72-27
EXHAUST EMISSION ANALYSIS
OF THE WILLIAMS RESEARCH GAS
TURBINE AMC HORNET
May 1972
Leonard D. Verrelli
and
Casimer J. Andary
Test and Evaluation Branch
Environmental Protection Agency
-------
Background
The New York City Air Resources Board was awarded an
EPA Grant for the purpose of evaluating field experience
with a turbine powered passenger car. The Test and Evaluation
Branch was given the task of testing this prototype vehicle
and developing simplified test procedures which could be. used
by New York City personnel. Initial testing of the vehicle
was performed in December 1971 using the 1970 Federal Test
Procedure to satisfy the NYC-Williams contractual obligations.
The vehicle was then delivered to the New York City Air
Resources Board on January 30, 1972, for acceptance and some
public demonstrations. The vehicle was returned to the EPA
laboratory for.more extensive testing to obtain complete
emission data and to develop a test procedure to be utilized
by New York ARB while the vehicle is on location in New York
City. The turbine car arrived at the EPA facility February 25,
1972. Testing was conducted through the month of March and
returned to Williams Corporation for a 60-hour teardown and
inspection as previously agreed with New York City ARB. The
vehicle was returned to EPA after the engine had b.een torn
down for inspection and after modifications were made to the
shut-down fuel drain system, combustor louver design and gear-
box breather system. It should be kept in mind that this
vehicle incorporates a "state of the art" engine and does
not represent the emission levels that can potentially be
achieved with gas turbine power plants.
Vehicle Tested '•.''"" '•
The Williams Gas Turbine, WR-26, regenerative turbine
automotive engine was mounted in a 1971 American Motors
Corporation Hornet test vehicle. The engine utilizes a
single power turbine to develop 80 horsepower. The engine
regenerator incorporates two large discs of a ceramic-
glass material for inlet air preheating. The standard
AMC automatic transmission is coupled to the engine through
a drop shaft gear system. Indolene Clear gasoline was the
test fuel used throughout the test program.
Test Program
Testing was accomplished using the 1970, 1972, and
1975 Federal Test Procedures. In addition, steady state
-------
•and proportional sampler tests.were conducted. As a
means of cross checks, continuous raw exhaust was'
sampled during sor.e of the aforementioned test procedures.
A.1200 CPM constant volume sampler (CVS) system was
used for the majority of the testing. Because of high
back pressure in the system a minor modification had to
be made to the CVS. The dilution air filters at the CVS
were blocked-off and air from the test cell was drawn
into a plenum collecting the vehicle exhaust. Mixing
occurred at the exit of the vehicle exhaust.
The cold start tests were made on a chassis dynamometer
after the vehicle had been parked in a 68-86° ambient condi-
tion for at least 12 hours.- Hot start data was taken after
the vehicle had been brought up to operating temperature.
It was then shut down and restarted according to the FTP
being performed.. ''
The concentrations of pollutants were measured using
both hot and cold flame, ionization detectors (FID), for
unburned hydrocarbon (KC). Non-dispersive infrared (NDIR)
analyzers were used for carbon monoxide (CO) and carbon •
•dioxide (CC^). A chemiluminescent analyzer was used to
determine both nitric oxide (NO) and -nitrogen dioxide
(NOp). The sum of NO and NO- is reported as NO . Concen-
trations obtained were then used to calculate the mass of
emission per mile of operation and reported as grams per •
mile (gpm).
To measure fuel consumption, a weigh scale system
was used and a carb-on balance technique was additionally
applied to provide a cross check. A significant difference
was noted throughout the test program between the actual
fuel weight data and a carbon balance calculation performed
using the emission data collected. As of this report an
explanation for this difference has not been found and
further investigation is continuing. (See table 6)
Running the vehicle with the turbine bypass in the
closed position was found to reduce the hydrocarbon (HC)
between 50 - 7053. This is due to the more even combustion
during deceleration requiring fewer relights. The .bypass
system for this particular vehicle has been set to operate
in the open position. However, it can easily be operated
in the closed position.by merely flipping a switch. Oper-
ating in the open mode provides some engine braking and
— 2 —
-------
cooler starting, which is not true in the closed position.
Steps are being taken to eliminate the bypass system from
future engines. Another method to achieve engine braking
.and cooler starting will be employed.
The purpose in performing numerous techniques and
test procedures was to try to establish an acceptable
test procedure for gas turbine testing for the New York
City Air Resources Board.
When sampling from low emission high flow vehicles
the hydrocarbons in the dilution air can contribute a
major portion of the total hydrocarbons and the correction
factor as called for in the Federal Register does not
represent a true picture. 'For the 1972 FTP tests, the
following calculation procedure was used:
Ce - (Ct - Cb - ) (1)
- - . .Vt Ve
Where:
Ce = concentration of undiluted exhaust
' .Ct = concentration of sample bag
Cb = concentration of background dilution air
Vt = volume flow rate of sample, or total CVS flow
Ve - volume flow rate of engine exhaust
Then the mass (gpm) of each pollutant emitted is calculated
from:
W = K Ce Vei (2)
Where:
K.= density of each pollutant
Vei = average volume of engine exhaust per. mile
(1) See emissions results, page 2, report 71-30, "Exhaust
Emission Analysis of 'the Williams Research Gas Turbine
Volkswagen", Leonard D. Verrelli, EPA.
-------
In all, eighteen (18) tests were performed according to
the 1972 FTP. Of the eighteen tests, six were cold starts
with the bypass open, six were hot starts with the bypass
open and three cold starts, three hot starts with the bypass
closed.
It was suggested by the manufacturer towards the end
of the test program, to try some emission tests with the
bypass closed. A few were accomplished. (See Table I).
Table III reflects the sane testing after engine modifications
were made.
Additionally there v;ere twelve tests run using the
• 1975 FTP. These tests used a different method to calculate •
mass emissions.
First a dilution factor is calculated according to
this equation:
DF = Vt/Ve
Where:
DF =. dilution factor
Vt = CVS volume flow rate
. Ve = engine volume flow rate
Then:
Cc = Ct-Cb (1 - i )
DF
Where:
Cc = corrected concentration of dilute exhaust
Ct = concentration of sample bag
Cb = concentration of background dilution air
Then the mass emissions in grams per mile are calculated
from;
Mass = K Cc- Vte
Where:
K = density, of pollutant
Vte = volume pumped through CVS during test
-------
Of these .eighteen 1975 tests, six were cold starts with
the bypass open, six were hot with the bypass open, four
cold with bypass closed and two hot with bypass closed.
This data is summarised in Table II. Additionally Table IV
reflects 1975 FTP after engine modification.
The proportional sar.pler data taken using the
7-mode driving cycle was calculated using the following
procedure:
X concentration X K = pollutant in GPM
Where:
CFM = airflow of the engine in cubic feet/min.
Miles = distance traveled over the cycle in a min.
Concentration = pollutant measured in PPM
K = density factor of % for HC = 16.33 X 10~6
CO =- 32.97 X ID"6
co2 = .5186
NOX = 54.16 X 10"6
The proportional sampler sampling method was explored
for two reasons, one, to develop a procedure that could be
adapted to the equipment the New. York City ARE had available,
secondly, to compare previous emission data obtained while
testing the Chrysler turbine as reported in SAE report
#680402, "Emissions From A Gas Turbine Automobile" dated
May 20-24, 1968. This data is summarized in Table V.
Conclusions
By positioning of the povrer turbine bypass in the
closed position it was demonstrated that low hydrocarbon
(HC) emissions are possible with this engine; however,
carbon monoxide (CO) and nitric oxides (NOX) were not
reduced sufficient.ly to be able.to meet 1975 or 1976
emission standards.
-------
As a result of our extensive emission testing, an
in depth driveability investigation with the vehicle
was not possible. Limited testing revealed slower than
normal accelerations, limite'd engine retardation and a
maximum speed of from about 78 mp~h.
. The various methods of testing used provided a good
data base for evaluating test methods and procedures.
The proportional sampler method of collecting emissions
proved to be the most practical for wor'k to be performed
at the New York City ARE.
- 6 -
-------
TABLE I
FY72 Federal Test Procedure
Before Engine Modification
"'"BPO cold BPO hot 2BPC cold BPC hot
5 tests 3 tests 1 test
HC GPM Hl8 '1.52 1.26
CO GPM 10.52 -9.^2 9.83
C02 GPM 1726 1676 .1769
NOX GPM 3.16 3.25 3.29
Bypass open
2Bypass closed
- 7 -
-------
rp l\ py -^ y y
I .-iDLji-i X J.
1975 Federal Test Procedure
Before Engine Modification
1BPO cold BPO hot 2BPC cold BPC hot
5 tests 4 tests 2 tests 1 test
EC GPM 2.84 3.33 1.23
CO GPM 9-12 8'. 56 ' 9.6? 9-10
CO GPM . 1659 1636 1634 1675
NOX GPM 2.97 2.79 , 2-86 3.00
•'•Bypass open
2Bypass closed
- .8 -
-------
TABLE III
After Engine Modification
1972 Federal Test Procedure
- Power Turbine Bypass Open
Cold Start Hot Start
Test * 12-2236 . 12-2237 12-2238 12-2243
HC* .97 ; .57 .71 .93
CO* 10.1 •...' 9.14 8.58 7.34
C02 1864.0 .". 1874.0 1704.0 1462.0
NOx 3.68 3.5 3.3 2.4
. Power Turbine Bypass Closed
Cold Start Hot Start
Test * 12-2239 . 12-2240 12-2241 12-2242 12-2245
HC* .23 .27 .31 .31 .26
CO* 7.9 6.84 7.02 7.50 7.8
C02 1599,0 1547.0 1559.0 1538.0 1454.0
NOx 2.64 2.59. 2.77 2.55 2.8
*GPM
- 9 -
-------
*GPM
TABLE IV
After Engine Modification
1975 Federal Test Procedure
Power Turbine Bypass Open
Test #
HC*
co2
NOx
Cold Start
12-2244
.62
7.43
1443.0
2.8
Hot Start
12-2246 12-2250
.74
7.77
1435.0
2.0
.72
6.92
1371.0
2.5
Power Turbine Bypass Closed
Test *
HC*
CO*
CO*
ft
NOx
Cold Start
•
12-2247 12-2248
.32 .23
6.34 - 7.41
1385.0 1454.0
2.7 3.07
Hot Start
12-2249
.23
6.33
1382.0
2.7
- 10 -
-------
TABLE V
Proportional Sampler Evaluation
Using 7-mode Cycle (Hot)
Bypass Onen
Before Engine Modification.
i
Test 9 EC gpm CO gpm j CO2 gpm NOx gpm
12-2188 .84 6.34 1323.0 2.77
Before Engine Modification
Test * HC gpm CO gpm C02 gpm NOx gpm
12-2251 .37 6.71 1567.0 2.64
Chrysler Turbine Data *
HC gpm CO gpm C02 gpm NOx gpm
.91 7.04 -9110,0 1.86
*Test data shown was calculated from the results reported in the
SAE report, "Emissions From A Gas Turbine Automobile", #68042,
May 1968.
- 11 -
-------
TABLE VI
Test Procedure
Fuel Consumption Data
(miles per gallon)
Before Engine Modifications
1972 FTP (cold start)
1972 FTP (hot start)
BP* Test No.
0
0
12-2208
12-2209
Actual
Weighed Data
6.15
6.28
Carbon Balance
4.91
3.65
1972 FTP
1972 FTP
1972 FTP
1972 FTP
1972 FTP
1972 FTP
1972 FTP
1972 FTP
1972 FTP
Fuel Consumption Data
After Engine Modifications
(cold start)
(cold start)
(hot start)
(hot start)
(hot start)
(hot start)
(hot start)
(hot start)
(hot start)
1975 FTP (cold start)
1975 FTP (hot start)
1975 FTP (hot start)
0
C
0
0
0
C
C
C
0
12-2236
12-2239
12-2237
12-2238
12-2243
12-2240
12-2241
12-2242
12-2245
0 12-2244
0 12-2246
C 12-2247
5.8
6.7
6.2
7.0
7.0
7.1
8.1
7.2
6.9
7.1
5.1
5.1
4.6
5.4
4.6
5.1
5.9
5.6
5.5
5.6
5.9
6.1
6.3
6.5
*BP - Bypass
0 - Open C- Closed
- 12 -
------- |