78-09
Technical Report
August,1978
Exhaust Emissions and Fuel Consumption of a
Heavy-Duty Gasoline Powered Vehicle
Over Various Driving Cycles
427 Cubic Inch 1977 California CMC 6500
by
Richard Nash
NOTICE
Technical Reports do not necessarily represent final EPA decisions or
positions. They are intended to present technical analysis of issues
using data which a,re 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.
Standards Development and Support Branch
Emission Control Technology Division
Office of Mobile Source Air Pollution Control
Office of Air and Waste Management
U.S. Environmental Protection Agency
-------
Abstract
This report presents exhaust emission and fuel economy measurements for
one heavy-duty gasoline fueled vehicle operated over various driving
cycles. These driving cycles were developed from actual in-use opera-
tional data collected in New York and Los Angeles under the CAPE-21
program. In each location, both freeway and non-freeway operational
parameters were recorded. A data matrix (relating speed, acceleration
and frequency of occurrence) was prepared for each city and class of
operation. Several different driving cycles were generated for each
matrix.
Evaluation of the concept of chassis testing for heavy-duty vehicles was
the major purpose of this project, The test program was designed to
measure the sensitivity of exhaust emissions and fuel economy to the.
various driving cycles and road load conditions. In addition, a brief
attempt was made to characterize cold start emissions and the effects of
increased vehicle frontal area. Three of the fully transient cycles
were "linearized" (steady state cruises and constant accelerations) to
see if a simpler type of transient operation could accurately predict
fuel economy. Finally, experiments were undertaken to measure the
instantaneous exhaust dilution ratio in order to assess the adequacy of
the CVS flow rate.
Several significant conclusions can be drawn. It is possible to test
vehicles up to approximately 15,000 kg inertia on a chassis dynamometer
using transient driving cycles. Emissions over different driving cycles
(representing the same category of operation) vary more than would be
expected from the test variability. "Linearized" cycles give lower HC
and CO emissions than their fully transient counterparts.
The following average emission and fuel consumption values were observed
for half load conditions:
g/km litre/100 km
HC CO NOx Fuel
New York Non-Freeway 3.6 137 3.8 73
Los Angeles Non-Freeway 1.5 84 4.1 53
New York Freeway 2.0 107 5.2 . 48
Los Angeles Freeway 0.9 75 6.3 47
-------
Table of Contents
Item Page
Table of Figures 1
I. Objectives 2
II. Summary of Results 2
III. Description of Experiment 3
A. Vehicle 3
B. Equipment 4
C. Driving Cycles 4
D. Test Matrix 5
E. Test Procedure 5
IV. Road Load 10
V. Results 15
A. Chassis 9-Mode Test 15
B. Driving Cycle Emissions and Fuel Consumption 15
C. Linearized Cycles 27
D. Cold Start Emissions 27
E. Dilution Ratios 27
F. Emission and Fuel Consumption Variation with Windage 29
G. Hydrocarbons Measured by HFID 29
VI. Recommendations 33
Appendices -
A. Raw Emission Data A-l
B. Driving Cycle Identification B-l
-------
-1-
Table of Figures
No. Title Page
1. . Driving Cycles 6
2. Driving Cycle Characteristics 7,8
3. Test Matrix 9
4. Measured Road Load Data 11
5. Theoretical Road Load 12
6. Dynamometer Road Load 13
7. Road Load Comparison 14
8. Chassis 9-mode Results 16
9. Average Driving Cycle Emissions and 17
Fuel Consumption
10. Driving Cycle Emissions, HC g/km 18
11. Driving Cycle Emissions, CO g/km 19
12. Driving Cycle Emissions, NOx g/km 20
13. Driving Cycle Fuel Comsump'tion, 1/100 km 21 .
14. Driving Cycle Emissions, HC g/kWh 22
15. Driving Cycle Emissions, CO g/kWh 23
16. Driving Cycle Emissions, NOx g/kWh 24
17. Driving Cycle Fuel Consumption, g/kWh 25
18. Average Emission Indexes 26
19. Cold Start Ratios 28
20. Average Dilution Ratios, chart 30
21. Dilution Ratios, graph 31
22. Windage Test Results 32
23. Windage Test, Emissions and Fuel 34
Consumption
-------
-2-
I. Objectives
The test program had the following six major objectives and questions to
answer:
1. The first major objective of this program was to evaluate the
capability of EPA's large tandem axle heavy-duty electric dyna-
mometer for emission testing purposes* The test vehicle's size
necessitated the use of a large 1.2 m /s CVS (Constant Volume
Sampler) unit. (Most of the remaining test equipment was identical
to that used in light-duty vehicle testing and did not need further
evaluation.) An attempt was also made to determine the relationship
between standard CVS hydrocarbon measurements and the HFID (Heated
Flame lonization Detector) system used for diesel vehicles.
2. Assuming that exhaust emissions can be accurately measured, it was
desired to evaluate the sensitivity of emissions and fuel consump-
tion to the various driving cycles and vehicle loads. The test
vehicle was run at three different road load conditions (simulating
empty, half and full loads) over the driving cycles. It was desired
to determine the variation among the cycles representing one cate-
gory (e.g., all New York non-freeway cycles) as well as the difference
between categories (e.g., New York non-freeway to Los Angeles
freeway).
3. The effect of cold start operation on emissions and fuel consumption
was also to be investigated.
4. Could "linearized" transient cycles, similar to the old light-duty
vehicle 7-mode test, accurately simulate the fuel consumption
observed on the full transient test cycle?
5. What is the effect of frontal area (or "windage") on emissions and
fuel consumption?
6. What size of CVS is necessary for testing large heavy-duty vehicles?
II. Summary of Results
The emission results obtained in this experiment are representative of
one truck only. It would be a grave mistake to make judgments based on
one vehicle whose characteristics might be significantly different than
the general population. This point cannot be over emphasized. Further
testing of different vehicles is necessary before any firm general
conclusions can be drawn. In light of this qualification, the following
results can be stated:
1. In spite of many teething problems, it is entirely possible to test
heavier gasoline fueled vehicles on EPA's large tandem axle dynamometer.
-------
-3-
It is certainly a more difficult process than automobile testing
due to the vehicle size and general configuration of the test cell.
The large CVS functions adequately; however, further work should be
done on calibration and maintenance procedures. A more accurate
method of determining and setting vehicle road load is also required.
Agreement between the integrated continuous HFID and standard CVS
hydrocarbon results was not very good. The HFID averaged 8% higher
with variability of + 40%. No explanation is available. More work
must be done with probe location, calibration, and integrator
operation to correct this difficulty.
2. Significant variations occur between driving cycles generated from
the same category (e.g., New York non-freeway). It is unclear why
this occurs, since each cycle passes the same statistical criteria
and is drawn from the same data matrix. As might be expected,
there are also significant differences from category to category.
3. Cold start operation causes a significant increase in emissions and
fuel consumption. Typically, a five minute cold start test will
show approximately five times as much hydrocarbon, twice as much CO
and NOx, and about twenty percent more fuel consumption than a
fully warmed-up five minute test. These effects generally dis-
appear after about ten minutes, or two warm-up cycles.
4. Linearized transient cycles (using straight acceleration and
deceleration ramps, and steady state cruises) do not accurately
duplicate the emissions measured over full transient cycles. Very
roughly, the linearized cycles gave half the HC and CO emissions;
while NOx and fuel consumption remained about the same.
5. As expected, increasing the simulated frontal area caused increased
fuel consumption and emissions. However, a slight decrease in NOx
was observed at the largest frontal area tested; this is possibly
due to increased EGR at high power levels.
6. A constant volume sampler with a capacity of 1.2 cubic metres per
second appears to be adequate for gasoline engine vehicles. This
flow rate was adequate even when testing at 12,800 kilograms of
simulated inertia.
III. Description of Experiment
A. Vehicle
The test vehicle was a 1977 model CMC 6500 cab-over chassis. Its GVWR
was 13,760 kilograms, it had an empty weight of 3725 kilograms. It was
equipped with a 7.0 litre V-8 engine with a 7.5 compression ratio and a
four barrel carburetor. This engine belongs to the GM 114 family, and
-------
-4-
was certified to the California emissions standards. The engine was
equipped with a canister, throttle return control (throttle kicker), air
pump, exhaust gas recirculation and sodium cooled valves. An Allison
four-speed automatic transmission was used (MD 1714) with a 7.17 rear
axle ratio. Four 11.00 x 20 tires were on the single rear driving axle.
B. Equipment
A heavy-duty Labeco dual roll chassis dynamometer was used for all
testing. (As the test truck had a single drive axle, the front dyna-
mometer rolls were disconnected.) This dynamometer has a roll diameter
of 1.02 metre, a mechanical inertia of approximately 5400 kilograms in
the single roll configuration and can electrically simulate inertia from
2700-50,000 kilograms. Road load force can be simulated by various
dynamometer circuits that control the constant, first and second order
speed contributions. The dynamometer has motoring capability and can be
used at speeds up to 100 km/h. A constant speed cooling fan was used
for all testing.
Operational data from the vehicle and dynamometer was recorded on
magnetic tape for later computer analysis. Data collected included
manifold vacuum, engine RPM, vehicle speed and roll speed. This data
was recorded at one second intervals during all testing.
Emission measurement equipment conformed with the light-duty vehicle
certification regulations. The only major exception was the use of a
1.2 cubic metre/second CVS (critical flow venturi unit) to handle the
increased exhaust flow of the heavy-duty truck. In addition to this
normal emission measurement equipment, a HFID and C0» analyzer were used
to monitor the dilute exhaust flow continuously. (This last equipment
was discontinued approximately half-way through the test sequence to
reduce testing manpower requirements.)
C. Driving Cycles
Driving cycles for this experiment were developed from actual in-use
data collected and analyzed under the CAPE-21 project. In-use vehicles
were instrumented in New York City and Los Angeles. Data was collected
for freeway and non-freeway operation; it was later organized into
separate data matrices. The combination of two cities and two types of
driving gives four operation categories.
For each category of operation, a data matrix was compiled. This matrix
contains information concerning speed, rate of change, and frequency of
occurance. (Several other parameters relating to engine operation were
also included in the data matrix; however, these are of no concern
here.) Since the data logger operated every 0.864 seconds, the data
matrix also reflected that time basis. Drivng cycles were generated
using computer programs developed for the CAPE-21 project.
The driving cycles created for this test program used a 1.0 second
interval between points. However, as the data matrix had a 0.864 second
-------
-5-
time basis a slightly "streched out"- driving pattern results. Speed
distributions will be identical, but all accelerations and decelerations
will be less severe than actual truck operation. This course was dictated
by the equipment and data transformation programs then available.
Three cycles, 11 through 13, were hand created linearizations of cycles
7 through 9. These linearized cycles are still transient, but the
acclerations and decelerations are all at a constant rate. Fluctuating
cruise conditions have been changed to steady state. Finally, cycle 14
is the light-duty certification driving schedule, better known as the
LA-4.
The fourteen driving cycles used in this test program are summarized in
Figure 1. The relationship between average speed and percent idle is
indicated in Figure 2A.
In reproducing the driving cycle on the dynamometer, a strip chart
showing desired vehicle speed versus time was prepared. The vehicle
driver would manually control the acceleration during the test to follow
the desired speed trace. Sample speed versus time charts for each of
the four cycle categories are shown in Figure 2B to illustrate the
driving patterns associated with each category.
D. Test Matrix
Originally it was planned to test the vehicle at three road conditions
(empty, half and full loads) on all driving cycles. For each combination,
three emissions tests were to be run. This was accomplished, but the
throttle kicker was maladjusted which resulted in erratic engine idle
speeds. After this problem was corrected, an abbreviated test matrix
was run. This included most driving cycles at half load with one driving
cycle for each major category at full and empty load conditions. The
final test matrix is indicated in Figure 3. All data collected when the
engine was maladjusted have been omitted.
Four cold start tests were run on the vehicle. These tests are also
indicated on Figure 3. In addition, a sequence of five tests was run to
simulate various frontal areas using drivng cycle 04 (New York freeway).
This was accomplished by adjusting the "windage" control on the chassis
dynamometer. Finally, a chassis version of the 9-mode engine certification
test was run.
E. Test Procedure
All emission testing, except for cold starts, was done with the dynamometer
warmed-up. The dynamometer would be operated for approximately 15
minutes at about 50 km/h until the dynamometer gear box reached opera-
ting temperature. The emission tests would be started with the engine
idling and the transmission in drive. The emission tests were run just
like the light-duty vehicle certification procedure, the only difference
-------
-6-
Figure 1
No.
01
02
03
04
05
06
07
08
09
10
11
12
13
14
Description
NY Non-Fwy
NY Non-Fwy
NY Non-Fwy
NY Fwy
NY Fwy
NY Fwy
LA Non-Fwy
LA Non-Fwy
LA Non-Fwy
LA Fwy
Linearized 07
Linearized 08
Linearized 09
Light-Duty LA-4
Driving
Length
0.998 km
1.094
1.014
3.975
3.927
3.895
2.012
2.140
2.108
25.123
1.878
2.097
2.076
12.038
Cycles
Time
302 sec
331
332
335
331
320
293
332
319
1225
300
300
300
1371
Idle
51.1%
52.7
52.7
14.9
15.1
15.7
30.1
28.8
29.6
2.3
37.3
31.3
25.3
18.9
Average Speed*
24.3 km/h
25.1
23.2
50.2
50.3
52.0
35.4
32.6
33.8
75.6
35.9
36.6
33.3
39.0
* Does not include idle time.
-------
—7—
Figure 2 A
Driving Cycle Characteristics
80
• LA Fwy
60 I
NY Fwy
*
13
-l
-------
-8-
Figure 2B
Driving Cycle Characteristics
100
100L
ion
60
120 180
seconds
240
300
0 ..-
NY Non-Fwy
100
60 120 180
seconds
240
300
60
120 180
seconds
240
300
-------
-9-
Figure 3
Driving Cycle
Description No,
NY Non-Freeway
NY Freeway
LA Non-Freeway
LA Freeway
Linearized
Light-Duty LA4
i
01
02
03
04
05
06
07
08
09
10
11
12
13
14
Test Matrix
Load Simulated
Empty Half Full
5840 Kg 7035 Kg 12800 Kg
CX
X
X XX
CX CX X
X
XX
X CX X
X
XX
X* X* X*
X XX
X
X
X* X*
Legend: X = Hot Start Test, 3 runs
C = Cold Start Test Sequence, 3 consecutive runs
* = For cycles 10 and 14, only 1 test run was made
-------
-10-
being the shorter driving cycles and larger CVS unit. Calibration
checks were done at the beginning of the day or after the cold start
test, and at the end of the day.
IV. Road Load
Road load measurements were taken with the vehicle in empty, half and
fully loaded conditions. Since this truck was a bare chassis, weights
were placed in a box bolted to the frame rails. No attempt was made to
give the vehicle a frontal area profile similar to an actual in-use
truck. Manifold vacuum and coast down times were recorded from back-to-
back runs on a proving ground straight-a-way. These data, along with
the calculated drag force, appear in Figure 4. Calculated theoretical
road load drag force for a truck with 9.3 square metres of frontal area
appear in Figure 5. (This area is typical for a delivery van.)
After obtaining the road load data, the vehicle was placed on the
chassis dynamometer where the on-road conditions were duplicated. This
was accomplished by adjusting the dynamometer "speed boost" and "windage"
controls until both the manifold vacuum and coast down times were duplicated.
Emission testing was done with the dynamometer thus adjusted.
At the end of the test program, coast down curves (speed vs. time) were
run for each load condition. These curves were differentiated and the
dynamometer road load calculated for various speeds. This data appears
in Figure 6. Drag forces from the measured, theoretical and dynamometer
road loads are compiled in Figure 7, Road Load Comparison. It can be
readily observed that the full load dynamometer curve varies significantly
from the theoretical. This deviation undoubtedly resulted from the
dynamometer setting procedure which essentially relied on only one speed
range. A more sophisticated dynamometer setting procedure, using several
speed ranges, should resolve the problem.
A discrepancy will be noted between the "empty" dynamometer road load
curve and the "empty" 72.4 kilometer calculated drag (from coastdown
data). This is due to the fact that the calculated drag was derived
from data on a vehicle with a mass of slightly less than 4000 kilograms.
When the vehicle was actually tested on the dynamometer, it was decided
to add the weight of a cargo box and bring the total simulated vehicle
mass to 5800 kilograms. The theoretical and dynamometer drag force
curves for the empty test configuration appear to be in very good agreement
for the "empty" load condition.
It should be pointed out that "half load" stands for approximately 1/2
of the GVWR, not halfway between "empty" and "full". (This resulted
from a misunderstanding when the actual road load determination was
being made.) For this particular vehicle, only 1200 kg separate "empty"
and "half" loads on the dynamometer.
-------
-11-
Figure 4
Measured Road Load Data
Condition
Empty
Half
Full
Mass
3995
7251
13,163
kg
kg
kg
64.
48.
49.
42.
4
9
7
0
Manifold
km/h
kPa
kPa
kPa
Vacuum
80.
41.
42.
35.
4
2
7
3
km/h
kPa
kPa
kPa
Coast-down
time
14.40s
17.35s
23.65s
Calculated Drag
72.4 km/h
1233
1857
2474
N
N
N
Average of several run pairs, 80.4 to 64.4 km/h
-------
-12-
Figure 5
Speed
0 km/h
20
40
60
80
100
Theoretical
Empty •
5,840 kg
861 N
941
1,181
1,580
2,140
2,860
Road Load
Drag Force
Half
" 7,035 kg
1,035 N
1,115
1,355
1,754
2,314
3,033
Full
12 -,800 kg •
1,883 N
1,963
2,202
2,602
3,162
3,881
2
Note: 9.3 m frontal area assumed.
Source: Study of Emissions from Heavy Duty Vehicles, May 1976, p. 30,
EPA-460/3-76-012
-------
-13-
Figure 6
Dynamometer Road Load
Test Condition/Inertia
Speed
24 km/h
34
43
53
63
72
82
92
Empty
5,840 kg
787 N
956
1,161
1,330
1,624
1,868
2,157
2,660
Half
7,035 kg
627 N
925
1,125
1,325
1,605
1,908
2,250
2,842
Full
12,800 kg
720 N
996
1,308
1,624
2,095
2,590
3,118
4,368
-------
Figure 7
Road Load Comparisons
4000 -
8 2000 _
M
O
00
n)
S-i
p
Road (Observed)
— Dynamometer
Theoretical
20
40
60
Speed - km/h
I
80
I
100
-------
-15-
V. Results
A. Chassis 9-Mode Test
A chassis 9-mode test was run on the vehicle to determine if the engine
had similar emissions characteristics to its certification data engine.
This was a somewhat complicated procedure as the vehicle had an automatic
transmission that did not permit steady speed operation around the 2000
RPM test point. However, in spite of this difficulty, an attempt was
made to run the 9-mode test. Emissions were collected using the CVS
during three minute test modes. Engine RPM and manifold vacuum were
recorded to permit calculation of the engine power output. (See Technical
Support Report for Regulatory Action, HDV 76-04, "Engine Horsepower Mod-
eling for Gasoline Engines," December 1976.) Results from this "modified"
chassis 9-mode test are given, mode by mode, in Figure 8. Also included,
are the certification emission data engine results and the 1977 California
standards.
Except for carbon monoxide, the results are fairly close to the emission
data engine. This discrepancy is explained by the difficulty in controlling
the vehicle with the automatic transmission as well as the possibility
of minor throttle fluctuations during the test. This test was run with
the driver monitoring a manifold vacuum gauge and attempting to control
the engine to the test point. It is also suspected that the carburetor
was not functioning properly, even after the adjustment earlier in the
test sequence.
B. Driving Cycle Emissions and Fuel Consumption
For all transient cycles run at half load, average emission and fuel
consumption values can be found in Figure 9. Results are presented in
terms of distance traveled (g/km, 1/100 km) as well as in terms of
engine power output (g/kWh). Power output was calculated from the
second by second RPM and manifold vacuum recordings using the gasoline
engine horsepower model. (See Technical Support Report for Regulatory
Action, HDV 76-04, "Engine Horsepower Modeling for Gasoline Engines",
December 1976.) Emissions and fuel consumption for the individual
cycles are listed in Figures 10 through 17. Figure 18 gives average
emissions in terms of fuel consumed.
All fuel consumption figures are based on an exhaust carbon balance and
not actual fuel measurement. This is the identical procedure used for
light-duty vehicle gas mileage numbers.
For a given cycle category and vehicle load, the various driving cycles
did not give the same emissions or fuel consumption. This observation
is readily apparent from a quick inspection of the results, Figures 10
through 17. For several cases an analysis of variance was performed to
determine the relative variability from test to test (same cycle) and
between the cycles. Our intuitive feeling was confirmed; the cycles do
give statistically different results. However, this is not to say that
there is any real practical difference. Raw results are included in the
appendix, the reader can draw his or her own conclusions.
-------
Mode
-16-
Figure 8
Chassis 9-Mode Results
Power HC
CO
NOx
1 Idle
2 16 in Hg
3 10
4 16
5 19
6 16
7 3
8 16
9 CT
Weighted
Average
HC
CO
NOx
0.0 kW
23.6
60. A
23.2
8.0
24.1
105.2
25.9
0.0
30.7
Test
Results
0.66 g/kWh
70.94
4.66
20.4 g/h
5.8
7.4
4.6
20.4
4.6
96.4
3.4
6.6
20.25
Certification
Data Engine
0.55 g/kWh
23.60
6.68
458
746
1,667
562
318
624
13,297
853
163
2,178
g/h 3.2 g/h
136.6
377.4
126.8
52.2
153.2
366.2
117.0
12.4
143.0
1977
California
Standards
1.34 g/kWh
33.53
10.06
-------
Figure 9
Average Driving Cycle Emissions and Fuel Consumption
Cycle
Category
NY-NF
NY-FWY
LA-NF
LA-FWY
Average
HC
3.64
2.02
1.54
0.90
2.03
g/km
CO
137
107
84
75
101
NOx
3.78
5.29
4.09
6.32
4.87
1/100 km
Fuel
72.8
47.8
53.4
46.6
55.2
HC
5.19
2.48
2.22
0.99
2.72
g/kWh
CO
196
130
121
82
132
NOx
5.38
6.48
5.91
6.95
6.18
Fuel
775
431
570
378
i
538 £
i
LDV-LA4 1.51 99 5.44 55.5 1.79 118 6.49 488
Note: This figure contains the averages for
all hot start transient driving cycles
run at half load. The LDV-LA4 results
are included for comparison.
-------
-18-
Figure 10
Driving Cycle Emissions
HC - g/km
Load
Cycle
Category
NY-NF
NY-FWY
LA-NF
LA-FWY
Linear 07
08
09
No.
01
02
03
04
05
06
07
08
09
10
11
12
13
Empty
2.84
2.61
2.42
1.15
0.76
0.49
0.80
Half
3.01
4.45
3.45
2.57
1.74
1.74
1.35
1.75
1.52
0.90
1.08
Full
4.86
4.11
3.47
2.03
1.49
LDV-LA4
14
1.08
1.51
-------
-19-
Figure 11
Driving Cycle Emissions
CO - g/km
Load
Cycle
Category
NY-NF
NY-FWY
LA-NF
LA-FWY
Linear 07
08
09
No.
01
02
03
04
05
06
07
08
09
10
11
12
13
Empty
122.2
145.4
89.4
118.8
43.5
33.0
49.3
Half
114.5
158.2
137.1
131.1
97.2
92.4
79.3
94.9
77.8
74.9
60.6
Full
221.4
235.3
141.7
214.6
87.8
LDV-LA4
14
76.7
98.8
-------
-20-
Figure 12
Driving Cycle Emissions
NOx - g/km
Load
Cycle
Category
NY-NF
NY-FWY
LA-NF
LA-FWY
Linear 07
08
09
No.
01
02
03
04
05
06
07
08
09
10
11
12
13
Empty
2.47
4.65
3.21
4.01
3.99
3.26
3.68
Half
3.74
4.62
2.97
5.11
5.32
5.45
3.82
4.32
4.14
6.32
3.51
Full
4.68
6.14
5.40
6.00
5.92
LDV-LA4
14
5.33
5.44
-------
-21-
Figure 13
Driving Cycle Fuel Consumption
litres/100 km
Load
Cycle
Category
NY-NF
NY-FWY
LA-NF
LA-FWY
Linear 07
08
09
No.
01
02
03
04
05
06
07
08
09
10
11
12
13
Empty
72.8
50.7
52.0
48.2
50.8
46.1
48.8
Half
70.8
75.0
72.6
49.1
47.2
47.1
52.6
55.4
52.3
46.6
51.1
Full
90.8
65.6
62.5
61.7
61.2
LDV-LA4
14
52.2
55.5
-------
-22-
Figure 14
Driving Cycle Emissions
HC - g/kWh
Load
Cycle
Category
NY-NF
NY-FWY
LA-NF
LA-FWY
Linear 07
08
09
No.
01
02
03
04
05
06
07
08
09
10
11
12
13
Empty
4.35
3.05
3.59
1.35
1.25
0.93
1.37
Half
4.49
5.71
5.38
3.04
2.29
2.10
1.93
2.45
2.29
0.99
1.51
Full
5.06
3.27
3.47
1.58
1.48
LDV-LA4
14
1.37
1.79
-------
-23-
Figure 15
Driving Cycle Emissions
CO - g/kWh
Load
Cycle
Category
NY-NF
NY-FWY
LA-NF
LA-FWY
Linear 07
08
09
No.
01
02
03
04
05
06
07
08
09
10
11
12
13
Empty
187.3
170.1
132.6
139.2
71.9
62.8
84.8
Half
170.8
202.7
213.7
154.8
124.5
111.6
113.1
132.6
116.6
82.3
85.0
Full
230.2
187.2
141.8
167.2
87.1
LDV-LA4
14
97.7
117.6
-------
-24-
Figure 16
Driving Cycle Emissions
NOx - g/kWh
Load
Cycle
Category
NY-NF
NY-FWY
LA-NF
LA-FWY
Linear 07
08
09
No.
01
02
03
04 '
05
06
07
08
09
10
11
12
13
Empty
3.78
5.44
4.75
4.75
6.60
6.22
6.34
Half
5.58
5.92
4.64
6.04
6.81
6.59
5.46
6.04
6.23
6.95
4.93
Full
4.88
4.89
5.41
4.68
5.88
LDV-LA4
14
6.79
6.49
-------
-25-
Figure 17
Driving Cycle Fuel Consumption
g/kWh
Load
Cycle
Category
NY-NF
NY-FWY
LA-NF
LA-FWY
Linear 07
08
09
No.
01
02
03
04
05
06
07
08
09
10
11
12
13
Empty
824.2
438.2
569.6
417.0
619.2
650.0
619.7
Half
779.5
709.4
835.9
428.2
445.9
420.0
555.5
572.4
581.4
378.4
529.4
Full
697.9
385.5
462.1
354.8
448.0
LDV-LA4
14
491.0
487.8
-------
-26-
Figure 18
Average Emission Indexes
g/kg Fuel
Pollutant Category
HC NY-NF
NY-FWY
LA-NF
LA-FWY
LDV-LA4
CO NY-NF
NY-FWY
LA-NF
LA-FWY
LDV-LA4
NOx NY-NF
NY-FWY
LA-NF
LA-FWY
LDV-LA4
Load
Empty
5.28
6.96
6.30
3.24
2.79
227.0
388.0
233.0
333.0
199.0
4.58
12.41
8.34
11.39
13.83
Half
6.70
5.33
3.92
2.62
3,67
253.0
280.0
209.0
217.0
241.0
6.94
13.94
10.46
18.37
13.30
Full
7.25
8.48
7.51
4.46
3.30
330.0
486.0
307.0
472.0
194.0
6.99
12.68
11.71
13.22
13.12
-------
-27-
C. Linearized Cycles
Linear cycle 11 is the simplified counterpart of Los Angeles Non-Freeway
cycle 07. Results from both can be compared for all three load conditions
in Figures 10-13. It is evident that the linearized cycle gives substantially
less hydrocarbon and carbon monoxide. Fuel consumption is also lower,
but to a much lesser degree. NOx emissions were about the same on
average.
Comparisons for the other linear cycles (12 & 13) cannot be made because
paired test runs were not made. (These cycles were run at "empty" load
while their transient counterparts were tested at half load.)
D. Cold Start Emissions
As indicated on the test matrix, four cold start test runs were made.
These tests occurred early in the morning after the truck had been
sitting overnight on the dynamometer. They were run in a similar manner
to the hot start tests, except that the dynamometer was obviously not
warmed-up. Three consecutive tests were run back-to-back with the truck
approaching hot start operation by the third test. In Figure 19 the
ratios of the cold start test results to the hot start averages are
indicated. Generally, by the third test run the vehicle has reached
stable operation. It can also be tentatively concluded that the more
severe the cycle, the more work done by the vehicle, the quicker it
warms up. This is evidenced by comparing half load cycles 01 (NY non-
freeway) and 04 (NY freeway). The more severe freeway operation shows a
much quicker warm-up.
The hydrocarbon values for the cycle 07, half load test, appear to be
suspicious. A preliminary, and very cursory examination of the raw test
data does not indicate why these values should be so high.
As mentioned previously, these cold start tests were run with the dyna-
mometer not warmed-up. This dynamometer has a rather large gear box
between the rolls and the power absorber. For normal testing, the
dynamometer is warmed up by operating the vehicle for 15-20 minutes at
40 km/h. This allows the gear box lubricant temperature to stabilize.
It is estimated that up to 20% greater drag force is experienced with a
cold dynamometer. It is possible to warm up the dynamometer by motoring
it. However, since the test vehicle had an automatic transmission,
motoring would require loosening the tie-downs and jacking the vehicle
up. (Vehicle operating instructions cautioned against towing without
removing the driveshaft or raising the rear wheels. This precluded a
simple motoring of the dynamometer.)
E. Dilution Ratios
One of the reasons for performing this experiment was to determine the
size of CVS required to test the larger heavy-duty trucks. If the CVS
flow is inadequate, condensation can form in the sample bags and, in
-------
-28-
Cycle
04
01
04
07
Load
Empty
Half
Half
Half
Figure 19
Cold Start Ratios
Run
HC
CO
NOx
Fuel Consumed
1
2
3
1
2
3
1
2
3
1
2
3
4.71
0.99
1.00
11.09
2.54
1.57
4.93
1.06
0.78
17.49*
2.76
2.75
1.14
0.75
0.83
3.41
1.46
1.34
1.31
0.93
0.96
2.22
0.97
0.99
1.85
1.22
1.07
2.28
1.49
1.26
1.84
1.30
1.19
1.91
1.29
1.04
1.17
0.99
0.98
1.49
1.14
1.11
1.20
1.10
1.08
1.18
1.01
0.99
Data are ratios of the cold start results to the hot start
averages.
* Suspicious HC data, no explanation available.
-------
-29-
extreme cases, sample can be lost when the volume from the vehicle
exceeds the CVS capacity. By examining the dilute exhaust constituents,
the average dilution ratio can be calculated for a test cycle. The
average power during the test cycle can be calculated using EPA's gasoline
engine model (see Technical Support Report for Regulatory Action, HDV
76-04, "Engine Horsepower Modeling for Gasoline Engines", December 1976)
from the RPM and manifold vacuum. This was done for all the hot start
tests in the test matrix, results appear in Figures 20 and 21. As
expected, the average dilution ratio is a very smooth function of the
average power level. This relationship can be used for sizing the CVS
unit if the average power level is known.
In addition to calculating the average dilution ratio, continuous CO
measurements were taken for several emission runs. By knowing the peak
C0_ level observed in the dilute exhaust stream during a cycle, the
lowest instantaneous dilution ratio can be estimated. (This is only a
rough check since the raw C0« level out of the exhaust pipe can vary
slightly with the mode of engine operation.) These estimated peak
dilution ratios are plotted as a function of the average dilution ratio
in Figure 21. The line drawn indicates the worst case condition for the
particular average dilution ratio. As expected, as the average dilution
ratio goes down (and correspondingly, the average power level rises),
the peak dilution ratio also goes down. However, it is interesting to
note that the margin between the average and peak dilution ratio decreases.
(Peak dilution ratios were calculated by assuming a constant 13.4% C02
in the raw exhaust.)
F. Emission and Fuel Consumption with Windage
A sequence of six tests was run to determine the variation of emissions
and fuel consumption with increased frontal area. These tests were run
with cycle 04, New York Freeway, under half load conditions with 7,035
kilograms of simulated inertia. The dynamometer was adjusted normally,
with the exception of the "windage" control. (This control is supposed
to control the effect of aerodynamic drag, i.e., drag force will increase
with the square of the vehicle speed.) The windage setting was varied
from zero up to a value approximately twice normal for this truck.
Results from this sequence of six tests are presented in Figures 22 and
23. Generally, emissions and fuel consumption increased with increased
frontal area. However, NOx decreased slightly at the highest windage
setting. As with earlier reversals of NOx emissions, it is suspected
that this is due to an increased EGR flow at the higher power levels.
G. Hydrocarbons Measured by HFID
During some of the test runs, an attempt was made to measure hydro-
carbons by continuously analyzing the dilute exhaust stream with a
heated flame ionization detector. The analyzer output was integrated to
give an average value. This system is very similar to that used for
light-duty diesel vehicles. The purpose behind this experiment was to
see if the bag sample and the dilute continuous sample would yield the
-------
-30-
Figure 20
Average Dilution Ratios
Run
1
3
4
5
6
7
8
9
11
12
13
15
16
17
19
20
21 & 22
23
24
25
26
27
28
29
30
Cycle
03
• 04
07
10
11
12
13
14
01
02
03
04
05
06
07
08
09
10
11
14
03
04
07
10
11
Load
Empty
Empty
Empty
Empty
Empty
Empty
Empty
Empty
Half
Half
Half*
Half
Half
Half
Half
Half
Half
Half
Half
Half
Full
Full
Full
Full
Full
Average Dilution
46.7
18.8
30.7
11.7
34.9
*Note: For this test program, half load
means half of the GVWR, not GVWR
plus half of the payload capacity.
34.
33.
24.
44.
12,
34,
23,
45.0
47.0
19.
20.
19.
30.6
30.7
32.0
.6
.5
38.
14.
25.9
9.1
28.4
Average Power
7.2 kw
,5
,7
36.
16.
63.0
13.7
13.2
14.5
24.8
8.0
9.2
7.0
36.2
33.4
36.2
17.3
16.6
15.8
67.3
16.
26.
.1
.5
10.
53.
24.
94.8
22.7
.5
,7
,7
-------
-31-
Figure 21
Dilution Ratios
60 -
40 -
O
•rl
JJ
20
20
I
40
I
60
I
80
100
Average Power - kw
-------
-32-
Figure 22
Windage Test Results
litre/100 km
Windage
Setting
0.00
0.06
0.12
0.173
0.24
0.36
Coast
Down
23.80 s
19.00
17.35
13.75
9.65
Drag 2
Force
1354 N
1696
1857
2343
3339
HC
1.32
2.16
2.50
2.57
2.54
3.54
g/km
CO
69.3
103.1
122.7
131.1
159.8
236.0
NOx
3.82
4.41
4.99
5.11
5.80
5.21
Fuel
Consump
38.3
43.7
48.0
49.1
55.9
63.7
80.4 to 64.4 km/h
At 72.4 km/h
Normal Condition
All tests, Cycle 04 NY Freeway, Half Load, 7035 kg inertia
-------
-33-
same results. Since it is assumed that exhaust hydrocarbons from gasoline
engines do not condense using the normal CVS technique, both methods
should give about the same analysis. Unfortunately they did not. The
heated flame ionization detector gave values from 40% below to 40% above
the bag sample. The reason is unknown. On top of the large variability
observed, there was an offset of 8% in favor of the HFID. That is, the
heated flame ionization detector, on average, recorded an 8% higher
hydrocarbon value. However, this is a relatively new system, and had
just been assembled for this experiment. The analyzer had not been
optimized nor had various probe locations been evaluated.
VI. Recommendations
This experiment was concerned with emissions and fuel consumption
measurements on one 1977 model year California heavy-duty truck. Care
should be taken not to extrapolate the results from this one vehicle
farther than good engineering practice will allow. Before any firm,
definite and far reaching conclusions can be drawn, many other types and
sizes of vehicles should be tested.
One particularly interesting point, which should be examined in more
detail, is the variation of emissions with the various driving cycles
from the individual categories. The reasons for these variations should
be investigated to determine what, if any, changes might be made in the
cycle generation procedure. Hopefully, cycles generated from the same
input matrix and screened to the same statistical level will yield the
same emission values. Also, the variation of emissions between the
various categories of operation might be more thoroughly defined with a
larger test fleet. This test fleet should consist of vehicles from the
major manufacturers representing all types and sizes of heavy-duty
engines.
In view of the variation between some of the dynamometer road load
curves and their theoretical counterparts, a more thorough investigation
of road load is justified. In addition to examining the vehicle road
load, a new dynamometer setting procedure might eliminate some of the
variation. No firm recommendations can be made in this vein, only the
suggestion that the entire matter warrants further investigation.
The final topic, HFID hydrocarbon measurement, is particulary trouble-
some. No ready explanation is available for the offset in emission
values observed or in the large variability. This will be a crucial
element for diesel vehicle testing. In the next phase of gasoline truck
testing, significant efforts should be expended to locate the cause of
these discrepancies and to correct them.
-------
Figure 23
Windage Test - Emissions and Fuel Consumption
6 -
2 -
T r—
2000
— 1
3000
Fuel litre/100 km /10
NOx g/km
HC g/km
CO g/km/100
4000
i
u>
Drag Force - N at 72.4 km/h
-------
APPENDIX A
Raw Emission Data
No. Cycle Load Run
1 03 Empty 1
2
3
Av
2 04 Empty 1 Cold
2
3
3 04 Empty 1
2
3
Av
4 07 Empty 1
2
3
Av
5 10 Empty
6 11 Empty 1
2
3
Av
7 12 Empty 1
2
3
Av
HC
2.87
3.06
2.58
2.84
12.30
2.59
2.62
2.39
2.58
2.86
2.61
2.51
2.35
2.41
2.42
1.15
0.80
0.71
0.76
0.76
0.53
0.47
0.46
0.49
g/km
CO
123.0
135.5
108.2
122.2
166.3
109.3
120.6
129.4
146.4
160.4
145.4
91.9
84.7
91.7
89.4
118.8
46.2
42.5
41.9
43.5
33.5
33.9
31.5
33.0
litre/100 km kWh
NOx
2.62
2.42
2.37
2.47
8.60
5.67
4.96
4.61
4.62
4.72
4.65
3.48
3.03
3.11
3.21
4.01
4.05
3.96
3.96
3.99
3.40
3.18
3.21
3.26
Fuel
74.
73.
70.
72.
59.
50.
49.
49.
51.
51.
50.
53.
51.
51.
52.
48.
51.
50.
50.
50.
46.
45.
45.
46.
2
4
9
8
4
4
7
6
0
6
7
1
3
7
0
2
6
3
4
8
5
8
9
1
Work
0.66
0.66
0.66
3.88
3.30
3.30
3.37
3.40
3.42
3.40 .
1.37
1.35
1.35
1.36
21.44
1.15
1.14
1.12
1.14
1.10
1.11
1.09
1.10
HC
4.40
4.69
3.95
4.35
12.60
3.12
3.15
2.82
3.02
3.32
3.05
3.68
3.50
3.59
3.59
1.35
1.30
1.17
1.28
1.25
1.02
0.89
0.88
0.93
CO
188
207
165
187
170
131
145
152
171
186
170
135
126
136
132
139
75
70
70
71
63
63
60
62
g/kWh
.5
.7
.8
.3
.4
.6
.2
.6
.1
.5
.1
.0
.2
.6
.6
.2
.4
.0
.3
.9
.7
.9
.9
.8
NOx
4.01
3.70
3.63
3.78
8.81
6.83
5.97
5.44
5.40
5.48
5.44
5.11
4.52
4.63
4.75
4.75
6.62
6.53
6.64
6.60
6.47
5.99
6.21
6.22
Fuel
839.7
830.6
802.3
824.2
449.2
448.1
442.0
431.9
440.1
442.7
438.2
575.7
564.4
568.8
569.6
417.0
622.0
611.7
623.8
619.2
652.8
637.0
654.1
650.0
-------
-2-
No.
8
9
10
11
12
13
14
15
Cycle Load Run
13 Empty 1
2
3
Av
14 Empty
01 Half 1 Cold
2
3
01 Half 1
2
3
Av
02 Half 1
2
3
Av
03 Half 1
2
3
Av
04 Half 1 Cold
2
3
04 Half 1
2
3
Av
HC
0.93
0.79
0.68
0.80
1.08
33.39
7.65
4.74
3.03
2.90
3.10
3.01
4.90
4.53
3.92
4.45
3.65
3.40
3.30
3.45
12.68
2.72
2.01
2.87
1.97
2.87
2.57
g/km
CO
67.4
42.4
38.1
49.3
76.7
390.0
167.5
153.7
113.4
109.1
121.0
114.5
161.2
163.9
149.3
158.2
143.4
139.3
128.7
137.1
172.2
122.2
125.4
141.8
117.1
134.3
131.1
litre/100
NOx
3.51
3.63
3.91
3.68
5.33
8.51
5.59
4.70
3.75
3.73
3.74
3.74
4.78
4.52
4.55
4.62
2.98
2.84
3.10
2.97
9.42
6.64
6.09
5.11
5.11
5.12
5.11
Fuel
49.
48.
48.
48.
52.
105.
81.
78.
70.
70.
71.
70.
75.
76.
72.
75.
73.
72.
72.
72.
59.
54.
53.
50.
48.
49.
49.
5
5
4
8
2
7
0
8
5
2
7
8
5
7
9
0
0
6
3
6
0
0
2
2
2
0
1
km kWh
Work
1.20
1.22
1.20
1.21
9.44
0.85
0.76
0.76
0.67
0.67
0.67
0.83
0.85
0.88
0.85
0.65
0.66
0.64
0.65
3.87
3.55
3.56
3.35
3.40
3.35
3.37
g/kWh
HC
1.62
1.34
1.16
1.37
1.37
39.01
10.12
6.19
4.51
4.33
4.62
4.49
6.44
5.81
4.88
5.71
5.69
5.22
5.23
5.38
13.01
3.05
2.24
3.41
2.31
3.41
3.04
CO
116.6
72.3
65.6
84.8
97.7
455.6
221.4
200.5
169.0
162.7
180.5
170.8
212.1
210.1
186.0
202.7
223.5
213.7
203.8
213.7
176.6
136.8
140.1
168.3
136.9
159.3
154.8
NOx
6.07
6.20
6.74
6.34
6.79
9.94
7.39
6.12
5.60
5.56
5.57
5.58
6.30
5.78
5.67
5.92
4.66
4.36
4.91
4.64
9.67
7.44
6.80
6.07
5.97
6.07
6.04
Fuel
632.0
611.2
616.0
619.7
491.0
911.6
790.4
759.2
776.2
772.9
789.4
779.5
733.0
725.2
669.9
709.4
840.0
822.2
845.5
835.9
446.9
446.3
438.8
439.7
416.1
428.9
428.2
-------
-3-
No. Cycle Load
16 05 Half
17 06 Half
18 07 Half
19 07 Half
20 08 Half
21 09 Half
22 09 Half
Average Runs 21 &
Run
1
2
3
Av
1
2
3
Av
1 Cold
2
3
1
2
3
Av
1
2
3
Av
1
2
3
Av
1
2
3
Av
22
HC
1.95
1.49
1.78
1.74
1.97
1.69
1.55
1.74
23.61
3.73
3.71
1.59
1.33
1.13
1.35
1.75
' 1.73
1.77
1.75
1.52
1.70
1.42
1.55
1.53
1.39
1.51
1.48
1.52
g/km
CO
106.6
84.1
101.0
97.2
102.6
90.1
84.4
92.4
175.9
76.9
78.2
91.2
75.2
71.3
79.3
95.0
87.7
101.4
94.9
71.5
74.9
68.9
71.8
82.1
82.6
86.3
83.7
77.8
litre/100 km kWh
NOx
5.40
5.38
5.19
5.32
5.52
5.43
5.41
5.45
7.29
4.93
3.98
3.74
3.98
3.75
3.82
4.37
4.36
4.21
4.32
3.96
3.92
4.04
3.98
4.40
4.09
4.41
4.30
4.14
Fuel
47.8
46.6
47.2
47.2
47.6
46.6
47.0
47.1
62.3
53.2
51.9
52.9
52.6
52.4
52.6
55.5
54.9
55.8
55.4
50.1
50.7
50.7
50.5
53.9
53.8
54.6
54.1
52.3
Work
3.06
3.11
3.04
3.07
3.25
3.20
3.22
3.22
1.59
1.35
1.33
1.44
1.42
1.37
1.41
1.55
1.53
1.50
1.53
1.30
1.31
1.29
1.30
1.49
1.50
1.54
1.51
1.40
HC
2.50
1.89
2.30
2.29
2.36
2.06
1.88
2.10
29.96
5.54
5.61
2.22
1.89
1.67
1.93
2.41
2.42
2.51
2.45
2.47
2.73
2.32
2.51
2.17
1.96
2.08
2.07
2.29
g/kWh
CO
136.6
106.3
130.6
124.5
122.9
109.8
102.1
111.6
223.2
114.2
118.2
127.8
106.7
104.8
113.1
130.8
122.8
144.2
132.6
115.8
120.4
112.6
116.3
116.5
116.1
118.5
117.0
116.6
NOx
6.92
6.80
6.71
6.81
6.62
6.61
6.55
6.59
9.25
7.33
6.01
5.24
5.65
5.51
5.46
6.01
6.11
5.99
6.04
6.42
6.31
6.60
6.44
6.24
5.75
6.06
6.02
6.23
Fuel
452.2
434.7
450.8
445.9
421.0
419.2
419.7
520.0
583.7
583.2
579.2
547.1
550.9
568.5
555.5
563.9
567.4
585.8
572.4
599.2
601.6
611.3
604.1
564.5
558.1
553.2
558.6
581.4
-------
-4-
No. Cycle Load
23 10 . Half
24 11 Half
25 14 Half
26 03 Full
27 04 Full
28 07 Full
29 10 Full
30 11 Full
Run
1
2
3
Av
1
2
3
Av
1
2
3
Av
1
2
3
Av
1
2
3
Av
HC
0.90
1.21
1.05
0.97
1.08
1.51
4.53
4.52
5.53
4.86
3.81
4.52
4.01
4.11
3.96
3.53
2.92
3.47
2.03
1.53
1.49
1.45
1.49
g/km
CO
74.9
69.6
59.0
53.2
60.6
98.8
245.1
207.7
211.4
221.4
227.0
237.1
242.0
235.3
161.6
135.3
128.1
141.7
214.6
90.1
88.0
85.3
87.8
litre/100 km kWh
NOx
6.32
3.41
3.53
3.59
3.51
5.44
3.92
4.88
5.24
4.68
6.91
5.96
5.56
6.14
4.75
5.68
5.76
5.40
6.00
6.22
6.04
5.51
5.92
Fuel
46.6
50.5
51.2
51.5
51.1
55.5
91.8
91.4
89.3
90.8
66.3
65.0
65.6
65.6
63.9
62.3
61.3
62.5
61.7
63.2
60.5
59.9
61.2
Work
22.90
1.35
1.34
1.32
1.34
10.11
0.99
0.97
0.97
0.97
4.96
5.02
5.01
5.00
2.02
2.02
1.99
2.01
32.26
1.92
1.86
1.90
1.89
HC
0.99
1.69
1.47
1.39
1.51
1.79
4.65
4.73
5.80
5.06
3.05
3.58
3.18
3.27
3.95
3.51
2.95
3.47
1.58
1.49
1.50
1.43
1.48
g/kWh
CO
82.3
96.7
82.7
75.6
85.0
117.6
251.9
217.4
221.4
230.2
181.9
187.9
191.9
187.2
161.1
134.8
129.6
141.8
167.2
88.1
89.0
84.2
87.1
NOx
6.95
4.74
4.95
5.11
4.93
6.49
4.04
5.10
5.50
4.88
5.54
4.72
4.41
4.89
4.74
5.65
5.83
5.41
4.68
6.08
6.11
5.44
5.88
Fuel
378.4
517.7
529.7
540.9
529.4
487.8
696.6
706.4
690.7
697.9
392.2
380.4
384.0
385.5
470.4
458.1
457.9
462.1
354.8
456.2
451.4
436.3
448.0
-------
APPENDIX B
Driving Cycle Identification
Cycle No. Identification No.
01 123 667 645 7
02 179 960 930 5
03 104 736 920 3
04 741 286 985
05 209 279 083 3
06 137 610 363
07 152 778 878 5
08 210 620 459 3
09 211 939 981 9
10 235 541 19
------- |