EPA/AA/TDG/94-01
Technical Report
Testing of an Electric Vehicle on a
Clayton Water-Brake Chassis Dynamometer
by
Ronald M. Schaefer
March 1994
NOTICE
Technical Reports do not necessarily represent final EPA
decisions or positions. They are intended to present technical
analysis of issues using data which are currently available. The
purpose in the release of such reports is to facilitate the
exchange of technical information and to inform the public of
technical developments which may form the basis for a final EPA
decision, position or regulatory action.
U. S. Environmental Protection Agency
Office of Air and Radiation
Office of Mobile Sources
Office of Regulatory Programs and Technology
Technology Development Group
2565 Plymouth Road
Ann Arbor, MI 48105
-------
Table of Contents
Page
Number
I. Summary l
II. Introduction 1
III. Description of Test Procedures 3
IV. Description of Test Vehicle 5
V. Test Facilities and Analytical Methods 6
VI. Discussion of Test Results 7
A. SAE J1634 Energy Consumption Testing 7
B. Individual City and Highway Energy Consumption
Testing. . 8
C. Petroleum-Equivalent Fuel Economy Results. .... 9
D. SAE J1634 Range Testing 11
VII. Future Efforts 11
VI11. Acknowledgements. . 11
IX. References 12
Appendix A A-l
-------
I.
A 1988 Ford Escort Wagon equipped with a lead-acid battery
propulsion system was tested for energy consumption and range on a
Clayton water-brake chassis dynamometer at the EPA National Vehicle
and Fuel Emissions Laboratory in Ann Arbor, Michigan.
Two different test procedures were used to determine energy
consumption. The first test procedure followed the newly developed
SAE Recommended Practice J1634 for the testing of electric
vehicles. Following this procedure will yield only a combined
energy consumption value for both the city and highway driving
cycles. Therefore, the vehicle was also tested separately over
city and highway driving cycles which more closely follow the
standard test procedure for conventional vehicles. From this
procedure, energy consumption values over the city and highway
cycles were determined independently of each other without
deviating greatly from the current test and calculation procedures.
Presented in this report are three power consumption values.
The first is System AC Energy Consumption, or the amount of power
from the electrical wall plug to the on-board vehicle charging
system. The second value is System DC Energy Consumption, or the
amount of power from the on-board vehicle charging system to the
batteries. Both these power values were measured during the
battery recharging period after driving the vehicle over various
cycles. The last power value measured during this program was
Vehicle Net DC Energy Consumption, or the actual amount of power
delivered from the batteries for propulsion of the vehicle. This
power value was obtained during the actual test procedure driving
cycles.
AC energy consumption values with the Clayton water-brake
dynamometer averaged 504 W-hr/mile over the SAE J1634 test
procedure. The values reported here are wall-power numbers and
would be the amount of power for which the consumer would need to
pay. The city energy consumption value was determined to be 656 w-
hr/mile, and the highway value was measured to be 489 W-hr/mile.
By applying the calculation method used in DOE rulemaking that will
allow the conversion of individual city and highway energy
consumption values into a combined petroleum-equivalent fuel
economy value, a fuel economy value of 133.5 MPG was obtained.
II. Introduction
The California Air Resources Board (CARB) has mandated that 2
percent of all vehicle sales in that state shall be zero-emitting
vehicles by the year 1998. [1] One possible means of obtaining zero
pollution at the tailpipe in a conventional passenger vehicle is
with the use of a battery-powered propulsion system.
-------
-2-
Since battery-powered vehicles behave differently than
conventional internal combustion engine (ICE) vehicles, it was
necessary to develop a new test procedure to accurately determine
energy consumption and range of electric-powered vehicles. A task
force was formed under the Society of Automotive Engineers (SAE)
Light-Duty Vehicle Performance and Economy Measurement Standards
Committee with the goal of developing a standard test procedure
that "establishes uniform procedures for testing electric battery-
powered vehicles which are capable of being operated on public and
private roads." This procedure is said to "allow for the
determination of energy consumption and range based on the Federal
Emission Test Procedure (FTP) and the Highway Fuel Economy Test
Procedure (HWFET)."[2]
EPA has membership in this committee and followed the
development of this SAE procedure. This procedure was developed
and finalized, however, without testing any electric vehicles over
it to determine if this procedure was reliable and repeatable.
As a result of the need to test an electric vehicle over this
procedure to obtain information regarding the procedure's
repeatability, a round-robin test program was formed involving the
U.S. Department of Energy (DOE), Idaho National Engineering
Laboratory (INEL), U.S. EPA, Ford Motor Company, and CARS. Each
site agreed to test an electric battery-powered 1988 Ford Escort
Wagon for energy consumption and range following the SAE J1634
procedure. EPA tested the vehicle on both a Clayton water-brake
dynamometer and a Horiba electric dynamometer. EPA also conducted
additional testing of the vehicle to obtain separate city and
highway energy consumption values. In its current form, the SAE
J1634 procedure will only yield a combined city/highway energy
consumption value. The electric dynamometer results are not
presented in this report but will be published at a later date.
Separate city and highway energy consumption values were
required to determine a petroleum-equivalent fuel economy value
based on the DOE'a rulemaking released on February 4, 1994.[3] The
DOE rulemaking allows a combined fuel economy value, to be used in
Corporate Average Fuel Economy (CAFE) calculations, to be
calculated from individual city and highway energy consumption
values. The rulemaking references the SAE J1634 procedure but also
requires individual city and highway energy consumption values.
This report contains energy consumption and range results when
testing the vehicle over the J1634 procedure. Individual city and
highway energy consumption values were also obtained so that a
petroleum-equivalent fuel economy value could be determined for
this vehicle. All results presented in this report were obtained
on a Clayton water-brake chassis dynamometer.
-------
-3-
III. Description of Teat Procedures
The main purpose of this round-robin test program was to test
a single electric vehicle at four different sites over the newly
developed SAE J1634 test procedure. This procedure addresses
battery conditioning, data to be recorded during testing, energy.
consumption testing, range testing, and coastdown testing.
The batteries were aged by INEL prior to testing at EPA. They
were aged to an equivalent mileage of between 2,000 and 6,200
miles, as the procedure requires. INEL discharged the batteries
both at a C/3 rate (46 amps for these batteries) on a battery test
stand and when equipped on the vehicle.[4] Both methods yielded
similar results and proved to be an acceptable method of verifying
or measuring battery capacity.
The test procedure requires a substantial amount of data to be
recorded from each test. INEL supplied EPA with a data acquisition
system that measured the following data requirements over all
testing described in this report. The data recorded for each test
cycle were:
1. Actual miles traveled;
2. System AC Energy Consumption—Watt-hours delivered from
the electrical wall socket to the on-board charging system;
3. System DC Energy Consumption—Watt-hours delivered from
the on-board charging system to the batteries; and
4. Vehicle DC Energy Consumption—Watt-hours delivered from
the batteries to the electric motor for propulsion of the vehicle.
The SAE energy consumption test procedure requires the vehicle
to be driven over two consecutive Urban Dynamometer Driving
Schedule (UDDS) cycles separated by a 10-minute soak with the key
switch in the "off position and the test cell fan not operating
during this soak period. Immediately following the second UDDS
cycle, two Highway Fuel Economy Test (HFET) cycles were driven
separated by a 15-second soak with the key switch in the "on"
position and the brake pedal depressed.
The SAE procedure requires energy consumption values to be
measured during the second HFET cycle only. It was, however, not
possible to adhere to this requirement, because the recharge data
will automatically reflect two HFET cycles. Therefore, EPA testing
deviated from this SAE requirement and measured energy consumption
over two UDDS and two HFET cycles.
-------
-4-
The SAE procedure also requires the measurement of a combined
UDDS/HFET driving range. The range test requires the vehicle to be
driven over two successive UDDS cycles followed by two HFET cycles.
A 15-second soak is required between the two HFET cycles followed
by a 10-minute soak after the second HFET cycle. The test cell fan
was shut off for the 10-minute soak but remained operational for
the 15-second soak. This test sequence was repeated until the test
termination criteria were met, at which point the vehicle was
quickly decelerated to a stop. Extended soaks between each cycle
(about 40 seconds), however, resulted due to the fact that the
video driver's aid could not be reloaded instantaneously.
The test termination criteria detailed in the SAE procedure
were somewhat different from current existing test procedures.
With the SAE J1634 procedure, the range test would continue even if
the vehicle could not meet the required speed profile provided the
vehicle is operated at the maximum available power output during
such occurrences. The criteria for termination that ended each
range test during EPA testing of the electric vehicle was the
requirement to stop the test if the vehicle does not reach 45 MPH
after 30 seconds from the 187-second mark and then hold a 45 MPH
speed until the 305-second mark of a UDDS cycle. Each range test,
therefore, ended at the 217-second mark of the UDDS cycle where a
45 MPH speed could not be attained.
The SAE procedure also describes how coastdown testing of an
electric vehicle shall be performed. EPA used coastdown data
supplied by INEL for setting up the dynamometer. The complete
coastdown curve for this vehicle is supplied in Appendix A of this
report. A Clayton water-brake dynamometer is not equipped to
handle an entire coastdown curve. Therefore, an EPA determination
was made to set the actual dynamometer horsepower based on the INEL
supplied 55 to 45 MPH coastdown time. The actual dynamometer
horsepower setting on the dynamometer control system was adjusted
until a time of 23.14 seconds was achieved for coasting from 55 to
45 MPH. The resultant actual dynamometer horsepower was 5.26 hp.
EPA also tested the electric vehicle over a test procedure
more similar to the conventional test procedure that would yield
separate city and highway energy consumption values so that a
petroleum-equivalent fuel economy value could be calculated, city
values were obtained by driving two consecutive UDDS cycles
separated by a 10-minute soak with the key switch in the "off"
position and the test cell fan off. Similarly, highway values were
obtained by driving the vehicle over two successive HFET cycles
separated by a 15-second soak period with only the brake pedal
depressed.
Both the city and highway cycles were started after soaking
the vehicle on charge overnight so that at the beginning of each
test cycle, the batteries were at a full charge, similar to the SAE
procedure. For the EV tested, the city energy consumption value
and the highway energy consumption value were obtained in a
slightly different manner than is done for conventional vehicles.
-------
-5-
Because it is not easy to obtain separate "hot" and "cold"
energy consumption values for an EV, the subject of how the two
UDDS cycles back=to-back compare with the cold/hot weighted
approach used for conventional testing should be discussed. In
essence, the approach used here has a weighting factor of 0.5
applied to the cold UDDS cycle and 0.5 applied to the hot UDDS
cycle. These values are not the same as the 0.43 and 0.57 factors
used for the conventional calculation methodology. Whether or not
the 50-50 weighting favors, is neutral to, or penalizes EVs with
respect to their energy consumption depends on the ratio of the
energy consumed by the EV in the cold start cycle to that consumed
in the hot start cycle. EVs have a lower ratio then CVs due to the
lack of the need for extra fuel (energy) to start the vehicle and
the substantially lower energy requirement to warm up the system to
operate more efficiently. For example, fuel (energy) rate
delivered to an engine is higher during the time it is warming up
and reduces to the lower "hot" rate only after the coolant
temperature exceeds a certain value. No similar losses exist for
EVs. Both vehicles suffer losses caused by excess friction in the
drivetrain during warmup. Because EVs have a more favorable cold
to hot energy consumption ratio, running two UDDS cycles back to
back allows this warmup benefit to be reflected in the test result.
If the energy consumed driving the latter part of the second
UDDS test ("Bag 4" in conventional vehicle nomenclature) is lower
than that consumed in the latter portion of the first UDDS cycle,
then the EV would be treated favorably by the way these tests were
run, since for conventional vehicles the assumption is made that
the two portions (Bag 2 and Bag 4) of the test are identical.
For the highway cycle fuel economy determination, conventional
vehicles are operated through two highway cycles. In order to
reduce variability and obtain the highest MPG value, only the
second of the two cycles is used for determining the MPG value even
though both cycles are run "hot." This is easy to do with
conventional vehicles since the fuel consumed is determined by a
carbon balance of the exhaust emissions and so only the second test
is sampled. With EVs, however, there isn't any convenient way to
not count the energy consumed during the first highway cycles, so
the way the tests were run is very slightly unfavorable to the EV.
Further study of these test procedure nuances may be warranted
if the SAE procedure is revised to permit the separate
determination of city and highway energy consumption. No
adjustments to the data in this report have been made, and if any
were contemplated, they would be quite small.
IV. Description of Test Vehicle
The test vehicle used for this round-robin test program was a
1988 Ford Escort Wagon equipped with a manual transmission and
radial tires. The vehicle was tested at an equivalent test weight
of 4,250 Ibs and an actual dynamometer horsepower of 5.26.
-------
The vehicle was converted to electric propulsion for DOE '-z\
Soleq Corporation. The system consists of 13 6-volt sealed lead-
icid Sonnenschein batteries (Model :?o. DF-150) with ^ sinai^
electric motor.
A picture of -his test vehicle is provided in Figure 1 below.
Figure 1
Ford Escort Wagon Test Vehicle
'^H***
V.
Test Facilities and Analytical Methods
EPA testing was conducted on a water-brake Clayton model ECE-
50 double-roll chassis dynamometer using a direct-drive variable
inertia flywheel unit and road load power control unit. The
vehicle was equipped with its own charging system. The vehicle was
recharged from a wall socket providing 125 volts at 20 amps. This
amount of recharge power was ample to recharge the battery pack to
a full state-of-charge during an overnight soak period.
The data acquisition system used in this test program to
measure power consumption and vehicle miles travelled was provided
to EPA with the test vehicle from the Idaho National Engineering
Laboratory. This system measured power consumed during both
vehicle testing and vehicle charging periods. Voltage and amperage
data was acquired every 100 milliseconds but was not stored at this
rate. The 100-millisecond data was averaged over the storage
period, which was 1 second for vehicle testing and 1-minute for
vehicle charging.
-------
-7-
VI. Discussion ef Teat Results
A. SAB J1634 Enerc
The Escort electric vehicle was first tested over the SAE
J1634 Energy Consumption Procedure. This testing yielded a
combined city/highway energy consumption value. Table 1 includes
the results from EPA testing on a Clayton water-brake dynamometer.
Table l
SAB J1634 Energy Consumption Test Results
Water-Brake Dynamometer:
Test #1
Test #2
Average
System AC
(W-hr/mile)
468
540
504
System DC
(W-hr/mile)
339
378
358
Vehicle Met DC
(W-hr/mile)
259
264
262
All results presented in Table 1 are in watt-hours per mile
(W-hr/mile). "System AC" represents the amount of energy supplied
from the wall outlet to the on-board recharging system. "System
DC" represents the amount of power delivered from the on-board
charging system to the batteries. "Vehicle Net DC" represents the
amount of overall power delivered from the batteries during driving
with the amount of energy generated from regenerative braking
subtracted.
From the water-brake chassis dynamometer results, the two
tests conducted differ substantially when considering System AC and
DC power usage. The two power levels differ by 15 and 12 percent
respectively between the two individual tests. The net DC results
correspond very well, however, with only a 2 percent difference.
From these results, it can also be seen that a substantial amount
of power from the wall socket is apparently lost in the on-board
charging system. An average of 18.4 kW-hr from the wall socket
results in only 13.1 kW-hr being supplied to the batteries. This
would seem to indicate that the charging system on this vehicle is
only about 71 percent efficient. This results in a substantial
penalty in a petroleum-equivalent fuel economy value since the
calculation is based on wall-power values. A more efficient
charging system would increase calculated fuel economy results.
The differences between "System DC" and "Vehicle Net DC" power
values result from battery inefficiency. The average "Vehicle Net
DC" power consumed over the SAE procedure was 9.5 kW-hr. Compared
to the wall-power usage of 18.4 kW-hr, the overall vehicle energy
efficiency from wall to battery-out is approximately 52 percent
-------
-8-
when considering charging system and battery inefficiencies. The
amount of energy delivered to the batteries during this driving
cycle from the use of regenerative braking was approximately 0.6
kW-hr and will be reflected in all three power consumption values.
B.
>tion Testing
The vehicle was tested at EPA over separate city and highway
test cycles to determine individually the city and highway energy
consumption values for this vehicle. The city values were obtained
by running two successive UDDS cycles separated by a 10-minute
soak. The results obtained from this testing are presented in
Table 2 below. Again, all values are in watt-hours per mile.
Table 2
City Cycle Energy Consumption Test Results
Water-Brake Dynamometer:
Test #1
Test #2
Average
System AC
(W-hr/mile)
670
642
656
System DC
(W-hr/mile)
428
395
412
Vehicle Met DC
(W-hr/mile)
303
303
303
The deviations in individual test results are less than those
acquired during the J1634 testing. System AC and DC results differ
between individual tests by only 4 and 8 percent respectively
accompanied by no deviation in vehicle net DC power usage between
tests. Again, large inefficiencies were noted during this testing
with the on-board recharging system and the batteries. The wall-
power used during this testing averaged 9.8 kW-hr, and the amount
of power supplied to the batteries during recharging averaged 6.2
kW-hr resulting in a recharging efficiency of 63.percent. During
this driving cycle, the "battery-out" power averaged 4.5 kW-hr.
When comparing this value to the wall-power, an overall battery
power efficiency of 46 percent is achieved.
The electric vehicle was next tested to obtain a highway
energy consumption value. The driving schedule used here was two
successive HFET cycles separated by a 15-second soak period. Table
3 below presents the results obtained from this testing. All
values are presented in watt-hours per mile.
-------
-9-
Table 3
Highway Cycle Energy Consumption Test Results
Water-Brake Dynamometer :
Test #1
Test #2
Test #3
Average
System AC
(ff-hr/mile)
564
463
441
489
System DC
(W-hr/mile)
346
287
286
306
Vehicle Net DC
(W-hr/mile)
222
220
217
220
A third test was conducted during this phase due to the rather
large discrepancy between the results from the first two tests.
All three tests, however, were included in presenting an average
value.
These results again indicate a rather large inefficiency in
the on-board charging system. The wall-plug power used for these
three tests averaged approximately 10.4 kW-hr. The corresponding
amount of power supplied to the batteries during recharging
averaged 6.5 kW-hr, a recharge efficiency of 62 percent.
When driving over two successive HFET cycles, the brakes are
only applied four times. Therefore, the amount of power supplied
to the batteries during driving by the regenerative braking system
should be very small. The overall battery efficiency obtained here
for "wall to battery-out" is approximately 45 percent. (The
"Vehicle Net DC" power used averaged 4.7 kW-hr for this testing.)
C.
Pafcirolatm—Eerulvalent Ttiel Bconomv Results
Using the recent DOE rulemaking entitled "Electric and Hybrid
Vehicle Research, Development, and Demonstration Program;
Equivalent Petroleum-Based Fuel Economy Calculation,"[3] a
petroleum-equivalent fuel economy value can be calculated for this
electric vehicle. The only requirements needed for this
calculation are the city and highway energy consumption values
presented in the previous section.
The DOE rulemaking references the SAE J1634 test procedure for
obtaining energy consumption values for an electric vehicle; SAE
procedure J1634 will only allow for a combined city/highway energy
consumption value to be measured. EPA tested the electric vehicle
separately over city and highway driving cycles so that these
energy consumption values could be determined independently of each
other.
-------
-10-
Presented here is a step-by-step petroleum-equivalent fuel
economy calculation resulting from EPA testing on the water-brake
chassis dynamometer and the use of the method outlined in the DOE
rulemaking.
City energy consumption = 0.656 kW-hr = 1.524 mile
mile kW-hr
Highway energy consumption = 0.489 kW-hr = 2.045 mile
mile kW-hr
City fuel economy = 1.524 mile x 33.44 kW-hr - 50.963 MPG gasoline-equiv.
kW-hr gallon
Highway fuel economy » 2.045 mile x 33.44 kW-hr = 68.385 MPG gasoline equiv.
kW-hr gallon
It should be noted that the value for the kW-hr of gasoline
gallon
used here differs from the value suggested by DOE. The value used
here is consistent with the value used by EPA in other fuel economy
rulemakings that involve fuel energy content.
After applying the factor for converting a kilowatt-hour value
into equivalent gallons of gasoline, a composite fuel economy value
for the electric vehicle can be determined based on the 55/45
method used for CAFE calculations.
MPGcomp08ite
M^city
= 57.562 MPG gasoline equiv.
50.963 68.385
Now, it is necessary to apply the Petroleum Equivalency Factor
(PEF) found in the DOE rulemaking for an overall petroleum-
equivalent fuel economy for this electric vehicle.
MPGeo-()0.it. x PEF
57.562 X 2.32 =- 133.5
Therefore, the value that would used for this electric vehicle
in any calculations of a manufacturer's average fuel economy would
be 133.5 miles per gallon.
-------
-11-
D. SAB J1634 Range Testing
The last phase of testing with this electric vehicle consisted
of range testing following the protocols outlined in the SAE J1634
test procedure. The vehicle was driven over successive UDDS and
HFET driving cycles until the test termination point was reached.
The driving cycles and test termination point are described in more
detail in Section III of this report.
The SAE procedure is written such that the range test shall
continue if the vehicle cannot meet the driving trace just as long
as the vehicle is operated at maximum power output. The criteria
that ended the range test at EPA was where the procedure states
that "if the vehicle cannot attain 45 miles per hour 30 seconds
after the 187-second in the UDOS cycle, the test shall be
terminated.1* At the 217-second mark of the third UDDS driving
cycle (after two previous UDDS cycles and two HFET cycles) , the
vehicle speed was not 45 MPH, the vehicle was quickly decelerated
to stop, and the range determined. The resultant range was
measured as 37.8 miles. INEL informally reported a similar range
result to EPA.
VII. Future Efforts -
The electric vehicle is currently being tested at EPA on a
Horiba 48-inch single-roll electric chassis dynamometer over the
same test sequence described in this report. These results will be
published in a separate EPA technical report when testing is
completed.
The vehicle will then be tested at Ford Motor Co. and GARB.
INEL will then publish a report describing all the results from
testing at the four sites included in this round-robin test
program.
VIII , Aclmoif
The authors would like to acknowledge the cooperation and
support of George H. Cole and Roger A. Richardson of the Idaho
National Engineering Laboratory for supplying EPA with the test
vehicle and the data acquisition system.
The authors also appreciate the efforts of James Garvey, Ray
Ouillette, Rodney Branham, Robert Moss, and Carl Fuller of the
Testing Support Branch who conducted the driving cycle tests and
data acquisition. The word processing and editing efforts of
Jennifer Criss and Lillian Johnson of the Technology Development
Group are also appreciated.
-------
-12-
IX. References
1. "Proposed Regulations for Low-Emission Vehicles and Clean
Fuels—Staff Report," State of California Air Resources Board,
August 1990.
2. "Electric Vehicle Energy Consumption and Range Test
Procedure," SAE Surface Vehicle Recommended Practice, SAE J1634,
1993.
3. "Electric and Hybrid Vehicle Research, Development, and
Demonstration Program; Equivalent Petroleum-Based Fuel Economy
Calculation," 10 Code of Federal Regulations. Part 474, Docket No.
EE-RM-94-101, February 4, 1994.
4. Letter from George H. Cole, Program Manager, Electric and
Hybrid Vehicle Program, INEL to Ronald M. Schaefer, Mechanical
Engineer, U.S. EPA, February 7, 1994.
-------
A-l
Appendix A
Ford Escort Electric Vehicle Coaatdovn Data
Spaed
(MPH)
60
59
58
57
56
55
54
53
52
51
50
49
48
47
46
45
44
43
42
41
40
39
38
37
36
Inertia
f ' 0 -
f '2 -
VI (MPH)
VO (MPH)
Time
(see)
0.00
1.88
3.80
5.77
7.77
9.82
11.92
14.06
16.24
18.48
20.76
23.10
25.48
27.92
30.41
32.96
35.56
38.22
40.94
43.73
46.57
49.47
52.44
55.47
58.57
Ht. = 4,250 Ibs
38.7385 lb
0.0186 lb/MPH*2
55
. 45
Speed
(MPH)
35
34
33
32
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
Miw =
Mdlc -
60
10
Time
(sec)
61.73
64.96
68.27
71.64
75.08
78.59
82.17
85.82
89.55
93.35
97.22
101.16
105.17
109.26
113.41
117.64
121.93
126.29
130.72
135.21
139.76
144.38
149.05
153.77
158.54
132.09423
1.9814135
20
10
t-tO (8) - 23.14 163.36 45.73
Hp - (f'O + f'2*V*2) * V * (5280/(3600*550))
flpeed (MPH)
60
50
40
30
20
10
Power (Ho)
16.91
11.36
7.31
4.44
2.46
1.08
(Power (lew)
12.61
8.47
5.45
3.31
1.84
0.81
------- |