1 J 0.10
5.8 i 0.17
2.3 ± 0.42
3.4 ± 0.16
2.7 1 0.12
1.9 ± 0.08
0.7 ± 0.05
2.5 t 0.07
4.8 ± 0.26
46.1 + 0.77
16.4 ± 0.66
33.0 1 1.02
4.5 ± 0.26
HDTO-Half -
3.4 .
5.7
2.5
2.8
2.6 .
1.8
0.7 .
2.5 .
4.7 .
47.8
16.9 .
30.6 .
4.7
Table B5. 1976 Ford Hydrocarbon Emissions Compositions
Weight Percentage of THC
DRR-Empty
Methane . .
Ethylene
Propylene
N-Butane
Isopencane
N-Pentane .
Isooctane . .
Benzene
Toluene
Paraffinic
Olefinic
Aromatic
Acetylenic
DRR-Half
6.1 ± 0.35
12.3 ± 0.18
5.3 * 0.17
2.9 ± 0.40
2.9 ± 0.43
2.2 ± 0.04
0.9 ± 0.22
4.2 i 0.07
5.1 ± 0.38
39.6 ± 0.31
26.8 ± 0.73
24.2 t 1.31
9.4 ± 0.31
HDTC-Empty .
6.3 .
11.2 .
5.2 .
2.7
2.7
1.7
0.6 .
3.7
4.4
41.3 .
26.3
24.?. . .
8.2
HDTC-Half
4.9
10.5
4.8
3.0
3.2
2.0
0.6
3.8
4.4
44.1
25.3
22.7
8.0
.
.
,
.
.
f
.
.
.
.
.
Thi* paper is subject !o revi\ion. Statt'mcnts and opinions ad-
vanced in papers or dhcuiiion arc the author'*-ar>d arc his
responsibility, not SAfc's; however, the paper has been edited
by SAl. fo- uniform styling and format. DHcimion will be
printed with ilie paper if it is published in SAl. Tranvactions.
I or permission to publish thi\ paper in full or in part, contact
the SA(. Publication* Division.
Person* wishing to submit papers to be considered for pre-
sentation Or publication through SAl- should send the manu-
script or a 300 word abstract of a proposed manuscript to:
Secretary, I'npint'cring Activity Hoard, SAl:.
Printed in U.S.A.
------- |