EPA-600/2-77-236
December 1977
Environmental Protection Technology Series
EVALUATION OF A PROPORTIONAL
SAMPLER FOR AUTOMOTIVE
EXHAUST EMISSIONS
Environmental Sciences Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the ENVIRONMENTAL PROTECTION TECH-
NOLOGY series. This series describes research performed to develop and dem-
onstrate instrumentation, equipment, and methodology to repair or prevent en-
vironmental degradation from point and non-point sources of pollution. This work
provides the new or improved technology required for the control and treatment
of pollution sources to meet environmental quality standards.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/2-77-236
December 1977
EVALUATION OF A PROPORTIONAL SAMPLER FOR
IAUTOMOTIVE EXHAUST EMISSIONS
by
Peter Gabele
Emission Measurement and Characterization Division
Environmental Sciences Research Laboratory
Research Triangle Park, North Carolina 27711
ENVIRONMENTAL SCIENCES RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
RESEARCH TRIANGLE PARK, NORTH CAROLINA 27711
-------
DISCLAIMER
This report has been reviewed by the Environmental Sciences Research
Laboratory, U.S. Environmental Protection Agency, and approved for
publication. Mention of trade names or commercial products does not
constitute endorsement or recommendation for use.
-------
ABSTRACT
A test program was conducted to evaluate a proportional sampler for use
in automotive exhaust gas emissions research. Automobile emissions test re-
sults obtained using the proportional sampler were compared with results
obtained using the conventional constant volume sampler.
Measurements obtained using the proportional sampler for hydrocarbons,
carbon monoxide, nitrogen oxides, and carbon dioxide are within 19, 46, 20,
and 14 percent, respectively, of measurements obtained using the constant
volume sampler. Such differences render the proportional sampler unaccept-
able as a quantitative research tool.
Inability of the exhaust gas flow meter to accurately measure pulsating
exhaust flow is cited as the principal cause of error in the proportional
sampler.
iii
-------
SECTION 1
INTRODUCTION
Light duty vehicle exhaust emission standards are expressed in units of
mass per vehicle mile traveled. The current emissions certification proce-
dure used in the United States utilizes a chassis dynamometer, a constant
volume sampler, and a set of exhaust gas analyzers. With these three basic
blocks of equipment, automobile exhaust emissions can be generated, represen-
tatively sampled, and measured.
The constant volume sampler (CVS) technique is well known and its
essential features are described in the Federal Register (1). Vehicle
exhaust gases are ducted into a CVS where the gases are cooled by dilution
air prior to sampling. Dilution ratios are significantly high to minimize
water condensation within the sampler and to minimize unwanted chemical
reactions. The combined exhaust plus diluent flow rate is maintained con-
stant in order to simplify both sampling and computational requirements.
From strictly a research point of view, the CVS, although reliable and
accurate, suffers from two primary drawbacks: it is a large, awkward piece
of equipment having virtually no mobility, and it dilutes the raw exhaust
sample resulting in low gas specie concentrations. Frequently in research
work it becomes desirable to examine the emission of non-regulated gases
having raw exhaust concentrations near the threshold sensitivity level. In
such cases the requirement for exhaust gas dilution becomes an unacceptable
drawback.
One alternative mass measurement technique utilizes a proportional
sampler for sampling raw exhaust gas. As the name implies, the proportional
sampler samples raw exhaust at a flow rate proportional to the total exhaust
gas flow rate. The concept of proportional sampling was first employed in
the early sixties with the development of a device which sampled automobile
-------
exhaust at a rate proportional to the measured air-flow into the carbureter
(2). This technique had as its best feature the capability of sampling the
vehicle during actual real world operation since the equipment was completely
portable. Although the technique showed great promise, the subsequent
arrival of engine air pumps and carburetor air bleeds rendered intake air-
flow measurement overly cumbersome. Consequently, the proportional sampler
was abandoned in favor of CVS techniques.
More recently, gas flow metering technology has reached the threshold
of engine exhaust gas measurement. Among the possible candidate flow meters,
the ultrasonic vortex meter has been one most highly recommended for engine
exhaust measurement (3). For this reason, a proportional sampler for automo-
bile emissions research was developed around an ultrasonic vortex flow meter.
This paper describes an evaluation performed on a prototype proportional
sampler developed under contract for EPA by Aeronutronic-Ford. A complete
description of the sampler and its development is contained in the contract
final report (4).
-------
SECTION 2
CONCLUSIONS AND RECOMMENDATIONS
An evaluation of a proportional sampler was completed to assess its value
as a tool for use in automobile emissions research applications. The evalua-
tion concerned itself with the sampler's ability to obtain representative
exhaust gas samples from four different automobiles. No attempt was made
to comprehensively assess the effect of sample degradation. Results from
continuous sample monitoring versus bag sample analysis implied that the raw
exhaust sample's integrity did not significantly deteriorate during the time
period between sampling and analysis.
The test results indicate that measurements obtained for hydrocarbons
(HC), carbon monoxide (CO), nitrogen oxides (NO ), and carbon dioxide (C09)
/\ £
using the proportional sampler are within 19, 46, 20 and 14 percent, respec-
tively, of measurements obtained using the constant volume sampler. The
magnitude of these differences renders the proportional sampler unacceptable
as a quantitative research tool in its present configuration. It should be
useful, however, for qualitative emissions testing.
The vortex flow meter was found to be the primary source of error within
the proportional sampler. Although the flow meter performed adequately when
metering non-pulsating type airflows, rather significant metering errors were
observed when operating under actual automobile exhaust flow conditions. No
detailed tests were carried out to determine what specific aspect of engine
exhaust flow was adversely affecting the measurement capabilities of the
vortex meter, but it is related to gas pulsations.
Future work involving integration of vortex flow meters into automobile
emissions measurement systems should attempt to resolve this engine exhaust
flow metering problem. Other types of flow meters being considered for use
in either proportional samplers or continuous mass measurement systems should
-------
be qualified in tests using actual engine exhaust. A test procedure similar
to that used in this study employing a Roots or turbine meter to measure and
record integrated flow volumes is a simple and effective method for evaluating
candidate flow meters.
-------
SECTION 3
PROPORTIONAL SAMPLER
GENERAL DESCRIPTION
A schematic of the proportional sampler fluid system is shown in Figure
1. The ultrasonic vortex flow meter, the sample valve, and the signal proces-
sor (not pictured) are the three main components of the system. Exhaust gas
from the vehicle enters the sampler and is immediately sampled. The sample
valve is controlled through processor signals to sample at a mass flow rate
proportional to the mass flow rate of total exhaust as measured by the vortex
flow meter. Both sample and total exhaust flow rates are corrected to stand-
ard conditions via signal processor with the aid of temperature and pressure
sensors. This insures that proportional mass flow is being maintained even
though actual exhaust to sample flow ratios may vary considerably.
Figure 2 shows a front view of the proportional sampler. On the front
panel are displays for cumulative exhaust volumes measured during the test
and analog outputs for actual and standard exhaust gas flow rates, sample
temperature, flow meter temperature, and sample valve frequency. Information
obtained through proper use of these outputs can be used effectively when
troubleshooting the device.
Sample Line
The sample line between the sample valve and the condenser coil in the
refrigerated water bath is temperature-controlled to 93° C. This precaution
is taken to avoid condensation of heavy hydrocarbons and water vapor in the
sample line. The condenser coil is also temperature-controlled to +2° and
the condensation is controlled at this cool temperature. After passing
through the condensate trap, the sample is pumped into a tedlar sample bag for
post-test analysis.
-------
Exhaust Gas Heat Exchanger
One rather obvious feature of the proportional sampler is the heat
3
exchanger tank. The tank occupies a volume of 0.24 m and is designed to
cool the exhaust gases before their entry into the ultrasonic flow meter.
Since flow velocity is propagated downstream from the engine at sonic speed,
the location of the vortex flow meter after the heat exchanger does not
introduce any problems associated with sample lag time. The sample valve has
been observed to respond almost instantaneously to changes in engine speed.
A test conducted using a dual channel oscilloscope to simultaneously monitor
airflow entering the carburetor and exhaust flow at the vortex flow meter
indicated a 125 millisecond response time. This test was run with the
standard 12-foot section located between the vehicle and the proportional
sampler.
-------
SECTION 4
EXPERIMENTAL PROCEDURES
GENERAL PROCEDURES
The experimental evaluation of the proportional sampler was carried out
in an automobile emissions laboratory equipped with a chassis dynamometer.
Routine gas (HC, CO, C09, AND NO ) analyses were carried out in accordance
£ X
with specifications given in the Federal Register (1). Chassis dynamometer
operation was also conducted in accordance with these specifications, one
exception being that a 24,000 CFM cooling fan was used during all tests.
Both the proportional sampler and the conventional constant volume
sampler were used to sample gaseous emissions from the test vehicles. The
samples were analyzed and emission measurements were obtained for each case.
Measurements obtained using the proportional sampler were compared to those
obtained using the CVS. These comparisons formed the basis for the propor-
tional sampler evaluation.
Test Vehicles
Tests were conducted on four different vehicles. Each of the vehicles
is described in Table 1. In order to study the effect of engine size on
proportional sampler operation, the vehicles were selected to purposely
represent a fairly broad range of engine displacement sizes. The vehicles
were tested in the following order: 1) Ford Mustang II, 2) Chevrolet Nova
(Gould catalyst car), 3) Ford Pinto, and 4) Honda CVCC.
Test Cycle
Each cycle consisted of the hot transient phase of the 1975 Federal Test
Procedure (75 FTP), specifically, the 505 second portion associated with bag
3 of the 75 FTP. The test cycle included a 10-minute hot soak period
between each test. Because the Honda developed an overheating problem, the
soak periods between its tests were increased to 20 minutes.
-------
The hot 505s were selected in order to rule out variabilities associated
with the cold transient phase. The cold transient phase has been identified
as the major source of variability for the 75 FTP because parameters that
affect engine and catalyst warm-up characteristics and carburetor choke
activity all come into play (5).
At least ten test cycles were completed for each vehicle on the CVS as
well as on the proportional sampler. Each vehicle was tested using the pro-
portional sampler and CVS techniques alternately over groups of five test
cycles. This procedure was followed in order to compensate for possible
drifts in vehicle emission rates as the testing progressed. A review of the
data at the completion of the test period, however, revealed that such drifts
did not occur to any significance. • -''->••-
• . , " -
Troubleshooting
A number of checks were run at the conclusion of the basis evaluation
program to identify the sources of error involved in the tests.
Sample Lead Time Investigation—
The discovery of rapid response times with regard to sample valve opera-
tion prompted a closer examination of the exhaust system between the engine
and the proportional sampler. Because a 12-foot section of 3-inch flexible
line had been used during the tests, gases from the engine had been lagging
behind the sample valve about (50/ J-.125 seconds, where x equals the actual
^
flow rate through the exhaust line in cubic feet per minute. This effec-
tively created a sample lead time error. For the larger displacement engines
the sample lead time was less because of the higher flow rates involved. But
for a vehicle having a small engine, such as the 1500cc Honda, the flow rates
observed over the test cycle were always below 100 ACFM, and therefore sample
lead time errors were probably significant.
Additional tests were conducted on the Honda using a short flexible con-
necting line between the vehicle and the proportional sampler. By reducing
exhaust residence time before the sample port, it was hypothesized that sample
lead times would likewise be reduced and a more representative sample would
be obtained.
8
-------
Sample Degradation Test—
A rather quick and simple test was performed to qualitatively determine
the extent of sample degradation occurring within the proportional sampler.
The Mustang II was operated in a steady state mode and gas samples were drawn
off using the proportional sampler. These samples were analyzed using two
different procedures. In the one procedure the sample was delivered directly
for analysis, while in the other procedure the sample was stored in a tedlar
bag for analysis following completion of the test. The analysis results
from these two separate sampling procedures were compared as a way of assess-
ing the extent of sample degradation in the tedlar bag sample.
Dilution Ratio Tests--
Dilution ratio tests were conducted to isolate and hopefully quantitate
the error contribution of the gas analyzers. The Ford Mustang II was oper-
ated at a steady state condition and gas samples were collected simultaneously
by the CVS and proportional sampler. Following analysis, the dilution ratio
of the CVS sample was calculated by dividing each proportional sampler specie
gas concentration (raw exhaust) by its corresponding CVS concentration. In
this manner four dilution ratio values were calculated. Comments regarding
the performance of the analyzers could be made by simply comparing the
different values obtained for the dilution ratio.
Vortex Flow Meter—
An examination of the ultrasonic vortex flow meter was conducted to
determine its accuracy in metering automobile exhaust gas flow. A Roots meter
with a cumulative volume indicator was chosen as the reference meter for
evaluating the vortex meter. The Roots meter had been checked out using a
recently calibrated laminar flow element and excellent agreement was obtained
for both devices over the range of flows to be examined. The Roots meter was
connected to the carburetor intake through a 0.2 cubic meter buffer used to
effectively dampen out flow pulsations to the Roots meter. An engine without
an air pump or carburetor air bleeds was chosen because the mass of fuel and
air entering the carburetor was equal to that of the exhaust gas leaving the
tailpipe. With the vehicle's exhaust pipe connected to the proportional
sampler, a series of steady and non-steady state tests were conducted and
volume flows at both the Roots meter and proportional sampler were recorded.
-------
Before the results could be directly compared, the flow volumes were
adjusted to account for fuel addition at the engine and water vapor removal
in the heat exchanger. Flow compensation at the Roots meter for downstream
addition of fuel at the engine was calculated by assuming an air-fuel ratio
of 14:1. Flow compensation at the vortex meter to account for water vapor
removal upstream was calculated using a hot, raw exhaust water vapor content
of 16.7 percent by volume. Since the exhaust gas exited the heat exchanger
at 80°F, the water vapor content was reduced to 5.8 percent by volume, an
overall reduction of almost 11 percent.
10
-------
SECTION 5
RESULTS AND DISCUSSION
BASIC EVALUATION
Emissions of hydrocarbons, carbon monoxide, nitrogen oxides, carbon
dioxide, and methane were measured from each of the test vehicles over numer-
ous test cycles. The measurement data was broken down by vehicle and grouped
according to the sampling method used when testing. Thus there were two dis-
tinct sets of data for each vehicle tested - one associated with the CVS
sampling technique and a second associated with the proportional sampling.
Each set of data was summarized by statistically calculating the arithmetic
means (x), coefficients of variation (CV), and 95 percent confidence inter-
vals (CI). The emissions measurement summaries for each vehicle are con-
tained in Tables 2-5.
To provide a good visual illustration depicting the extent of agreement
between the two sampling methods, the 95 percent confidence intervals are
graphically featured in Figures 3-7 for each gas specie measured. Where
confidence intervals fail to overlap, a difference between the two sampling
methods statistically exists. This difference varies for the different specie
gases examined. The only consistent trend observed was that hydrocarbon
measurements were high by the proportional sampler for each vehicle tested.
Carbon monoxide results showed the best agreement in that only one of four
displayed any significant difference. Both C0? and NO measurements were
£ TV
low by the proportional sampler in three out of four cases, while methane was
high by the same majority. The "odd-ball" cases for C0? and NO both occurred
^ rt
when testing the Honda, which led to the speculation that some aspect asso-
ciated with testing this vehicle was possibly affecting the results.
Results of Sample Lead Time Investigation
Because the most obvious feature of the Honda was its small engine size,
11
-------
sampling irregularities derived from this physical aspect were suspected.
A hypothesis was formed that sample lead times, a phenomenon discussed in the
previous section, were responsible for the proportional sampler's failure in
obtaining a representative sample. To investigate this hypothesis, addi-
tional tests were run using a 2-foot flexible exhaust line between the
vehicle and sampler in lieu of the regular 12-foot line. Because the 2-foot
line substantially reduced the exhaust gas residence time in the line between
the vehicle and the sample port, it was hypothesized that sample lead times
would be minimized. Test results shown in Table 6, indicated that the
measurement results were not as sensitive to changes in exhaust line length
as hypothesized. Although this test did not rule out the existence of sample
lead time errors altogether, it did lower the suspicions that this problem
was the sole cause of the measurement differences noted in Figures 3 - 7.
Sample Degradation Test Results
The two sets of analyses carried out separately on .the directly sampled
exhaust gas and on the bag sample yielded the same concentration levels.
These results indicated that storage of the sample over the 505 second test
interval was not having any significant effect on sample integrity. The raw
exhaust hydrocarbon and nitrogen oxide concentrations for the vehicles tested
averaged 750 ppmc and 800 ppm, respectively. These levels are considerably
lower than those observed in reference 2 where tests showed the disappearance
with time of hydrocarbons and nitrogen oxides in raw exhaust bag samples.
For the levels observed in this study, the reaction rates would be signi-
ficantly lowered and thus it was not expected that sample integrity would be
jeopardized.
Dilution Ratio Test Results
The dilution ratio test results indicate that the analyzers performed
extremely well (Table 7). The average dilution ratio calculated over the
steady state test was 9.37 with a coefficient of variation equal to 4 percent.
During testing the raw exhaust volumes through the proportional sampler and
diluted volumes through the CVS were also measured. The calculated dilution
ratio based on these measurements was 8.85, a value about 6 percent lower
than the dilution ratio calculated usiny specie gas ccr.cer.troticr.s. This
12
-------
suggested that the raw exhaust volumes being measured by the proportional
sampler were somewhat high, assuming, of course, that the CVS was not in
error.
Vortex Flow Meter Examination Results
The examination of the vortex flow meter was conducted by first using a
test vehicle, and then later using a variable speed blower for exhaust gas
generation through the meter. A statistical summary of the examination
results, which consists of a linear regression analysis, is shown in Table 8.
2
The values listed in the R column are the squares of the multiple correla-
2
tion coefficient R , which is defined as
R2 _ Sum of squares due to regression
Total (corrected) sum of squares.
2
The closer R comes to equaling one, the better the fitted equation explains
the variation in the data (6).
Use of Engine Exhaust--
Flow measurements were made with the engine operating over steady state
and non-steady state modes of operation. The steady state speeds were
adjusted to obtain flow measurement data over a range extending from 20 to 80
SCFM. The regression analysis data indicate that the truly strong correlation
2
required between the two flow meters is lacking. Values for R below 0.980
and slope values below 0.90 reflect less than adequate agreement between the
two flow metering devices. Extremely poor values are apparent when the
engine was tested over various non-steady states but these were a result more
of bunched data than gross differences between the measurements. On the
average, the non-steady state measurements differed by about 12 percent with
a coefficient of variation equal to 56 percent. These results are still less
than desirable.
In an effort to reduce flow pulsations upstream of the vortex meter, a
0.2 cubic meter buffer was added between the car and the proportional sampler.
Again some steady state tests were run and, although some improvements were
noted in the data, agreement between flow meters was less than desirable.
Use of Blower Exhaust--
Agreement between vortex and Roots meter readings was greatly improved
13
-------
when the blower was used for exhaust gas generation. As shown in Table 8,
2
the steady state data produced results having an R value and slope of 0.998
and 0.96, respectively. For each of the five tests conducted over the 20 to
80 SCFM range of flow rates examined, agreement between flow meters was
always within 1.5 percent. The non-steady state results showed equal
2
agreement. Values of R and slope values were 0.999 and 0.97, respectively,
and agreement over each of five tests was within 1 percent.
These results verify that the vortex flow meter in the proportional
sampler experiences some difficulty in metering actual automobile exhaust.
The rather significant errors observed are likely the major contributor to
the differences obtained with the CVS and proportional sampler. Also of
interest was the observation that the vortex meter measurements were neither
consistently higher nor lower than those of the Roots meter. Both positive
and negative differences were noted over the steady state tests using the
engine.
14
-------
EXHAUST
FLOW STRAIGHTENER"
FIOWMETER TUBE
(EXTERNAL)
V12
(/•—-x ^ T
}^Al7\ ©4^
I TANK J*"""^\ FILTER'
\ I MWR / X\ X
r*ir -i
I 8 !
i j
EXHAUST
FLOWMETER
WATER BATH &
CONOENSATE TRAP
INLET DUCT
(FROM TAILPIPE)
EXIT DUCT
•(TO CVS OR VENT)
BAG SAMPLES
(TO ANALYZERS)
_ TRAP &
OBATH DRAIN
Figure 1. Proportional sampler fluid system schematic.
-------
COOLINGWA
CONNECTIONS
Figure 2. Proportional sampler console (front view).
-------
MUSTANG II
GOULD NOVA
CVS
! P.S.
CVS
P.S.
PINTO
[ CVS
i P.S.
HONDA CVCC
P.S.
I I I I I I 1 I 1 I T
I i
JO 0.5 1.0 1.5 2.0 2.5 3 3.5 4.0 i 4.5 5.0 5.5 6.0
GRAMS PER TEST
! Figure 3. Hydrocarbon emission measurement comparisons - CVS vs
: proportional sampler.
MUSTANG II
GOULD NOVA
PINTO
HONDA CVCC
CVS
P.S.
CVS
P.S.
CVS
P.S.
CVS
P.S.
I I I I T I I I I I
I I
0 10 20 30 40 50 60 70 80 90 100 110
GRAMS PER TEST
Figure 4. Carbon monoxide emission measurement comparisons -
CVS vs proportional sampler.
17
-------
MUSTANG II
I
GOULD NOVA
i PINTO
HONDA CVCC
CVS -
P.S. -
CVS -
P.S. -
CVS -
CVS -
P.S. -
I 1 1 I 1 1 1 1 1 1
— M
— ^1
1 l
— n • —
_ — ,
I I i I I I I 1 1 1
l i
[0 1 2 3 4 5 6 7 8 9 10
15
,20
25
GRAMS PER TEST
Figure 5. NOx emission measurement comparisons - CVS vs proportional sampler.
MUSTANG II
GOULD NOVA
PINTO
HONDA CVCC
CVS
P.S.
CVS
P.S.
CVS
P.S.
CVS
P.S.
I I II
I I
I I I I
11 12 13 14 15 16 17 18 19 20 21 22 23
GRAMS PER TEST (10-2)
Figure 6. Carbon dioxide emission measurement cumpaiibuni - CVS vs proportions! ssmpEsr.
18
-------
MUSTANG II
GOULD NOVA
PINTO
HONDA CVCC
CVS
: P.S.
JCVS
!P.S.
| CVS
I P.S.
(CVS
JP.S.
_L
I
I
10.10 0.20 10.30 10.40
I GRAMS PER TEST
0.50
0.5
| Figure 7. Methane emission measurement comparisons - CVS vs proportional.sampler.
Table 1. TEST VEHICLE DESCRIPTIONS
Vehicle
Engine Size
& Configuration
Primary Emission
Control System
1977 Ford Mustang II
1976 Chev. Nova
(Prototype)
1977 Ford Pinto
1976 Honda
CVCC
4.9 Liter
Conventional V-8
2 bbl
5.7 Liter
4 bbl
Conventional V-8
2.3 Liter
4 cylinder
2 bbl
1.5 Liter
4 cylinder
3 bbl
Oxidation catalyst,
Air Pump
Experimental Gould
Oxidation Catalyst,
Air Pump
Oxidation Catalyst
Stratified
Charge
19
-------
Table 2. FORD MUSTANG II
DATA SUMMARY
Specie
Gas
Statistic
Constant Volume Sampler
Proportional Sampler
HC
CO
NO.
CO,
CH
4
x
CV
95% CI
x
CV
95% CI
x
CV
95% CI
x
CV
95% CI
x
CV
95% CI
5.27 girt/test
10%
4.91-5.63 gm/test
20.51 gm/test
8%
19.41-21.61 gm/test
7.80 gm/test
3%
7.65-7.95 gm/test
1986 gm/test
2%
1959-2013 gm/test
.241 gm/test
8%
.227-.253 gm/test
5.33 gm/test
6%
5.14-5.52 gm/test
21.97 gm/test
13%
20.24-23.70 gm/test
6.92 gm/test
5%
6.71-7.13 gm/test
1750 gm/test
5%
1698-1802 gm/test
.270 gm/test
4%
.264-.276 gm/test
Table 3. CHEVROLET NOVA (GOULD CAR)
DATA SUMMARY
Specie
Gas
HC
CO
NO.
CO,
CH,
Statistic
x
CV
95% CI
x
CV
95% CI
x
CV
95% C!
x
CV
95% CI
x
CV
95% CI
Constant Volume Sampler
0.58 gm/test
9%
0.55-0.61 gm/test
1.44 gm/test
38%
0.87-2.01 gm/test
5.89 gm/test
5%
5.71-6.07 gm/test
2272 gm/test
4%
2217-2327 gm/test
0.219 gm/test
9%
0.207-.231 gm/test
Proportional Sampler
0.71 gm/test
7%
0.68-0.74 gm/test
0.66 gm/test
35%
0.53-0.79 gm/test
4.50 gm/test
4%
4.40-4.60 gm/test
1905 gm/test
3%
1873-1937 gm/test
0.170 gm/test
6%
U.lb4-u.l/6 gm/test
20
-------
Table 4. FORD PINTO
DATA SUMMARY
Specie Statistic
Gas
Constant Volume Sampler
Proportional Sampler
HC
CO
NO.
CO,
CH,
x
CV
95% CI
x
CV
95% CI
x
CV
95% CI
x
CV
95% CI
x
CV
95% CI
3.43 gm/test
5%
3.32-3.54 gm/test
61.78 gm/test
6%
59.3-64.3 gm/test
10.76 gm/test
8%
10.18-11.34 gm/test
1491 gm/test
7%
1421-1561 gm/test
0.230 gm/test
9%
0.217-0.243 gm/test
4.41 gm/test
oa
O*
4.16-4.66 gm/test
95.03 gm/test
10%
88.2-101.7 gm/test
8.45 gm/test
7%
8.03-8.87 gm/test
1294 gm/test
4%
1257-1331 gm/test
0.460 gm/test
11%
0.424-0.496 gm/test
Table 5. HONDA CVCC
DATA SUMMARY
Specie
Gas
HC
CO
NO.
CO,
CH,
Statistic
x
CV
95% CI
x
CV
95% CI
x
CV
95% CI
x
CV
95% CI
x
CV
95% CI
Constant Volume Sampler
2.91 gm/test
8%
2.76-3.06 gm/test
22.15 gm/test
5%
19.20-25.10 gm/test
17.45 gm/test
21%
16.90-18.00 gm/test
1204 gm/test
7%
1151-1257 gm/test
0.110 gm/test
7%
0.104-0.116 gm/test
Proportional Sampler
3.75 gm/test
8%
3.58-3.92 gm/test
21.29 gm/test
5%
20.19-22.39 gm/test
20.80 gm/test
9%
20.17-21.43 gm/test
1377 gm/test-
5%
1339 - 1415 gm/test
0.120 gm/test
5%
0.114-0.126 gm/test
21
-------
Table 6. COMPARISON OF EMISSION MEASUREMENT
RESULTS FOR SHORT VERSUS LONG EXHAUST LINE
Specie
Gas
Statistic
Short Line
Long Line
HC
CO
NO.
CO.
x
CV
95% CI
x
CV
95% CI
x
CV
95% CI
x
CV
95% CI
3.79 gm/test
6%
3.60-3.98 gm/test
21.13 gm/test
10%
19.28-22.98 gm/test
20.25 gm/test
4%
19.64-20.86 gm/test
1380 gm/test
2%
1357-1403 gm/test
3.71 gm/test
10%
3.40-4.02 gm/test
21.43 gm/test
9%
19.81-23.05 gm/test
21.29 gm/test
6%
20.26-22.32 gm/test
1348 gm/test
7%
1290-1406 gm/test
Table 7. CALCULATED DILUTION RATIOS
Gas Specie
Dilution Ratio
HC
NOV
x
CO
CO,
2
x
CV
8.95
9.83
9.58
9.12
9.37
4%
CVS
Volume
Raw Exhaust
Volume from
Proportional
Sampler
8.85
22
-------
Table 8. ULTRASONIC VORTEX METER
VERSUS BOOTS METER LINEAR REGRESSION ANALYSIS
VORTEX (CFM) = a + b [ROOTS(CFM)]
a = INTERCEPT
b = SLOPE
GAS GENERATOR
ENGINE
ENGINE
ENGINE
BLOWER
BLOWER
MODE
STEADY STATE
STEADY STATE
W/BUFFER
NON-STEADY
NON-STEADY
STEADY-STATE
a
5.58
12.63
17.17
1.40
1.34
b
0.86
0.75
0.59
0.97
0.96
R2
0.9333
0.986
0.410
0.999
0.998
23
-------
REFERENCES
1. Federal Register, Vol. 37, No. 221, November 1972.
2. Smith, R., Rose, A.M., and Kruse, R. An Auto-Exhaust Proportional
Sampler. Paper presented at the Air Pollution Control Association
Annual Meeting, Detroit, Michigan, June 1963.
3. Olson Laboratories, Inc. Study of an Ultrasonic Vortex Flowmeter for
Measurement of Mobile Source Mass Emissions-. Final Report CRC-APRAC
Project No. CAPE-22-72. Anaheim, California, 1976. -.;•-;••
4. Haskins, H.J. Development of a Proportional Sampler for Automobile
Exhaust Emissions Testing. EPA-600/2-76-169, U.S. Environmental
Protection Agency, Research Triangle Park, North Carolina, 1976. 106 pp.
5. Juneja, W.K., D.D. Horchler, and H.M. Haskew. A Treatise on Exhaust
Emissions Test Variability. SAE Paper 770136, February 1977.
6. Draper, N.R., and H. Smith. Applied Regression Analysis. John Wiley
& Sons, Inc., New York, 1966.
24
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/2-77-236
2.
3. RECIPIENT'S ACCESSION>NO.
4. TITLE AND SUBTITLE
EVALUATION OF A PROPORTIONAL SAMPLER FOR AUTOMOTIVE
EXHAUST EMISSIONS
5. REPORT DATE
December 1977
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Peter A. Gabele
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Environmental Sciences Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
10. PROGRAM ELEMENT NO.
1 AD 712 BA-51 (FY-77)
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
Iri-house
14. SPONSORING AGENCY CODE
EPA/600/09
15. SUPPLEMENTARY NOTES
16. ABSTRACT
A test program was conducted to evaluate a proportional sampler for use in
automotive exhaust gas emissions research. Automobile emissions test results
obtained using the proportional sampler were compared with results obtained
using the conventional constant volume sampler.
Measurements obtained using the proportional sampler for hydrocarbons,
carbon monoxide, nitrogen oxides, and carbon dioxide are within 19, 46, 20, and
14 percent, respectively, of measurements obtained using the constant volume
sampler. Such differences render the proportional sampler unacceptable as a
quantitative research tool.
Inability of the exhaust gas flow meter to accurately measure pulsating
exhaust flow is cited as the principal cause of error in the proportional sampler.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b. IDENTIFIERS/OPEN ENDED TERMS
c. COS AT I Field/Group
* Air pollution
Automobiles
* Exhaust emissions
* Samplers
* Evaluation
13B
13F
21B
14B
18. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19. SECURITY CLASS (ThisReport]
UNCLASSIFIED
21. NO. OF PAGES
29
20. SECURITY CLASS (Thispage)
UNCLASSIFIED
22. PRICE
EPA Form 2220-1 (9-73)
25
------- |