EPA-AA-SDSB-80-12
Technical Report
Evaluation of the Contamination Effects of Diesel Exhaust
on a Critical Flow Sample System
by
Eugene Danielson
May 1980
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 devel-
opments 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, Noise and Radiation
U.S. Environmental Protection Agency
-------
-2-
Abstract
Testing has been conducted to determine 1) the effectiveness
of complete diluted diesel exhaust stream filtering in preventing
contamination of a CFV sample system, and 2) the effect of this
filtering on diesel gaseous emission meaurements. Contamination
was evaluated by observing the repeatability of gaseous emissions
measurements from a gasoline vehicle which was operated between
each of several diesel test sequences. Special attention was given
to detecting hydrocarbon contamination by 1) measuring the concen-
tration of injected propane (between diesel tests), and 2) by
comparing concentrations of hydrocarbons in background air sampled
through the CFV (exposed) sample system to the measured concen-
trations of hydrocarbons in background air sampled through the
background (unexposed) sample system. These latter comparisons were
also made between diesel tests.
The effect of exhaust stream filtering on the levels of CO,
C02 and NOx in diesel exhaust was determined by comparing these
levels with and without filtering.
The results of this testing indicate that complete filtering
of diluted diesel exhaust immediately before the CFV inlet is
effective in preventing CFV contamination. Further, filtering
of the diesel exhaust before the CFV inlet does not appear to
effect the measured levels of CO, C02 and NOx in diesel exhaust.
(HC is measured before the CFV filter.)
-------
-3-
Introduction
EPA's new interim transient heavy-duty diesel site at MVEL
(Ann Arbor) will utilize a CFV sample system. Considerable concern
has been expressed about the possibility of contaminating this
sample system with the heavy hydrocarbon and particulate material
contained in diesel exhaust. The only known laboratory where an
unproteded CFV is used for diesel testing is the DOT laboratory in
Cambridge, Massachusetts. Conversation with personnel at that
facility indicate that the main CFV nozzle must be cleaned once
every four hours (test time) in order to maintain a valid flow
calibration. Such frequent cleaning would dramatically delay test
programs at MVEL, and hence, an alternative was sought.
Other laboratories (notably GM labs) filter the complete
exhaust stream when testing diesels on a CFV system. This is known
as bulk stream filtering, and has the advantage of directly pro-
tecting the CFV system from diesel particulate contamination. This
means that potential periodic shut downs for nozzle cleanings can
be eliminated.
However, data providing information on the effectiveness of
the concept were not available. In addition, it was unknown
whether or not bulk stream filtering of the diesel exhaust would
affect the measured levels of the bagged emissions (CO, C02, and
NOx). Therefore, a test program aimed at evaluating the effec-
tiveness of bulk stream filtering and its impact on the bagged
emissions was conducted using the part scale double-dilution system
(i.e., a light-duty vehicle operated on chassis dynamometer and a
350 cfm CFV unit connected to the primary tunnel of a double-
dilution system).
Objective
This study was conducted primarily to evaluate the effective-
ness of using bulk stream filtering to protect a CFV sample
system when testing diesel vehicles. A secondary objective of this
study was to obtain an indication of the impact of bulk stream
filtering on the bagged diesel exhaust emissions.
Test Procedure
Before an analysis and discussion of the data is presented, a
detailed description of the test sequence and equipment is neces-
sary.
Events 1 through 4, listed in Table 1, indicate the general
test sequence and corresponding CFV configuration used to evaluate
contamination (primary objective). The basic philosophy used in
this program was to expose the CFV to diesel exhaust (Event 1), and
then measure the effect on gasoline vehicle emissions (Event
2). In addition, propane and background hydrocarbon samples were
taken to provide additional check on hydrocarbon contamination
-------
-4-
in the CFV. After establishing a baseline of gasoline vehicle
emissions and propane levels, and taking initial background mea-
surements, it was intended to repeat the sequence five times
without interruption.
However, at the end of the second sequence, it was apparent
that the CO and HC emissions from the gasoline vehicle (a 1976
Chevette) were not repeating. It was suspected that this was being
caused by a malfunction in the "hot stove" valve that controls the
flow of heated manifold air entering the carburetor. However,
possible contamination could not be ruled out. Some confirmatory
testing was necessary. This confirmatory testing consisted of
measuring the emissions from another vehicle (a 1977 Chevette)
using an unexposed (to diesel exhaust) CFV and then subsequently
measuring its emissions using the exposed CFV. With this vehicle,
there were essentially no differences in the levels of measured
emissions between the two CFVs. It was concluded from this, that
the CO and HC nonrepeatability was vehicle related. Therefore, the
1977 Chevette was used to finish the testing.
This final series of tests with the '77 Chevette provided data
which were used to evaluate diesel contamination of the CFV. This
was the primary objective of the study. A secondary objective was
to quantify the effect of bulk stream filtering on the bagged
diesel emissions (CO, C02, and NOx). This evaluation consisted
of comparing filtered diesel exhaust emission levels to unfiltered
diesel exhaust emission levels. The filtered exhaust data came
from measurements taken during Event 1. The unfiltered exhaust
data came from measuring the bagged emissions during a series of
tests in which no bulk stream filter was used. These nonfiltered
exhaust measurements were taken using a conventional PDP-CVS
because it was expected that unfiltered diesel exhaust would very
likely contaminate a CFV system.
Results
Contamination Evaluation
The emission levels of the two Chevettes plus the propane
injection and background levels are presented in Figures 1, 2, and
3. All of the gasoline vehicle emission levels were measured over
the hot LA-4 cycle.
Any contamination of the CFV by diesel exhaust was expected to
primarily affect HC measurements. In this regard, the HC data
presented in Figure 1 is of primary interest. This data indicates
that there is no HC contamination of the CFV system as measured by
any of the comparisons available. The measured HC levels of both
Chevettes remained unchanged (except for the second sequences of
the 1976 Chevette discussed in the test procedure section) through-
out the complete sequence of events. Similarly, the measured HC
level during propane injections also remained constant. Finally,
-------
-5-
the background EC levels sampled through the CFV sample line are
essentially equal to the background HC levels sampled through the
background line. This is significant because the sample lines were
exposed to the diesel exhaust while the background lines were not
exposed to diesel exhaust. From these comparisons, it was con-
cluded that there were no residual diesel hydrocarbons contam-
inating the CFV sample system.
Exposing the CFV sample system to filtered diesel exhaust was
not expected to have any appreciable effect on the CO, NOx, and
CC>2 gasoline vehicle measurements. The data presented in Figures
2 and 3 confirm that these measurements were in fact unaffected.
That is, there were no steady trends (either increasing or de-
creasing) in these measurements. If there had been deterioration a
definite trend should have appeared for the number of tests that
were conducted.
Effect of Bulk Stream Filtering on Diesel Emissions (CO, NOx,
and CC-2)
The question has been raised as to whether or not bulk stream
filtering affects the measurements of CO, NOx, and C02 in the
diesel exhaust. To help address this question, measurements of the
CO, C02 and NOx levels in the filtered exhaust were taken during
the diesel events (Table 1). Immediately after the contamination
evaluation testing was completed, unfiltered exhaust gaseous
emission measurements were made using a conventional PDP-CVS.
Table 2 presents a comparison of the filtered exhaust to unfiltered
exhaust levels of CO, C02 and NOx. From this comparison, it
appears that there is no appreciable change in the measured bagged
emissions levels.
It should be noted that hydrocarbon and particulate measure-
ments were also taken during both the CFV and PDP-CVS diesel
testing. These measurements were taken in order to quantify the
respective errors due to non-proportional sampling (i.e., using the
CFV without a heat exchanger). The analysis and results of this
testing are presented in a yet to be published Technical Report
entitled "Hydro-carbon and Particulate Sample Errors Due to Non-
Proportional Sampling".
Discussion and Conclusion
The results stated previously indicate that there was no
apparent contamination of the CFV sample system from diesel ex-
haust. Contamination of the CFV sample system would have been
expected to be cumulative. However, none of the Chevette tests,
nor any of the propane and background measurements indicated any
-------
-6-
contamination. The bulk stream filtering seems to have suffi-
ciently deterred contamination. In addition, the bulk stream
filtering of the exhaust did not appear to have any appreciable
effect on the measurements of diesel CO, C02 and NOx. Therefore
it was concluded from this testing that bulk stream filtering is
one viable method of protectng a CFV sample unit from diesel
exhaust contamination.
Recommendat ion
It is recommended that bulk stream filtering be allowed for
use in protecting CFV sample systems from diesel contamination.
-------
-7-
Table 1
CFV Diesel Contamination Testing
Sequence of Events
Event
Sample Type
Diesel Vehicle Testing
(4-Bag FTP + 2 Hot LA-4s)
CFV Configuration
CFV connected to 18-inch tunnel
per FR Vol.44, No.23, Feb. 1, 1979
Figure B81-4 except: 1) no heat
exchanger and 2) a bulk stream
filter was installed between the
dilution tunnel and the CFV.
Gasoline Vehicle Testing
(2 LA-4s) I/
Standard light-duty set-up per FR
Vol. 42, No. 124, June 28, 1977,
Figure B78-2.
Propane Injection
(Replicate samples)
Background HC Check;
Replicate samples taken
with main CFV sample
system and with back-
ground sample system.
I/ For the initial baseline testing, 3 LA-4s were driven.
-------
-8-
Table 2
Gaseous Emissions Test Comparison
Peugeot 504D
Hot LA-4 Driving Cycle
CO C02 NOx
Measurement System (gm/mi) (gm/mi) (gm/mi)
Unfiltered Exhaust - PDF 1.36 _+ 3.7% 361 _+ 2.7% 1.26 +_ 2.0%
CVS connected to 18-inch
dilution tunnel per FR Vol.
44, No. 23, Feb. 1, 1979,
Figure B81-3.
Filtered Exhaust - CFV 1.42 +_ 1.1% 367 +_ 1.6% 1.26 +_ 3.1
connected to 18-Inch
dilution tunnel per FR Vol.
44, No. 23, Feb. 1, 1979,
Fig. B81-4 Except: 1) No
heat exchanger, and 2) A
bulk stream filter was
installed between dilution
tunnel and CFV.
Difference + 4.4% + 1.7%
NOTE: 1) The bulk stream filter was changed after each diesel
test sequence;
2) The pressure drop across the filter was monitored
during all CFV diesel testing to be sure that the
upstream CFV pressure did not change too much. (A
large change in this pressure would effect the CFV
flow rate). The maximum pressure drop observed was
only 1 inch I^O. This small pressure change would
have a negligible effect on the CFV flow.
-------
-9-
SYMBOLS ABBREVIATIONS
- Denotes tests with CFV sample unit
exposed to diesel exhaust.
- Denotes tests with CFV sample unit not
exposed to diesel exhaust.
T5T5 - Denotes Diesel Test Sequence
-------
FIGURE. |
CFV-CVS QAS CA^ CONTROL RUN
HVETTE
HOT LA-4 CYCLE:
D
Vlp^ *TI*a^3 C^3*
ETT&
fc=» U I *t=
HOT LA-4- CYCLE:
^ ?? i
& Q 1
,, ,
TEST
too
BO
hO
4*0
SEQUENCE:
,. CFV-CVS
Q
x
,» f
X X
X X
X X
X X
I lit L 1 l_l i t 1 I 1 1 I I I i II I L_i__l 1
0
ft
X X
XXX
U R
X X X X
X X X X
X X
1 i i ,t , I I L 1 i .. .._i . .. L L i 1. .. t Lilt i i i i i 1
TEST SEQUENCE
-------
FIGURE, a .
TALL TESTS HOT LA-4 CYCLE.)
-11-
CHEVETTE.
A A
H K
Q Q
T"EST
CHEVETT
h
Q
K 1-=.
O Q
IS
',5
1,0
O,5
CHEVETTH
|977 CHEVETTE
vfl £!
1 1 i IJ t
o
~t ID
TEST SEQUENCE.
-------
CALL TESTS HOT LA-4 CYCLE)
-12-
V3
0
o
o
CHEMETTE.
A A A A
A
CHEVETTE
Q B
4 + + 4-AAAAAX*XA
' » '
M H P
d- *s-\ t-C
\
B
(r^
S
5 IO
TEST
IS
------- |