EPA-AA-EOD-80-12
Technical Report
December, 1980
EVALUATION OF THE
BECKMAN 951A ATMOSPHERIC
NOx ANALYZER
by
Sherman D. Funk
NOTICE
Technical reports do not necessarily represent final EPA decisions or
positions. Their publication or distribution does not constitute any
endorsement of equipment or instrumentation that may have been evalu-
ated. 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
improvements in emissions measurement.
Engineering Staff
Engineering Operations Division
Mobile Source Air Pollution Control
Environmental Protection Agency
2565 Plymouth Road
Ann Arbor, Michigan 48105
-------
Abstract
The primary method used for the measurement of NOx emissions by
EPA and the automotive industry is by the chemiluminescence proc-
ess. Several improvements have been made in these instruments
since the original vacuum type NOx analyzers. EPA has taken the
opportunity to evaluate some of these instruments in an effort to
refine its analysis capability.
Although other instruments have been evaluated in various degrees
by EPA and the industry, this report discusses the Beckman 951A
which has a history of wide usage and previous evaluations.
Bench testing and operational comparisons were completed in the
EPA laboratory over a period of six weeks. Although the 951A
still demonstrates need for improvement in certain areas, it met
all EPA requirements and advertised specifications. The test
results correlated to within _+0.5% of full scale to the presently
used Teco 10A. The 951A has been found to be an acceptable
alternative to the vacuum type chemiluminescence analyzer.
-------
I. INTRODUCTION
The NOx analyzer evaluation project on which this report is writ-
ten is the result of earlier investigations into improved state of
the art instrumentation for the analysis of vehicle NOx emis-
sions. These investigations led to the Beckraan atmospheric NO/NOx
analyzer. Two Beckman versions (the 951 and the 951A,) have been
evaluated at this laboratory. Since the 951A has been tested to a
greater degree it will be the one addressed in this report.
II. BACKGROUND
EPA EOD has used the vacuum type NOx analyzers for Light Duty
vehicle emission testing since 1972. The vacuum pumps associated
with the' Teco 10A models have been a constant source of main-
tenance problems and a potential cause of leaks. In addition, the
use of pure oxygen for ozone generation has always been a safety
concern. These problems and concerns as well as the development
of the atmospheric NOx analyzers, prompted the search for a better
instrument to measure NOx.
In the past few years, a number of atmospheric NOx analyzers have
been advertised and considered by the EPA laboratory. However,
only a few have ever had extensive evaluation. One was the
Philco-Ford version. The report, completed in June 1974 indicated
it was an acceptable instrument for light duty vehicle emissions
testing. Inquiries to industry laboratories revealed that the
Beckman 951 had also been extensively tested and used in the
industry. These data, coupled with our limited resources to eval-
uate instruments in general, prompted us to select the 951 for a
more thorough evaluation. A Beckman 951 was purchased and was
tested for a period of time. However, during the evaluation
period the newer version the 951A, was developed. Although the
basic detectors are unchanged, the 951A utilizes improvements in
flow control and regulation. The company claimed the 951 could
not be upgraded to a 951A configuration, but provided a 951A for
additional testing. The following test plan was developed to
assess the performance of the 951A relative to our current
instrumentation.
III. TEST PLAN
A test plan to evaluate this instrument was designed and imple-
mented in three parts.
A. Preliminary Checkout
A basic analyzer checkout procedure was preformed in a speci-
fically designed equipment test rack by the following steps.
1. Stability, Noise, Speed of Response, and Repeatability
checks were made by performing the following:
-2-
-------
a. Select 0-100 PPM range on the instrument and an
appropriate span gas.
b. Alternatively pass zero and span gas through the
instrument while in the NOx position of the converter
for a period of 75 seconds each.
c. Using a stopwatch, measure the time from a gas
input change until reading indicates 90% of full scale.
d. During these tests observe traces for noise,
stability, and repeatability.
2. Using procedure given in manual, set pots for range
attenuation compatability.
3. Perform standard procedure for converter efficiency
checks.
4. Run a curve check on the 0-100 PPM and the 0-250 PPM
ranges and process data.
5. In order to check flow sensitivity, perform the fol-
lowing tests.
a. With the instrument adjusted for normal operation,
pass a span gas through the instrument for 45 seconds.
b. Record reading on the chart recorder.
c. Adjust instrument by-pass flow up to 2400 CCM and
down to 400 CCM in 200 CCM steps.
Comparison Tests
The Beckman unit was plumbed into the auxiliary by-pass line
of the Teco 10A system in Analysis Site A001. A series of ten
FTP and two HWFET was run. These produced 64 sample bag com-
parisons from actual certification tests.
C02 Interference Test
Two sets of four mixtures of NOx and C0£ were blended and
then measured on the Teco 10A and the 951A both in the NOx and
the NO modes (see Appendix C). The Teco 10A was assumed to be
interference free.
IV. SUMMARY OF RESULTS
The results of our tests can be summarized as follows:
A. Speed of response - Time from external valve change (zero to
span) to 90% FS-Fast response switch position - 6.3 sec; Slow
response switch position - 10.4 sec. Installed in Analyzer
bench - Fast response switch position - 11.4 sec. Slow
response switch position - 15.8 sec.
-3-
-------
B. Attenuation Accuracy - Ability to match measurements from one
range to another was less than _+0.5 of full scale error.
C. Converter efficiency at final temperature setting (601° F)
98.15%. Converter has a fairly flat response curve within the
operating range (See Appendix E).
D. Linearity - -0.4% non-linear on 0-100 PPM range; +0.1% non-
linear on 0-250 PPM range.
E. By-Pass Flow Sensitivity - No change in output while by-pass
flow was changed from 400 CCM to 2400 CCM in 200 CCM steps.
Sample pressure remained the same. However, a 0.1 PSIG change
in sample pressure caused a 1.0% of full scale change in out-
put of analyzer in the same direction.
F. Precision - Zero and span settings were repeatable to less
than ±0.5% of full scale.
G. Noise (degree of random and periodic deviations of output)
average peak to peak - noise was no greater than 0.3% of full
scale.
H. Stability - Zero and span drifts were less than _+!% of full
scale for a period of four hours of continuous runs.
I. Comparison tests to Teco 10A - 64 comparisons produced a mean
difference of +0.33 PPM. These were no greater differences
than 0.7 PPM (See Appendix B). Corrected sample
(Sample-Background) values from these same tests were compared
and correlated to a mean difference of +0.04 PPM (See Appendix
B 1). A series of 10 blended SAC bag measurements resulted in
differences of less than +0.'j PPM.
J. Response to C02 - Response to CC>2 was less than 1 PPK at
concentrations near the top of the C02 operating range for
light duty emissions. (See Appendix C).
K. There are a number of potential problem areas with the Beckman
951A:
1. The extensive use ot miniature thin wall teflon tubing in
the unit, some areas in which (capillary tubing,
especially) inadvertent pinching or damage could cause
abnormal operation. Capillaries should be made of some
kind of hard material.
2. It is difficult to make the NO bypass flow balance adjust-
ment using the NC^-free NO gas method since opening and
closing the door cause a 0.5% difference in output. The
door must be opened and the HV interlock set to make an
adjustment with the bypass compensator valve. It is not
known why this door causes a difference in output.
-4-
-------
Another area, which may be part of the problem, is that
the sample capillary is positioned so that closing the
door presses against it. This could possibly affect
capillary flow. This has not been proven, but the design
should be improved.
3. Higher temperatures of the converter cause different
responses at different concentrations causing some
apparent non~;inearity. It is not known why this occurs.
It is recommended to run at the lower end of the tempera-
ture operating range.
4. The NO bypass valve knob kept coming loose although it was
re-installed a number of times. Also, the NO/NOx mode
switch kept coming loose even though it was tightened
repeatedly.
5. The sample flow rate came no where near 200 CCM at 4 PSIG
as the manual recommends. It also turns out that 200 CCM
is the incorrect recommended flow as per Al Roddan of
Beckman. We recommend a sample flow rate of 160 CCM.
This should be set with a calibrated rotameter. Our pres-
sure setting turned out to be 4.35 PSIG on the gauge.
6. This analyzer also exhibits seme of the same problems as
Teco 10A.
a) Noise was measured up to an average of 0.3% of full
scale peak to peak, sometimes a little higher.
b) Generally, chemiluminescent analyzers have relatively
slow response in comparison to other emissions
analyzers primarily due to sample flew. Even though
it is about 3 seconds faster than the Teco 10A the
951A would still slow down testing. Response time for
analyzer alone from gas change at inlet port to 90% of
full scale is 6.3 seconds. When installed in the
analytical bench, this increased to 11.4 seconds as
compared to the Teco 10A response of 14.8 seconds.
c) Similiar to the 10A there are sluggish problems after
a period of non-use (2 to 3 days) for the first few
hours. If used continuously there are less problems.
d) There also appears to be some sluggishness when
returning to a span reading after reading a bag,
although this is not serious.
VI. DISCUSSION
The Beckman 951A continuously analyzes a flowing gas sample and
determines levels of nitric oxide (NO) or Oxides of Nitrogen (NOx)
-5-
-------
(NO + N02 = NOx). It utilizes the chemi luminescent method of:
detection. To determine NO, sample NO is quantitatively converted
into N02* by oxidation with molecular ozone produced within the
analyzer from air or oxygen supplied by an external cylinder.
Approximately 10% of the NOX molecules are elevated to an
electronically excited state accompanied by the emission of
photons. These photons strike the photomultiplier detector, gen-
erating a low level DC current. The current is amplified to drive
an output indicating device, (recorder and/or meter). NOx deter-
mination is identical to that described above except that, before
entry into the reaction chamber, the sample is routed through a
converter where the N02 component is converted to NO.
After receipt of this instrument it was installed in a specifi-
cally designed checkout bench for preliminary tests. It was then
installed in the A001 exhaust analytical bench as shown in
Appendix D. During these tests a number of problems and issues of
concern arose.
During initial checkout we found very poor sensitivity and occa-
sional loss of zero. It was found that the ozone and exhaust
lines were reversed on the reaction chamber fittings. This
apparently had been done by the Beckman service man. We also
found that the high voltage was 100 volts higher than nominal. He
had also raised this to increase sensitivity. When we corrected
the reversal, the sensitivity problem was corrected. The higher
than necessary voltage created excessive dark current. When we
lowered the voltage to nominal, this corrected the zero (dark cur-
rent) problem.
We found that the recommended sample pressure of 4 PSIG only pro-
duced 138 CCM of sample capillary flow rather than the 200 CCM
that the manual indicated. After additional investigations with
GM and the Beckman design engineer, John Harman, it was finally
decided to run at the 160 CCM flow rate that was ttie originally
recommended value since no one seemed to know where the 200 CCM
recommendation originated. GM is running at the 160 CCM rate
also. We set this flow from a calibrated rotameter and marked it
on the pressure gauge. It reads approximately 4.35 PSIG. Bypass
flow was set at 2 liters per minute.
It was found that after running a few converter efficiency checks
that the pressure matching method for NO bypass flow balance was
inadequate since some readings were higher in the NO mode than the
NOx mode. This was corrected by using a known N02 free cylinder
of NO and setting the flow balance until the readings were the
same in both modes.
Initial curve checks on the instrument on ranges 0-100 PPM and
0-250 PPM were .less than 1% non-linear. Checks on 0-50 PPM were
-1.2% non-linear and on 0-5000 PPM were 6% non-linear. However,
gases on these ranges are non-dependable and no emphasis was
placed on these checks. However, after installing the instrument
-6-
-------
in the analytical bench and during a series of converter tempera-
ture adjustments to find the best converter efficiency it was
found that the higher temperature settings near the top end of the
recommended range resulted in some abnormal curve plots, primarily
on the 0-100 PPM range. It was really never determined why these
higher temperatures caused more non-linearity on that one range.
Possible causes could be N0£ in the cylinders with some unknown
reaction occurring in the converter. This same phenomenon was
encountered in an EPA evaluation in March 1973 of a Ford modified
Teco 10A and is discussed in excerpts from that report in Appendix
G. These problems were not investigated since time did not
allow. Since the higher temperatures 650° - 700°F not only caused
this negative response but affected the stability to some degree
it was decided to run the converter on the low end of recommended
operation (605°F) that would give 98%+ converter efficiency and
still result in acceptable curve linearity. It should be noted
that each time the converter temperature was changed it was neces-
sary to re-adjust the NO bypass balance.
During our check out previous to the actual comparison tests. The
converter heater blanket element opened. The Beckman service rep
was contacted. He then brought out a new one and it was installed
and worked satisfactorily.
A final converter efficiency of 98.1% was accepted with a tempera-
ture of 607°F. Although it was found that 100% efficiency could
be achieved at 691°F, 99% at 689°F and 98.4 at 697°F that the low-
er setting was more acceptable for all-around operation. More than
sixteen settings and efficiency checks were made (See Appendix E).
The bag comparison tests were completed by the Light Duty team on
actual certification tests. (Results are on Appendix B and B~l.
Although the one to one comparisons of all bags, sample and back-
ground indicate a slight bias for the 951A to read in the positive
direction (Mean difference of +.33 PPM with a standard deviation
of 0.168), the corrected samples (sample minus background) corre-
lated to a mean difference of +0.04 PPM with a standard deviation
of 0.152. All readings were on 0-100 PPM range. The definite
positive bias of the 951A could not be explained. However the
magnitude is so small that it is not considered significant.
C02 interference checks were made by C&M personnel with blended
bags of CC>2, NOx and air. (See Appendix C tor results.) The
different results of the two tests can be explained by the fact
that the interference level is below standard instrument
variability.
VII. CONCLUSIONS/RECOMMENDATIONS
A. The Beckman 951A is an acceptable instrument to measure light
duty vehicle NOx emissions. It meets all EPA requirements and
Federal Register specifications. However, for testing on
-7-
-------
ranges above 0-1000 PPM, in order to achieve the required
linearity, different combinations of sample and ozone capil-
lary flow plus the use of oxygen for ozone may have to be used.
Since, (a) the 951A uses the less expensive and safer air
rather than oxygen for ozone, (b) it is an atmospheric type
analyzer thereby eliminating the need for a vacuum pump, (c)
it provides a slightly faster response than the 10A, (d) its
NOx emission measurement results correlate satisfactorily with
the presently used Teco 10A, (e) its configuration and
plumbing requirements are compatible with our analytical
benches, and (f) there is a significant amount of historical
data available on these units from other laboratories, it is
recommended as an equivalent, if not an improved alternative
to the vacuum type NOx analyzers and could be used to replace
them in Light Duty Certification and E&D Light Duty analytical
benches.
VIII. BIBLIOGRAPHY
1. Interferences in Chemiluminescent Measurement of NO and N02
Emissions from Combustion Systems, R. Matthews, R. Sawyer, and
Robert Schefer, Dept. of Mechanical Engineering, University of
California, Berkley, Calif. Nov. 1977.
2. Ford Motor Company Intra-Company Memo, Atmospheric Pressure
Chemiluminescent NOx Analyzer, R. Ford and J. Westveer,
July 25, 1972.
3. Memo from Thermo-Electron Corp to D. Paulsell, EPA, NOx
Converter Efficiency Test Expression, John Dunlay, Sept. 17,
1973.
4. Pressure Quenching in the Chemiluminescent Nitric Oxide-Ozone
Reation, William Zolner, Ph.D., Thermo-Electron Corp.
5. EPA Inter-Office Memo, Report on NO/NOx Analyzer Investigation
Progress, S.D. Funk, March 28, 1980.
6. Beckman Sales Literature, 1980.
7. EPA EOD Engr. Staff Project Sub-Task Summary, Atmospheric NOx
Analyzer Evaluation, S.D. Funk, Sept. 2, 1980.
8. Technical Report, Evaluation of Philco-Ford Chemiluminescent
NOx Analyzer, June 1974.
9. Technical Evaluation of Ford Motor Co. Alternate Test
Equipment and Procedures, D. Paulsell, March 1973.
-8-
-------
IX. APPENDICES
A. Beckman 951A and 951 Schematic Flow Diagram. A-l Beckman 951A
and 951 Differences
B. NOx Exhaust Sample Bag Comparison Table B-l Corrected NOx
Sample Comparisons
C. C0£ Interference Table 0^2 Graphs
D. Beckman 951A Plumbing Connections in Analytical Bench
E. Converter Efficiency Graph
F. Excerpt from an EPA Technical Evaluation of a Ford Modified
Teco 10A.
-9-
-------
1-25-81
APPENDIX A
Olc>nil«J Aif Of O«Y9-:A Icf
Chim'ilumin»«enl Itexiioi Wimlvv.
. O»-'->—-Les
/? I S,.n?l,
T- I /I IJOM«J; Opiliir
now LJ ^—p*^^; "i
ftji.xxe T «••»-' I
Vj'.vt.—. \. I T
~x—"
SAMPLE • N0/N0 M^
B«k.P-c»u,. BVPASS ^"sol-ndd
BrjuU.o/ Flowmew ^
FLOW DIAGRAM OF BECKMAN 951A
.SL,S I ij-s-m f^v..-'=i
FLOW DIAGRAM OF BECKMAN 951
-------
1-25-81 APPENDIX A-l
Differences in Beckman 951A and 951
In the Beckman 951A:
1. A type J thermocouple output of converter temperature
is provided.
2. Pressure settings and flows are visible thru the new
front panel window.
3. Trim pots to separately trim each range have been added.
4. The span and zero solenoids are removed.
5. An led indicator showing converter cycling has
been added.
6. Converter temperature adjust is now from front panel.
7. Improved flow balancing thru use of needle valve.
The use of a needle valve has improved flow balancing.
8. Internal ozonator on/off switch has been added.
9. Internal changes of flow configuration and pressure
regulation have been made.
-------
1-25-81
APPENDIX E
NOx EXHAUST SAMPLE BAG COMPARISONS
Diff. = 951A - 10A
BECKMAN VS. TECO
READINGS = PPM
DATE
11-20-80
11-20-80
11-20-80
11-20-80
11-21-80
11-21-80
] 1-21-80
11-21-80
11-20-80
11-20-80
11-21-80
11-21-80
TEST #
80-6816
80-6823
80-6819
80-6820
80-6825
80-6833
80-6830
80-6828
80-6824
80-6864
80-6869
80-6865
BAG 1 and 4
BG
95 1A 10A
0.7 0.6
0.7 0.2
0.6 0.1
0.8 0.3
0.6 0.1
0.7 0.1
0.5 0.2
0.5 0.2
0.6 0.1
0.4 0.2
0.7 0.2
0.7 0.2
DIFF
+0.1
+0.5
+0.5
+0.5
+0.5
+0.6
+0.3
+0.3
SAMP
951 A 10A
27.9 27.9
60.3 60.3
2.5 2.2
21.9 21.6
22.0 21.6
13.5 13.3
16.5 16.3
14.7 14.4
+0.5 80.8 80.1
+0.2
-+0.5
+0.5
26.4 26.2
41.7 41.4
2.3 1.8
DIFF
0.0
0.0
+ 0.3
+0.3
+0.4
+0.2
BAG 2
BG
951A 10A
0.8 0.2
0.6 0.1
0.7 0.3
0.6 0.1
0.5 0.0
+0.2_[0.5 0.2
+0.3
+0.7
+0.2
+0.3
+0.5
0.5 0.2
0.4 0.2
0.4 0.2
0.5 0.3
DIFF^
+0.6
+0.5
+0.4
+ 0.5
+0.5
+0.3
+0.3
+0.2
+0.2
+0.2
and 5 BAG 3 and 6
SAMP
951A 10A
9.7 9.1
13.8 13.1
7.1 6.7
4.3 3.8
5.1 4.6
7.1 6.8
4.1 3.7
; 8.5 8.4
10.3 9.9
0.8 0.6
DIFF
+0.6
+0.7
+0.4
+0.5
+0.5
+0.3
+0.4
+0.1
+0.4
+0.2
BG
951A 10A
0.3 0.1
0.3 0.0
0.7 0.3
0.3 0.0
0.3 0.0
0.3 0.1
0.3 0.1
0.2 0.1
0.4 0.2
0.6 0.3
DIFF
+0.2
+0.3
+0.4
+0.3
+0.3
+0.2
+0.2
+0.1
+0.1
+0.3
SAMP
951A 10A
19.5 19.1
43.7 43.6
11.3 10.9
12.4 12.1
11.6 11.3
17.9 17.8
5.7 5.6
12.. 9 12.8
18.1 17.9
2.1 1.9
DIFF
+0.4
+0.1
+0.4
+0.3
+0.3
+0.1
+0.1
I
+0.1
l
+0.2
i
+0.2
MEAN DIFFERENCE +0.33 PPM or
0.33% of
FULL SCALE ON 0-100 PPM RANGE
STD DEV. 0.168
-------
1-29-81
APPENDIX B-l
DATE
TEST #
CORRECTED NOx SAMPLE COMPARISONS
95]A (SAMP. - BG) - 10A (SAMP -BG)
CORRECTED SAMPLES
11-20-80
11-20-80
11-20-80
11-20-80
11-21-80
11-21-80
11-21-80
11-21-80
11-20-80
11-20-80
11-21-80
11-21-80
80-6816
80-68>23
80-6819
80-6820
BAG 1-4
TRANS COLD
951A 10A
27.2 27.3
59.6 60.1
1.9 2.1
21.1 21.3
1
80-6825 J21.4 21.5
80-6833
80-6830
80-6828
80-6824
80-6864
80-6869
80-6865
12.8 13.2
16.0 16.1
14.2 14.2
80.2 80.1
20.2 26.0
41.0 41.2
1.6 1.6
DIFF
-0.1
-0.5
-0.2
-0.2
+0.1
-0.4
-0.1
0.0
+0.1
0.0
-0.1
0.0
BAG 2-5
STABLE COLD
951A 10A
8.9 8.9
13.2 13.0
6.4 6.4
3.7 3.7
4.6 4.6
6.6 6.6
3.6 3.5
8.1 8.2
9.9 9.7
-0.3 -0.3
DIFF
0.0
+0.2
0.0
0.0
0.0
0.0
+0.1
-0.1
+0.2
0.0
BAG 3-6
TRANS. HOT
95^A lOA
19.2 19.0
43.4 43.6
10.6 10.6
12.1 12.1
11.3 11.3
17.6 17.8
5.4 5.5
12.7 12.7
17.7 17.7
1.5 1.6
DIFF
+0.2
-0.2
0.0
0.0
0.0
-0.2
-0.1
0.0
0.0
-0.1
READINGS = PPM
MEAN DIFF -0.12
STD DEV. .18
MEAN DIFF +.04
STD. DEV .10
MEAN DIFF.
STD. DEV.
-0.04
0.11
TOTAL TESTS
MEAN DIFF =-0.043 PPM
STD DEV = 0.152
-------
APPENDIX C
C02 INTERFERENCE
(NOx ANALYZERS BECKMAN 951A vs. TECO IDA)
RANGE 0-100PPM
BAG 1
2
3
4
5
6
7
8
CALC.
BLEND '
PPM NOx CO 2%
52.3 0
52.3 1.5
52.3 2.5
52.3 4.81
52.3 0
52.3 1.5
52.3 2.51
52.3 4.81
C02%
0
1.47
. 2.45
4.77
0
1.45
2.47
. 4.77
MEASURED
PPM
NOx
95 1A 10A
51.6 50.8
51.9 51.2
49.6 49.5
51.1 50.8
51.1 50.2
51.2 50.5
30.9 50.2
51.3 50.7
NO
951A 10A
41.9 40.9
41.2 40.4
34.0 33.9
38.9 38.6
41.6 41.3
40.7 40.6
40.2 39.9
41.3 41.0
OFFSET
951A-10A
NOx NO
+0.8 +1.0
+0.7 +0.8
+0.1 +0.1
+0.3 +0.3
+0.9 +0.3
+0.7 +0.1
+0.7 +0.3
+0.6 +0.3
BIAS
@ 0% C02
951A-10A
NOx NO
+0.8 +1.0
!
i
1
+0.9 +0.3
5
J
I
t
\
INTERFERENCE
OFFSET - BIAS
NOx NO
0 0
-0.1 -0.2
-0.7 -0.9
-0.5 -0.7
0 0
-0.2 -0.2
-0.2 0.0
-0.3 0.0
NOTE: Bias is defined as the difference between the 951A and the 10A readings at 0% C02- 10A
is considered with no interference.
Offset is defined as the difference between the 951A and the 10A with interference and
bias combined.
Interference is calculated by subtracting the bias from the offset at various levels of
C02.
-------
APPENDIX C-l
1-25-81
-------
APPENDIX C-2
1-25-81
-------
TO ATMOSPHERE
NOTE.:
I. VACUUM
2.
FL-OOR
.3.. ENCLOSURE DENOTED AS
P2. L-OOATSTO l/Sj E>C
.•£
CTYR
-------
APPENDIX E
-------
V
1 25 81
APPENDIX F
the back pressure with precision during the measurement
period.
The converter temperature and flow were studied to
assess their affect on measured values. A valve is posi-
tioned in the bypass line of the converter to simulate the
pressure drop of the converter and thus achieve balanced
flows and resultant equivalent readings as per above dis-
cussion. See Fig. 1C.
One perplexing characteristic concerning the converter
v/as the lower values obtained for the span gas in the con-
vert mode than in the bypass mode. Initially it was thought
to be a flow unbalance, but tests showed increeised loss of
NO reading as the converter temperature increases. There
are a series of reactions which may occur to produce this
effect :
a) 2N02 ~~
°2
b) O2 + 2C — - 2CO
c) 2NO + 2CO — »- N2
2CO2
Reaction (c) requires a reducing atmosphere and is enhanced
in the +400°C temperature range. In the case of bag analy-
sis from CVS, the absence of the reducing atmosphere makes
this reaction unlikely. These reactions are speculative r
but theoretically could cause the effects; however, opera-
tion at the recommended temperature of 475°C produces neg-
ligible losses as well as high conversion efficiency. See
Fig. ID.
E. Several miscellaneous aspects of 'operation are
mentioned here as areas of possible improvement. In this
particular instrument, the sample pump bypass valve, N2,
was teed to the sample pressure regulator bleed off tube,
both connected to the exhaust fan. The influence of the
pump backflow on the regulator changed a span val^te from
90.7 to 88.5, or -2.5%. This effect simply reflects a
change in sample flow. The bypass, bleed off, and reaction
chamber exhaust should all be independently vented to the
exhaust fan. See Fig. 1C.
A gage to monitor the sample capillary back pressure
to ± .5'!H20 and a finer control metering valve should be
installed in the sample line to assure flow equivalence
during all measurements.
------- |