Performance Test of
Copper Smelter for Monitoring
ASARCO Double Contact Process
Acid Plant
at El Paso, Texas
SCOTT RESEARCH LABORATORIES, INCORPORATED
-------
Final Report
EPA 73-CUS-l
SRL 2113 05 0873
Volume 1
Performance Test of
Copper Smelter for Monitoring
ASARCO Double Contact Process
Acid Plant
at El Paso, Texas
Submitted To:
Environmental Protection Agency
Research Triangle Park, North Carolina 27711
Contract 68-02-0233
Task Order Number 10
Robert M. Martin, Project Officer
September 25, 1973
SCOTT RESEARCH LABORATORIES, INC.
2600 Cajon Boulevard
San Bernardino, California 92411
-------
SRL 2113 05 0873
1.0 INTRODUCTION
Scott Research Laboratories, Inc., performed source emissions
tests at the ASARCO Acid Plant, El Paso, Texas on Contract 68-02-0233
Task Order 10.
The objective of the program is to perform continuous monitor-
ing of acid plant emissions over a six-month period and to investigate
performance of the SO,, monitoring system using methods specified by EPA.
. -Prior--to .this testing .period Scott-Research Laboratories con-
ducted presurveys at the ASARCO facility and at the Phelps-Dodge New Cornelia
2
plant in Ajo, Arizona. Scott coordinated design, procurement and
installation of continuous monitoring and recording instrumentation.
Coordination of data processing requirements of the continuous data was
also accomplished.
This report describes tests performed during the period of
July 9 to July 13, 1973. Data acquired during the previous performance
evaluation testing period from May 29 to June 4, 1973 is also included.
2.0 SUMMARY
A DuPont 460 S0_ Analyzer was selected for the ASARCO plant
program and installed on May 17, 1973. It successfully completed 168
hours of operation without maintenance on May 25, 1973.
A total of nine duplicate modified Method 8 tests were performed
between July 9 and July 13, 1973 to complete evaluation of the DuPont in-
strument. In addition, three Acid Mist Tests using the modified Method 5,
or Monsanto, Procedure, a 24 hour zero drift, and a 24 hour calibration drift
test were performed during this test period. Nine calibration gas analyses
conducted on May 31, 1973 during the previous performance evaluation test-
ing period are reported herein.
Method 12, "Performance Specifications and Specification Test Procedures
for Continuous Sulfur Dioxide Measurement System for Non-Ferrous Smelters."
SRL 2113-01-0473, "Presurvey., ASARCO and Phelps-Dodge Acid Plants,"
W. F. Fuller, April 10, 1973.
SCOTT RESEARCH LABORATORIES. INC.
-------
-2-
SRL 2113 05 0873
Portions of the test program as specified in the draft copy of
Method 12 that were completed in the initial test period are:
Calibration Gas Values
Calibration Error
Repeatability
Response Time
Span Gas
Medium Gas
Low Gas
522 ±13 ppm 291.0 ±0.2 ppm 142.5 ±6.9 ppm
-6.6 ±0.02% -1.2 ±0.0.% -0.5 ±0.08%
-0.02 ±0.3% -0.02 ±0.04% +0.02 ±0.13%
32.8 ±0.4 seconds
All of the above parameters met requirements specified in the
draft copy of Method 12 and further tests were not performed in these
areas. The results of the nine duplicate modified Method 8 tests for S0_
are shown in Table 1. The range of values measured are:
Range of Average SO- values measured (ppm) Method 8 DuPont
12.5 to 239 19.9 to 224
Mean--difference: -(percent -of-span) "+0 .'083 +0.47%
The results of the three Acid Mist tests conducted on July 13, 1973 are:
Date and Time
7-13-73
0925
7-13-73
1318
7-13-73
1532
Average
Analyses
mg_HJ30,/SCF
£. ~~*f
0.0040
0.0045
0.0023
* NOTE: See Section 5.10
SCOTT RESEARCH LABORATORIES. INC.
-------
-3-
SRL 2113 05 0873
Table 1
Comparison of S0_ Measurements
Date and Time
7-9-73
1617
7-10-73
1011
7-10-73
1418
7-10-73
1602
7-10-73
1745
7-11-73
0816
7-11-73
1000
7-12-73
1627
7-12-73
1805
Average Wet
Analysis
ppm
12.5
122.0
21.0
117.5
53.0
19 .-5
49.5
239
22.5
Du Pont
Average Reading
ppm
19.9
121.2
22.1
116.3
48.5
22.2
51.4
224.3
23.1
SCOTT RESEARCH LABORATORIES. INC.
-------
-4-
SRL 2113 05 0873
Several other measurements were made as required by the draft copy
of Method 12. The values for these parameters are:
2 hours
24 hours
Zero Drift:
Calibration Drift (Span Gas) :
Calibration Drift (Medium Gas):
Calibration .Dri-ft (Low Gas) :
-HO.098 ±0.022%*
-0.25 ±1.93%*
-0.075 ±0.003%**
+0.017 ±0.006%**
+0.27 +2.21%*
-0.12 ±0.18%*
* Combination of data from both performance evaluation test periods
** Data from the second performance evaluation test period only.
Raw data obtained during the initial test period is shown in
Appendix A. Data obtained during the second test period is shown in Appendix
B.
3.0 PROGRAM SCHEDULE
The basic program schedule is indicated in Figure 1. The schedule
has been flexible necessarily in response to such factors as plant operations,
shutdowns and vacations; delivery and installation schedules; and the
developmental nature of the methods being employed. Current estimates of
major milestones are as follows:
4/2/73 Presurvey (complete)
4/18/73 Contractors Meeting (complete)
5/18/73 Instrumentation Installa-
tion at ASARCO
7/6/73
8/1/73
Recorder Instrumentation
at Ajo complete
ASARCO performance
test report
10/1/73 Phelps-Dodge Performance
Test Report
12/1/73 Continuous Monitoring
at ASARCO Completed
3/1/74
4/30/74 Final Report Issued
Phelps-Dodge Continuous
Monitoring Complete
(completed 6/15/73)
(complete)
SCOTT RESEARCH LABORATORIES, INC.
-------
9e
w
u>
tn
sa
o
B
o
S3
o
o
Apr
Presurvey
Contractors Meeting
ASARCO
Instrumentation Installation
Warmup Period
Performance Test Periods
Continuous Monitoring Period
Tape Cartridge Removal Dates
PHELPS-DODGE
Recorder Installation
Performance Test Period
Continuous Monitoring Period
Tape Cartridge Removal Dates
Plant Vacation
Reporting
Presurvey
Performance Test Period Report
Data Progress Report
Final Report
A
A
A
1973
May June1 July Aug Sep Oct Nov Dec
A
ESJ
E
A
8
A AA
1 ll-te
Hfc."1
A
Z.I
A
INT, faft
A
ssir
:A
AA
o
A
A
VHMKM^M
A
A
Z3
A
A
AAAA
CO
A
AAAA.
AP-D
A
A
A
A
A
1974
Jan Feb Mar
A
A
*-
A
A
i
A
A
JraftJ
Apr
i
i
i
\
i
t
A
Issue .
C/J
NJ
M
M
CO
O
O
oo
OJ
I
Ul
Figure 1 Program Schedule
-------
-6-
SRL 2113 05 0873
Times of tape cartridge removal and data progress reports are
also shown in the program schedule. Tape cartridge dates are tentative
because weekly data have been requested for an unspecified period during
the continuous monitoring.
The plant operating log for both ASARCO (from 5-17-73) and
Phelps-Dodge (from 6-15-73) is indicated in Figure 1. The initial per-
formance testing period at ASARCO was interrupted on 6-4-73 due to plant
shutdown for maintenance, and the dates for continuing this task were
again rescheduled when the plant shut down for maintenance on 6-25-73.
The remaining performance tests were performed during the week of July 9,
1973.
4.0 INSTRUMENT INSTALLATION AND OPERATION
A DuPont 460 continuous SO- outlet monitor was selected for the
ASARCO acid plant on the basis of similarity of types in use at other copper
smelters. Installation was completed on May 17, 1973 and verified by a
DuPont "representative. A 'West'inghouse E2B environmental recording package
and a dual pen strip chart recorder were installed at the same time. Both
the recorders accepted input from the S0« outlet monitor (DuPont 460) from
2000 hours on May 17, 1973. The inlet instrument was a Beckman S0_ Analyzer
owned and serviced by ASARCO. This input was not wired in immediately due
to potential interference of the recording system with plant control functions.
A revision of the input/output block of signal conditioning units in the
Westinghouse E2B recorder permitted recording the inlet data beginning
May 30, 1973. The need for recording stack volume flow data has been de-
termined and the necessary additional signal conditioning equipment has been
installed. Data will be taken upon the installation of a pneumatic signal
transducer. With the exception of strip chart pen inking problems, the
recording instrumentation has functioned very well.
The recording instrumentation shown functionally in Figure 2,
includes a strip chart recorder and a Westinghouse E2B environmental re-
cording package. The E2B is a tape recording system and is used with the
Westinghouse Environmental Systems Center's translation and reporting services,
SCOTT RESEARCH LABORATORIES, INC.
-------
Stack Outlet
Probe
Sample Flow
DuPont 460
S0~ Analyzer
0-2000 ppm S02
Acid Plant Inlet
Sample Flow
Beckman 315
S0_ Analyzer
0-12% S0n
Signal
Control
DuPont
Control Unit
Zero Control
Set
Zero
0-10 mv
up
down
Zero Backset
Switches
Channel "C"
Honeywell-Brown
Strip Chart Recorder
Channel "B1
Signal
;onverter,
ACROMAG
mv/v
0-lv
0-10 mv
Pneumatic
Pressure
Transducer
Adjustable
Attenuator
10-15 ma
non-
linear
Response
Signal
Conditioner/
Linearizer
Moore Ind.
ma/v
0-5 v
Linear Response
Signal
londitioner,
tfestinghouse
A/D
0-100
Signal
Conditioner
Westinghouse
A/D
0-100
Pulses/min
Signal
Conditioner
Westinghouse
A/D
0-100
Westinghouse
Tape Recorder
Pulses/min.
Pulses/min.
Channel "C"
Outlet S02
loncentration
Channel "A"
Outlet Stack
Volumetric
Flow
Channel "B11
Inlet S02
Conditioner
Figure 2 Functional Diagram, Equipment Installation - ASARCO, El Paso
-------
-8-
SRL 2113 05 0873
Four channels of analog signals; time, SO inlet, SO- outlet,
and stack volumetric flow (to commence mid-August) can be converted to
proportionate pulse form and recorded on magnetic tape at the test site.
The continuously moving tape has the capability to record a data pulse
every 0.6 second (100 pulses a minute) with a maximum capacity of 6,000
pulses per hour per channel for 36 days. In any 15 minutes of time, up to
1500 pulses are put on the tape with a recording accuracy of plus or
-minus -two .pulses (O.,133%).
Once a week the tape cartridges are removed from the recorder in
the field and are returned to the service center for processing. An identi-
fication field data cartridge form accompanies each tape cartridge to the
processing center. The translation service provides the following:
15 minute, hourly, and 6-hour running averages for SO- outlet and S0_
inlet; and a pollution frequency distribution report for a 6-hour running
average for the SO- outlet data.
The DuPont 460 Analyzer successfully completed the required 168
.hours of operation without maintenance on.May 25, 1973. Routine maintenance
was omitted successfully to demonstrate that weekly routine checks by Scott
personnel would be sufficient during the six month continuous monitoring
period. However, span changes were made during the preliminary test week
by exercising external manual control to accommodate higher levels of plant
SO- output than were anticipated by presurvey data analysis.
5.0 PERFORMANCE TEST RESULTS
The performance tests were conducted in two phases. The first
phase was conducted from May 29 to June 4, 1973. Data obtained during that
period included the testing of nine calibration gas samples using Method
8 procedures, and DuPont instrument stability checks for 2-hour zero and
calibration drifts.
Additional performance tests were conducted from July 9 to July
13, 1973. These included additional DuPont instrument stability tests,
nine duplicate Method 8 tests, and three Acid Mist Measurements with accom-
panying visual opacity readings of the acid plant stack outlet.
SCOTT RESEARCH LABORATORIES. INC.
-------
-9-
SRL 2113 05 0873
Only one probe positioned in one port was used for the duplicate
Method 8 Tests, the sample line being split with a "T" behind the probe to
the two sample trains and meter boxes. The probe was stationary during
the Method 8 tests but was traversed across the stack (24 points) during
the Acid Mist Tests. Method 8 procedure is shown in Appendix C and Acid
Mist Procedure is shown in Appendix D.
The DuPont instrument probe was located at the same level as the
wet chemical test probe but entered the stack from a port at right angle
to the first probe. The DuPont instrument performance tests were conducted
as described in the draft copy of Method 12 and shown in Appendix E. De-
tails of the testing are described in the following subsections.
5.1 CALIBRATION GAS ANALYSIS BY METHOD 8
Calibration gas cylinders containing S0? analyzed in the laboratory
at levels of 547 ppm (span), 287 ppm (medium), 142 ppm (low), and less than
1 ppm (zero gas) were installed on the sampling platform. Except for the
zero gas, each of the calibration gases was analyzed in the field to check
the laboratory analysis. The procedure used was Method 8 sampling train
connected to the regulator of the gas cylinder by a 3/8 inch diameter Teflon
line. The approximate length of the Teflon lines was 30 feet.
The results of the field analysis are summarized in Table 2. The
EPA test number in the table corresponds to the sample analysis number. The
Scott test number identification indicates the level of SO- concentration
in the gas cylinder being tested. Of nine tests only one was discarded
as an outlyer (Test Number S-73-001 204). The average values and confi-
dence intervals at the 95% level were:
522 ppm ±13 ppm Span Gas
291.0 ppm ±0.2 ppm Medium Gas
142.5 ppm ±6.9 ppm Low Concentration
The confidence interval for the low gas was affected primarily
by discarding one of the three tests. The field analysis values as
listed above were then adopted as the reference values for the calibration
gas in spanning the instrument and in performing of further calculations
as specified in the draft copy of Method 12.
SCOTT RESEARCH LABORATORIES. INC.
-------
©
SCOTT RESEARCH If
CO
O
33
«».
LTORIES,
p
Test Date
31 May 1973
31 May 1973
31 May 1973
31 May 1973
31 May 1973
31 May 1973
31 May 1973
31 May 1973
31 May 1973
Table 2 - Calibration Gas Check with Method 8
Scott Test Number EPA Test Number Results in PPM
2113-Low-166-l
2113-Low-166-2
2113-Low-166-3
2113-Med-287-l
2113-Med-287-2
2113-Med-287-3
2113-Span-547-l
2113-Span-547-2
2113-Span-547-3
S-73-bOl 201
8-73-001 204
S-73-bOl 203
S-73-001 206
5-73-001 207
S-73-bOl 208
S-73-bOl 202
S-73-001 205
8-73-001 209
144
44
141
292
288
293
521
507
538
Train Remarks
B
A Discarded
A
A
B
A
B
B
A
NJ
O
01
O
oo
CO
-------
-11-
SRL 2113 05 0873
5.2 CALIBRATION ERROR
The calibration error for the DuPont 460 S0_ Analyzer was
determined by admitting calibration gases in a random sequence at the base
of the instrument probe approximately 12 inches outside the stack skinline.
The sequence of this procedure yielded three sets of five data points, one
set for each of the calibration gases tested. In addition, the zero gas
was admitted at random times yielding a total number of 21 data points for
the procedure. ..The .sequence used .and the data acquired are summarized in
Table 3. The data shown were taken at a constant sample flow rate of 4 SCFH.
The differences between the instrument reading and the known calibration gas
values are shown in Table 3 for the span, the medium and the lower concen-
trations. The mean difference for the span gas was -5.8 ppm. The mean
difference for the medium and low concentrations were -11.9 and -4.9 ppm,
respectively. Confidence intervals at the 95% level were calculated to be
±0.8 ppm for the span gas, ±0.1 ppm for the medium gas and ±0.8 ppm for
the lower concentration gas. By direction of the EPA project personnel,
the span to be used for comparison of these values is the full scale span
of the DuPont instrument which was 1,000 ppm. As a percentage of span, the
mean differences are -0.6 ±0.02% for the span gas, -1.2 ±0.01% for the medium
gas, and -0.5 ±0.08% for the low gas. Method 12 permits means differences of
±5 ±10% as maximums. Consequently the DuPont instrument meets the calibra-
tion error performance required.
5.3 REPEATABILITY
Repeatability of the instrument was determined by further analysis
of the calibration error data. Each calibration gas concentration was
treated as a set of five points yielding four successive differences. Table
4 indicates the successive span readings, the resultant differences, the
computed mean difference and the confidence interval at the 95% level for
each of the calibration gases. As a percent of span the mean differences
are -0.02 ±0.03% for the span gas, -0.02 ±0.004% for the medium gas and
0.02 ±0.13% for the low gas. Each of the sets of data meets the repeat-
ability criterion stated in the draft of Method 12 which was ±2 ±4% of
span. The DuPont instrument therefore meets specifications.
SCOTT RESEARCH LABORATORIES, INC.
-------
©
SCOTT RESEARCH LABORA
O
2
5
o
Sequence-
1
2
3
4
5
6
7
8
9
10
11
12
13
14
.15
16
17
18
19
20
21
JLCLU.LC -J V^a-LJULTi-Cl L.O.W11 JLjJ- J.VJI-
Data Taken 5/30/73, 1000 hours, Sample Flow =
Instrument Reading,
Corrected for Zero Calibration
Gas Reading Gas Differences, ppm
Span Med Low Zero Volume Span Med Low
x Results of
280 Methods
Analysis
x
137 - 5.5
517 From - 5.0
Table 2
137.5 Span is - 5.0
522
514 5 - 7 5
-'-l.'T J i i *> / J
+ 13 ppm
280.5 -10.5
137.5. Medium is - 5.0
Slfi s 291'° s s
D'° + 0.2 ppm " D0
x
278 Low is -13.0
138 J^o'5 - 4.5
+6.9 ppm
517 . - 5.6
x
278 -13.0
516 - 6.0
138 - 4.5
279 -12.0
x
4 SCFH
Mean
Difference
as p:pm
Span
-5.8
*
Medium
- 11.9
Low
- 4.9
Confidence
Interval
@ 95%
Span
.205
Medium
.102
Low
.824
Mean .
Difference
As % of Span
Span
0.58 +0.02%
Medium
-1.19 +0.01%
Low
-0.49 +0.08%
CO
OJ:
O
01
o
oo
w
N»
-------
,gj\
^&?
ff
n
o
H
H
tn
w
58
<->
s
f
CO
o
ye
o
2
w
z
p
Table 4 - Repeatability Data
Successive
Span Read-
ings, ppm
517
514.5
516.5
517
516
i
Differ-
ences ,
ppm
-2.5
2.0
0.5
-1 0
J. W
May 30, 1973 - 1000
Mean Successive Differ-
Dif f er- __ Medium Read- ences ,
ence 95 ings, ppm ppm
280
0.5
280.5
-2.5
278
o.o
278
1.0
-0.25 -0.34 279
ppm j i ;
Hours
Mean
Differ-
ence
-0.25
Successive
Low Read-
95 ings, ppm
;
t
137
137.5
137.5
138
.037 138
ppin j i
Differ- Mean
ences y Differ- c_
ppm ence 95
0.5
0.0
0.5
0.0
0.25 1.30
ppm ;
CO
to
i
I*
LO
O
Ln
O
00
OJ
CO
-0.02 +0.03% Span
-0.02 +0.004% Span
0.02 +0.13% Span
-------
-14-
SRL 2113 05 0873
5.4 RESPONSE TIME
The response time of the instrument was determined by admitting
zero gas at the base of the probe and performing five repetitions of an
upscale to span and downscale to zero gas sequence. The times required
to obtain 95% of the span gas and the zero gas readings were then recorded
as shown in Table 5. While not required, lag. times for the gas to be trans-
ported from the base of the probe through the sample line to the instrument
are also shown. The' upsc"al"e total response ~time was 29.0 ±0.5 seconds.
The downscale total response time was 32.8 ±0.4 seconds. Both response
times meet the stated requirement of less than 5 minutes maximum in the
draft of Method 12 and further the downscale and upscale response times
are different by 11.6% and therefore meet the additional criterion of being
less than or equal to 15% different. The stated response time for the in-
strument is 32.8 ±0.4 seconds.
5.5 ZERO DRIFT (TWO HOUR)
The draft of Mefchod 1-2 -requires 15 -consecutive readings-of the
instrument zero at two-hour intervals to form a valid statistical set of
data. Nine consecutive intervals were obtained during the first performance
evaluation testing period prior to the plant shutdown for maintenance.
The six remaining consecutive intervals were obtained during the second
sampling sequence. The results of both test periods are shown in Table
6. Note that only two-hour consecutive readings were used in this analysis.
Several zero gas readings were taken during each calibration check of the
instrument and were averaged to a mean zero gas reading before the difference
over the interval was calculated. The differences yield a mean difference
of +0.98 ppm (+0.098% of span) for 15 data points and a confidence interval at
the 95% level of 0.22 ppm (+0.022% of span). These values meet the acceptance
criterion of +5 +10% of span for the instrument.
5.6 CALIBRATION DRIFT (TWO HOUR)
The Calibration drift for two-hour intervals was determined in
the same way as the zero drift by using span calibration gas concentration.
Table 7 indicates that the mean difference is -2.47 ±19.32 ppm or as a
SCOTT RESEARCH LABORATORIES, INC.
-------
o
O
H
SB
W
99
n
B
B9
O
33
>
O
a
O
Total
30
29
29
29
28
Upscale, Sec.
Table 5 - Response Time, Sample Flow 4 SCFH
Downscale, Sec.
Average
Lag (Total)
16
15
15
29.0
CI95
(Total)
+0.5
Total
34
33
33
32
32
Lag
19
19
18
17
Average
(Total)
32.8
CI95
(Total)
+0.4
CO
p
ro
o
oo
u>
Ul
I
Sample comparison:
32.8 - 29 v inn Percent difference in , ./.. c
-------
-16-
SRL 2113 05: 0873
Table 6 - Zero Drift, 2 Hours
Zero Zero
Date Time Reading Reading
Mean Interval
A Zero Difference @ 95%
5/31/73 0845 0.5
0.5
0.4
0.4 0.45
1115 1.2
1.4
1.5
1.5 1.40
+0.95
6/2/73 0915 0.9
0.8
0.9
1.0 0.90
1100 0.8
1.0
1.0
1-1 0.98
1300 0.5
0.6
0.6
0.5 0.55
1445 1.2
1.2
1.3
1.2 1.22
1700 0.4
0.5
0.5
0.5 0.48
+0.08
-0.43
+0.67
-0.76
6/3/73 0900 1.2
^A
1.3
1.1
1.0 1.15
1045 0.9
1.0
1.0
1.0 0.98
1245 1.2
1.5 1.35
1445 1.0
.8 0.90
.1645 1.2
1.3 1.25
-0.17
+0.38
-0.45
+ 0.35 ^
f \
[4 SCOTT RESEARCH LABORATORIES, INC.
-------
SRL 2113 05 0873
-17-
Table 6 - Zero Drift, 2 Hours
(Continued)
Zero
Date Time Reading
7-10-73 0900 0.7
0.8
0.7
0.7
1115 1.2
1.2
1.3
1.4
1330 1.2
1.4
1.4
1.4
1515 0.7
0.9
0.9
0.7
1715 1.5
1.8
1.8
1.8
7-11-73 0915 0.6
0.9
1.0
1.1
1115 0.8
0.9
0.9
0.9
1315 1.0
1.0
0.2
0.8
Zero
Reading
0.725
1.275
1.350
0.80
1.725
0.900
0.875
0.750
Mean
Mean
Reading A Zero Difference
Confidence
Interval
@ 95%
+0.55
+0.075
-0.550
+9.925
-0.125
+0.098%
0.98
+0.022%
+0.22 ppm
SCOTT RESEARCH LABORATORIES, INC.
-------
-18-
SRL 2113 05 0873
Table 7 - Calibration Drift, 2 Hours
Date Time
5/31/73 0845
1115
6/2/73 0915
1100
1300
1445
1700
6/3/73 0900
1045
1245
1445
1645
7/10/73 0900
1115
1330
1515
1715
7/11/73 0915
1115
1315
$Arf SCOTT RESEARCH
Confidence
Span Mean Interval
Reading A Span Difference @ 95%
521.0
- 3.0
518.0
536.0
- 5.0
531-0
+ 0.5
531.5
+ 9.0
540.5
- 9.0
531.5
537.5
- 7.0
530.5
+ 6,0
536.5
- 2.5
534.0
--4.5
529.5
520.5
+ 1.5
522.0
- 2.0
520.0
521.0
+ 1.0
515.5
- 5.5
518.5
- 2.0
516.5
t V
-14.5 V
502.0 ' -2.47 ppm +19.32 ppm
-0.25% +1.93%
LABORATORIES, INC.
-------
-19-
SRL 2113 05 0873
percentage of span -0.25+1.93%.. The acceptance criteria for calibration
drift at two-hour intervals is ±3 ±6% of span at the 95% confidence level.
The instrument meets the requirement.
During the second period of testing the low (142.5 ppm) & medium
(291 ppm) calibration gases were also checked for 2 hour drift. The re-
sults are shown in Tables 8 and 9. A mean difference from six data points
for low and medium gases is +0.017 +0.064% and -0.075 +0.032% respectively
which supports the span gas calibration drift data.
5.7 ZERO DRIFT (24 Hour)
In accordance with Method 12 sufficient data were collected over
the two sampling periods to form a valid statistical set for determinating
24-hour zero drift. This set is compiled from three, 24-hour intervals.
Several zero gas readings were taken during the calibration checks at the
start and end of each 24-hour interval. The zero readings from each check
were averaged to a mean zero reading before the difference over the interval
was calculated. The interval differences yield a mean difference of +0.27
±2.21%. These results, shown in Table 10, meet Method 12 requirements of
±5 ±10%.
5.8 CALIBRATION DRIFT (24 HOUR)
An evaluation of calibration drift for a 24-hour period was
performed coincident to the 24 hour zero drift sequence. The results, shown
in Table 11, are from data taken during three 24 hour intervals, using
the span gas. A calibration check at the start and end of each interval
yielded a difference. The differences for each interval were calculated
to a mean difference of -0.12 ±0.17%. Method 12 requirements are ±5 ±10%.
5.9 RELATIVE ACCURACY
Relative accuracy of the DuPont instrument was determined during
simultaneous sampling using two separate Method 8 trains "Teed" to one
sample probe. The DuPont instrument probe was located at the same level
in the stack but entered through a different port and at right angles to
SCOTT RESEARCH LABORATORIES. INC.
-------
-20-
SRL 2113 05 0873
Table 8 - Calibration Drift, 2 Hour
Low Gas
Date
7-10-73
7-11-73
Time
0900
1115
1330
1515
1715
0915
1115
1315
Zero Corrected
Chart
Reading
137.0
137.5
138.0
137.0
135.0
136.0
136.0
139.0
Confidence
Mean Interval
A Span Difference @ 95%
+ 0.5
+ 0.5
- 1.0
- 2.0
0.0
+ 3.0 ^
f
/
4-0.17 ppm +0.64ppm
0.017% +0.064%
SCOTT RESEARCH LABORATORIES, INC.
-------
SRL 2113 05 0873
-21-
Table 9
Calibration Drift, 2 Hour
Medium Gas
Date Time
7-10-73 0900
1115
1330
1515
1715
7-11-73 0915
1115
1315
Zero Corrected
Chart
Reading
281.5
280.5
279.0
280.0
278.0
279.5
278.0
279.0
Confidence
Mean Interval
A Span Difference @ 95%
- 1.5
- 1.5
+ 1.0
- 2.0
- 1.5
N
- 1.0 ^
f V
-0.75 ppm +.32 ppm
0.075% T.032%
SCOTT RESEARCH LABORATORIES, INC.
-------
-22-
SRL 2113 05 0873
Table 10 - Zero Drift, 24 Hour
Date
6-2-73
6-3-73
7-9-73
7-10-73
7-12-73
7-13-73
Time
0915
0900
1730-
1745
1715-
1730
1730-
1745
1715-
1730
Mean
Zero Mean Zero Difference Confidence
Reading Reading AZero A Zero Interval @ 95%
0.9
0.8
0.9
1.0 0.90
+ 0.25
1.2
1.3
1.1
1.0 1.150
+ 0.25
0.9
1.2
1.3
1.3 1.175
+ 0.550
1.5
1.8
1.8
1.8 1.725
+6.550
1.3
1.5
1.5
1.4 1.425
0.0
1.4
1.5
1.4 N
-
f >
t
lt4 1* 0.0 2.67ppm 22.1ppm
+0.27% +2.21%
SCOTT RESEARCH LABORATORIES, INC.
-------
-23-
SRL 2113 05 0873
Table 11- Calibration Drift, 24 Hours
Date
6-2-73
6-3-73
7-9-73
7-10-73
7-12-73
7-13-73
Time
0915
0900
1730
1715
1730
1715
Span
Reading
536.0
537.5
514.5
515.5
518.5
512.5
Mean
Difference
A Span A Span
+ 1.5
+ 1.0
- 6.0
>
1
X
-1.17ppm
-0.12%
Confidence
Interval
@ 95%
Y
-1.71ppm
-0.17%
SCOTT RESEARCH LABORATORIES, INC.
-------
-24-
SRL 2113 05 0873
the first probe. The probe tips were located near the centroid of the
stack with the standard wire mesh probe filter attached to the DuPont
instrument probe and an open-ended glass tube for the sampling trains.
Nine duplicate Method 8 samples were taken from July 9 to July 12,
1973 and analyzed by barium perchlorate titration in the field. An EPA
chemist performed the analysis consisting of two each titrations for each
train and for each run. Table 12 summarizes this data.
The draft of Method 12 requires nine duplicate samples to form
a valid statistical set of data for analysis. Values labeled in Table 12
as DuPont Instrument are planimeter averaged values from the strip chart
record. These values are more accurate than 6-point field averages taken
onsite. Baced on the nine duplicate samples obtained, an analysis was
made which indicated (as shown in Table 12) that the mean difference was
+0.083 ±0.47% of span. The complete set of data therefore meets the require-
ment stated in the draft of Method 12 for relative accuracy.
The five duplicate tests conducted on June 2, 1973, during the
preliminary test period did not provide data that were repeatable or con-
clusive. A summary of that data is provided in Appendix C.
5.10 ACID MIST DETERMINATION
Three acid mist determinations per the draft copy furnished by
EPA was taken during this testing period. A velocity traverse of the stack
conducted during the initial test period indicated strong asymmetric
flow and the possibility of recirculation of external gases in the stack.
A 24-point traverse was specified for this series of tests. During the two
hour duration of these tests, the EPA Project Officer visually observed the
plant stack emissions and recorded data. The chemical analysis is shown in
Table 13. Opacity data are shown in Appendix A. The sample box was heated
during these tests due to incorrect interpretation of the draft copy. Con-
sequently the data is suspect.
5.11 ORSAT ANALYSIS
Orsat analysis as prescribed in the Federal Register was performed
during the initial performance evaluation test. The analysis indicated
SCOTT RESEARCH LABORATORIES, INC.
-------
Table 12 - Relative Accuracy, Data In Performance Evaluation Tests
0fy
^5«?
CA
n
o
H
x
u>
r>
B
B
O
a
^.
TORIES. IN
r>
Date
& Test Test
Time Type Number
7-9-73 Method S73
1617 8 S02
Two Trains
7-10-73 off of
1011 1 Probe
7 10 73 See App* A
1418 for ,D;:
scription
7-10-73
1602
7-10-73
1745
7-11-73
0816
7-11-73
1000
7-12-73
1627
7-12-73
1805 N
r ^
001 301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
f 318
Train
Number
1A
IB
2A
2B
3A
3B
4A
4B
5A
5B
6A
6B
7A
7B
8A -
8B
9A
9B
Wet
Analysis
ppm
14.8
10.2
121.0
123.0
21.0
21.0
116.0
119.0
46.0
60.0
19.0
20.0
49.0
50.0
241
237
22
23
Avg. Wet
Analysis
ppm
12.5
122.0
21.0
117.5
53.0
19.5
49.5
239
22.5
Test
DuPont
Avg. Reading
ppm
19.9
121.2
22.1
116.3
48.5
22.2
51.4
224.3
23.1
P
N9
Results M
Difference, to
Dupont to wet Mean CI o
Analysis, ppm Difference @ 95% Remarks ^
CO
-7.4 NOTE: Train A was W
teed off of train B.
Train B was directly
+ .8 connected to the
probe.
DuPont avg. readings
obtained by planimet-
., , er measurement of
~r_L » t- t , \
strip chart. N>
r Ol
i
+4.5
-2.7
-1.9
+14.7
-.6 +0.83ppm -4.68ppm
_i n r\n o o/ . r\ /. ~i v
-------
SRL 2113 05 0873
-26-
Date and
Time
7-13-73
925
7-13-73
1318
7-13-73
1532
Table 13
Acid Mist Determination
Test
Number
AM #1
AM //2
AM #3
Wet Analysis
mg
0.0040
0.0045
0.0023
SCOTT RESEARCH LABORATORIES. INC.
-------
-27-
SRL 2113 05 0873
the stack composition to be approximately that of air at standard conditions.
The data are presented in Appendix A.
6.0 DATA PROCESSING STATUS
As of the date of this report eight tapes have been submitted
to Westinghouse for processing. Programming difficulties at Westinghouse,
previously mentioned in the interim report, have been rectified by Westing-
house .personnel excep±~for provision of a ..daily summary pollution frequency
distribution report. Scott is currently removing tape cartridges at '
the ASARCO facility on a weekly basis except as noted further in this
section.
Data reduction at Scott related to the periods for the performance
evaluation test has been completed. All data from these records are discussed
in the interim letter report number 2113 02 0773 and other sub-sections of
this report. Each strip chart covering performance evaluation test periods
was examined for instrument problems, zero adjustment, and for periods
which should be omitted due to instrument calibration or routine maintenance.
Further, to obtain accurate values for calculating relative accuracy of
the wet test method and DuPont instrument trace, Scott performed planimeter
averaging of all modified Method 5 test periods covered by the data in
Table 12 and Appendix B. These averages replace the six point field averages
which were calculated and used at the time of testing in the field.
SCOTT RESEARCH LABORATORIES, INC.
------- |