FINAL REPORT FOR
FINAL CERTIFICATION OF THE CONTINUOUS
EMISSION MONITORING SYSTEM AT
ALLEGHENY POWER SYSTEMS
PLEASANTS NO. UNIT
CORPORATION
-------
CORPORATION
DON 81-222-018-05-21
Radian Contract No. 222-018-05
FINAL REPORT FOR
FINAL CERTIFICATION OF THE CONTINUOUS
EMISSION MONITORING SYSTEM AT
ALLEGHENY POWER SYSTEMS
PLEASANTS NO. 1 UNIT.
Contract No. 68-02-3542
Work Assignment 5
Prepared for:
Peter R. Westlin
U.S. Environmental Protection Agency
Office of Air Quality Planning & Standards
Emissions Measurement Branch
Prepared by:
L.O. Edwards
January 13, 1982
8501 Mo-Pac Blvd./P.O. Box 9948/Austin, Texas 78766 / (512)454-4797
-------
CONTENTS
Section Page
1 INTRODUCTION 1
2 SUMMARY OF RESULTS 3
3 DESCRIPTION OF FGD PROCESS AND SAMPLING LOCATIONS 6
4 DESCRIPTION OF THE GEMS 9
5 CERTIFICATION METHODS AND RESULTS 11
5.1 Conditioning and Operational Test Periods 11
5.2 Response Time 12
5.3 Calibration Error 13
5.4 Drift Test 13
5.5 Relative Accuracy 13
5.5.1 Inlet 15
5.5.2 Outlet 17
6 QUALITY ASSURANCE/QUALITY CONTROL 25
APPENDIX_Aj_" RESULfS~OF "CERTIFICATION TESTING " 29
APPENDIX B: REPRODUCTION OF CERTIFICATION TESTING RAW
DATA SHEETS (EPA file copies .only) 46
-------
FIGURES
Number Page
3-1 Schematic of Modular Ductwork Downstream of Scrubber 7
4-1 Schematic of the GEMS . . 10
TABLES
Number Page
2-1 Summary of Certification Test Results for GEMS Performance
at Pleasants //I 4
2-2 Summary of Reference Method Results 5
5-1 Daily Zero, Calibration and Span During Certification
Period 14
5-2 Pleasants //I, Inlet, Relative Accuracy Determination 16
5-3 Pleasants #1, Module A, Relative Accuracy Determination ... 19
5-4 Pleasants #1, Module C, Relative Accuracy Determination ... 20
5-5 Pleasants #1, Module D, Relative Accuracy Determination ... 21
5-6 RA Results with Best Corrections 18
5-7 Comparison of SOa Values 24
6-1 Concentrations of Certified Gases Introduced Through
Delivery Lines 26
6-2 Reference Method 3 and 6 Results of Certified Bottles Gases . . 26
6-3 Comparison Between Thorin and Ion Chromatography Results
of Randomly Selected Sulfate Samples 27
ii
-------
SECTION 1
•
INTRODUCTION
The U.S. Environmental Protection Agency (EPA) through its Office of
Air Quality Planning and Standards (OAQPS) is required by the Clean Air Act
Ammendments of 1977 to review individual New Source Performance Standards
(NSPS) every four years following promulgation. On June 11, 1979, 40 CFR 60,
Subpart Da was promulgated and established SOX, NOX and opacity standards
for utility steam generators. In preparation for the four year review, EPA
is gathering continuous sulfur dioxide emission data from representative high
efficiency flue gas desulfurization (.FGD) to determine the performance
characteristics for the best demonstrated emission controls. The results of
this and other studies will be analyzed to provide recommendations regarding
future revisions of the SOa control requirements of Subpart Da.
Radian Corporation, under EPA Contract No. 68-02-3542, conducted an
extended study of an enhanced lime FGD system on a coal-fired steam genera-
tor. The work was carried out at the Pleasants Power Station Unit //I
(Willow Island, West Virginia) from June 1 - October 10, 1981. The existing
continuous emission monitor (CEM) for S02 was upgraded, an oxygen monitor
added, and a data processor installed. An integral part of the program was
to conduct certification testing as proposed in Performance Specifications
2 and 3 of 40 CFR 60, Appendix B of October 10, 1979. This document is a
report of that certification testing at the end (September 29 - October 10)
of the data gathering effort. Only the data pertinent to the certification
are included herein; complete details of the extended monitoring program
will be given in a separate, final report for the project.
A Du Pont 400 SOz monitor, a Thermox Oz monitor and a DART microprocessor
(the DTD system) were activated about June 15, 1981, and run continuously
-------
until a scheduled unit outage on October 10, 1981.
In addition to the problems of operating a modified GEM system, several
other difficulties were encountered at this site:
the physical location of the extraction probe and sampling ports
do not conform to the guidelines in Performance Specifications
2 and 3,
very low levels of SOa were encountered on the FGD outlets, and
only one SOa monitor was available to measure in both the high-
range (inlet, about 2000 ppm) and low-range (outlets, about
100 ppm) .
Because of these difficulties, the certification testing served as
much to establish the accuracy and precision limits of the GEM system as
for meeting all the criteria of Performance Specifications 2 and 3.
-------
SECTION 2
SUMMARY OF RESULTS
Since the CEM had been operating continuously for several months, no
conditioning period was required. The relative accuracy testing (RA) was
more comprehensive than required for certification, and the operational test
period lasted for about 250 hours.
The results of the various tests are summarized in Table 2-1. Because
much of the 862 monitoring was carried out in a low concentration range (less
than 10 percent of the span values) , calibrations in this range were done and
are included in the table even though they are not required. All required
standards were met except the relative accuracy criterion.
Because of the three reasons given in the last paragraph of Section 1
•and because of extensive S02 stratification in the outlet ducts, the RA
results are much greater than the 20% standard. Correction factors which
could adjust for some of these problems (especially stratification) could be
developed as was done in the initial certification report. The resulting
outlet RA values would be 0.027, 0.051 and 0.035 Ibs/MMBtu for A, C, and D
respectively; see Table 5-6. Although these results are near or below 10 per-
cent of the emission standard (0.048 Ibs/MMBtu), such corrections were shown
not to be consistent over the 120 day monitoring period and are of question-
able use or validity.
The CEM SOa and 02 results were both consistently higher than the Ref-
erence Method (RM) results. In spite of some effort to identify the source
of the bias, none could be found. When the results were converted to
Ibs/MMBtu units, the systematic discrepancy was larger. For example, on the
inlet, the RA error was 7.7% (in ppm SOa) but 22.7% (in Ibs/MMBtu). Thus, a
-------
SB
TABLE 2-1. SUMMARY
Procedure
Conditioning Period
Operational Test Period
Response Time (min)
Calibration Error
Calibration Drift
(2-hour)
Calibration Drift
(24-hour)
Relative Accuracy
Relative Accuracy (absolute.
Ibs/MHBtu)
OF CERTIFICATION TEST RESULTS
Specification Monitor
>168 hours SO, and 0,
>168 hours SO, and 0,
415 mln SO,
»J
• <5Z 0, high-level
0, mid-levelt
<5Z SO, high-level
<5Z SO, mid-level
SO, low-levelt
<0.4Z 0, 0, high-level
0, mld-levelt
<2Z of span SO, high-level
SO, mld-levelt
SO, low-levelt
<10.5Z 0, 0, high-level
0, mid-level
<2Z of span SO, high-level
SO, mid-level^
SO, low-levelt
20Z S0,-0, System
<0.047
FOR CEMS PERFORMANCE AT
Inlet or Module
System A C
on- go Ing
242
0.60 0.47 0.47
3.07 2.10 2.57
0.9
4.7
2.3
0.7
10.0
0.26
0.24
0.23 (0.36)*
0.23 (0.60)*
0.16 (4.01)*
0.77
0.68
0.80 (1.25)*
0.32 (0.84)*
0.30 (7.32)*
22.7 (20.3)* 85.2 174.5
1.09 0.169 0.168
PLEASANTS f/1 3
|
i
Passed
D Certification?
see text
see text
0.50 yes
2.80 yes
yes
yes
yes
yes
yes
yes
yes
38.8 No
0.065 No
^r
I
I
I
I
fNot required for certification.
•Calculated from best 9 of 12.
Z of calibration gas.
-------
RADIAN
substantial fraction of the discrepancy in the CEM and KM comparision is due
to the poor agreement in the oxygen values. In spite of the systematically
high CEM results, in 25 of the 27 outlet SOa determinations, the CEM was less
than 55 ppm (0.19 Ibs/MMBtu) higher than the RM results. Copies of the raw
data are in the Appendix and a more complete description of the work will be
given in the project final report.
Finally, in order to get some estimate on the scrubber performance, the
reference method data only displayed in Table 2-2 show that the SOa(out)/
S02(in) = 0.187/4.72 = 4.0%, or a scrubbing efficiency of 96.0%, with a range
of from 92.5 to 98.3%, or approximately 96±3%.
TABLE 2-2. SUMMARY OF REFERENCE METHOD RESULTS (Ibs/MMBtu)
Module
Amount of
Data Used
Range
Average
(Ibs/MMBtu)
Low (Ibs/MMBtu) High
Inlet
Outlet A
Outlet C
Outlet D
Total
Outlet
11/12
12/12
6/6
9/9
27/27
4.72
0.199
0.097
0.167
0.187
4.41 (-0.31)
0.13K-0.68)
0.081(-0.016)
0.108(0.059)
0.081 (.0.106)
5.06 (+0.34)
0.356(4-0.157)
0.117(40.020)
0.209(0.042)
0.356(0.169).
-------
SECTION 3
DESCRIPTION OF FGD PROCESS AND SAMPLING LOCATIONS
A complete description of the Pleasants Power Station, Unit //I flue gas
desulfurization (FGD) system was given in the Pretest Site Survey and the
QA/QC Project Plan reports. It is briefly redescribed here for convenience
and continuity of this report.
Eastern bituminous coal (3.0% sulfur) is pulverized and burned in a
620 raw boiler. The effluent stream passes through an air preheater and a
modular EPS and is recombined in the scrubber inlet plenum. At this point,
the flue gas is separated into four parallel, identical S02 scrubber modules.
Each scrubber module has an inlet venturi quencher and a spray tower for S02
removal. Immediately above the spray nozzles, each module is divided into two
identical, parallel mist eliminators; see Figure 3-1. Above these, the flue
gas streams are recombined, the bypass reheat gas is added, and the product
is ducted into a 1000 foot stack.
The sampling ports on the inlet side of each module are located about
100 feet downstream of the booster fan that draws the gas from the scrubber
inlet plenum. The inlet gas to all four modules was shown to be equivalent
(±5%) using a portable Theta Sensor Series 7000 SOa meter. Thus, one module
(A) was selected for all inlet gas characterizations. The sampling ports on
the outlet ducts are located about ten feet above (.downstream from) the mist
eliminators, and each half of each module is serviced by four ports; see
Figure 3-1. Thus, there are 32 total sampling ports on the outlet ducts.
The extractive Du Pont probes are not located at any of the sampling
ports. The one probe extracting the inlet gas sample is inserted into the
scrubber inlet plenum. Each of the four modules has a separate probe to
-------
Du Pont
Extraction —•
Probe
Sampling
Ports
Mist
Eliminator
Spray
Nozzles
Scrubber Inlet Plenum From ESP
*
t
I
Figure 3-1. Schematic of Modular Ductwork for Pleasants Unit No. 1
-------
to sample each of the separate outlet gases. The ducts are about ten feet
across and the Du Pont probe inserts 50 inches into, the stream about
40 feet downstream of the mist eliminators. The sampling point is also at
the reunion of the two parallel modular ducts, but effectively samples only
one of the two ducts; see Figure 3-1. Ten feet further downstream are the
flow control louvers and, another 20 feet downstream, the bypass reheat
enters.
Therefore, cross-sectional sampling across the duct in the same plane as
the Du Pont probes is not possible in any case. Frequently, large flow dif-
ferences are found from one module to another. Severe SOa stratification at
both the outlet sampling ports cross-sectional plane and at the Du Pont cross-
sectional plane were documented; see the initial certification report or the ,
project final report for S02 stratification maps and duct profiles. Some of the
outlet stratification is due, at least in part, to uneven contactor fluid
flow through the spray nozzle matrix.
-------
SECTION 4
DESCRIPTION OF THE CEMS
The details of the continuous emission monitoring system were also given
in previous documents. Briefly, the system was build around one central
Du Pont 400 S02 analyzer; see Figure 4-1. Six heat traced delivery lines,
operated by a common air aspirator, fed into the central unit. A series of
six valves in parallel permitted flow in only one of the six sample lines at
a time. That flow was divided into two parallel streams, one passing into
the Thermox Oa analyzer (separately flow controlled) and the other into the
Du Pont UV cell where the SOz concentration was measured.
The signals, along with the status information, were fed into the DART
(Data Acquisition and Retrieval) microprocessor and to two strip chart re-
corders. The DART averaged the data and both printed a hard copy and stored
the hourly averages on a floppy disc. The system was zeroed, spanned and
calibrated daily at the central unit. Additionally, unit process data (e.g.,
load, flow) were collected manually from the charts and gauges in the scrubber
control and main control areas.
-------
Du Pont
Extraction
Probe
Sample-Purge
Valve
Heat Traced
Delivery Line ^
Uater Trap
Du Pont
S02
Analyzer
•iiv'cei'i
^•/Thermocouple Lii
H20 Vent
Temperature
Du Pont
Info
Center
Aspirator
Pump,
S02 Concentration
Vent
DART
Microprocessor
)2 Concentration
Thermox
02
Analyzer
Strip Chart
Recorder
Thermox
Control
Unit
Figure 4-1. Schematic of the CEMS
-------
SECTION 5
CERTIFICATION METHODS AND RESULTS
5.1 CONDITIONING AND OPERATIONAL TEST PERIODS
On about June 15, the Du Pont, Thermox and DART (DTD), the GEM system, .
became operational. From that time until the scheduled unit shut-down on
October 10, .which marked the completion of the long-term monitoring effort,
the CEM was kept operational. This did not mean that the system continuously
drew, conditioned and measured a sample of the flue gas. Indeed, interrup-
tions were frequent due to such things as:
• probe and delivery line pluggage,
• aspirator pluggage,
• electronic failures (blown fuses),
* mechanical valve failures,
• signal drift,
* failure of heat tracing, and
• other failures for unknown reasons.
In all cases, repair was effected and the system function restored. Thus,
such repairs, occurring almost daily, came to define normal operating proce-
dure.
Since the CEM had been operating "normally" immediately before the
certification testing, no additional specific conditioning period was needed.
During the operational test period, as many Method 3 and 6 runs as possible
were carried out (up to twelve per sampling site). This RA testing, along
with the other certification procedures, took about 250 hours and serves to
define the operational test period.
11
-------
5.2 RESPONSE TIME
The system response time is defined as the time that is required for the
system to reach an operational value from either zero or the span value. A
stable value is equivalent to a change of less than one percent of span for
30 seconds or less than five percent of the measured average for two minutes.
The greater of the upscale or downscale response time is defined as the
system response time.
To correctly measure the system response time, the sample must be intro-
duced as near as possible to the probe intake; it certainly must include the
sample delivery line(s). At Pleasants #1, the calibrated gas bottles were
taken to the probe-delivery line interface and introduced at that point. A
"T"-valve was installed to facilitate rapid switching from calibration gas
to process gas.
In all cases, calibration gas was introduced at the interface and passed
through the line and analyzers until a stable value was obtained. Then, with
coordination of an operator at the DTD central location via two-way radio,
the valve was turned so that the DTD drew in process gas in a normal manner.
DART printouts, chart recordings and manual recording of voltages were all
initiated when the switch was made; agreement between the three modes was
always satisfactory.
The procedure described above was carried out for each operational
sampling line (inlet and outlet modules A, C, D) using zero and span gases
for both SOa and Oz. The oxygen system was much slower because of the time
required for the aspirator system to flush out the Thermox sensing cell. The
inlet oxygen response time was 184 seconds; all outlet oxygen response times
were below 180 seconds. All SOa response times were well below 60 sec. The
system had a data acquisition delay of 180 sec with a 15 sec collection
period. The performance specifications require a response time of less than
15 minutes.
12
-------
5.3 CALIBRATION ERROR
The calibration error test is designed to measure the accuracy and pre-
cision capabilities of the system. A series of certified gases are introduced
into the DTD system, the values recorded and the error (as defined by the
certification procedure) calculated. The results are required to differ from
the gas certification values by less than 5%.
All required calibration criteria were met; see Table 2-1. However, the
low-level SOa calibration did show a 10% error; a low-level calibration is
not required. Also, the 10% at these levels represents an error of 16 ppm,
and this represents only 0.4% of span (4000 ppm).
An idea of the actual, day-to-day calibration error can be gained from
an examination of the daily zero-calibration-span readings taken during the
operational test period; see Table 5-1.
5.4 DRIFT TEST
The drift test involves measuring a calibrated gas, waiting for a period
of time (e.g., 2 or 24 hours), and then remeasuring the same gas. The sum
of the average difference and the 95 percent confidence interval divided by
the calibration value of the gas (for oxygen, 1.0 is to be used) defines the
calibration drift in percent. See Table 2-1 for results.
All required drift tests were within the performance specification
criteria. However, drift tests were also run for mid-level 02 and mid- and
low-level SOa. All of these tests also met the standards, but because of
the low reference gas value of 162 ppm in the low range, percent errors
appear to be more significant. The two-hour, low-level drift was 4.0%
(6.5 ppm) and the 24-hour, low-level drift was 7.3% (11.9 ppm).
5.5 RELATIVE ACCURACY
The crucial procedure in the certification method is the relative ac-
curacy test. Simply, the SOz and 02 measured by the GEMS are compared to the
13
-------
TABLE 5-1. DAILY ZERO, CALIBRATION AND SPAN DURING
CERTIFICATION PERIOD
SO 2
Date
9/28
9/29
9/30
10/1
10/1*
10/2
10/2*
10/3
10/4
10/5
10/5*
10/6
10/7
10/8
10/8*
10/9
(0.0)
Zero
2
1
13
4
'
2
-
1
-7
-8
-
-1
-1
-5
-
4
(162)
Low
176
170
161
148
160
152
152
180
172
158
-
151
160
158
-
175
(1527)
Mid
1530
1535
1520
1485
1525
1520
1535
1535
1530
1510
-
1525
1525
1540
-
1535
(2560)
High
2510
2505
. 2510
2445
2500
2435
2510
2520
2530
2485
-
2505
2505
2525
-
2540
(0.0)
Zero
0.0
-0.1
-0.3
-0.1
-
-0.2
-
-0.8
1.4
1.4
0.0
-0.3
-0.3
1.5
0.0
-1.8
02
(6.3)
Mid
6.3
6.3
6.3
6.3
-
6.1
-
5.9
7.5
7.5
6.5
6.3
6.3
7.6
6.3
.5..0
(20.9)
High
21.1
20.7
21.0
20.7
-
20.9
-
21.0
21.3
20.9
20.8
21.2
20.9
21.0
21.0
. 20.8 .
*Readjusted values
14
-------
EPA reference methods (6 and 3, respectively) directly, and in terms of
Ibs/MMBtu. Ideally, the reference methods are run at the same sampling point
and simultaneously with the GEMS. The percentage difference added to the 95%
confidence interval should differ by less than 20%.
5.5.1 Inlet
After the flue gas pases through an ESP, it is ducted into a common
plenum. From that very large duct, four ducts draw off flue gas to four
separate scrubber modules. The GEMS probe inserts vertically into the plenum,
but the sampling ports (four sets) are at the inlet to each scrubber, about
one hundred feet from the extractive probe and separated by a booster fan.
Since the flue gas has been shown to be well mixed at the inlet sampling
locations (see Initial Certification Report), only 3-point linear traverses
or one-point extractions were used for Method 3 and 6 sampling; no differences
in the results were discernible using these two methods.
The results (see Table 5-2) indicate that the monitor (i.e., M or GEMS)
was biased high. The source of this systematic discrepancy could not be well
defined, but the following points were considered:
• Calibration—Table 5-1 gives the daily zero and calibration data.
The SOa zero was normal and accurate, but the higher calibration
values did show some discrepancy. For example, during the RA in-
let testing on 10/6 and 10/7, the instrument was calibrated to
the mid-range reference gas at 1527 ppm and read 1525. But the
2560 ppm reference gas read 2505, or 55 ppm low. This calibration
difference of about 30 ppm at 2000 does not fully correct for
the M-RM differences which averaged 96.7 ppm (Table 5-2). If the
30 ppm correction were applied, the ppm relative accuracy would
improve from 7.7% to 6.3%.
• The oxygen readings on the CEM were consistently about 1.5% 0
high. This substantially increases the M-RM difference when
converted to Ibs/MMBtu. Although the oxygen meter did show
periods of 1.5% Oa drift, during the reference method testing
it behaved well. The CEM drew samples from each point for 210 sec
and collected data for the 15 second interval from 180-195 seconds.
The response time testing showed the slowest inlet response to
be 184 seconds, so that any low flow condition [there was frequent
plugging in the probe (once per week) and in the aspirator (once
per week)] would result in higher readings.
15
-------
TABLE 5-2. PLEASANTS #1 INLET
RELATIVE ACCURACY DETERMINATION
Run
No.
1
2
3
4
5
6
7
8
9
10
11
12
Date and
time
1006/1252
1006/1335
1006/1418
1006/1455
1006/1607
1006/1651
1006/1727
1007/1041
1007/1117
1007/1158
1007/1231
1007/1307
Average
S02
RM
M | Diff
a , , .
ppm (dry)
2155
2125
2065
1949
1958
1991
1979
2187
2569
2254
2199
2195
2135
2218
2186
2155
2133
2030
2069
2111
2516
2433
2378
2278
2279
Confidence Interval
Accuracy^
+63
+61
+90
+184
+72
+78
+132
+329
-136
+124
+79
+84
+96.7
66.8
7.7%
02
RM
M
%a dry
5.8
5.75
5.8
5.9
5.9*
6.2
6.15
5.6
5.8
5.8
5.8*
5.45
7.9
7.5
7.6
7.7
8.4
8.1
8.1
7.3
7.4*
7.5
6.9
7.5
S02
RM | M
Diff
lbs/106 Btu
4.84
4.76
4.64
4.41
4.43
4.60
4.55
4.85
5.77
5.06
4.94
4.82
4.81
5.79
5.54
5.50
5.48
5.51
5.48
5.60
6.28
6.11
6.02
5.52
5.77
0.95
0.78
0.86
1.07
1.08
0.88
1.05
'1.43
0.34
0.96
0.58
0.95
0.91
0.18
22.7%
All data are reported on a dry basis and use 9%
r
7Average of reference method tests results used for reference value.
Oxygen value missing; average value for that day used.
16
-------
• Air leaks would result in higher 02 readings. The system was
constantly checked. Reference gas was introduced at the probe-
delivery interface and the GEM gave SOz readings of no more than
35 ppm below the value for that gas introduced at the instrument
and Oa readings of 0.1% for oxygen free gases.
Sources of inaccuracies were constantly checked, but during the reference
testing, none were identified.
The consistently high CEM readings (higher than the RM values) appear
to be systematic rather than random. Indeed, the precision is somewhat
better than the accuracy, especially if the 30 ppm calibration difference is
subtracted from the CEM readings. No source of error large enough to account
for the nearly 100 ppm 862 difference could be isolated. The high CEM oxygen
readings make the comparison in Ibs/MMBtu poorer.
5.5.2 Outlet
At Pleasants //I, there are four parallel, identical (in design) S02
scrubbers, and one GEMS extractive probe has been installed on each outlet
(four probes total). The modules may be operated idenpendently. However,
each scrubber module is further divided into two parallel mist eliminators
just above the top (outlet) of each spray tower. The available sampling
ports are located just past the mist eliminator units (eight sets of sampling
ports), but the GEMS probes are at the reunion of the mist eliminator ducts
for each module; see Figure 3-1. The GEMS extractive probes projected fifty
inches into the twenty-foot reunion area. Thus, only one-half of each mist
eliminator unit was sampled by the CEM and RA testing was done on .only that
half of each module.
At Pleasants //I, the sampling ports are about 40 duct feet removed from
the extractive probe location. Also, the outlet ducts were known to be
stratified in S02 concentration. Therefore, all Method 3 and 6 sampling was
done using an integrated 12-point traverse. Since the flue gas was also
stratified at the one-point extraction probe, comparisons of the reference
method and GEMS results were used to define the relationship of the GEMS
17
-------
readings to the actual flue gas SOa and Oa concentrations (reference methods).
Therefore, meeting the certification criteria, especially for the outlet ducts,
was less important than providing a quantified picture of these relationships
and defining an error band for the GEM data.
Six runs were made on outlet Module C which had just recently come on-
line. Eleven were made on outlet Module D, but during the last two, the CEM
data indicated that a leak was developing on that channel. The source of
the trouble was traced to a hole burned in the delivery line, and the last
two runs were invalidated; thus, nine are reported. Twelve valid runs were
completed on outlet Module A.
The results of the relative accuracy testing are given in Tables 5-3
through 5-5. On all three modules operating during the RM testing period,
the precision is much better than the accuracy. Because of the stratifica-
tion at the extraction site, the RM procedure is probably only calibrating
the offset from the true (RM) value introduced by the single point extrac-
tive monitor. If such a calibration were reapplied to the CEM data set (that
is, the average M—RM difference were zero), the defined relative accuracies
would be as shown in Table 5-6.
TABLE 5-6. RA RESULTS WITH BEST CORRECTIONS
Module
A
C
D
RA (ppm) %
10.5%
49.8%
22.0%
RA (Ibs/MMBtu)
%
13.5%
52.6%
21.0%
RA (Ibs/MMBtu)
Absolute
0.027
0.051
0.035
The defined errors still exceed the limits set for certification passage
(20% or 0.048 Ibs/MMBtu) on at least one module.
18
-------
TABLE 5-3. PLEASANTS #1 MODULE A
RELATIVE ACCURACY DETERMINATION
Run
No.
1
2
3
4
5
6
7
8
9
10
11
12
Date and
time
1001/1353
1001/1523
1002/1022
1002/1148
1002/1522
1002/1638
1003/1008
1003/1120
1003/1411
1008/1100
1008/1159
1008/1537
Average
S02
RM
M '
Diff
ppm (dry)
58
60
58
78
64
57
140
144
152
67
69
66
84.4
107
106
108
132
112
103
192
220
205
109
104
99
Confidence Interval
Accuracy^
+49
+46
+50
+54
+48
+46
+52
+76
+53
+42
+35
+33
48.7
6.9
' 65.9%
02
RM
M
%* dry
6.25
6.05
6.3
6.8
6.4
6.1
6.7
6.7
6.4
6.4
6.4
6.4
6.5
6.4
8.3
7.4
7.4*
7.4*
8.0
8.2
8.1*
7.5*
7.5*
7.5*
S02
RM
M
Diff
lbs/106 Btu
.1343
.1371
.1348
.1877
.1498
.1307
.3345
.3441
.3557
.1568
.1615
.1544
.1985
.2521
.2480
.2908
.3318
.2815
.2589
.5050
.5878
.5434
.2760
.2633
.2507
.1178
.1109
.1560
.1441
.1317
.1282
.1705
.2437
.1877
.1192
.1018
.0963
.1423
.0268
85.2%
All data are reported on a dry basis and use 14% H20.
Average of reference method tests results used for reference value.
*0xygen value missing; average value for that day used.
19
-------
TABLE 5-4. PLEASANTS //I MODULE C
RELATIVE ACCURACY DETERMINATION
Run
No.
1
2
3
4
5
6
7
8
9
10
11
12
Date and
time
1004/1056
1004/1231
1004/1507
1004/1635
1005/1044
1005/1149
Average
S02
RM
M
Diff
ppm (dry)
35
39
40
51
38
49
42.0
82
45
91
93
89
63
Confidence Interval
Accuracy^
+47
+6
+51
+42
+51
+14
35.2
20.9
133.6?
02
RM | M
%* dry
6.25
5.95
6.0
6.1
6.2
6.3
8.3
8.3*
8.3*
8.3
6.9
6.0
S02
RM | M
Diff
lbs/106 Btu
.0811
.0885
.0911
.1169
.0877
.1139
.0965
.2208
.1212
.2451
.2504
.2157
.1435
.1397
.0327
.1540
.1335
.1280
.0296
.1176
.0508
174.5%
All data are reported on a dry basis and use 14% H20.
Average of reference method tests results used for reference value.
*0xygen value missing; average value for that day used.
20
-------
TABLE 5-5. PLEASANTS #1 MODULE D
RELATIVE ACCURACY DETERMINATION
Run
No.
1
2
3
4
5
6
7
8
9
10
11
12
Date and
time
0928/1700
0929/1030
0929/1553
0929/1720
0930/1055
0930/1220
0930/1528
0930/1718
1001/1017
Average
S02
RM
M | Diff
ppm (dry)
48
69
67
59
92
82
72
59
58
67.3
63
107
82
85
74
78
70
47
84
Confidence Interval
Accuracy6
+15
+38
+15
+26
-18
-4
-2
-12
+26
9.3
14.8
35.8%
02
RM
M
%a dry
5.85
6.0
6.05
5.95
5.95
5.5
5.8
5.9
6.2
6.7
6.7
7.5
7.1
6.9*
6.9*
6.9*
6.9*
6.5
S02
RM | M
Diff
lbs/106 Btu
.1082
.1571
.1531
.1339
.2088
.1807
.1618
.1335
.1339
.1667
.1505
.2557
.2076
.2090
.1793 •
.1890
.1697
.1139 •
.1979
.0423
.0986
.0545
.0751
.0295
.0083
.0079
•.0196
.0640
.0297
.0350
38.8%
All data are reported on a dry basis and use 14% H20.
Average of reference method tests results used for reference value.
*0xygen value missing; average value for that day used.
21
-------
Alternatively, one may consider all the outlet RA results as a whole.
If all 27 values are used (Ibs/MMBtu) average KM = 0.1652, average
difference = 0.0973, 95% CI = 0.0255, and the accuracy = 74.3%. If the
best corrected values for each module are used (as defined in the paragraph
above) the accuracy is about 25%.
Rather than dwell on correction factors or other methods to "improve"
the appearance of the results (i.e., nearer the certification criterion),
the data should be considered as representative of a comparison between the
CEM and the reference methods. It should be pointed out that most of the
CEM values are higher than the RM values (25 of 27). However, all but one
of the comparisons were within 55 ppm S02 or 0.19 Ibs/MMBtu. The source of
the systematic error may be the location of the extraction probes, but sim-
ilar ratios were not found during the initial certification studies. A more
complete treatment of the data will be given in the project's final report.
It should be pointed out that the CEM oxygen values were consistently high,
and efforts such as:
• repeated leak checks,
• passing calibrated gases through delivery lines,
• monitoring oxygen meter fluctuations, and
• checking concentrations of calibration gases with reference
Method 3.
did not reveal any consistent source of error. In almost all cases, the
dilution correction (using oxygen) served to increase the M-RM difference
by up to 15%.
Finally, in another effort to ensure the reliability of the results,
some RM runs were accompanied by Method 6 testing on the gas being delivered
to the CEM. Such sampling was done from a "T" at the probe-delivery line
interface while the CEM was continuously sampling. The results are presented
22
-------
in Table 5-7. The final report will make use of these data, and for the
purposes of this report, the only comments are:
• The RM values on the probe gases are only slightly closer to
the CEM values than the RM values obtained from the 12-point
traverses.
• Similar results were obtained in the initial certification
study.
• Since the gas sampled from the probe by the RM and the CEM
was essentially identical, the differences represent a limit
on the accuracy of either the reference method, the CEM, or
a comparison of the two.
23
-------
TABLE 5-7. COMPARISON OF S02 VALUES (ppm, dry)
Module
A
A
A
A
A
C
C
C
D
D
D
D
Date
1002
1002
1008
1008
1008
1004
1004
1005
0930
0930
0930
0930
RM 6,
Ports
[1]
64
57
67
69
66
39
40
49
92
82
72
59
RM 6
Probe
[2]
72
46*
74
75
79
54
60
36
110
113
84
82
CEM
[3]
112
103
109
104
99
45
91
63
74
78
70
47
[l]/[2]
0.89
1.24*
0.91
0.92
0.84
0.72
0.67
1.26
0.84
0.73
0.86
0.72
Ratio
[l]/[3]
0.57
0.55
0.61
0.66
0.67
0.87
0.44
0.78
1.24
1.05
1.03
1.26
[2]/[3]
0.64
0.45*
0.68
0.72
0.80
1.20
0.66
0.57
1.49
1.45
1.20
1.74
*Possible small leak, <10% error.
24
-------
SECTION 6
QUALITY ASSURANCE/QUALITY CONTROL
As called for in the task work plan, considerable effort was put into
the QA/QC task. Before any field work was begun, the QA/QC Project Plan was
written in strict accordance with EPA guidelines. As set out in that docu-
ment, all thermocouples, gas meters, pressure gauges and pitot tubes were
calibrated at Radian prior to going into the field. Many of these were used
in the certification procedures. They will also be calibrated upon request.
Once in the field, a quality control program was initiated. This con-
sisted of daily zero, span and calibration procedures on the Du Pont-Thermox-
DART system with control charts being kept. In addition, control charts were
also maintained for the balance, (NHi^SOij control standard, and a QC S02
and Oz gas (see QA/QC Project Plan for procedures and examples of quality
control charts). If values outside the limits were encountered (which have
occurred only in conjunction with the DTD system), corrective action was
taken. If a major variance was found, a malfunction report was initiated and
carried through the lines of dissemination as set out in the malfunction flow
chart in the QA/QC Project Plan.
Another quality control measure that was frequently used was to intro-
duce certified gases directly into the DTD system and then put the same gas
through the delivery lines. This was the primary method to leak check the
delivery lines. Examples of results are given in Table 6-1. No major dis-
crepancy was found for delivery lines in proper working order.
25
-------
TABLE 6-1. CONCENTRATIONS OF CERTIFIED GASES INTRODUCED
THROUGH DELIVERY LINES
Module
Inlet
A
C
D*
Low SO 2
—
172 (166)
—
142 (166)
High SO 2
2470 (2505)
2465 (2480)
2505 (2520)
2415 (2490)
High 02
20.6 (20.9)
20.8 (20.9)
20.8 (20.9)
20.8 (20.9)
NOTE: Numbers in parentheses are the values measured by injection of the
certified gas directly into the Du Pont and Thermox calibration.
*A section of heat trace line was not heating during this period, and some
water was in the delivery line.
Certified gases were used to calibrate the DTD system; the certifica-
tion letters are reproduced in Figure 6-1, 6-2 and 6-3. In addition, to
check the gases' certification and the EPA Method 3 and 6 procedures, several
Method 3 and 6 runs were done on the bottled gases; Table 6-2 presents typical
results.
TABLE 6-2. REFERENCE METHOD 3 AND 6 RESULTS
OF CERTIFIED BOTTLE GASES
Date
10/8
10/8
10/9
10/9
10/9
10/9
10/9
10/9
Species
S02
S02
S02
S02
S02
S02
02
02
Certification Value
1527.5 ppm
162.0 ppm
2560 ppm
162.0 ppm
1527.5 ppm
2560 ppm
6.17%
5.41%
EPA Method Value
1552 ppm
165.6 ppm
2599 ppm
168.7 ppm
1553 ppm
2576 ppm
6.22%
5.38%
26
-------
All qualitative laboratory glassware was class A quantative (e.g., burets) ,
and care was taken to be sure the volumetric fluids were measured at the
temperature designated on the glassware. All chemicals were of the highest
quality available.
Titrations were run on the available deionized water and the small blank
sulfate contribution accounted for. The isopropyl alcohol was checked for
residual peroxide by the spectrophotometric iodine method recommended in the
Method 6 procedure. On several occasions, the first three impingers of the
Method 6 train were titrated separately, and in all cases, over 95% of the
sulfate was found to be in the second (first peroxide) impinger. Also, three
times a follow-up peroxide impinger was inserted before the silica gel trap,
but no sulfate (<0.5%) was found.
As a Radian internal check on the titration procedure, arbitrary samples
were sent to Radian's Austin lab for sulfate determination by ion chromatog-
raphy. The results of one such audit are given in Table 6-3. The Thorin
results are biased slightly high except for the QC [(.NHiJaSOiJ sample.
TABLE 6-3. COMPARISON BETWEEN THORIN AND ION CHROMATOGRAPHY RESULTS
OF RANDOMLY SELECTED SULFATE SAMPLES
ID Number
PDS-0928-1
PDS-1001-1
PDS-1003-1 x
PIS-1006-6
PIS-1001-2
PXS-1009-4
QC Std [(NHOaSOiJ
Ion Chromatography (mm/1)
0.54
0.85
2.17
4.00
5.52
4.05
0.47
Thorin (mm/1)
0.57
0.87
2.19
4.06
5.59
4.13
0.48
27
-------
During the initial certification period, an independent Radian auditor
visited the Pleasants Power Station and conducted both system and performance
audits. A brief report of that audit was given in the initial certification
report, and the complete QA/QC results will be given in the project final
report. Thorin results on EPA audit samples showed no discrepancies larger
than 1.06%. No other irregularities were found. While a separate audit was
not conducted during the final certification, all the established QC proce-
dures were maintained during this period. For example, the (NHOaSOi, QC
sulfate standard was titrated daily; again, no irregularities were found.
28
-------
APPENDIX A
Certification Testing Composite Data Sheets not included elsewhere in text of
report (i.e., all tests except relative accuracy):
• Response Time Results - Pleasants #1
—Inlet, S02
—Inlet, 02
—A Module Outlet, S02
—A Module Outlet, 02
—C Module Outlet, S02
—C Module Outlet, 02
—D Module Outlet, S02
—D Module Outlet, 02
• Calibration Error Determination
—Du Pont - S02 System
—Thermox - 02 System
• 2-Hour Zero and Calibration Drift
—S02: High-Range (2560 ppm)
—S02: Mid-Range (1527.5 ppm)
—S02: Low-Range (162 ppm)
—02: High-Range (20.9%)
—02: Mid-Range (6.17%)
• 24-Hour Zero and Calibration Drift
—S02: High-Range (2560 ppm)
—S02: Mid-Range (1527.5 ppm)
—S02: Low-Range (162 ppm)
—02: High-Range (.20.9%)
—02: Mid-Range (6.17%)
29
-------
RESPONSE TIME RESULTS
Plant: Pleasants No. 1 Duct or Stack: Module - Inlet
Species:_
SO 2
_High Range: 2560 Final Range; 2050 Date; 9/29/81
ppm ppm
Test Run
1
2
3
Average
Upscale
Min.
.5
.8
.5
A = .60
Downscale
Min.
.5
.6
.6
B = .57
Slowest System Response Time
Comments: Du Pont flow at 4.3 CFH
.60
min.
RESPONSE TIME RESULTS
Plant: Pleasants No% 1 Duct or Stack: Module - Inlet
Species: Oa
_High Range; 20.9% Final Range; 5.8% Date: 9/29/81
Test Run
1
2
3
Average
Upscale
Min.
2.9
2.7
2.3
A = 2.63
Downscale
Min.
3.1
3.1
3.0
B = 3.07
Slowest System Response Time
Comments: Thermox flow at 2.7 CFH
3.07
min.
30
-------
RESPONSE TIME RESULTS
Plant: Pleasants No. 1 Duct or Stack: Module A - Outlet
Species:_
SOz
_High Range; 2528 Final Range; 7Q ppmDate; 10/1/81
ppm
Test Run
1
2
3
Average
Upscale
Min.
.3
.4
.3
A = .33
Downscale
Min.
.5
.4
.5
B = .47
Slowest System Response Time =
Comments: Du Pont flow at 6.7 CFH
.47
min.
RESPONSE TIME RESULTS
Plant: Pleasants No. 1 Duct or Stack; Module A - Outlet
Species; 02
_High Range; 20.9% Final Range: 5.9% Date: 10/1/81
Test Run
1
2
3
Average
Upscale
Min.
2.3"
1.7
1.7
A = 1.9
Downscale
Min.
2.6
2.0
1.7
B = 2.1
Slowest System Response Time
Comments: Thermox flow at 3.9 CFH
2.1
min.
31
-------
RESPONSE TIME RESULTS
Plant: Pleasants No. 1 Duct or Stack: Module C - Outlet
Species:_
SOa
_High Range; 2528 Final Range; 70 ppmDate; 10/9/81
ppm
Test Run
1
2
3
Average
Upscale
Min.
.6
.4
.4
A = .47
Downscale
Min.
.6
.4
.3
B = .43
Slowest System Response Time
Comments: Du Pont flow at 7.Q CEH
.47
min.
RESPONSE TIME RESULTS.
Plant: Pleasants No. 1 Duct or Stack: Module C - Outlet
Species:_
_High Range: 20.9% Final Range: 5.8% Date: 10/9/81
Test Run
1
2
3
Average
Upscale
Min.
2.3
2.3
2.3
A = 2.3
Downscale
Min.
2.8
2.3
2.6
B = 2.6
Slowest System Response Time
Comments: Thermox flow at 3.0 CFH
2.6
mm.
32
-------
Plant:
Species:
RESPONSE TIME RESULTS
Pleasants No. 1 Duct or Stack: Module D - Outlet
S02
High Range:2528 ppmFinal Range; 70 ppm Date: 9/30/81
Test Run
1
2
3
Average
Upscale
Min.
.33
.33
.40
A " .35
Downscale
Min.
.50
.50
.50
B « .50
Slowest System Response Time
Comments: Du Pont flow at 6.0 CFH
.50
min.
RESPONSE TIME RESULTS
Plant: Pleasants No. 1 Duct or Stack: Module D - Outlet
Species:
_High Range; 20% Final Range: 5.7% Date;9/30/81
Test Run
1
2
3
. Average
Upscale
Min.
2.3
2.3
2.2
A - 2.3
Downscale
Min.
2.8
2.7
2.9
B = 2.8
Slowest System Response Time
Comments: Thermox flow at 2.7 CFH
2.8
min.
33
-------
CALIBRATION ERROR DETERMINATION
Du Pont - SOa System
9/28/81
Run
No
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Calibration Gas
Concentration
ppm
A
0
2560
1527
2560
0
1527
0
2560
0
1527
2560
1527
0
2560
1527
Measurement System
Reading
ppm
B
1
2500
1515
2515
1
1525
-1
2515
-7
1520
2495
1510
-7
2500
1525
Arithmetic Mean
Confidence Interval
Calibration Error
Arithmetic
Differences
ppm
A-B
Mid
12
2
7
17
2
+8
2.9
0.7%
High
60
45
45
65
60
+55
4.2
2.3%
NOTE: Low-level = 162 ppm and readings were 173, 175, 172, 156, 168
(-6.8, 9.4, 10.0%); mid-level =• 1527.5 ppm; high-level = 2560 ppm.
34
-------
CALIBRATION ERROR DETERMINATION
Thennox - Oa System
9/28/81
Run
No
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Calibration Gas
Concentration
ppm
A
20.9
6.2
0.0
6.2
20.9
0.0
20.9
0.0
6.2
20.9
0.0
20.9
6.2
0.0
6.2
Measurement System
Reading
ppm
B
20.9
6.5
0.2
6.2
20.9
-0.2
20.8
-0.1
6.6
20.9
-0.1
20.5
6.5
-0.1
6.3
Arithmetic Mean
Confidence Interval
Calibration Error
Arithmetic
Differences
ppm
A-B
Mid
-0.3
0.0
-0.1 ,
-0.4
-0.3
-.22
.07
4.7%
High
0.0
0.0
0.0
-0.4
-0.10
0.08
0.9%
NOTE: Mid-level = 6.17%; high-level = 20.
35
-------
2-HOUR ZERO AND CALIBRATION DRIFT
: High-Range (2560 ppm)
Data
Set
No.
1
2
3
4
5
6
7
8
9
10
11
12
13
IV
15
Date
10/5
10/5
10/5
10/6
10/6
10/6
10/6
10/7
10/7
10/7
10/7
10/8
10/8
10/8
10/8
Time
Begin
1017
1240
1455
0845
1045
1300
1555
0846
1052
1308
1530
820
1048
1314
1525
End
1240
1455
1703
1045
1300.
1555
1820
1052
1308
1530
1736
1048
1314
1525
1730
Zero Rd.
Init.
A
6
-1
2
-1
1
1
-7
0
1
6
-7
-5
6
11
-5
Fin.
B
-1
2
6
1
1
-7
10
1
6
-7
-5
6
11
6
-1
Arithmetic Mean
Confidence Interval
Zero drift*
Zero
Drift
C»B-A
-7
3
4
2
0
-8
17
1
5
-13
12
11
5
-5
4
1.4
4.14
0.14%
Hi-Range Rdg.
Init.
D
2525
2520
2510
2505
2505
2485
2485
2505
2490
2520
2500
2525
2545
2530
2520
Fin.
E
2520
2510
2505
2505
2485
2485
2500
2490
2520
2500
2505
2545
2530
2520
2545
Span
Drift
F=E-D
-5
-10
-5
0
-20
0
15
-15
+25
-20
5
20
-15
-10
-1-25
Calibration drift*
Calib.
Drift
G=F-C
2
-13
-9
-2
-20
+8
-2
-16
20
-7
+3
+9
-20
-5
+21
-2.07
7.13
0.23%
Calibration drift calculated using the 4000 ppm span value as RV; using
2560, CD = 0.36%.
NOTE: All data collected during 2-week certification period are presented.
36
-------
2-HOUR ZERO AND CALIBRATION DRIFT
S02: Mid-Range (1527.5 ppm)
Data
Set
No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Date
10/5
10/5
10/5
10/6
10/6
10/6
10/6
10/7
10/7
10/7
10/7
10/8
10/8
10/8
10/8
Time
Begin
1017
1240
1455
845
1045
1300
1555
846
1052
1308
1530
820
1048
1314
1525
End
1240
1455
1703
1045
1300
1555
1820
1052
1308
1530
1736
1048
1314
1525
1730
Zero Rd.
Init.
A
6
-1
2
-1
1
1
-7
0
1
6
-7
-5
6
11
-5
Fin.
B
-1
2
6
1
1
-7
10
1
6
-7
-5
6
11
6
-1
Arithmetic Mean
Confidence Interval
Zero drift*
Zero
Drift
C=B-A
-7
3
4
2
0
-8
17
1
5
-13
12
11
+5
-5
4
1.4
4.14
0.14%
Hi-Range Rdg.
Init.
D
1535
1525
1530
1525
1515
1515
1520
1525
1535
1535
1495
1540
1550
1540
1535
Fin.
E
1525
1530
1525
1515
1515
1520
1525
1535
1535
1495
1515
1550
1540
1535
1540
Span
Drift
F=E-D
-10
+5
-5
-10
0
+5
+5
+10
0
-40
+20
+10
-10
-5
+5
Calibration drift*
Calib.
Drift
G=F-C
-3
+2
-9
-12
0
+13
-12
+9
-5
-27
+8
-1
-15
0
+1
-3.4
5.74
0.23%
Calibration drift calculated using the 4000 ppm value as RV: using 1527.5
ppm, CD = 0.60%.
NOTE: All data collected during 2-week certification period are presented.
37
-------
2-HOUR ZERO AND CALIBRATION DRIFT
S02: Low-Range (162 ppm)
Data
Set
No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Date
10/5
10/5
10/5
10/6
10/6
10/6
10/6
10/7
10/7
10/7
10/7
10/8
10/8
10/8
10/8
Time
Begin
1017
1240
1455
845
1045
1300
1555
846
1052
1308
1530
820
1048
1314
1525
End
1240
1455
1703
1045
1300
1555
1820
1052
1308
1530
1736
1048
1314
1525
1730
Zero Rd.
Init.
A
6
-1
2
-1
1
1
-7
0
1
6
-7
-5
6
11
-5
Fin.
B
-1
2
6
1
1
-7
10
1
6
-7
-5
6
11
6
-1
Arithmetic Mean
Confidence Interval
Zero drift*
Zero
Drift
C=B-A
-7
3
4
2
0
-8
17
1
5
-13
12
115
+5
-5
4
1.4
4.14
0.14%
Hi-Ranee Rdg.
Init.
D
171
151
151
151
150
157
160
155
107
158
158
166
158
161
Fin.
E
151
151
167
151
150
157
178.
155
167
158
153 '
166
158
161
165
Span
Drift
F=E-D
-20
0
+16
-1
+7
+21
-5
+12
-9
-5
+8
-8
+3
+4
Calibration drift*
Calib.
Drift
OF-C
-13
-3
+12
-1
+15
+4
-6
+7
+4
-17
-3
-13
+8
0
-1.0
5.5
0.16%
Calibration drift calculated using the 4000 ppm value as RV; using 162
as RV, CD = 4.01%.
NOTE: All data collected during 2-week certification period are presented.
38
-------
2-HOUR ZERO AND CALIBRATION DRIFT
Oxygen: High-Range (20.9%)
Data
Set
No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Date
10/5
10/5
10/5
10/6
10/6
10/6
10/6
10/7
10/7
10/7
10/7
10/8
10/8
10/8
10/8
Time
Begin
1017
1240
1455
845
1045
1300
1555
846
1052
1308
1530
820
1048
1314
1525
End
1240
1455
1703
1045
1300
1555
1820
1052
1308
1530
1736
1048
1314
1525
1730
Zero Rd.
Init.
A
-0.4
0
0
-0.4
-0.3
-0.1
-0.2
-0.3
-0.3
-0.3
-0.4
0
-0.8
-1.2
-1.1
Fin.
B
0
0
0
-0.3
-0.1
-0.2
-0.2
-0.3
-0.3
-0.4
-0.8
-0.8
-1.2
-1.1
-2.0
Arithmetic Mean
Confidence Interval
Zero drift*
Zero
Drift
C=B-A
0.4
0.0
0.0
0.1
0.2
-0.1
0.0
0.0
0.0
-0.1
-0.4
-0.8
-0.4
+0.1
-0.9
-0.13
0.20
0.33
Hi-Range Rde.
Init.
D
20.3
20.9
21.2
21.2
21.0
21.1
20.9
20.9
20.6
20.7
20.2
21.0
20.5
20.6
20.6
Fin.
E
20.9
21.9
20.7
21.0
21.1
20.9
20.9
20.6
20.7
20.2
20.5
20.5
20.6
20.6
20.3
Span
Drift
F=E-D
+0.6
+0.3
-0.5
-0.2
+0.1
-0.2
0.0
-0.3
+0.1
-0.5
+0.3
-0.5
+0.1
0.0
-0.3
Calibration drift*
Calib.
Drift
G=F-C
0.2
0.3
-0.5
-0.3
-0.1
-0.1
0.0
-0.3
+0..1
-0.4
+0.7
+0.3
+0.5
-0.1
+0.6
0.06
0.20
0.26
*1.0 used for reference value.
39
-------
2-HOUR ZERO AND CALIBRATION DRIFT
Oxygen: Mid-Range (6.17%)
Data
Set
No.
1
2
3
4
5
6
7
8
9
10 N
11
12
13
14
15
Date
10/5
10/5
10/5
10/6
10/6
10/6
10/6
10/7
10/7
10/7
10/7
10/8
10/8
10/8
10/8
Time
Begin
1017
1240
1455
845
1045
1300
1555
846
1052
1308
1530
820
1048
1314
1525
End
1240
1455
1703
1045
1300
1555
1820
1052
1308
.1530
1736
1048
1314
1525
1730
Zero Rd.
Init.
A
-0.4
0
0
-0.4
-0.3
-0.1
-0.2
-0.3
-0.3
-0.3
-0.4
0
-0.8
-1.2
-1.1
Fin.
B
0
0
0
-0.3
-0.1
-0.2
-0.2
-0.3
-0.3
-0.4
-0.8
-0.8
-1.2
-1.1
-2.0
Arithmetic Mean
Confidence Interval
Zero drift*
Zero
Drift
C=B-A
0.4
0.0
0.0
0.1
0.2
-0.1
0.0
0.0
0.0
-0.1
-0.4
-0.8
-0.4
+0.1
+0.9
-0.13
0.20
0.33
Hi-Range Rdg.
Init.
D
6.0
6.4
6.5
6.3
6.5
6.4
6.3
6.2
6.1
5.6
7.6
5.9
5.7
5.6
Fin.
E
6.4
6.5
6.5
6.3
6.5
6.4
6.3
6.2
6.1
5.6
5.7
5.9
5.7
5.6
4.8
Span
Drift
F=E-D
0.4
0.1
0.0
0.2
-0.1
-0.1
-0.1
-0.1
-0.5
0.1
-1.7
-0.2
-0.1
-0.6
Calibration drift*
Calib .
Drift
G=F-C
0.0
0.1
0.0
0.0
0.0
-0.1
-0.1
-0.1
-0.4
0.5
-0.9
+0.2
-0.2
+0.3
-0.05
0.19
0.24
*1.0 used for reference value.
40
-------
24-HOUR ZERO AND CALIBRATION DRIFT
S02: High-Range (2560)
Data
Set
No.
1
2
3
4
5
6
7
8
9
10
11
12
Date
9/28
9/29
9/30
10/1
10/2
10/3
10/4
10/5
10/6
10/7
10/8
10/9
Time
Begin
930
840
850
1007
1015
909
1005
830
845
846
908
851
End
840
850
830
950
909
1005
845
846
820
851
1223
i
Zero Rd.
Init.
A
2
1
13
4
0
1
-7
-8
-1
-1
-5
3
Fin.
B
-1
+12
-9
-2
1
-8
Inv
-1
-1
-5
3
8
Arithmetic Mean
Confidence Interval
Zero drift*
Zero
Drift
C-B-A
-1
+12
-9
-2
1
-8
ilid
+7
0
-4
+8
+5
+0.82
4.46
0.13%
Hi-Ranee Rdg.
Init.
D
2510
2535
2510
2500
2510
2520
2530'
2485
2505
2505
2525
2540
Fin.
E
2535
2510
2445
2435
2520
2530
Inv
2505
2505
2525
2540
2500
Span
Drift
F-E-D
+25
-25
-65
-65
+10
+10
ilid
+20
0
+20
+15
-40
Calibration drift*
Calib.
Drift
G-F-C
+26
-37
-56
-63
+9
+18
+13
0
+24
+7
-45
-9.45
22.67
0 . 80%
*Calibration drift calculated using the 4000 ppm span value as RV; using 2560
ppm as RV, CD = 1.25%.
NOTE: All data collected during 2-week certification period are presented.
41
-------
24-HOUR ZERO AND CALIBRATION DRIFT
S02: Mid-Range (1527.5 ppm)
Data
Set
No.
1
2
3
4
5
6
7
8
9
10
11
12
Date
9/28
9/29
9/30
10/1
10/2
10/3
10/4
10/5
10/6
10/7
10/8
10/9
Time
Begin
930
840
850
1007
1015
909
1005
830
845
846
820
851
End
840
850
830
950
909
1005
830
845
846
820
851
1223
Zero Rd.
Init.
A
2
1
13
4
0
1
-7
-8
-1
-1
-5
3
Fin.
B
1
13
4
2
1
-7
Invj
-1
-1
-5
3
8
Arithmetic Mean
Confidence Interval
Zero drift*
Zero
Drift
C-B-A
-1
+12
-9
-2
+1
-8
lid
+7
0
-4
+8
+5
+0.82
4.46
0.13°
Hi-Ranee Rdg.
Init.
D
1530
1535
1520
1525
1535
. 1535
1530'
1510
1525
1525
1540
1535
Fin.
E
1535
1520
1485
1520
1535
1530
1510
1525
1525
1540
1535
1535
Span
Drift
F-E-D
+5
-15
-35
-5
0
-5
-20
+15
0
+15
-5
0
Calibration drift*
Calib.
Drift
G-F-C
+6
-27
-26
-3
-1
+3
Invalid
+8
0
+19
-13
-5
-3.5
9.4
0.32%
Calibration drift calculated using the 4000 ppm span value as RV; using
1527.5 ppm as RV, CD = 0.84%.
NOTE: All data collected during 2-week certification period are presented.
42
-------
24-HOUR ZERO AND CALIBRATION DRIFT
S02: Low-Range (162 ppm)
Data
Set
No.
1
2
3
4
5
6
7
8
9
10
11
12
Date
9/28
9/29
9/30
10/1
10/2
10/3
10/4
10/5
10/6
10/7
10/8
10/9
Time
Begin
930
840
850
1007
10.15
909
1005
830
845
846
820
851
End
840
850
830
950
909
1005
830
845
846
820
851
1223
Zero Rd.
Init.
A
2
1
13
4
0
1
-7
-8
-1
-1
-5
3
Fin.
B
1
13
4
2
1
-7
Invs
-1
-1
-5
3
8
Arithmetic Mean
Confidence Interval
Zero drift*
Zero
Drift
C-B-A
-1
+12
-9
-2
+1
-8
lid
+7
0
-4
• +8
+5
0.82
4.46
0.13%
Hi-Ranee Rdg.
Init.
D
176
170
161
160
152
180
172 '
158
151
160
158
175
Fin.
E
170
161
148
152
180
172
158
151
160
158
175
158
Span
Drift
F-E-D
-6
-9
-13
-8
+28
-8
-14
-7
+9
-2
+17
-17
Calibration drift*
Calib.
Drift
G-F-C
-5
-21
-4
-6
+27
0
[rival id
-14
+9
+2
+9
-22
-2.27
9.59
0.30%
*Calibration drift calculated using the 4000 ppm span value as RV: using 162
as RV, CD = 7.3%.
NOTE: All data collected during 2-week certification period are presented.
43
-------
24-HOUR ZERO AND CALIBRATION DRIFT
02: High-Range (20.9)
Data
Sec
No.
1
2
3
4
5
6
7
8
9
10
11
12
Date
9/28
9/29
9/30
10/01
10/2
10/3
10/4
10/5
10/6
10/7
10/8
10/9
Time
Begin
930
840
850
1007
1015
909
1005
830
845
846
908
851
End
840
850
830
950
909
1005
830
845
846
820
851
1223
Zero Rd.
Init.
A
0.1
-0.1
-0.3
0.1
-0.2
-0.8
-1.4
-0.4
-0.5
-0.3
0.0
-1.8
Fin.
B
-0.1
-0.3
-0.1
-0.2
-0.8
-1.4
-1.4
-0.5
-0.3
1.5
-1.8
0.1
Arithmetic Mean
Confidence Interval
Zero drift*
Zero
Drift
C-B-A
-0.2
-0.2
+0.2
-0.3
-0.6
-0.6
0.0
-0.1
+0.2
+1.8
-1.8
+1.9
+0.02
0.53
0.55
Hi-Ranee Rdg.
Init.
D
21.1
20.7
21.0
20.7
20.9
21.0
21.3'
20.3
21.2
20.9
21.0
20.8
Fin.
E
20.7
21.0
20.7
20.9
21.0
21.3
20.9
21.2
20.9
21.0
20.8
21.4
Span
Drift
F-E-D
-0.4
+0.3
-0.3
+0.2
+0.1
+0.3
-0.4
+0.9
-0.3
+0.1
-0.2
+0.6
Calibration drift*
Calib.
Drift
G-F-C
-0.2
+0.5
-0.5
+0.5
+0.7
+0.9
-0.4
1.0
-0.5
-1.7
+1.6
+0.5
0.20
0.57
0.77
*1.0 used for reference value.
NOTE: All data collected during 2-week certification period are presented.
44
-------
RADIAN
24-HOUR ZERO AND CALIBRATION DRIFT
02: Mid-Range (6.17%)
Data
Set
No.
1
2
3
4
5
6
7
8
9
10
11
12
Date
9/28
9/29
9/30
10/1
10/2
10/3
10/4
10/5
10/6
10/7
10/8
10/9
Time
Begin
930
840
850
1007
1015
909
1005
830
845
846
908
851
End
840
850
830
950
909
1005
830
845
846
820
851
1223
Zero Rd.
Init.
A
0.0
-0.1
-0.3
0.1
-0.2
-0.8
-1.4
-0.4
-0.5
-0.3
0.0
-1.8
Fin.
B
-0.1
-0.3
-0.1
-0.2
-0.8
-1.4
-1.4
-0.5
-0.3
1.5
-1.8
0.1
Arithmetic Mean
Confidence Interval
Zero drift*
Zero
Drift
C-B-A
-0.2
-0.2
40.2
-0.3
-0.6
-0.6
0.0
-0.1
+0.2
+1.8
-1.8
+1.9
0.02
0.53
0.55
Hi-Ranee Rdz.
Init.
D
6.3
6.3
6.3
6.3
6.1
5.9
7.5 '
6.0
6.3
6.3
6.3
5.0
Fin.
E
6.3
6.3
6.3
6.1
5.9
7.5
7.5
6.3
6.3
7.6
5.0
6.7
Span
Drift
F-E-D
0.0
0.0
0.0
-0.2
-0.2
+1.6
0.0
0.3
0.0
+1.3
-1.3
+1.7
Calibration drift*
Calib.
Drift
G-F-C
0.2
0.2
-0.2
+0.1
+0.4
+2.2
0.0
+0.4
-0.2
-0.5
+0.5
-0.2
.24
.44
.68
*1.0 used for reference value.
NOTE: All data collected during 2-week certification period are presented.
45
------- |