EPA-600/2-78-007
January 1978 Environmental Protection Technology Series
EVALUATION OF A SULFUR DIOXIDE MASS
EMISSION RATE MONITORING SYSTEM
^ sr^
Environmental Sciences Research Laboratory
Office of Research and Development
% U.S. Environmental Protection Agency
Research Triangle Part, North Carolina 27711
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development. U S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the ENVIRONMENTAL PROTECTION TECH-
NOLOGY series. This series describes research performed to develop and dem-
onstrate instrumentation, equipment, and methodology to repair or prevent en-
vironmental degradation from point and non-point sources of pollution. This work
provides the new or improved technology required for the control and treatment
of pollution sources to meet environmental quality standards.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/2-78-007
January 1978
EVALUATION OF A SULFUR DIOXIDE
MASS EMISSION RATE MONITORING SYSTEM
by
Roosevelt Rollins
Emissions Measurement and Characterization Division
Environmental Sciences Research Laboratory
Research Triangle Park, NC 27711
ENVIRONMENTAL SCIENCES RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
RESEARCH TRIANGLE PARK, NC 27711
-------
DISCLAIMER
This report has been reviewed by the Environmental Sciences Research
Laboratory, U.S. Environmental Protection Agency, and approved for publi-
cation. Mention of trade names or commercial products does not constitute
endorsement or recommendation for use.
-------
ABSTRACT
An evaluation was conducted to determine the capabilities and limitations
of a commercially available monitoring system that provides sulfur dioxide
mass emission data as a direct output. The monitoring system was operated
continuously for extended periods at a coal-fired power plant and at a sul-
furic acid production facility. Additional testing was performed at a Simu-
lated Stationary Source Facility to confirm some deficiencies noted during
field operations. The system performance was verified by comparing its output
data with the results obtained by EPA reference methods tests.
Results are presented for three performance tests at each field site.
For the power plant tests, the monitoring system agreed within ± 20% of the
accepted reference method. In the case of the acid plant, the system accuracy
was as poor as 58 percent. Generally, the monitoring system performed reli-
ably throughout the extended test program. The system remained operational
greater than 90 percent of the time during the approximately 4-month test
program.
This report covers a period from August 11, 1976, to July 1, 1977, and
work was completed as of June 8, 1977.
iii
-------
CONTENTS
Abstract iii
Figures vi
Tables vi
Acknowledgements vii
1. Introduction 1
2. Conclusions 2
3. System Description 4
Sample interface 4
SCL analyzer 4
Velocity subsystem 8
Analog computer 8
4. Field Evaluation 10
Test sites 10
Test procedures 11
Results and discussion 13
5. Stationary Source Simulator Tests 24
Test facility 24
Results and discussion 24
-------
FIGURES
Number Page
1 Block diagram of the S(L mass emission rate
monitoring system 5
2 Photograph of the monitoring system 6
3 Block diagram of the S02 analyzer 7
TABLES
Number Page
1 Power Plant Test Data for Test Period #1 14
2 Power Plant Test Data for Test Period #2 15
3 Power Plant Test Data for Test Period #3 16
4 Acid Plant Test Data for Test Period #1 18
5 Acid Plant Test Data for Test Period #2 . .' 19
6 Acid Plant Test Data for Test Period #3 20
7 Zero Drift Data for the S02 Analyzer 22
8 SSSF Test Data 25
9 Zero Drift Data for SSSF Tests 27
10 SO and Emission Rate Values Data corrected
fof Zero Drift Errors 28
-------
Acknowledgments
We are grateful to Carolina Power & Light Company and Texasgulf, Inc.
for their cooperation and assistance in the use of their facilities during
the evaluation program.
The evaluation at the power plant as well as the first test series at
the acid plant was conducted by Scott Environmental Technology, Inc. under
EPA Contract No. 68-02-1400, Task 26. Additional testing at the acid plant
was performed by Northrop Services, Inc. under EPA Contract No. 68-02-2566.
Analyses of Method 8 samples obtained from the acid plant were made by
Engineering-Science under EPA Contract No. 68-02-1406, Task 32.
vii
-------
SECTION 1
INTRODUCTION
The Emissions Measurement and Characterization Division of the Environ-
mental Sciences Research Laboratory is engaged in programs to develop and
evaluate measurement techniques and monitoring systems for pollutants emitted
from stationary sources. Program activities include the evaluation of com-
mercially available monitoring systems when a need is anticipated or known to
exist. These activities are carried out primarily to support the implemen-
tation of performance standards for stationary sources.
According to regulations promulgated by the Environmental Protection
Agency (EPA), certain stationary sources in designated industries must meet
specific emission limitations. In addition to requiring these sources to
comply with the emission limitations, EPA regulations call for continuous
stack monitoring and the reporting of excess emissions on a mass-rate basis.
Conventional monitoring systems provide emission data in terms of gas volume
concentrations. This necessitates supporting measurements of volumetric flow
rates and production rates in order to relate the concentration data to
applicable standards. Errors may result if all measurements are not made
concurrently.
More recently, the EPA has promulgated regulations that allow mass
emission data to be calculated by using gas volume concentrations and pre-
determined conversion factors. The conversion factor would vary with plant
operations but is assumed to remain relatively constant between successive
determinations.
An alternate method for generating mass emission information is to use a
monitoring system that provides real time mass emission rate data as a direct
output. This approach should be more accurate and also greatly ease the bur-
den of collecting emission data as required by EPA regulations.
This report presents the results of an evaluation of a commercially
available monitoring system that provides sulfur dioxide (SO,) mass emission
rate data as a direct output. The monitoring system was designed for con-
tinuous monitoring applications at combustion sources. To determine system
accuracy and reliability over an extended period, evaluations were conducted
at a coal-fired power plant and at a sulfuric acid production facility.
During the evaluation program, EPA reference methods were used to verify
system performance.
-------
SECTION 2
CONCLUSIONS
A monitoring system for sulfur dioxide (S0?) mass emission rate (MERMS)
has been successfully evaluated at a coal-fired power plant and at a sulfuric
acid production facility.
The MERMS performed reasonably well during field operations at the power
plant. For each of three test periods, the monitoring system agreed within
20% of the reference method. The mass emission rate errors were largely due
to significant differences in both the SO concentration and velocity measure-
ments. In comparison with reference meth6d values, the MERMS outputs were con-
sistently 15-20% higher for S02 concentrations and about 10% lower for veloci-
ties. Since these percentage errors were nearly constant for all individual
measurements, system accuracy can be improved by applying appropriate cor-
rection factors during calibration of the SO^ analyzer and velocimeter sub-
systems.
During field operations at the acid plant, the MERMS relative accuracy
was as poor as 58 percent. The poor performance was due primarily to severe
zero drift in the S02 analyzer subsystem. Field tests for zero drift showed
that the analyzer drift was typically 20-25 ppm per hour. The zero drifts
were usually negative and thus resulted in lower S0? concentration readings.
The zero drift problem was critical at the acid plant because of the low S02
concentration levels experienced during the test program.
An "automatic zero" option is available for the basic SO^ analyzer used
in the MERMS, but this optional feature was not incorporated Tn the system
tested. It is likely that such an option may be necessary in order for the
MERMS to meet current minimum performance specifications prescribed for moni-
tors of S02 from stationary sources.
Despite the minimum attention provided, the MERMS operated reliably
throughout the extended field test program. During almost 2 months of
continuous operation at the power plant, the system required corrective main-
tenance actions on only two occasions. On one of these occasions, the prob-
lem was attributed to poor installation procedures rather than to a system
weakness.
Upon installation of the MERMS at the acid plant site, a major mal-
function occurred. The differential pressure sensor used in the velocity
measuring subsystem was damaged when a backpurge solenoid failed, thus ex-
posing the sensor to 50 psi of instrument air. Considering the delicate
-------
nature of the pressure sensor, a more dependable method of switching backpurge
air to the pi tot lines should be incorporated. After becoming fully opera-
tional at the acid plant site, the MERMS operated continuously for over 2
months without any serious problem. Whenever a malfunction did occur, minimum
efforts were required to diagnose and correct the problem.
It should be noted that the monitoring system tested was manufactured in
1972. The manufacturer claims that some modifications have been made to im-
prove overall system performance. No attempt has been made to substantiate
these claims.
-------
SECTION 3
SYSTEM DESCRIPTION
The S02 Mass Emission Rate Monitoring System (MERMS) was manufactured by
Western Research and Development, Ltd, Calgary, Alberta, Canada. It measures
the effluent temperature, velocity, and S02 concentration independently and
simultaneously. With these three measured variables plus appropriate con-
stants, the system computes the instantaneous mass emission and volumetric
flow rates on a continuous basis.
The system, shown schematically in Figure 1, has five basic subsystems:
sample interface, S0« analyzer, velocimeter, temperature transmitter, and
analog computer/remote control station. The measuring instrumentation is
housed in two temperature-controlled enclosures designed to be located near
the stack. The analog computer/remote control station was designed to be
located in a control room. Photographs of the MERMS are shown in Figure 2.
The sample probe, sample line, and analyzer sample cell may be heated up
to 250 C. Their temperatures are individually controlled by electronic
circuitry which also provides automatic protection shutdown when any tempera-
ture falls outside a preselected range. Provisions are included for auto-
matic backpurge of both the sample and Pitot lines. Calibration of the S02
analyzer and velocimeter may be performed from the remote control station.
SAMPLE INTERFACE
The sample interface consists of a combination probe assembly that in-
corporates a thermocouple, a Pitot tube, and a gas sample extraction probe.
The gas sample probe is fitted with a sintered stainless steel filter for
removal of coarse particulates. A heat-traced Teflon line connects the
sample probe to the S02 analyzer. The probe assembly is 4 m long and about
7.6 cm in diameter.
S02 ANALYZER
The concentration of S02 in the sample is measured by a DuPont Model 400
Photometric Analyzer. This analyzer, shown schematically in Figure 3, is a
dual beam single path instrument utilizing the principle that gases absorb
radiation at characteristic wavelengths. Light from an ultraviolet source,
after passing through a sample cell, is split into a reference and a measuring
beam. These beams are directed through appropriate bandpass filters before
reaching the detectors. The wavelength of the measuring beam filter is
selected for strong absorption by S02 while the reference wavelength is
selected from minimum absorption by S02< The analyzer used in the MERMS
-------
SOjANALYZER
PITOT
TUBE
FLOW METER
BY-PASS
VALVE
THERMOCOUPLE
f| SOLENOID
IMS}—
TEFLON
ASPIRATOR
VENT
ANALOO COMPUTER
ft
REMOTE CONTROL STATION
2-PEN RECORDER
FOR MASS EMISSION RATE (ki/br)
AND VOLUMETRIC FLOW RATE (m'/mln)
Figure 1. Block diagram of the S02 mass emission rate monitoring system
-------
-
Figure 2. Photograph of the monitoring system
-------
SAMPLE OUT
PHOTOTUBE
BEAM SPLITTER
PHOTOTUBE
U.V.
LAMP
Figure 3. Block diagram of the SCL analyzer
-------
employs a measuring wavelength of 280 nm and a reference wavelength of 578
nm. The output from each detector is amplified, and the difference between
their signals is proportional to the S02 concentration in the sample.
The sample gas is pulled through the analyzer by an air-driven Teflon
aspirator. The analyzer may be calibrated by introducing into the sample
cell a gas whose concentration is known or by using an optional optical filter
to simulate a known concentration.
VELOCITY SUBSYSTEM
Continuous measurement of effluent velocity is accomplished by measuring
both the effluent temperature and velocity head near the point at which the
sample is extracted.
temperature is measured by a Type J (iron-constantan) thermocouple in-
terfaced with a Leeds & Northrup temperature/millivolt transmitter. This
instrument is a solid state amplifier whose output varies linearly with the
millivolt signal from the thermocouple in the probe assembly.
The velocity head is measured by an "S" type Pi tot tube coupled with a
Western Research & Development Model 301 Velocimeter. The velocimeter employs
a differential pressure sensor and converts the differential pressure mea-
sured by the Pitot tube into an electrical signal.
The temperature transmitter is calibrated by disconnecting the thermo-
couple inputs and using a millivolt source to simulate zero and span voltages
as read from a thermocouple conversion chart. The velocimeter is calibrated
by applying a known differential pressure at the pressure transducer and
making suitable electronic adjustments.
ANALOG COMPUTER
The analog computer receives the input signals representing SOo concen-
tration, differential pressure, and temperature. With these three variables
plus appropriate constants, the effluent velocity, volume flow rate and
pollutant mass emission rate are calculated on a continuous basis. Adjust-
ments for required constants, including barometric pressure, specific
gravity, pitot factor, and stack cross-sectional area, are made on the front
panel.
Analog signals representing volume flow rate and mass emission rate are
available for strip chart recording. The continuous status of SOp concen-
tration, temperature and velocity are displayed on digital panel meters on
the front panel. Analog outputs are also available for these three readouts.
The computer is calibrated by using a digital voltmeter to adjust
various circuit components to provide specified values at given test points.
A check of computer operation is then made by introducing simulated inputs
and comparing the outputs with precalculated results.
8
-------
The computer is programmed to solve the following equations
> x 7^-
f x V SG x
V$ = 4.8415 x P, " /u'r' x Ts
Qs = 5145.715 x Pf x A x v/ ^Px * F
E_ = 0.8192 x Ccn x Pf x A x ' ^-^-^- ^
__ - \J.U\3L. A V/CO AF.pAnA/ c7> 3 T
m 502 Y SG x T
where: Vg = Effluent velocity (at stack conditions) in m/sec
o
Qs = Volume flow rate (at standard conditions) in m /min
Em = Mass emission rate (at standard condition) in kg/hr
Pf = Pi tot factor (may include a profile factor)
D.P. = Differential pressure in inches of H«0
TS = Stack temperature in °R
SG = Specific gravity of effluent (ratio of the molecular weight
of the effluent to air)
PS = Absolute stack pressure in inches of Hg
2
A = Cross-sectional area of stack in m
CCn = S00 concentration in ppm
OUn C.
The basic monitoring system may be custom designed to meet specific re-
quirements. The system manufactured for the EPA was designed for the fol-
lowing conditions:
Mass Emission Rate 0-15,920 kg/hr (at standard conditions)
Volume Flow Rate 0-100,000 m3/min (at standard conditions)
Velocity 0-30 m/sec (at stack conditions)
SOp Concentration 0-1000 ppm (on a wet basis)
Differential Pressure 0-2.5 inches H20
Temperature 30°-250°C
Area (Stack) 0-100 m2
-------
SECTION 4
FIELD EVALUATION
TEST SITES
Field testing of the MERMS was conducted at a power plant and at a sul-
furic acid procudtion facility, both located in the south eastern United States.
The power plant unit was used primarily during peak load periods and
had a capacity for generating approximately one million pounds of steam per
hour which produced about 150 MW of electrical power. Emissions from the
coal -fired boiler was control ed by cyclones located after the economizer
followed by a Buell electrostatic preci pita tor located after the air heaters.
An induced draft fan, following the preci pi tator, exhausted the effluent gases
through a 60-meter high concrete stack. Stack conditions at the sampling
point were as follows:
Temperature
Moisture ^7%
S0« Concentration 500-800 ppm
0 " 5-6%
C02 " 12-13%
CO "
Velocity 12-14 m/sec
Static Pressure ^-10 in H20
Area (Stack) 8.55 m2
The MERMS probe was located in a 1.2-m x 7.5-m duct on the outlet of
the electrostatic precipitator. The instrumentation, including remote con-
trol station, was located near the duct and on a platform under the precipi-
tator hoppers.
The sulfuric acid facility was a conventional double absorption/double
contact process plant burning elemental sulfur. The unit was designed for a
converter inlet SOp concentration of 10 percent and a production capacity of
10
-------
1500 tons of acid per day. Control of sulfur compound emissions involves
recycling gases from a first absorption tower through catalytic stages before
going to the final absorber. Mist eliminators are installed in the ab-
sorption towers for removal of any acid mist. Stack conditions at the
sampling point were as follows:
Temperature 70°-80°C
S02 Concentration 50-200 ppm
Velocity 27-31 m/sec
Moisture
Static Pressure ^1 in FLO
Area (Stack) 2.14 m2
The MERMS probe was located in a 2-m circular stack following the
final absorption tower. Sampling ports were about 2 m below stack exit.
The instrumentation was located on an unsheltered platform 6 m below
the sampling ports. The remote control station and data recorders were
located in a small room at ground level.
TEST PROCEDURES
The monitoring system was operated continuously from August 9, 1976,
through October 8, 1976, at the power plant and from January 5, 1977, through
March 19, 1977, at the acid plant.
Reference methods 6 and 8 were conducted at the power plant and acid
plant, respectively. At both facilities, the reference tests were performed
during three test periods, separated by approximately 2-weeks, intervals.
Between performance test periods, the monitoring system operated unattended.
Plant personnel checked the system occasionally to confirm that it was
operational. Maintenance actions were performed only after it became obvicus
that a malfunction had occurred. No attempt was made to validate data gen-
erated or system performance during periods of unattended operation.
Performance test procedures were based on those specified in "Perfor-
mance Specifications and Specifications Test Procedures for Monitors of S02
and NO from Stationary Sources", Federal Register, Volume 40, Number 194,
October 6, 1975. Some of the prescribed test procedures were slightly modi-
fied to accommodate the MERMS and plant operations. For example, specified
procedures for zero and span are based upon a system whose output is in terms
of volume concentration. With the MERMS, zero and span checks/adjustments
are performed for each subsystem independently rather than for the system as
a whole. The regulations also stipulate that no more than one reference mea-
surement shall be performed in any one hour. However, due to the limited and
uncertain operating schedules of the power plant boiler, sampling times were
abbreviated to allow for a maximum number of samples during a test period.
11
-------
At the power plant, field tests for zero and span drift were not per-
formed because of the brief time period available for each test series. A
performance test period was generally conducted during an 8-to-10 hour span
and thus did not allow for the accumulation of enough drift data to be
conclusive.
Single point sampling was employed throughout the test program. To
insure that possible stratification had no effect on comparative test data,
the reference and monitor probes were positioned less than 8 cm apart. For
the power plant tests, the reference probe was attached to the monitor probe
and shared a common port. For the acid plant test, the reference probe was
located in an adjacent port. The probes could be seen through another port
and positioned so that all probes sampling the same area in the stack.
At both facilities, the MERMS spans for S02 concentration, temperature
and velocity were set at their designed full-scale ranges. Spans for mass
ewission rate and volumetric flow rate were set at 1592 kg/hr and 10,000
m /min, respectively. These values were based upon maximum expected outputs
as determined by preliminary site measurements. The monitor probe and sample
line operating temperatures were 125 C at the power plant and 95 C at the
acid.plant. The S02 analyzer sample cell was operated at 125 C at both
test sites.
Following initial setup at both facilities and at the beginning of each
performance test period, the MERMS was calibrated according to manufacturer's
instructions. The nature of the system dictates that each subsystem be cali-
brated separately and independently.
The S02 analyzer was first calibrated with blends of 549 and 964 ppm SOp
in nitrogen and nitrogen zero gas. Instrument responses were adjusted for
the sample cell pressure difference between sample and calibration modes. An
internal optical filter is incorporated in the monitor for checking the cali-
bration remotely without gases. The simulated S02 concentration given by
insertion of this filter was determined relative to the span gas responses
during original calibration. This span filter was then used for calibration
checks throughout the remainder of the performance test period.
The velocimeter was initially calibrated by applying known differen-
tial pressures at the pressure sensor. A syringe was used as the source, and
the pressure was read on an inclined water manometer. The velocimeter also in-
corporates provision for remote calibration by electronically simulating a
known differential pressure. This simulated pressure response was then re-
lated to the initial calibration responses and used for calibration checks
during the Pi tot tube in the calibration. A Pi tot calibration factor had
been established in the laboratory prior to commencing the field test program.
The temperature transmitter was calibrated by disconnecting the thermo-
couple inputs and using a millivolt source to simulate zero and span voltages
12
-------
as read from a thermocouple conversion chart. Appropriate electronic adjust-
ments were made to produce proper output voltages.
Computer operations were checked at the beginning of each performance
test period by introducing simulated inputs and comparing the outputs with
precalculated results. A complete calibration of this subsystem was neces-
sary only upon initial installation at each test site and involved making
electronic adjustments to give specified voltage readings.
With the system calibrated and continuously recording data, concurrent
measurements of SOp concentration, effluent velocity, and temperature were
performed according to EPA reference methods procedures. At the power plant,
EPA Reference Method 6 tests for SOp concentration were run for about 15
min each, during which approximately 1 ft of sample was collected. At the
acid plant.gEPA Reference Method 8 tests were run for 40 min, collecting
about 35 ft of sample.
In addition to recording the MERMS primary outputs of mass emission and
volume flow rates, the SOp concentration and velocity outputs were also con-
tinuously recorded for informational purposes. The monitor temperature
output was periodically read from the digital panel meter and logged for
future reference.
RESULTS AND DISCUSSION
Performance test results are listed in Tables 1-3 for the power plant
and Tables 4-6 for the acid plant. All MERMS outputs, except temperature,
were read from strip chart recordings. The values shown for MERMS temper-
atures are approximations in that the temperature was not continuously re-
corded. Reference method S0« concentrations, velocities, and volume flow
rates were calculated according to procedures outlined in the Federal
Register, Volume 36, Number 247, December 23, 1971.
Power Plant Tests
Tables 1, 2, and 3 show the data obtained during testing at the power
plant during the three test series. The values for SOp concentration, volume
flow rate and emission rate are reported on a wet basic and at standard
conditions. For all samples, mass emission rates determined by the monitor
were higher than those calculated from reference methods data. Analyses of
the raw data show that the SOp concentration was the variable primarily re-
sponsible for these differences. For each of the three test periods, the
MERMS SOp concentrations were about 15-20% higher than the corresponding
reference method values. It is also evident that a consistent disagreement
existed between velocity values as determined by the two methods, the
MERMS velocities ran about 10% lower than reference method values. Since the
emission rate is directly proportional to both SOp concentration and velocity,
the low velocity values offset the high SOp values, resulting in good agree-
ment between emission rate values.
13
-------
TABLE 1. POWER PLANT TEST DATA FOR TEST PERIOD #1
DATE
9/1/76
ii
ti
n
ii
n
n
n
9/2/76
ii
ii
Sampl e
#
1-1
-2
-3
-4
-5
-6
-7
-8
-9
-10
-11
Emission Rate
(kg/hr)
Ref.
174
230
452
379
474
496
428
426
602
593
661
MERMS
248
307
515
492
510
530
527
521
617
655
670
Flow
Rate
(m-Vmin)
Ref.
2954
3000
4818
4937
5148
4973
5127
5071
5219
5346
5399
MERMS
2840
2954
4575
4550
4700
4940
5075
5075
4930
5050
5000
SO,, (ppm)
c.
Ref.*
370
483
589
482
578
627
525
528
724
697
769
MERMS
538
643
713
678
660
665
641
619
772
802
820
Vel (m/sec)
Ref.
6.9
7.1
11.9
12.5
13.2
12.8
13.1
13.0
13.5
13.7
13.8
MERMS
6.5
6.7
10.9
11.3
11.8
12.5
13.1
13.1
12.7
13.0
13.1
Temp (°C)
Ref.
83
87
100
110
115
116
116
116
118
116
116
MERMS**
74
75
88
96
no
113
113
113
113
116
116
Corrected to 7.0%
**Approximate value
-------
TABLE 2. POWER PLANT TEST DATA FOR TEST PERIOD #2
en
DATE
9/21/76
ii
H
ii
H
n
n
n
n
n
Sample
#
2-1
-2
-3
-4
-5
-6
-7
-8
-9
-10
Emission Rate
(kg/hr)
Ref.
391
376
328
394
378
430
387
400
411
412
MERMS
459
451
393
459
450
482
447
447
462
465
Flow
Rate
(n)3/min)
Ref.
5292
5400
4907
5414
5473
5301
5057
5407
5376
5416
MERMS
4700
5093
4700
5230
5250
5090
5075
5230
5250
5250
S0? (ppm)
c.
Ref.*
464
437
420
458
434
510
481
465
480
478
MERMS
600
550
517
538
525
588
545
533
542
550
Vel (m/sec)
Ref.
14.3
14.6
13.2
14.3
14.5
14.0
13.4
14.3
14.2
14.4
MERMS
12.8
13.7
12.9
14.0
14.1
13.6
13.5
13.9
14.0
14.3
Temp (°C)
Ref.
120
120
120
113
113
112
112
113
112
116
MERMS**
116
116
116
116
116
113
113
113
113
113
Corrected to 7.0%
**Approximate value
-------
TABLE 3. POWER PLANT TEST DATA FOR TEST PERIOD #3
DATE Sampl e
#
10/6/76 3-1
-2
-3
10/7/76 -4
-5
-6
-7
Emission Rate
(kg/hr)
Ref.
483
446
446
512
508
466
476
MERMS
490
462
490
574
567
542
523
Flow
Rate
(m3/min)
Ref.
5768
5450
5501
5684
5651
5294
5077
MERMS
5500
5270
5230
5487
5387
5106
4806
S02 (ppm)
Ref.*
524
512
507
563
562
550
586
MERMS
551
559
577
650
648
653
679
Vel (m/sec)
Ref.
14.7
13.9
14.0
14.3
14.3
13.4
12.8
MERMS
14.4
13.4
13.6
14.0
13.9
13.1
12.6
Temp (°C)
Ref.
122
121
121
117
118
120
119
MERMS**
119
119
119
116
116
116
116
Corrected to 7.0%
**Approximate value
-------
System relative accuracy data for the three power plant test periods
is summarized below.
Test Mean Absolute 95% System
Period Reference Mean Confidence Relative
Number Va1ue(kg/hr) Difference(kg/hr) Interval(kg/hr) Accuracy
1 446.8 61.5 22.0 18.7%
2 390.7 60.8 6.9 17.3%
3 476.7 44.4 23.1 14.2%
Relative accuracy is reported as the sum of the absolute value of the mean
difference and the 95% confidence interval of the differences expressed as a
percentage of the mean reference method value. All three of the results are
better than the 20% value currently specified as minimum performance specifi-
cation for monitors of SOp from stationary sources.
Field tests for zero and span drift were not performed during testing at
the power plant. However, general observations revealed a potential analyzer
drift problem. The MERMS has four major components subject to drift: the
SOp analyzer, velocimeter, temperature transmitter, and computer. Of these,
the SOp analyzer was the only component observed to be Drone to significant
drift, primarily zero drift. This was exemplified by the drift adjustments
required before and after each performance test period. Zero drift for other
components and span drift for all components were noted to be minimal, even
after several days of unattended operations.
The operational period was greater than 168 hr for each test period at
the power plant. In fact, during the almost 2-months program, only two
incidents occurred that required corrective maintenance action. On the first
occasion, the sample probe temperature sensor failed, causing the system to
automatically shut down when the operating temperature fell below set-point.
The second failure was caused by the temperature sensor not making good
physical contact with the probe and again triggering automatic shutdown.
Both of these problems were easily diagnosed and corrected, resulting in
minimum system downtime.
Acid Plant Tests
Tables 4, 5, and 6 show the test data obtained during testing at the
acid plant. The first series of tests were performed on the first day fol-
lowing system setup and thus before the normal 168-hr conditioning period.
During this test period, the MERMS emission rate values compared favorably
with the reference method values. However, during the second and third
test periods, the monitor did not perform as well. The combination of low
S02 concentrations and low velocities resulted in significant disagree-
ment between emission rate values.
17
-------
TABLE 4. ACID PLANT TEST DATA FOR TEST PERIOD #1
DATE Sample
#
1/4/77 1-1
-2
-3
-4
- 1/5/77 -5
00 " -6
-7
-8
-9
Emission Rate
(kg/hr)
Ref.
74.4
71.3
78.3
80.4
77.8
73.0
76.5
73.4
81.8
MERMS
85.
81.
80.
78.
83.
81.
81.
81.
78.
0
9
9
7
1
9
9
9
7
Flow
Rate
(m^/min)
Ref.
3292
3292
3323
3323
3323
3323
3292
3292
3292
MERMS
3110
3200
3200
3200
3200
3075
3150
3200
3150
S00 (opm)
c.
Ref.
142
136
148
152
147
138
146
140
156
MERMS
157
160
157
155
160
164
160
157
153
Vel (m/sec)
Ref.
30.6
30.6
30.9
30.9
30.9
30.9
30.6
30.6
30.6
MERMS
28.7
28.6
29.2
29.2
29.4
28.1
20.1
29.3
29.0
Temp (°C)
Ref.
79
79
79
79
79
79
79
79
79
MERMS*
79
79
79
79
79
79
79
79
79
*Approximate value
-------
TABLE 5. ACID PLANT TEST DATA FOR TEST PERIOD #2
DATE
2/23/77
2/24/77
2/25/77
it
2/26/77
n
ii
n
Sampl e
#
2-1
-2
-3
-4
-5
-6
-7
-8
-9
Emission Rate
(kg/hr)
Ref.
48.4
55.0
53.7
57.4
52.4
48.6
39.2
46.9
50.1
MERMS
47.8
62.6
55.5
51.5
32.8
51.7
33.0
19.0
13.6
Flow
Rate
(m3/min)
Ref.
2962
2991
3022
3022
3013
3015
3012
3015
3015
MERMS
2540
2602
2601
2570
2570
2655
2655
2680
2697
S09 (ppm)
c.
Ref.
103
115
112
120
no
102
82
98
105
MERMS
116
132
160
132
84
120
84
44
32
Vel (m/sec)
Ref.
27.6
28.4
28.2
28.2
28.3
28.2
28.3
28.2
28.2
MERMS
24.5
22.2
24.5
24.5
24.8
24.1
24.4
.2
24.1
Temp (°C)
Ref.
77
76
75
75
77
77
78
77
77
MERMS*
70
75
75
75
75
75
75
75
75
*Approximate value
-------
TABLE 6. ACID PLANT TEST DATA FOR TEST PERIOD #3
DATE Sample
Emission Rate
# (kg/hr)
3/18/77 3-1
-2
-3
-4
-5
^ 3/19/77 -6
o » _y
-8
Ref.
28.
26.
37.
32.
34.
30.
29.
38.
4
6
1
2
8
8
3
1
MERMS
26.9
18.8
13.6
13.6
13.6
29.9
27.3
25.9
Flow Rate
(m3/min)
Ref.
2942
2942
2942
2800
2800
2881
2881
2881
MERMS
2553
2553
2553
2590
2590
2511
2494
2494
S02 (ppm)
Ref.
61
57
79
72
78
67
64
83
MERMS
76
48
48
44
44
88
80
76
Vel (m/sec)
Ref.
27.8
27.8
27.8
26.5
26.5
26.9
26.9
26.9
MERMS
23.6
23.2
23.4
23.4
23.4
22.8
22.8
22.5
Temp
Ref.
76
76
76
76
76
73
73
73
(°C)
MERMS*
75
75
75
75
75
75
75
75
*Approximate value
-------
System relative accuracy data for these three test periods is summarized
below:
Test Mean Absolute 95% System
Period Reference Mean Confidence Relative
Number Value(kg/hr) Difference(kg/hr) Interval(kg/hr) Accuracy
1 76.3 6.2 2.9 11.7%
2 50.2 12.5 9.6 43.9%
3 35.2 11.0 7.5 57.5%
Again, the S02 concentration was the variable primarily responsible for
the poor system accuracy. Note that for samples 2-5, -8, -9 (Table 5) and
samples 3-2, -3, -4, -5, -8 (Table 6), the MERMS S02 values were considerably
lower than the corresponding reference method values. These measurements
were made a few hours after daily zero adjustments and include large errors
due to severe zero drift in the SOp analyzer.
Table 7 shows the zero drift data obtained for the S02 analyzer during
two test periods. Both 2 hr drift results exceed the 2% value specified for
continuous monitors of SOp from stationary sources. Note that the amounts
of zero drift experienced were of the same order of magnitude as the average
SO, concentration levels during the testing. Zero drift for other subsystems
ana span drift for all subsystems, including the S02 analyzer, were observed
to be negligible throughout the acid plant test program.
Upon installation of the MERMS at the acid plant site, a major break-
down occurred. The differential pressure sensor in the velocimeter was
damaged when a backpurge solenoid malfunctioned and allowed 50 psi of in-
strument air to damage the sensor. The sensor was removed and returned to
the manufacturer for repair. After replacing the pressure sensor and re-
turning the system to operation on January 5, 1977, no other major mal-
function was experienced through completion of the test program on March 19,
1977.
Some minor problems were encountered during the first part of the test
program. Most of them involved the system's temperature controllers. In one
instance, it was found that during some backpurge cycles the cold instru-
ment air cooled the sample probe below the set-point and resulted in auto-
matic system shutdown. This problem was solved by simply decreasing the probe
temperature from 125 C to 95 C. On other occasions, a faulty cabinet heater
apparently contributed to erratic responses by the temperature control
electronic circuitry, usually during periods of severe cold weather. After
replacing the defective heater with a heat lamp, these problems disappeared.
One recurring problem was experienced throughout the acid plant test
program. This problem was due to the water-logged signal cable harness that
21
-------
TABLE 7.
ZERO DRIFT DATA FOR THE S02 ANALYZER
Date
hr A
(ppm)
4 hr A Z
(ppm)
2/23
2/24
2/25
2/26
-30
0
-23
-46
-24
-24
-50
-46
-37
-30
-70
-84
2 hr
Mean difference = 31.1 ppm
95% Confidence Interval = 11.83 ppm
Drift (% Span) = 4.3%
4 hr
Mean difference =61.3 ppm
95% Confidence Interval =51.5 ppm
Drift (% Span) = 11.3%
3/18
3/19
-46
-12
-30
0
-42
-19
-58
-30
-61
2 hr
Mean difference =24.8 ppm
95% Confidence Interval = 17.75 ppm
Drift (% Span) = 4.3%
4 hr
Mean difference =49.7 ppm
95% Confidence Interval =31.4 ppm
Drift (% Span) =8.1%
22
-------
connected the measuring instrumentation to the remote control station.
During initial setup of the MERMS, a portion of the cable harness had been
left lying on the floor of the building where the analog computer/remote con-
trol station was located. Upon return to the test site a few days later, the
cable was found submerged. This condition contaminated the signal cables
and resulted in leakage between signals being fed to the computer. The
temperature signal input was the only one noticeably affected.
The remaining problems were caused by the hostile environment of an acid
plant. Corrosion of some components necessitated frequent cleaning and, in
one case, the replacement of the thermocouple connector and wire. After
about 30 days of operation, the Pitot tube became clogged with a greenish,
clay-like substance from the effluent. This substance also accumulated on
the probe filter element. Backpurge of the Pitot and sample lines did not
effectively remove this substance.
23
-------
SECTION 5
STATIONARY SOURCE SIMULATOR TESTS
During the period of June 1-8, 1977, the MERMS was further evaluated at
the EPA Stationary Source Simulator Facility (SSSF). This test program was
conducted to confirm system deficiencies noted during the field evaluation.
Tests for relative accuracy, zero drift, and span drift were performed.
TEST FACILITY
The SSSF is basically a closed-loop wind tunnel with provisions for
providing tailored and well-controlled air pollutant atmospheres. Con-
trollable parameters include gas stream velocity, temperature, particulate
size and loading, humidity, and gas concentration. The tunnel test section
is fitted with a number of ports to allow for sampling using EPA reference
methods and various types of monitors. A complete description of the SSSF
may be found in the report, EPA-650/2-75-015, Fabrication and Installation
of the Stationary Source Simulator, January 1975.
The SSSF was operated to simulate acid plant conditions. The S02 con-
centration was varied from 50-150 ppm while maintaining the temperature and
velocity at approximately 75 C and 25 m/sec, respectively. Cylinders of
pure SOp were used to charge the tunnel to desired levels. Gas concen-
trations were maintained by manual control of injection values. EPA refer-
ence method tests were performed to establish actual SOp concentration levels
and gas stream velocities. The reference and MERMS prooes were located in
opposing ports on the first test section. The stack cross-sectional area
at the sampling point was approximately 0.557 m^.
The MERMS was set up, calibrated, and left running continuously for a
7-day period. Zero and span adjustments were made at the beginning of each
test day. The SSSF was operated only during the normal 8-hr work day.
RESULTS AND DISCUSSION
Table 8 shows the data obtained during testing of the MERMS at the SSSF.
As noticed during the field test program, the monitoring system SO concen-
trations were consistently higher and velocities lower than corresponding
reference values. Relative accuracy data is enumerated below for each of
the recorded outputs.
24
-------
TABLE 8. SSSF TEST DATA
ro
en
Date
6-1-77
6-2-77
6-6-77
Sample
1
2
3
4
5
6
7
8
9
Emission Rate
(kg/hr)
Ref.
10.4
10.2
10.4
10.3
5.6
4.2
9.2
6.3
11.9
MERMS
11.2
13.1
15.7
15.7
6.7
8.1
11.3
9.4
17.6
Flow Rate
(m3/min)
Ref.
627.5
627.4
626.5
625.5
627.7
628.3
625.2
620.7
613.8
MERMS
620
600
620
610
620
620
610
600
580
S02 Cone.
(ppm)
Ref.
106
103.7
105.3
105.3
56.6
42.2
93.5
64.7
123.1
MERMS
no
131
153
153
61
77
108
90
180
Velocity
(m/sec)
Ref.
22.4
22.4
22.4
22.5
22.4
22.4
22.5
22.3
22.1
Temperature*
(°C)
MERMS
21
21
21
21
21
21
21
21
20
.9
.5
.9
.8
.9
.9
.9
.4
.0
74
74
74
76
74
73
76
74
74
Average temperature measured by the MERMS and used in the reference calculations
-------
Mean Ref.
Value
8.7
624.7
100.6
22.4
Mean
Difference
3.36
15.8
18.6
0.7
95% Confidence
Interval
1.39
7.35
10.32
0.19
Relative
Accuracy
54.6%
3.7%
28%
4%
System
Output
Emission Rate
(kg/hr)
Volume Flow Rate
(rrr/min)
SOp Concentration
(ppm)
Velocity (m/sec)
The poor relative accuracy for emission rate was due largely to the SCL
variable. As expected, the SO? analyzer zero drift performance accounted
for much of the error noted. Table 9 shows the drift data obtained for the
SCL analyzer during the test period. Note that the 2-hr zero drift was
better than the 2% value currently specified for SO monitors. Table 10
shows the monitor SO and emissions rate values corrected for approximate
zero drift errors. Relative accuracy determinations using these corrected
values are shown below.
S02 Concentration, cor- Emission Rate, cor-
rected for zero drift rected for zero drift
Mean Ref. Value 100.6 ppm 8.7 kg/hr
Mean Difference 8.1 ppm 1.2 kg/hr
95% Confidence Interval 3.4 ppm 0.34 kg/hr
Relative Accuracy 11.5% 17.6%
These results show that despite being relatively small, zero drift errors
seriously affected system accuracy.
26
-------
TABLE 9. ZERO DRIFT DATA FOR SSSF TESTS
DATE TIME 2 hr AZ (ppm)
6-1-77 0930 - 1130 +7
1130 - 1330 +10
1330 - 1530 +11
1530 - 1730 +13
6-2-77 0830 - 1030 +5
1030 - 1230 +11
1230 - 1430 -3
1430 - 1630 -13
6-3-77 0815 - 1015 -18
1015 - 1215 +23
1215 - 1415 +2
1415 - 1615 +8
6-6-77 0815 - 1015 -2
1015 - 1215 -14
6-7-77 0815 - 1015 -17
Mean Zero Drift (2 hr) = 10.5 ppm
95% Confidence Interval =3.4 ppm
Drift (%Span) = 1.4%
27
-------
TABLE 10.
S02 AND EMISSION RATE VALUES CORRECTED FOR ZERO DRIFT ERROR
l\5
00
Sample MERMS
# S09(pDm)
c.
i no
2 131
3 153
4 153
5 61
6 77
7 108
8 90
9 180
Outputs
E(kg/hr)
m
11.2
13.1
15.7
15.7
6.7
8.1
11.3
9.4
17.6
Approximate
Zero Drift
Error (ppm)
+2
+16
+35
+35
_ _ _
+20
+8
+20
+50
S02 Corrected
for Zero Drift
(ppm)
108
112
118
118
61
57
100
70
130
Em Corrected
for Zero Drift
(kg/hr)
11.0
11.2
12.1
12.1
6.7
6.0
10.5
7.3
12.7
-------
TECHNICAL REPORT DATA
(1'lease read Instructions on the reverse before completing)
RLPORT NO.
EPA-600/2-78-007
2.
4. TITLE AND SUBTITLE
EVALUATION OF A SULFUR DIOXIDE MASS EMISSION RATE
MONITORING SYSTEM
3. RECIPIENT'S ACCESSION>NO.
5. REPORT DATE
January 1978
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Roosevelt Rollins
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Environmental Sciences Research Laboratory - RTP, NC
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
10. PROGRAM ELEMENT NO.
1AD712 BA-18 (FY-77)
11. CONTRACT/GRANT NO.
12. SPONSORING AGENCY NAME AND ADDRESS
Environmental Sciences Research Laboratory - RTP, NC,
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711 '
13. TYPE OF REPORT AND PERIOD COVERED
Final 8/76 - 7/77
14. SPONSORING AGENCY CODE
EPA/600/09
15. SUPPLEMENTARY NOTES
16. ABSTRACT
An evaluation was conducted to determine the capabilities and limitations of
a commercially available monitoring system that provides sulfur dioxide mass emission
rate data as a direct output. The monitoring system was operated continuously for
extended periods at a coal-fired power plant and a sulfuric acid production facility.
Additional testing was performed at a Simulated Stationary Source Facility to confirm
some deficiencies noted during field operations. The system's performance was
verified by comparing its output data with results using EPA emissions measurement
reference methods.
Results are presented for three performance tests at both field sites. For
the power plant tests, the monitor agreed within 20% of the accepted reference
method. In the case of the acid plant, the system accuracy was as poor at 58%.
Generally, the monitoring system performed reliably throughout the extended test
program. The system remained operational greater than 90% of the time during the
four-month test period.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS C. COSATI Field/Group
*Air pollution
*Sulfur dioxide
*Mass
^Emission
*Monitors
*Evaluation
13B
07B
19. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19. SECURITY CLASS (ThisReport)
IINf.l ASSTFTFH
21. NO. OF PAGES
37
20. SECURITY CLASS (Thispage)
UNCLASSIFIED
22. PRICE
EPA Form 2220-1 (9-73)
29
------- |