EPA-600/4-77-022
April 1977
Environmental Monitoring Series
SURVEY OF CONTINUOUS SOURCE EMISSION
MONITORS: Survey No. 1
NOX and S02
Environmental Monitoring and Support Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, 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 MONITORING series.
This series describes research conducted to develop new or improved methods
and instrumentation for the identification and quantification of environmental
pollutants at the lowest conceivably significant concentrations. It also includes
studies to determine the ambient concentrations of pollutants in the environment
and/or the variance of pollutants as a function of time or meteorological factors.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/4-77-022
April 1977
SURVEY OF CONTINUOUS SOURCE
EMISSION MONITORS: Survey No. 1
NOX and S02
by
Michael C. Osborne and M. Rodney Midgett
Quality Assurance Branch
Environmental Monitoring and Support Laboratory
Research Triangle Park, North Carolina 27711
ENVIRONMENTAL MONITORING AND SUPPORT LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
RESEARCH TRIANGLE PARK, NORTH CAROLINA 27711
-------
DISCLAIMER
This report has been reviewed by the Environmental Monitoring and
Support Laboratory, U.S. Environmental Protection Agency, and approved for
publication. Mention of trade names or commercial products does not con-
stitute endorsement or recommendation for use.
ii
-------
CONTENTS
Tables iv
Acknowledgment v
1. Introduction 1
History 1
Purpose of This Survey 2
2. Survey Preparation 5
Industries and Pollutants 5
The Survey Reference Materials 6
Data Requirements 8
Finding Monitors for the Survey 11
Monitor Descriptions 11
3. Survey Procedures 17
Scope of Activities .... 17
Conducting the Field Measurements 19
4. Results 21
Monitor Performance 21
Monitor History 25
5. Discussion of Results 31
Monitor Performance 31
Monitor History 33
6. Conclusions and Recommendations 34
References 36
Technical Report Data and Abstract 37
m
-------
TABLES
Number Page
1 SPNSS Continuous Monitoring Requirements 5
2 Monitoring System History 10
3 Continuous Emission Monitor Survey Results 22, 23,
24
iv
-------
ACKNOWLEDGMENTS
The authors wish to acknowledge with appreciation the assistance
of Mr. Berne I. Bennett and his co-workers in the Quality Assurance
Branch for their assistance in analyzing the large number of gas cylinders
used in this survey. Appreciation is also extended to each of the survey
participants who cooperated in every aspect of the survey. Their names
cannot be mentioned here for the reasons stated in the report.
-------
SECTION 1
INTRODUCTION
HISTORY
On September 11, 1974, the U.S. Environmental Protection Agency publish-
ed in the Federal Register the proposed Standards of Performance for New
Stationary Sources (SPNSS) and State Implementation Plans (SIP) requirements
for continuous emission monitoring. On October 16, 1974, a separate continu-
2
ous emission monitoring method, Method 12, was proposed for primary copper,
zinc and lead smelters. A slightly revised version of the proposed SPNSS and
SIP requirements for continuous emission monitoring was promulgated on Octo-
3 4
ber 6, 1975. In January of 1976, Method 12 was withdrawn, and smelters
were added to the list of industries required, under SPNSS, to adhere to the
monitoring specifications established on October 6, 1975. Therefore, the
most significant regulatory document on continuous emission monitoring is the
o
Federal Register of October 6, 1975, which lists system performance specifi-
cations, operating procedures, data reduction, and reporting requirements.
The new, or existing, sources that have been, or are being, required
to install, calibrate, maintain, and operate continuous gaseous emission
monitors are, initially, from five basic industries:
- Fossil Fuel-Fired Steam Generators
- Nitric Ac-id Plants
- Sulfuric Acid Plants
- Petroleum Refineries
-------
- Primary Copper, Zinc, and Lead Smelters
At present, the only one of these industries in which monitors may be
used to determine compliance with the standards of performance is the smelter
industry.
Although federal regulations requiring continuous emission monitors are
fairly new, the use of these monitors in the aforementioned industries goes
back several years. Some of the monitors included in this survey were in-
stalled 6 or 7 years ago, while the majority went into operation since 1973.
Very few monitors that are on-line today have actually gone through a
4
performance test as outlined in the October 6, 1975 Federal Register. There
are two basic reasons for this. The first and primary reason is that most
monitors in plant operation today are on processes that are not classified as
new or modified sources according to SPNSS specifications. Instruments in-
staTi_d prior to October 6, 1975, are required to under go the Performance
Test only at the request of the administrator. The second reason is that
much time, effort, and money is required to conduct one of these 168-hour per-
formance tests. Reports from the few performance tests which have been con-
ducted indicate that most source samplers do not clearly understand many of
4
the details outlined in the October 6, 1975 Federal Register for conducting
a performance test. Therefore, few industries want to be the test case for a
consultant to learn how to conduct a performance test. As more experience is
obtained in performance testing and as more pressure is placed on the indus-
tries to conduct the tests, a significant number of these tests may be ex-
pected.
PURPOSE OF THIS SURVEY
In most cases, when a survey is conducted, insufficient data is the
-------
reason for the survey. The survey of continuous source emission monitors is
no exception. Although several different types of source emission monitors
had been evaluated by the EPA, prior to this survey, no continuing program of
emission monitor evaluation could be found within or outside the Agency. The
concern within the Quality Assurance Branch (QAB) of the Environmental Moni-
toring and Support Laboratory (EMSL) was that the monitors that were meeting
initial design specifications were possibly failing to perform satisfactorily
after months of continuous in-plant usage. Some question existed as to
whether continuous emission monitoring technology had progressed far enough
to develop a long-term maintainable system. Complaints about monitor failure,
particularly from electric utility companies, tended to lend credibility to
these questions. Were the industries expected to obtain valid monitoring
data from instrumentation that could not be expected to adequately perform
without excessive maintenance problems?
Another area of concern was which monitors worked well over extended
periods and which did not. The EPA policy on continuous source emission
monitoring has been, and still is, to provide basic performance requirements
for the continuous monitoring systems and leave it at that. No specific
monitors will be recommended, nor will EPA recommend a particular method or
mechanism over any other. Yet, the Quality Assurance Branch holds to the
position that long-term operation data on specific continuous monitors can,
and should, be made public knowledge. This practice will be followed without
officially recommending any product or method. Also, the names of the speci-
fic industries where this information was obtained will not be mentioned.
Their participation in this survey was totally voluntary and these industries
should, in no way, be penalized for cooperating.
-------
Additional and more extensive surveys will be required to obtain compre-
hensive answers to the questions: (1) Is emission monitoring technology
sufficiently advanced to provide a long-term maintainable monitoring system
for industry? and (2) How well do specific continuous emission monitors per-
form over an extended period of time? Therefore, this report should only be
considered a preliminary investigation and should be coordinated with sub-
sequent survey reports to provide a more complete answer to these questions.
-------
SECTION 2
SURVEY PREPARATION
INDUSTRIES AND POLLUTANTS
The industries which were considered for in-plant continuous emission
monitor evaluation are the same as those listed in the introduction as having
to meet the October 6, 1975, monitoring specifications. Table 1 lists these
industries and the pollutants which must be continuously monitored.
TABLE 1. SPNSS CONTINUOUS GASEOUS MONITORING REQUIREMENTS
(Industry) (Pollutant)
Fossil Fuel-Fired Steam Generators S02, NO, 02 or C02
Nitric Acid Plants N02
Sulfuric Acid Plants S02
Petroleum Refineries S02
Smelters: Primary Copper, Zinc, Lead S02
After a review of the list of possible industries and pollutants which
could be included in the initial survey, it was concluded that a survey in-
volving all of these industries and all of these pollutants would be too
extensive for a first effort in this subject area. There were too many un-
knowns with regard to possible survey tools, number of monitors available,
access to the plants and the monitors, and the cost of such a survey. The
problem became one of.properly reducing the number of industries and pollut-
ants so that the maximum beneficial information could be obtained without
exceeding time and budgetary allotments or current technical capabilities.
5
-------
In the process of making this decision about which industries and pollut-
ants were to be included, several significant considerations became apparent:
1) A satisfactory survey tool to evaluate the performance of opacity monitors
has not been adequately tested; 2) The number of available nitric acid plants
with continuous emission monitors is severely limited; 3) Op and COo emissions
are only indirectly related to other pollutant emissions in contrast to S02
and NO emissions, which are primary pollutants; 4) More monitors could be
A
readily located in power plants and sulfuric acid plants than in the other
listed industries; 5) A much lower range of SOp survey gases is required in
petroleum refineries than could be used in either power plants or sulfuric
acid plants; and 6) Smelters were of special interest since, in this industry,
the monitor is used to determine compliance with emission standards during
the performance test.
With these considerations in mind, the decision was made to limit the
initial survey to Fossil Fuel-Fired Steam Generators, Sulfuric Acid Plants,
and Smelters. The pollutants to be surveyed were limited to sulfur dioxide
(S09) and nitrogen oxides (NO ). The other industries and pollutants were to
£ X
be added in future surveys as time and resources became available, enabling
them to be included.
THE SURVEY REFERENCE MATERIALS
One additional limitation on the first survey was that only extractive
monitors were included. Ten to twenty percent of the S09 and NO monitors
£ /\
encountered when setting-up this survey were in-si'tu units. The in-situ
units could not be checked with calibration gases like the extractive units,
and calibration cells were not readily available to test the in-situ units.
-------
Since the beginning of the survey, Lear Siegler, a manufacturer of in-situ
monitors, has developed a way of introducing calibration gases into the probe
and around the sensor so that calibration cells would not be necessary.
Future surveys of S07 and NO monitors will include both extractive and in-
Cf 3\
situ monitors, using gas cylinders when possible and calibration cells on all
other units.
The October 6, 1975, Federal Register clearly states that extractive
monitors must be checked by zero and span gases, i.e., gas cylinders. There-
fore, the logical reference material for continuous extractive emission moni-
tors is cylinder gases. Some may think that the best reference would be EPA
Test Methods No. 6 and 7. It is generally agreed that these are reference
methods; however, it has been proven through collaborative testing that these
methods have within-laboratory standard deviations of about 4.0 and 6.6 per-
cent, respectively. At the 95 percent confidence level, this variability
would, of course, be considerably greater. On the other hand, gas cylinders
are readily certified by the gas vendors to be accurate within 2 percent, and,
with NBS certified or traceable calibration gases as back-up standards, the
concentrations may be further verified. The practicality of using gas cylin-
ders instead of the reference methods was another major consideration. Cylin-
der gases were, therefore, used as the reference.
Four cylinders of S02 in nitrogen, having a concentration of 100, 250,
600, and 850 ppm,* were purchased. These cylinders were then sent to the
*EPA policy is to express all measurements in Agency documents in metric units.
When implementing this practice would result in undue cost or lack of clarity,
ERC/RTP is providing conversion factors for the particular nonmetric units used
in the document. For this report the factors are: 1 ppm SO = 0.376 mg/m.
Jn
1 ppm NO = 0.802 mg/n
-------
National Bureau of Standards (NBS) for analysis at the beginning of the survey
and again at the end. The cylinders were used to calibrate a continuous
source SC^ monitor, which was then used to measure S02 in 25 survey cylinders.
In this way, the survey cylinders were traceable to the cylinders analyzed by
the NBS.
Four cylinders of nitric oxide (NO) in nitrogen were purchased directly
from the NBS. These were Standard Reference Materials (SRMs) at the concen-
trations of 100, 250, 500, and 1000 ppm. As in the above case, the SRMs were
used to calibrate an NO source monitor. Seventy-five NO survey cylinders
were then analyzed on the source monitor, thereby once again establishing
traceability.
All of the cylinders were guaranteed stable for a period of 12 months
from the date of shipment from the vendor's plant. The cylinders were made
of troj»ted alutiinum, fully pressurized with 30 cubic feet of gas per cylin-
der and weighed less than 25 pounds per cylinder.
DATA REQUIREMENTS
The information obtained from each survey participant was in two
general categories: 1) measured cylinder concentration data and 2) monitor
history. The concentration data were simply the monitor reading and equiv-
alent concentration in parts per million for each survey cylinder gas that
was introduced into the monitor. At least three different pollutant concen-
trations were introduced into each monitor, yielding at least three differ-
ent data points for each monitor.
The information classified as monitor history was considerably more
extensive. Answering the question, "How well do continuous emission monitors
perform over an extended period of time?" required delving into the more
8
-------
involved questions of "Why do monitors fail to perform?" and "What types of
problems do monitor users encounter?" To obtain answers to these questions
a "Monitor System History" questionnaire (Table 2) was prepared.
On the day of the field testing, the questionnaire was filled out by
the individuals in the plant who were most familiar with the monitoring
equipment. Most of the questions were intended to show not only whether the
monitors performed satisfactorily, but also to pinpoint why the user was or
was not pleased with the operation of the monitor.
-------
TABLE 2. MONITORING SYSTEM HISTORY
1. Monitor description, including model number.
2. Date that the monitor became operational at its present location.
3. Dates and descriptions of modifications to the monitoring system.
4. How satisfied are you with the services provided by the equipment supplier?
5. How satisfied are you with the overall operation of the monitoring system?
6. What is the normal maintenance schedule for this equipment?
7- For each time that the monitoring system was inoperative during the past
6 months, list the date, number of hours, and reasons:
Date No. of Hrs. Reason
8. What special maintenance problems have you had during the past six months
with:
a) the probe
b) the delivery system
c) the monitor
d) the recorder
9. What was the span of the instrument on the day of the test?
10. What was the concentration range of the emissions on the day of the test?
10
-------
FINDING MONITORS FOR THE SURVEY
The perfect continuous emission monitor survey would seek out a sample
of X number of identical monitors from each of Y number of different monitor
designs. Unfortunately, the survey covered in this report was hampered from
following the perfect experimental design. The major obstacle was locating
enough S09 or NO extractive continuous emission monitors to conduct a survey.
£ A
An initial list of nine utilities, sulfuric acid plants, and smelters with
some type or types of S09 or NO monitors was submitted by the Emission
c. x
Measurement Branch of the Office of Air Quality Planning and Standards. Each
of these nine plants was contacted, and the planned monitor survey was dis-
cussed. Of the nine plants on the original OAQPS-suggested list, only two
finally participated in this first survey. Reasons for nonparticipation in-
cluded: 1) The monitors are not operating at this time, 2) A compressed gas
cannot be readily introduced into the monitor, 3) The monitor has been
returned to the manufacturer for repairs, 4) The monitors have been turned
off until data collected earlier may be evaluated, and 5) Plant employees
are not allowed to talk with EPA representatives at this time. Since two
plants were an insufficient number of participants for the survey, suggest-
ions for additional participants were sought through correspondence with
instrument suppliers, EPA Regional Offices, State air pollution control
agencies, and through the industries themselves. About 20 companies were
contacted to obtain the 8 different companies that were finally included in
the survey. Fortunately, some of the companies that participated had several
monitors among a number of plants.
MONITOR DESCRIPTIONS
A total of 19 monitors were included in this initial survey. This
11
-------
included 6 NO or NOX monitors and 13 S02 monitors. There were four different
types of NO or NO monitors surveyed -- Dynascience, MSA, Envirometrics, and
/\
Thermo Electron. S02 monitors were also of four different types - DuPont,
Leeds and Northrop, Dynascience, and Thermo Electron.
The only reasons that these monitors were selected over others were that
they met the requirement of being extractive NO or S09 emission monitors and
A £
that they were in operation in the plants of the companies participating in
the survey.
The following description of each monitor covered by the survey should
provide an insight into the differences that existed between the monitors
surveyed. Literature from the monitor venders is the primary source of this
information. Some difficulty in comparing system performance specifications
may be encountered because each vender reported on different specifications.
DuPont nude! 400 Photometric Analyzer (SOo)
This photometric monitor operates in the UV-visible range at a "measur-
ing" wavelength of 302 nm and a reference wavelength of 546 nm. Changes in
S02 concentration cause the difference in absorption between 302 nm and 546
nm to vary. The light signal is detected by the monitor and displayed as
changes in SOo concentration. The manufacturer claims that high linearity
permits the use of widely different pollutant ranges with one analyzer. Each
S02 monitor contains an optical filter that mimics the light absorption
characteristics of SOo and is used for daily calibration, once the filter is
calibrated against a known source. DuPont offers the option of purchasing
just the basic Model 400 Analyzer - or a complete custom-engineered system
that includes the Model 400 and a sampling system. Some system specifications
include:
12
-------
- Precision: +_ 1% full scale
- Accuracy: + 2% full scale
- Zero Drift: 1% full scale in 24-hr for most applications
- Sample Temperature Range: up to 1000°F (S38°C)
Leeds and Northrop 7864 Infrared Analyzer (SOp)
Essentially, the unit is the combination of two simple infrared analyz-
ers. One senses all of the infrared-absorbed gases in the sample; the other
senses all but the gas you want to measure, i.e. S02- The range of the
particular instrument tested was 0 to 10,000 ppm. Calibration gases can be
used; however, a three-way solenoid valve for selection of sample or cali-
brating gas must be specified as a separate option. Similarly, sample con-
ditioning equipment must be either ordered separately, ordered as a standard
sampling system, or otherwise provided by the user. Some system specifica-
tions include:
- Accuracy: + 1% of span or 30 ppm (whichever is greater)
- Reproducibility: +_ 1.5% of span (24-hr) or 30 ppm (whichever is
greater)
- Zero-drift: +_ .7% of span (24-hr) or 20 ppm (whichever is
greater)
- Analyzer Temperature: Maximum of 140°F (46°C), thermostated
Dynascience -PSC-01, Self Contained Dual Monitor and Sample Aquisition System
for NOX/S02
The Dual Monitoring Model utilizes the EPD Selectro Cells. These electro-
chemical transducers convert the gas concentrations into electric current out-
puts for simultaneous readings of the two pollutants being measured. Typical
concentration ranges of the instrument go up to 2000 ppm. Calibration of the
13
-------
system with compressed gases is standard with the option of automatic zero
and span calibration for unattended use. A complete "turnkey" system capable
of continuously acquiring, conditioning, analyzing, and recording pollutant
concentrations car, be provided. Some system specifications include:
- Sensitivity: +_ 2% of full scale
- Repeatability: +_ 2% minimum
- Zero Drift: +_ 2% of full scale
- Operating Temperature Range: 40-43°C (Temperature controlled
at 43°C)
Thermo Electron SO,, Pulsed Fluorescent Analyzer - Model 40
In this analyzer, a gas sample is submitted to a source of pulsed ultra-
violet illumination through a monochomatic filter. S02 molecules, energized
by the high-intensity pulsed-light source, emit an S02 specific illumination,
which, through a narrow band filter, impinges upon the sensitive surface of a
photomultiplier tube. The emitted light is linearly proportional to the con-
centration of SOp molecules in the sample.
Five standard sensitivity ranges are available, going up to 5000 ppm.
Calibration gases are used for calibration purposes, but care must be taken to
note the percentage of quenchable compounds in the compressed gas cylinder. A
separate sample conditioning and sample collection system is available from
the manufacturer. Some system specifications include:
- Accuracy: +_ 0.5%
- Reproducibility: +0.5%
- Zero Drift: + 1% (7 days)
- Operating Temperature Extremes: 0-40°C
14
-------
Environmetrics Series N-76 Analyzer (NO )
/\
Each monitor utilizes FARISTOR cartridge sensors to monitor NOX over
various sensitivity ranges. Inside the FARISTOR is a liquid-state device,
containing a chemically sensitive activating surface layer upon which pollut-
ant molecules are strongly absorbed by catalytic action. This action results
in a change of oxidation state, producing a charged surface. The magnitude
of the charge is determined by the rate at which the gas molecules reach the
activating surface, this rate in turn being directly proportional to the
pollutant concentration. Typical operating ranges for the monitor extend
upward to 10,000 ppm. The system calibrates against a gas of known concen-
tration in a nitrogen carrier gas. Sampling and conditioning systems must be
obtained separately. System specifications are said to meet or exceed exist-
ing EPA standards. This statement was made in 1972, when there were no
existing EPA standards on continuous emission monitors.
Thermo Electron Series 10A NO-NO Chemiluminescent Analyzer
A
This chemiluminescent analyzer includes an NO-NO converter for the con-
y\
version of NO to NO for subsequent measurement via the chemiluminescent
A
process. The instrument has eight different concentration ranges all the way
up to 10,000 ppm. Regular calibration of the monitor requires the use of
calibration gas cylinders. As with the TECO S02 analyzers, a separate sample
conditioning and sample collection system is available from the manufacturer.
Some system specifications include:
- Accuracy: +_ 0.5%
- Reproducibility: +0.5%
- Zero Stability: +_ 1 ppm (24-tirs)
- Operating Temperature Extremes: 0-40°C
15
-------
MSA Lira Infrared Analyzer - Model 202 (NO)
This analyzer operates on the principal of nondispersive infrared (NDIR)
detection. The NDIR analysis method is used on the hypothesis that most
nitrogen oxides (NO ) emitted from power plants are in the form of nitric
/\
oxide (NO). Typical analytical ranges are 0-2000 ppm NO. Calibration
selector values for using calibration gases are built into the system. The
complete sampling acquisition and conditioning system consists of a specially
designed 5-micron shielded filter probe assembly for installation in the stack
and a heat-traced insulated Teflon or stainless steel sample line condition-
ing system. Some system specifications include:
- Accuracy: 1.0%
- Repeatability: 1.0%
- Zero Drift: 1.0% (24-hr)
- Operating Temperatures: 0-48°C
16
-------
SECTION 3
SURVEY PROCEDURES
SCOPE OF ACTIVITIES
In the complete picture of what is involved in conducting a survey of
continuous emission monitors, the actual field measurements make up only a
small part. The following is the chronological order for this first survey
and should be looked upon as the cycle which each additional survey is likely
to follow.
1. Identified companies with monitors
This step was described earlier, including the difficulties which
were encountered.
2. Initial contact with a knowledgable employee
Few plant managers knew whether their monitors were extractive, the
make of the monitor, or the operating range of the monitor. An individual
who worked directly with the instrument on a day-to-day basis was contacted.
3. Sent letters requesting the participation of the company in the
survey
Official permission to perform tests in a given industry almost
always required a written request.
4. Initial calibration of survey cylinders
This step needed to be taken within a few weeks of the actual field
measurements to minimize chances of degradation taking place.
5. Receipt of confirmation of willingness to participate
The survey was held up until this letter was received.
17
-------
6. Second conversation with plant contact
At this point, information was obtained concerning the span of the
instrument, an agreeable testing date, and a confirmed mailing address for
the gas cylinders.
7. Shipped cylinders
Available means of shipping compressed gas cylinders was limited.
While S02 cylinders could be sent through the U.S. Mail, NO cylinders were
limited to only a few commercial trucking companies. Even then, delays were
common because the cylinders could not be placed on trucks hauling clothing.
8. Mailed final notification of visit
To be certain that no misunderstandings existed between the surveyor
and the industry, the plant manager or supervisor received a final confirma-
tion of the date of the proposed visit and the extent of the planned testing.
The letter was received approximately a week before the actual field testing,
to ensure that it was not forgotten.
9. Determined if cylinders had arrived
The surveyor did not go into the field until he was certain that all
necessary equipment was on site.
10. Conducted the field measurements
This step will be expanded upon later.
11. Returned cylinders and re-analyzed concentrations
If the cylinders' concentrations are re-analyzed quickly, there is
little opportunity for the cylinders to degrade.
12. Supplied participating industries with tentative results
The courtesy of letting the industries know, without delay, how they
performed has created a willingness on their part to participate in the next
18
-------
survey. It was often possible to provide the industries with this information
during the actual field testing. The concentrations did not change signifi-
cantly between pre- and post-test laboratory evaluations.
13. Combined The results of the independent surveys and published the
data
Information was compiled about both instrument accuracy, based on
field measurements, and instrument maintenance problems, based on the monitor
history questionnaire.
Conducting the Field Measurements
Each company has its own security procedures and it is, therefore,
difficult to generalize about them. In most cases, however, it was found
that the best approach was to proceed to the front gate and ask to see our
principal contact. In some small plants this was the plant manager, or the
local plant supervisor, but in most cases our contact was the instrument man
or a plant engineer.
After a discussion of the planned testing, the next task was to locate
the shipped gas cylinders. After the gas cylinders were taken to the monitor
location and the proper regulators attached, a determination was then made as
to where in the sampling line to introduce the gas. Ideally, the gas should
be introduced immediately after the sampling probe so that the majority of
the sampling and conditioning system would also be evaluated. It was
observed in this survey, however, that most installations were not prepared
for introducing the gas immediately after the probe, so a less suitable site,
such as just prior to the monitor, was usually selected. A wide range of
stainless steel fittings and flexible tubing was available to ensure a leak-
free connection to the sampling line.
19
-------
At least three different concentrations of reference gas were introduced
at each monitor, and the instrument reading was recorded in each case. No
attempt was made to calibrate or zero the instrument before introducing the
gas. Care had to be taken with some monitors to prevent over-pressurizing
the sample lines. After obtaining the readings from the monitor for the
various gases, the concentration of the stack gases being recorded on the
strip chart or other recorder was noted. After the gas line was disconnected
and the cylinders were re-sealed for return shipment, the appropriate plant
personnel were asked to complete the monitor history forms. Usually, there
was limited information available to document items on the survey form. As
a result, most interviews lasted only a few minutes.
The last task was the shipment, or arranging the shipment, of the
cylinders back to the laboratory for re-analysis. Leaving this task to the
plant personnel would have significantly delayed the completion of the
project.
20
-------
SECTION 4
RESULTS
MONITOR PERFORMANCE
Table 3 shows the survey results from introducing three or more survey
gases into each continuous emission monitor. Names of participating compan-
ies have been omitted for reasons stated previously. In order to be able to
identify individual sources when evaluating the data, however, a number and
letter have been designated for each participating company. The letters used
are £, A, and S^ corresponding to power plants, sulfuric acid plants, and
smelters, respectively. The number in front of the letter designates which
power plant, etc. In some cases, more than one of the same type of monitor
were evaluated at a given site. When this happened, the monitors were also
numbered.
A comparison of actual vs. measured concentration is included as the
basic evaluation criteria. This value is called the percent accuracy; i.e.,
% accuracy = concentration measured - actual concentration
actual concentration x 100,
where the "actual concentration" is defined as the value established by the
Quality Assurance Branch.
The average percent accuracy that is included in the results is the
absolute average, giving no consideration to the sign of the individual
values. One exception to the procedure is the results from the survey of the
TECO NO monitor. The reason for this exception is explained in the section
A
on "Discussion of Results."
21
-------
TABLE 3.
Pollutant Plant Monitor Cylinder
NO IP Dynascience FF 3152
FF 3264
FF 3275
NO 2P MSA FF 1121
FF 1138
FF 1106
NO 3P Envirometrics FF 1226
x FF 1192
FF 1108
NOV 3P TECO (No. 1) FF 1226
x FF 1192
FF 1232
NO 3P TECO (No. 2) FF 1226
x FF 1192
FF 1232
NO 3P TECO (No. 3) FF 1226
x FF 1192
FF 1232
SO, 3P DuPont FF 1073
* FF 1176
FF 1065
Concen.
measured @
plant ppm
520
1120
1640
240
400
420
215
350
370
235
390
425
245
388
392
235
390
400
140
385
630
Actual
concen
ppm
471
975
1459
236
404
406
246
395
406
246
395
407
246
395
407
246
395
407
150
367
782
V
h
Accuracy
+9.9
+14.9
+15.7
AVE: 13.5
+0.8
-1.0
+1.9
AVE: 1.2
-12.6
-11.4
+8.9
AVE: 11.0
-4.5
-1.3
+4.4
AVE: 3.4
-0.4
-1.8
-3.7
AVE: 2.0
-4.5
-1.3
-1.7
AVE: 2.5
-6.7
+4.9
-19.4
AVE: 10.3
22
-------
TABLE 3.
Pollutant Plant Monitor Cylinder
S09 1A Leeds and FF 3234
Northrop FF 3259
FF 3248
S09 1A DuPont FF 3234
c FF 3259
FF 3248
SO- IS DuPont FF 1487
* (No. 1) FF 1385
FF 1458
S0? IS DuPont FF 1487
(No. 2) FF 1385
FF 1458
S07 IP Dynascience FF 1184
* FF 1086
FF 1494
S02 2A TECO F 312*
F 126*
i
FF 1166
FF 1258
FF 1175
Concen.
measured @
plant ppm
2150
4450
6150
2600
5300
7750
6,400
10,900
15,800
4,950
9,700
14,600
130
330
727
108
340
193
408
427
Actual
concen
ppm
2440
5180
7750
2440
5180
7750
5,560
10,800
16,300
5.650
10,800
16,300
151
364
805
147
453
158
362
378
%
Accuracy
-11.9
-14.1
-20.6
AVE: 15.5
+6.6
+2.3
0
AVE: 3.0
+13.3
+0.9
-3.1
AVE: 5.8
-14.4
-10.2
-10.4
AVE: 11,0
-13.9
-9.3
-9.6
AVE: 10.9
-26.5
-24.9
AVE. 25.7
+22.2
+12.7
+13.0
AVE. 16.0
TABLE 3 (continued)
23
-------
TABLE 3.
Pollutant Plant Monitor
SO- 4P DuPont
* (No. 1)
S09 4P DuPont
* (No. 2)
S09 4P DuPont
* (No. 3)
S0? 4P DuPont
* (No. 4)
SO, 4P DuPont
* (No. 5)
S02 3A DuPont
Cylinder
FF 1079
FF 1186
FF 1269
FF 3234
FF 1079
FF 1186
FF 1269
FF 3234
FF 1079
FF 1186
FF 1269
FF 3234
FF 1079
FF 1186
FF 1269
FF 3234
FF 1079
FF 1186
FF 1269
FF 3234
FF 1258
FF 1175
F 126*
F 314*
F 162*
Concen.
measured @
plant ppm
792
816
808
2252
720
736
756
2188
756
764
792
2284
732
752
772
2192
844
848
860
2332
355
400
485
645
770
Actual
concen
ppm
759
778
804
2440
759
778
804
2440
759
778
804
2440
759
778
804
2440
759
778
804
2440
362
378
453
615
760
%
Accuracy
+4.3
+4.9
+0.5
-7.7
AVE: 4.3
-5.1
-5.4
-6.0
-10.3
AVE: 6.7
-0.4
-1.8
-1.5
-6.4
AVE: 2.5
-3.6
-3.3
-4.0
-10.2
AVE: 5.3
+11.2
+9.0
+7.0
-4.4
AVE: 7.9
-1.9
+5.8
+7.1
+4.8
+1.3
AVE: 4.2
*denotes air is the carrier gas instead of nitrogen.
TABLE 3 (continued)
24
-------
MONITOR HISTORY
In trying to establish the overall reliability of the monitors surveyed,
investigators used the series of questions in Table 2 to develop a record of
the history of the monitoring system. Some of the participating industries
had not maintained enough records on their monitoring equipment to answer all
of the questions. Therefore, information which would have been helpful in
evaluating some of the monitors was not available. Specifically, background
information on the Envirometrics (N02) monitor could not be obtained from the
company that owned the monitor.
The following is a question-by-question summary of the information
obtained by use of the monitoring-system history questionnaire:
1. Monitor description including model number. The answers to this
question are in the earlier sub-section on "Monitor Descriptions".
2. Date that the monitor became operational at its present location.
A summary of the years in which the monitors became operational is listed be-
low:
Year 1970 1971 1972 1973 1974 1975
No. of Monitors 130833
3. Dates and descriptions of modifications to the monitoring system.
The dates when modifications were made to the monitoring systems were not
available at most plants that were visited. Several significant modifications
were made, however, and are listed below:
Monitor Modifications
4P-DuPont No. 1-5 (S0«) 1. Added a new design particulate filter.
2. Shortened probe.
3. Increased sample line from 1/4 inch to 5/8
inch
25
-------
Monitor Modifications
4. Added a fuse to prevent shorting out in 220-
volt heat traced line.
5. Made probe of 316 stainless instead of 304.
4P-DuPont No. 1-3 (802) 1. Added switch to purge system during downtime.
IS-DuPont No. 1 & 2 (S02) 1. Added temperature gauge at the monitor.
2. Devised a filtering system to keep out
moisture and particulates.
2P-MSA (NO) 1. Replaced sample lines.
IP-Dynasciences (NOX & S02) 1. Changed from a "glued-together" sample
conditioner to a "screwed-together" unit.
a
2. Replaced a Teflon piece in the diaphragm pump
with a rubber one.
3. Changed blow back valves from a 10-minute
cycle to a 20-minute cycle.
4. Replaced Teflon seats with nylon seats and an
0-ring.
2A-TECO (S02) 1. Began making their own particulate filters.
3P-TECO (NO ) 1. Have added heat tracing to the sampling line.
J\
and DuPont (1) (S02)
4. How satisfied are you with the services provided by the equipment
supplier? There were no complaints registered with regard to any of the equip-
ment suppliers.
5. How satisfied are you with the overall operation of the monitoring
system? Most operators were pleased with the monitoring system as a whole.
However, three companies (2A-TECO, 2P-MSA, and 4P-DuPont (1-5) indicated that
26
-------
there had been.unreasonably poor performance of the pumps, sampling lines, and
probes.
6. What is the normal maintenance schedule for this equipment? Responses
to this question are listed below:
Monitor Maintenance Schedule
IP-Dynascience (NOX and S02) 1. Calibrate once per week with calibration
gases.
2. Check flow and battery daily.
3. Clean SCU scrubbers once every two weeks
to one month.
4. Rebuild pumps and compressors twice a
year.
5. Valve maintenance 3-4 times per year.
1. Daily calibration with calibration gases.
1. Cleaning of capillary tubes and change
oil in the vacuum pump once every 3 months.
1. No maintenance schedule available yet.
1. Check out system once per week.
2. Calibration check once per month or soon-
er with optical filters (DuPont).
3. Clean lenses once per month.
1. Calibrate each morning with optical
filter.
2. Particulate filter change once per week.
1. Daily zero and calibration check with
calibration gases. If error exists, then
maintain.
2P-MSA (NO)
3P-TECOO-3) NO
A
3P-DuPont (S02)
lA-Leeds and Northrop and
DuPont (S02)
IS-DuPont (2)(S02)
2A-TECO (S02)
27
-------
Monitor Maintenance Schedule
3A-DuPont (S02) 1. Check it 3 times a day with optical
filter.
4P-DuPont (1-5)(S02) 1. Observed daily.
2. Weekly check with optical filter.
3. Complete calibration monthly with cali-
bration gases.
7. For each time that the monitoring system was inoperative during the
past 6 months, list the date, number of hours, and the reason. This was the
most difficult question on which to obtain answers. The purpose of this
question was to determine the availability of the instrument, exclusive of
plant shut-downs or other times when the instrument was not needed. A rough
estimate of the amount of time each monitor was non-functional during the 6-
month period prior to the survey follows:
Monitor Approximate Time Non-Functional Over 6 Mos.
IP-Dynasciences (SCL & NO ) - less than 3 weeks
Cm. A
2P-MSA (NO) - 10 days
3P-DuPont (S0?) & TECO (1)(NOJ - 6 weeks
£ t\
3P-TECO (2&3)(NO ) - functioned continuously
y\
4P-DuPont (1-5)(S02) - 2-4 weeks each
lA-DuPont & Leeds & Northrop (S02) - less than 1 day each
2A-TECO (S02) - 1 week
3A-DuPont (S02) - less than 1 day
IS-DuPont (1)(S02) - 2 days
-DuPont (2)(S02) - 1 day
28
-------
8. What special maintenance problems have you had during the past 6-
months with the probe, the delivery system, the monitor, or the recorder?
Plants reporting problems with clogged probes include:
- 2P - MSA (NO)
- 4P - DuPont (1-5)(S02)
- 1A - DuPont (S02)
- 2A - TECO (S02)
- 3A - DuPont (S02)
- IS - DuPont (1 & 2)(S02)
Problems with the delivery system were of three types: clogging, condensation,
and leaks. Those plants reporting problems with clogging were:
- 4P DuPont (1-5)(S02)
- 2A TECO (S02)
Condensation problems were reported by:
- 2P MSA (NO)
- 3P DuPont (S02)
- 3P TECO (NOJ
X
- IS DuPont (1 & 2)(S02)
The problem of leakage was reported by:
- IP Dynasciences (NO & S09)
/\ £
Problems that were reported with the monitor, itself, included:
4P DuPont (1-5)(S02) - 1) cracked housing on solenoid
2) dirty optics
3) defective diaphragm on vacuum breakers
IS DuPont (1 & 2)(S02) - 1) fluorides etch lenses
1A Leeds and Northrop (S02) 1) cracked lenses
2) electronic problem with power supply
29
-------
2A TECO (S02) 1) dirty optics
3A DuPont (S02) 1) dirty optics
Recorders were often not a part of the purchased monitor package and
were also generally not a problem.
9. and 10. What were the span of the instrument and the concentration range
of the emissions on the day of the test? The purpose of these questions was
to place the survey gases in perspective with the operating range of the
instrument, the concentration at which it was calibrated, and the normal
operating level of the instrument. The results of the questions are as
fol1ows:
Monitor
IP Dynascience (NO )
A
IP Dynascience (S02)
2P MSA (NO)
3P TECO (1-3) (NOJ
^
3P DuPont (S02)
4P DuPont (1-5) (S02)
1A DuPont & Leeds and
Northrop (S02)
2A TECO (S02)
3A DuPont (S02)
IS DuPont (1&2) (S02)
Calibration
Check ppm
not known
not known
888
900
not known
3233
5000
300
2300
9800
Full scale
ppm
600
1500
1000
1000
600
4000
10000
500
5000
20000
Concen. range
ppm
200-400
300-600
80
200-400
0
1200-2600
1500-2000
20-30
1700
4500-9000
30
-------
SECTION 5
DISCUSSION OF RESULTS
MONITOR PERFORMANCE
Before a discussion of how the monitors performed as a whole, it is
necessary to clarify the results of the survey of the TECO (S02) monitor at
plant 2A. As was noted in the section on "Monitor Descriptions", the TECO
($02) monitor is a pulsed fluorescence instrument which is sensitive to
quenchable compounds in the survey gas. We were aware of this problem when
the survey was conducted and, therefore, used both cylinders of S02 in nitro-
gen and SOp in normal dry air (20% 02). The results of the survey show that
the monitor consistently read the S02 in air cylinder concentrations low and
the S02 in nitrogen cylinder concentrations high. An analysis of the sulfuric
acid plant stack gas indicated that there was 8.3 percent oxygen present.
Assuming that the amount of quenching was directly proportional to the oxygen
concentration of the stack gas, an estimate of the average percent accuracy
based on the survey results would indicate that the TECO monitor was reading
low by only 1.7 percent. This calculation is obviously questionable since
additional testing would be required to measure how much oxygen quenching
really occurred. The approximation obtained by ratioing the oxygen present
gives, however, a much better representation of how well the monitor was
functioning than either the test performed with S02 in nitrogen or the test
performed with S02 in air. Further evaluation of this type of monitor, using
more representative survey gases, is anticipated during our next performance
31
-------
survey.
Another point which should be considered before any gross generaliza-
tions about the findings of the survey are made, relates to the concentration
level of the survey gases. In some of the installations, the monitor range
encountered at the site was much higher than anticipated. Therefore, the
concentration levels that were introduced were occasionally on the lower end
of the monitor concentration read-out range. Since few monitors are cali-
brated specifically for the low end of the scale, some of the monitors in the
survey may seem to perform worse than they would if we had always used survey
gases which were at least mid-range. Closer attention will be paid to this
point in future surveys.
In general, the monitoring systems surveyed performed better than had
been expected on the basis of earlier reports from industrial monitor users.
Nine of the 19 monitors surveyed had an average percent accuracy of less than
5 percent. This is exceptionally good, considering the age of most of the
monitors and the limited regular maintenance which they had been given. Only
six of the 19 monitors failed to come within 10 percent of the actual concen-
trations, and the worst one of these was out by only 15.5 percent. Therefore,
there were no really poor monitor performances; only some that might be
classified marginal. Without making reference to any specific monitor, the
data clearly show that certain monitoring systems were out-performing others.
Insufficient data are available to determine if the larger percent accuracy
was caused by failure within the instrument or by problems with the cylinder
gases used to calibrate the instrument. No specially good or bad performance
was observed in any one of the three industrial categories. It should be
noted, though, that at least one monitoring system worked exceptionally well
32
-------
in each of the three industrial categories.
MONITOR HISTORY
Each individual monitoring system that was evaluated in the survey was
found to have had one or more maintenance problems. The significant finding
was that the probe, the delivery system, and the monitor were all consistent-
ly contributing to these maintenance problems. The most prominent problem
experienced by the majority of operators was clogging of either the probe or
the sample lines. The second greatest problem was related to the first, in
that condensation in the sample line often is the cause of, or at least con-
tributes to, the clogging.
Several problems were listed for some installations. This fact may,
however, be more closely related to the disposition of the operator who
provided this information than to the quality of the monitor.
In general, most participants were at least satisfied with the overall
operation of their monitor, and, as far as could be determined, no one was
displeased with the services of the equipment supplier.
Several monitoring systems experienced a significant amount of downtime
during the 6 months prior to the survey. This fact was partially due to the
low priority attached to keeping the monitoring system going, compared to
other types of instrumentation in the plant. Since downtime was usually an
estimate, it may have been over-exaggerated. Until plants start keeping a
record of this information, it will be impossible to obtain good monitor
availability data.
33
-------
SECTION 6
CONCLUSIONS AND RECOMMENDATIONS
Most of the survey participants were pleased to learn that information
on the reliability and accuracy of continuous emission monitors was being
gathered for public distribution. Such information can only enhance quality
control in this area of instrumentation.
The initial survey produced few surprises. Several general conclusions
from the survey were, however, possible.
1) Continuous emission monitors can perform exceptionally well, and
some did in this study.
2) Some types of continuous emission monitors consistently perform
better than others.
3) The industrial category in which the monitor is placed has no effect
on how well or how poorly the monitor performs.
4) The probe, the delivery system, and the monitor all contributed to
regular maintenance problems.
5) The two most prominent maintenance problems encountered were
clogging of the probe and sample lines and condensation.
Recommendations for future surveys included:
1) Expand the survey to include:
a. more different types of monitors
b. in-situ monitors
c. monitors in nitric acid plants, petroleum refineries, and other
34
-------
types of smelters
d. 662, ^2* anc' opacity monitors
e. monitors that have passed the 168-hour performance test
2) Repeat the survey on the same 19 monitors at regular intervals.
3) Use survey gases that are at the 50 percent and 90 percent of the
full-scale concentration levels that are suggested in the October 6, 1975,
4
Federal Register for calibration purposes.
4) Conduct evaluations of the pulsed fluorescence instrument at oxygen
and C02 levels that are equal to those in the stack gas.
5) Introduce the survey gas at the probe exit whenever possible.
35
-------
REFERENCES
1. U.S. Environmental Protection Agency. Standards of Performance for New
Stationary Sources. Federal Register, 39:32852-32860, September 11, 1974.
2. U.S. Environmental Protection Agency. Standards of Performance for New
Stationary Sources. Federal Register, 39:37040-37049, October 16, 1974.
3. U.S. Environmental Protection Agency. Standards of Performance for New
Stationary Sources. Federal Register, 40:46250-46270, October 6, 1975.
4. U.S. Environmental Protection Agency. Standards of Performance for New
Stationary Sources. Federal Register, 41:2332-2341, January 15, 1976.
5. Hamil, H. F., and D. Camann. Collaborative Study of EPA Methods 5, 6,
and 7 in Fossil Fuel-Fired Steam Generators: Final Report. U.S. Environ-
mental Protection Agency Technical Report No. EPA-650/4-74-013. May 1974.
(Available from National Technical Information Service, 5285 Port Royal
Road, Springfield, Virginia 22161.)
36
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/4-77-022
2.
3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE'
SURVEY OF CONTINUOUS SOURCE EMISSION MONITORS: SURVEY
NO. 1 - NOY AND S09.
/\ £.
5. REPORT DATE
April 1977 issuing date
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
8. PERFORMING ORGANIZATION REPORT NO.
Michael C. Osborne and M. Rodney Midgett
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Quality Assurance Branch
Environmental Monitoring and Support Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, N.C. 27711
10. PROGRAM ELEMENT NO.
1HD621
11. CONTRACT/GRANT NO.
12. SPONSORING AGENCY NAME AND ADDRESS
Environmental Monitoring and Support Laboratory-RTP, NC
Office of Research and Development
U.S. Environmental Protection Aqency
Research Triangle Park. NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
Final
14. SPONSORING AGENCY CODE
EPA/600/08
15. SUPPLEMENTARY NOTES
16. ABSTRACT
The Quality Assurance Branch (QAB) of the Environmental Monitoring and Support
Laboratory has undertaken the task of evaluating the accuracy and reliability of
continuous monitors which have been on-line for six months or longer. Only extractive
monitors installed in power plants, sulfuric acid plants, and smelters were included
in the first survey. The program was designed so that additional industries and types
of monitors could be included in later surveys.
The plan for the initial survey involved sending NBS-traceable gas cylinders of
varying concentrations of SO, or NO to the participating industries. The gases from
the cylinders were introduced into the monitoring systems and the resulting data made
available to both the QAB and the industry.
Results from this survey of eight different companies, ten different plant sites,
and nineteen different monitors revealed that less than half of the monitors failed to
achieve a performance accuracy of + 10 percent. The instruments which failed most
often were the ones that have seenHess wide-spread use. In general, the S02 monitors
and the NO monitors performed equally well. However, more information was available
on S0_ monitors since they outnumbered NO monitors in this survey almost two to one.
£ 3\
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b. IDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Monitors
Sampling
Sulfur dioxide
Nitrogen oxide
Air pollution
Chimneys or Emission
Continuous emission
monitors
13B
18. DISTRIBUTION STATEMENT
Release to Public
19. SECURITY CLASS (ThisReport)
Unclassified
21. NO. OF PAGES
42
20. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
37
------- |