vvEPA
United States
Environmental Protection
Agency
Environmental Sciences Research
Laboratory
Research Triangle Park NC 2771 1
EPA-600/2-78-178
August 1978
Research and Development
Automatic Interfacing
System for Sampling
Total Mercury in
Stationary Source
Emissions
-------
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-600/2-78-178
August 1978
AUTOMATIC INTERFACING SYSTEM FOR
MERCURY IN STATIONARY SOURCE EMISSIONS
by
D.J. Sibbett and T.R. Quinn
Geomet, Inc.
2814 A Metropolitan PI.
Pomona, Ca. 91767
Contract No. 68-02-1789
Project Officer
Roy L. Bennett
Emissions Measurement and Characterization Division
Environmental Sciences Research Laboratory
Research Triangle Park, North Carolina 27711
ENVIRONMENTAL SCIENCES RESEARCH 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 Sciences Research
Laboratory, U.S. Environmental Protection Agency and approved for publication.
Approval does not signify that the contents necessarily reflect the views and
policies of the U.S. Environmental Protection Agency, nor does mention of
trade names or commercial products constitute endorsement or recommendation
for use.
fi
-------
ABSTRACT
This program was initiated with the objective of developing an automatic
instrumental interface which will permit the determination of low levels of
mercury vapor while simultaneously removing relatively high levels of inter-
ferences from stationary emissions.
The system as designed, fabricated and tested included sample conditioner,
dilutor and pump modules. The conditioner decomposes mercury compounds and
scrubs out particulates and interfering gases. The diluter adds cleaned air
to the sample to adjust elemental mercury vapor concentrations to levels which
can be monitored with a DuPont Photometric Analyzer. The pump module draws
sample through the system and maintains a constant gas pressure in the photo-
meter.
Laboratory tests under conditions of high contamination confirmed the
utility of the system. Utilizing resuspended fly ash at levels from 8.7 to
15.4 g/m3 and 220 to 2175 PPM S02, replicate tests were conducted in the labor-
atory at mercury levels ranging from 0.038 to 1.70 mg/m3. Removal of particu-
lates and S02 was quantitative. No loss of mercury vapor was detected.
Field tests were conducted at Unit No. 2 of the John E. Amos Power Genera-
tion Plant of American Electric Service Corporation at St. Albans, W. Va., a
coal-fired steam generating plant. The sampling probe assembly was installed
through a port between the electrostatic precipitator and the 900 foot (high)
stack in a duct (6.1 x 61 meters). Sample drawn at 22 1pm through a glass
lined, SS probe were divided between bubblers and the interface-photometer
system. In the majority of comparative tests run, the bubbler and instrumental
procedures agreed very well in determination of the amounts of mercury in the
exhaust. Values ranged from 1.7 to 7.0 jug/m3 by the instrumental system; the
range was 1.7 to 7.3 jug/m3 by the bubbler method. The detection limit was
0.4-0.5 ug/m3. Performance by the instrumental system was stable and reliable.
This report was submitted in fulfillment of Contract No. 68-02-1789 by
Geomet, Incorporated under the sponsorship of the U.S. Environmental Protection
Agency. The report covers the period from 5 May 1975 to 30 April 1976.
iii
-------
CONTENTS
Disclaimer ii
Abstract iii
Figures vi
Tables vii
Acknowledgment viii
1. Introduction 1
2. Conclusions 2
3. Recommendations 4
4. Description of Automatic Interface 6
General Description 6
5. Tests of the Interface System 12
Laboratory Tests 12
Field Tests . . 34
Appendices
A. Operating Instructions 52
B. Sampling Procedures 84
Analytical Procedure 86
-------
FIGURES
Number Page
1 Interface System Showing Components
of Modules 5
2 Conditioning Module 7
3 The Diluter Module 9
4 The Pump Module 11
5 Calibration of Du Pont 400 Photometric Analyzer
(X10 Scale) 13
6 Schematic Diagram of Assembly Full System
Tests in Laboratory 14
7 Full System Performance 18
8 Particulate Test Assembly . 25
9 Particulate Test Assembly 26
10 Sulfur Dioxide Test Assembly 30
11 Combined Sulfur Dioxide and Particulate
Test Assembly 33
12 Schematic Diagram of Sampling Port Location 35
13 Probe Alignment 36
14 Detailed Schematic of Measurement System 38
VI
-------
TABLES
Number Page
—-^———-—•^* ^ • i ••••
1 Full System Tests 15
2 Test of Oiluter 22
3 Particulate Collection 24
4 Particulate Removal During Conditioner/
Prefilter Tests 28
5 Sulfur Dioxide Removal During Conditioner Tests 31
6 Removal of Sulfur Dioxide and Particulates
by Prefilter Conditioner 32
7 Tests of Automated Interface Instrumentation 41
8 Calibration and Direct Bubbler Measurements 43
9 Definition of Terms 49
Vll
-------
ACKNOWLEDGEMENTS
We wish to acknowledge support from the Environmental Sciences Research
Laboratory, U.S. Environmental Protection Agency. This program was conducted
under the direction of Dr. Roy L. Bennett, Project Officer. Discussions with
Dr. Bennett were of considerable assistance during the program.
We also wish to acknowledge the assistance rendered by the American
Electric Power Service Corporation in permitting us to test our instrumentation
at the J.E. Amos Power Generation Plant in St. Albans, W. Va. In particular,
the assistance of Mr. T.T. Frankenberg and the plant management is gratefully
noted.
vm
-------
SECTION 1
INTRODUCTION
The need for techniques to measure relatively low levels of mercury vapor
in stationary source emissions presents a unique problem in removal of potential
interferences from sample streams. Generally, each of the interferences is
present at concentrations which may be as high as one million times that of the
mercury vapor. An example of an industrial exhaust gas contained: 1-6 jug/m3 of
mercury in the presence of 0.5-1.0 g/m3 particulates, approximately 1.0-1.6 g/m3
fo sulfur dioxide and up to 1.1 g/m3 of oxides of nitrogen (as N02). In order
to measure the mercury vapor concentration by a photometric technique, the
levels of these potential interferences must be reduced to low values ;
The technical objective of this project was to develop a relatively
ticated automatic interfacing system which will sample total mercury in source
streams and suitably condition, dilute and transport the sample to a mercury
measuring instrument.
Subtasks in development of the interface system included:
1. Design and construction of interface instrumentation so that a repre-
sentative total mercury sample will be obtained from the source stream. This
system shall provide for conditioning the sample and for dynamic dilution to
reduce high mercury concentrations which are encountered at some sources to
levels within the calibration range of the measuring instrument. Conditioning
includes the conversion of all mercury compounds to an elemental state and the
reduction of gaseous and particulate interferences to levels which do not inter
fere with the photometric procedure.
2. Test of the system in the laboratory and the field. The field test
shall be conducted utilizing a source having high particulate loading and
high sulfur dioxide concentrations.
3. Evaluation and interpretation of laboratory and field data including
estimation of the following factors: accuracy, precision, sensitivity, stabi-
lity, response time, interferences and reliability.
4. Delivery of the interface system to EPA at the end of the study toget-
her with complete operating procedures.
-------
SECTION 2
CONCLUSIONS
An interface system consisting of three main modules has been designed,
fabricated and tested. It consists of a conditioner, a diluter and a pump
module. These components may be assembled into several sample processing
configurations depending on the requirements of the source sample. All test-
ing of the interface in the laboratory and the field was carried out in
conjunction with a DuPont 400 Photometric Analyzer.
The conditioner contains two main functions, a furnace for heating the
gases and a liquid scrubber. The furnace is designed so that mercury com-
pounds may be thermally decomposed at temperatures to 1000°C, if required.
It is usually operated at 300-400 C. The unit also contains a liquid scrub-
bing system for removal of particulates and interfering gases such as sulfur
dioxide and nitrogen dioxide. The scrubber utilizes concurrent processing
of the sample gas flowing at 2-2.5 liters/minute with solutions of sodium
bicarbonate or sodium and ammonium bicarbonate usually pumped at the rate of
5 ml/min. through a twelve foot coil. This procedure removes all measurable
traces of S02 and N02.and>97% of particulates even at particle loadings in
the range of 10-12 g/m3. At very high particle loadings a heated prefilter
may also be added. These three functions may be used separately or in combin-
ation.
The diluter module which regulates sample flow to the DuPont Photometer
can be operated In two modes: (1) to dilute the sample with clean air by any
ratio from 1/5000 to 1/19, or (2) without any dilution. Gas flow regulation
is accomplished by use of two Hastings mass flowmeters. One unit controls
the total flow at rates to 5 liters per minute. The second establishes the
fraction of the sample gas which will be analyzed.
The pump module draws the sample gas through the system, exhausts the
waste liquid into a reservoir from the scrubber and maintains a constant
pressure in the photometric analyzer. This latter feature is required for
reproducible measurements.
Laboratory testing of the full interface system was carried out at parti-
culate levels ranging from 8.7-15.4 g/m3 of fly-ash and at 220 PPM sulfur
dioxide. This level of fly-ash necessitated the use of a heated prefilter to
achieve complete removal of the interference from particulates. The furnace
was operated at 400°C in these tests; the scrubber utilized 0.2M NaHCOa cir-
culated at a rate of 5 ml/minute. Tests were replicated at mercury concentra-
tions varying from 0.038 to 1.70 mg/m3.
-------
These tests demonstrated a number of operating features of the interface
system:
(1) Particulates at high concentrations can be reduced to undetectable
limits by the prefilter and scrubber components. The scrubber alone
can remove >97% of particulates even at very high loadings. By
removing a major portion of the particulates in a prefilter, the
scrubber-prefliter system achieves 100% efficiency.
(2) Sulfur dioxide was removed with nearly 100% efficiency by the
scrubber.
(3) Mercury losses were undetectable.
cv
In tests with dimethyl mercury and with the furnace operating at 400 C
the system quantitatively yielded elemental mercury. No mercury losses were
detected.
Tests of the diluter operating with dilution ratios between 1/21 and 1/110
showed that dilutions can be controlled readily by use of the two flowmeter
control console. At the lower dilution ratios, experimental measurements of
diluted concentrations agreed with the values set on the controls within - 2%;
at the highest dilutions the agreement was * 41.
Field tests of the interface system were carried out at Unit No. 2 of the
J.E. Amos Power Generation Plant of the American Electric Service Corporation
at St. Albans, W. Va. The test system was positioned on top of a duct between
the electrostatic precipitators and the 900 foot stack. Operating conditions
within the duct were: (1) Temperature, 280-320°F; (2) Air Velocity, 26.8-35.8
m/sec; (3) Gas Flow Rate, 1000-1300 m3/sec; (4) The estimated exhaust composi-
tion was NOx^SOO PPM, S02 400-600 PPM, particulates 0.5-1.0 g/m3.
A heated, six-foot, glass lined, stainless steel probe was extended 41.9
cm into the duct at a position 1.9 meters from the duct wall. Gas was drawn
from the duct at 22 liters/minute but only 2.5 liters/minute was passed to the
instrumentation. A zero baseline for the system was established by passing
air through a small bed of silver-alumina tablets. Analyses were obtained
utilizing the full interface system in conjunction with the DuPont Photometric
Analyzer operating at the 253.7 nm wave length and by use of the conventional
bubbler collectors followed by analysis by the standard reduction procedure.
An end-to-end calibration procedure was also employed. The two sets of data
were compared in 37 sets of measurements.
A good correlation between the instrumented analyses obtained by use of
the interface and the DuPont Photometric Analyzer and the reference manual
procedure was obtained. Concentrations of mercury vapor in the exhaust gases
were low. Instrumented results ranged from 1.74 to 6.96 ug/m3 with a mean
value of 4.23 ug/m3. Reference method results varied from 1.60 to 7.25 ug/m3
with a mean result of 4.66 ug/m3. These results demonstrate that the interface
system rejects the interfering exhaust components adequately allowing mercury
vapor to pass through the system for photometric analysis.
-------
SECTION 3
RECOMMENDATIONS
The automated interface system has been fabricated and tested in a variety
of configurations. Its performance has been proven under relatively stringent
field conditions. However, some beneficial modifications may be visualized.
The primary area in which improvements seem required is related to field
calibration of the system during monitoring operations. For example, during
the tests at the J.E. Amos Power Generation Plant a method for end-to-end
configuration of the system calibration was established. The procedure was
relatively time-consuming to carry out and required close attention from the
operator. In essence, the procedure required (1) establishment of a clean
air baseline signal, (2) introduction of a volumetric charge of mercury vapor
by syringe immediately behind the probe and (3) determination of system re-
sponse by the instrumental and bubbler procedures. As part of this procedure
the three way valve shown in Figure 1 , page 5 is adjusted to draw ambient air
through the silver-alumina tablet bed and thence through the photometer to
establish a baseline response. In continuous monitoring operations, this pro-
cedure might be automatically controlled and programmed to take place for a
short interval, such as five minutes, during every two hours. This would
confirm the consistent performance of all components of the monitoring system.
An easily manipulated reference mercury source which can be handled in
awkward physical situations would be highly desirable. Preferably, the source
should operate to introduce a precalibrated mercury vapor sample into the inlet
to the interface. ,This calibration standard would remove the necessity for
time-consuming measurements utilizing the reference* bubbler collection techni-
que. Its introduction might also be programmed and automated.
For operations in which sulfur trioxide is present, it would be desirable
to establish a technique for easy replacement of the filter placed immediately,
downstream from the conditioner. This filter removes sulfuric acid mist aero-
sol which otherwise interferes with, the operation of the Photometric Analyzer.
A housing containing more than a single filter is needed so that sample gas
streams may be readily switched between in-place filters. This change might
also be automated in continuous operations.
These improvements would complement the system in a wide variety of moni-
toring applications.
* Federal Register, Volume 38 INo. 66), pp. 8820-8850, April 6, 1973
-------
N/A.CAJUM
AMBIENT D\ \JLTn ON
Figure 1. Interface System showing components of modules.
-------
SECTION 4
DESCRIPTION OF AUTOMATIC INTERFACE
GENERAL DESCRIPTION
The Automated Mercury Interface instruments are designed to assist spec-
trophotometric determinations of the mercury content of gas streams. Normally,
a spectrophotometer cannot utilize samples taken directly from a source because
of the presence of interferences. The Automated Interface which is interposed
between the source probe and a photometer serves to remove the common interfer-
ences of particulate matter and sulfur and nitrogen oxides. It also automati-
cally regulates the sample gas flow to the optical cell, and if the mercury
content is too high, the interface system is used to quantitatively dilute
the sample with mercury-free air.
The interface system is composed of three main modules. These are the
conditioning module, the diluter module, and the pump module. As it is com-
posed of three units it is adaptable to the testing requirements of various
sources. The three units may be connected in several modes or only two of the
three units may be used to meet different testing requirements. The Interface
System is also portable because of its unitized design. It permits testing of
a considerable number of sources for mercury by photometric measurements, at
the 253.7 nm absorption peak.
Figure 1 is a block diagram which, illustrates a characteristic mode of
operation of the interface system when used in conjunction with a DuPont
400 Photometric Analyzer. The 10% silver/alumina pellets (lower right) are
used to scrub mercury from the ambient atmosphere or for comparative measure-
ments when interferences are present. The principles involved are described
below.
Conditioning Module
The conditioning module which is shown photographically in Figure 2 pre-
pares the gas for photometric measurements. Its functions are shown in the
left part of Figure 1, a schematic of the entire Interface System. The condi-
tioner is responsible for the conversion of mercury compounds to elemental
mercury and for the removal of interferences such as sulfur dioxide and parti-
culate matter.
The initial process: After passing through the intake line, the sample
gas enters the furnace located in the conditioning module,. The furnace which
may be set to automatically hold temperatures up to 1000 C is responsible for
-------
Figure 2. Conditioning Module
-------
the decomposition of mercury compounds. This section Is usually operated in
the range 300-400°C. The sample is heated as ft passes through a coil of 10
feet of stainless steel tubing (0.635 mm, i.d. x 0.953 mm, o.d.). This heating
decomposes mercury compounds to the elemental state. This process may be by-
passed if it is not needed.
The scrubber process: After exiting from the furnace, the sample gas goes
through a cooling tube on the outside of the unit and enters a glass scrubber
coil. (The gas sample may enter the conditioner untt at this point if the
decomposition process is not required). In the coil which is made of 12 ft.
of pyrex glass tubing (5mm, i.d. x 7mm, o.d.), the gas is mixed with, a solution
of 80% 0.2 M sodium bicarbonate and 20% of 0.2 M ammonium bicarbonate which
also contains 2 ml of Triton X100 per liter. This solution is usually pumped
through the coil at a rate of 5 ml/min. The scrubbing operation which is
carried out by concurrent passage of the gas and liquid through the coil,
removes all measurable sulfur dioxide and>97% of the particulate matter. The
scrubber fluid is fed to the coil by a variable stroke piston pump. The flow
can be varied from~l to 7 ml/min. A check valve in the line prevents excess
fluid from being drawn into the coil by the vacuum. A 2 gallon container serves
as a reservoir for the scrubber fluid. The waste fluid is separated from the
gas in the waste separator. It passes with 0.5 1/min. of the sample gas to
a 2 gallon waste container. The two gallon containers allow a full 24 hours
between servicing. The processed sample gas, at a flow rate of 2 liters/min-
ute, exits from the module to a condenser where excess water is removed.
Gas flow through the conditioning module is controlled by the pump module
as shown in Figure 1. The pump module draws the gas into the conditioning
module at approximately 2.5 1/min.
In cases where mercury compounds are not of concern, the furnace may be
bypassed. This is easily accomplished by connecting the sample source to the
inlet to the scrubber section which is located on the right side of the module,
or a 12/5 socket joint may be fitted to the sample line (or probe) and then
connected by tubing to the inlet to the scrubber portion of the module. By
this method, the unheated gas goes directly into the scrubber coil without
passage through the furnace. This technique should be used in sampling oper-
ations where mercury compounds or vapors have been previously exposed to high
temperatures.
The Diluter Module
The diluter module which is shown in Figure 3 automatically regulates gas
flow to the photometer, and it may simultaneously be used to dilute high mer-
cury concentrations to levels applicable to the calibration range of the
analytical instrument. Gas flow regulation is accomplished using a 0.5 1/min.
single set point Hastings Mass Flowmeter and an automatic valve which is con-
trolled by the flowmeter. Any desired flow rate up to 5 1/min (* 1%) may be
obtained by placing the set point of the total flowmeter at the desired level.
The valve is automatically controlled to achieve the indicated total flow rate.
When dilution is required it is accomplished by using the lower flow range
(0-100 ml/min.) Hastings Mass Flowmeter and a second automatic valve to divert
8
-------
Figure 3. The Diluter Module
-------
a portion of the processed sample stream into a clean air stream. The scheme
is shown in Figure 1, center section. The indicator of the low flow range
meter (Meter No. 2) is used to select the desired flow rate of the sample
from the total stream. With this sensor in use, the two-way valve is open.
The three-way valve is open to the air from the mercury filter, but closed to
sample air. Excess sample is exhausted through the vacuum pump. In this
operation dilutions from 1/5000 to 100/1900 may be obtained. After dilution,
the gas is discharged from the rear of the unit to the photometer for analysis.
In the event that the diluter module is used for total flow control, rather
than dilution, the total sample passes through the three-way valve and enters
the flowmeter transducer of the high range valve (No. 1). It is here that the
measurement of the total flowrate is made. The flowrate indicated by the
meter on the front panel is set at the desired condition. Generally the flow
rate is 1.5-2.0 1/min. Once through the transducer, the gas passes through
the automatic servo valve controlled by the flowmeter, and then exits from the
unit, passing to the photometer.
The 10% Silver-Alumina Pellets
The 10% silver-alumina tablets (1/8" diam.) positioned between the diluter
and the photometer can be used during calibration of the system. After leaving
the diluter, the sample gas can either go directly to the photometer or through
the silver-alumina absorbent and then to the photometer. This absorbent is
used to remove mercury from the gas and to establish the zero or baseline. By
use of the two modes of operation two signals are obtained. The difference
between the signals is proportional to the amount of mercury in the sample.
This technique is useful in confirming system performance in a variety of tests.
The Pump Module
The pump module which is shown in Figure 4 draws the sample gas through
the system. It has three functions. It provides suction to the sample and
waste Tines, and it regulates the pressure in the photometer cell. This is
required for reproducible measurements. A magnehelic gauge indicates the
vacuum present in the system.
The pump module allows the spent scrubber fluid to be drawn from the waste
separator in the conditioner. A flowmeter on the module monitors the sample
gas flow which is drawn off with the scrubber fluid. A Thomas air pump provides
the vacuum for the waste and sample lines. The sample gas enters at the back
of the unit from the DuPont photometer and from the waste container, and it is
exhausted through the pump. Figure 1 shows the procedure schematically.
OPERATING INSTRUCTIONS
The detailed operating instructions for the system as established in the
tests as described in Section 5 are included in Appendix A.
10
-------
VACUUM
FBiP MODULE
Alt HOW
Figure 4. The Pump Module
-------
SECTION 5
TESTS OF THE INTERFACE SYSTEM
Prior to the field tests a series of development tests of the interface
system were conducted in the laboratory. These have been described as: (1)
Full System Tests, (2) Tests of the Diluter and (3) Tests of the Conditioner
Subsystem.
LABORATORY TESTS
Full System Tests
A series of repetitive tests were run with the full system including the
prefilter in order to establish: (1) whether mercury hold-up was negligible and
(2) whether particulates and sulfur dioxide were reduced to levels at which
interference with mercury determinations was eleminated.
Prior to commencement of these tests the DuPont 400 Analyzer was calibrated
using the mercury vapor generator and the EPA bubbler technique to establish
its response. Figure 5 shows a typical calibration. During these procedures
the mercury vapor generator was connected directly to the analyzer. As a
check on the retention of mercury in the conditioner-diluter, each series of
tests of the full interface system was preceded by a check of the calibration
utilizing a bypass around the conditioner-diluter.
The test assembly used in the full system tests is shown schematically in
Figure 6. The techniques involved in operating this assembly were identical
with those outlined in Appendix A.
Each test was carried out for thirty minutes. Sulfur dioxide was metered
into the system utilizing a needle valve and rotameter. No separate S02 analy-
ses were made. The fly-ash level was determined separately in five minute
determinations prior to each 30 minute test. The fly-ash utilized was a coal
ash which had been seived through a 200 mesh screen. The gas sampling rate was
2.5 liters/minute. The prefilter was heated to 400°C; the furnace temperature
was also 400°C.
Mercury vapor concentrations were measured before and after each test,
during a period when fly-ash and S02 were not added to the system as well as
during the test period when both potential interferences were present. The
data from six sets of runs replicated five times each are compiled in Table 1.
The mean values for each set of five tests are shown directly under each series
of replicates. This table is a record of the data which were obtained for
analyses of six levels of mercury in the presence of 11.8-12.1 g/m3 (mean values)
of fly-ash at an S02 concentration of 220 ppm. Mercury concentrations were
12
-------
Sensitivity: X10 Scale
0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0
Mercury Concentration
(mg/m3)
Figure 5. Calibration of Du Pont 400 Photometric Analyzer (X10 Scale),
13
-------
Air
rilter
[Pump
3-Way
Valve
Stirred
Particle
Source
Hg
Source
Pump
filter
Filter
S-wav
iPrefilter
I 3 -way
Valve
Rotameter
) Valve
SO2
Sourc
Sample
Conditionei
Air
Figure 6. Schematic diagram of assembly full system tests in laboratory.
-------
TABLE 1. FULL SYSTEM TESTS
cn
Test
No.
1
Z
3
4
5
Mean
6
7
8
9
10
Mean
11
12
13
14
15
Mean Re
Before
Test
0.90
0.85
0.83
0.87
0.80
0.85
3.60
3.55
3.85
3.80
3.95
3.75
6.80
7.05
7. 15
7.20
6.90
corder Valut
During
Test
0.93
0.83
0.83
0.85
0.83
0.85
3.70
3.60
3.80
3.70
3.95
3.75
6.90
7. 10
7. 10
7.20
6.95
ss (mv)
After
Test
0.90
0.83
0.83
0.87
0.80
0.85
3.70
3.55
3.70
3.65
3.95
3.71
6.90
7.05
7.05
7.30
6.90
Mean M
Before
Test
0.088
0.082
' 0. 080
0.085
0.078
0.083
0.362
0.358
0.395
0.385
0.403
0.381
0.880
0.960
0.995
1.015
0.905
ercury Cone
During
Test
0.090
0.080
0.080
0.082
0.080
0.082
0.376
0.362
0.385
0.376
0.403
0.380
0.905
0.975
0.975
1.015
0.925
(mg /m J )
After
Test
0.088
0.080
0.080
0.085
0.078
0.082
0.376
0.358
0.376
0.370
0.403
0.377
0.905
0.960
0.960
1.055
0.905
Hg Passed
thru
System
102.3
98.8
100.0
96.5
102.6
100.0
101.9
101. 1
99.7
99.6
100.0
100.5
100.3
101.6
99.7
98. 1
102.2
Parti culate
Level
(g/m3)
10.6
10. 1
12.9
11.7
13.8
11.8
12.2
13.3
13.2
11. 1
10.5
12. 1
11.3
15.4
12.7
13.6
8.7
S02
Cone.
(PPM)
220
220
220
220
220
220
220
220
220
220
220
220
220
220
220
220
220
System
Pressure
' (cm H20)
808
808
808
808
808
808
808
808
808
808
808
808
808
808
808
808
808
Mean
7.02
7.05
7.04
0.951
0.959
0.957
100.4
12.3
220
808
(continued)
-------
TABLE 1 (continued)
Teat
No.
16*
17*
18*
19*
20*
Mean
Zl
22
23
24
25
Mean
26
27
28
29
30
Before
Test
3.6
3.9
3.7
3.4
3.6
3.6(4)
2. 1
2.0
2.3
2.3
2.2
2.1(8)
8.3
8.2
8.5
8.0
8.2
During
Test
3. S
3.8
3.7
3.4
3.6
3.6(0)
2. 1
2.0
2.3
2.3"
2.2
2. 1(8)
8.3
8. 1
8.6
8.0
8.2
After
Test
3.5
3.8
3.7
3.4
3.6
3.6(0)
2.0
2.0
2.3
2.3
2. 1
2. 1(4)
8.3
8. I
8.5
8.0
8.2
Before
Test
0.038
0.042
0.040
0.036
0.038
0.038(8)
0.23
0.22
0.25
0.25
0.24
0.23(8)
1.76
1.68
1.87
1.52
1.68
Du r i ng
Test
0.037
0.041
0.040
0.036
0.038
0.038(4)
0.23
0.22
0.25
0.25
0.24
0.23(8)
1.76
1.60
1.94
1.52
1.68
After
Test
0.037
0.041
0.040
0.036
0.038
0.038(4)
0.22
0.22
0.25
0.25
0.23
0.23(6)
1.76
1.60
1.87
1.52
1.68
Mg massed
thru
System
98.7
98.8
100.0
100.0
100.0
99.5
102.2
100.0
100.0
100.0
102.1
100.9
100.0
97.6
103.7
100.0
100.0
f articulate
Level
(g/m3)
11.6
12.7
12. 1
13.3
11.9
12.3
13.1
12.3
13.4
11.6
10.5
12.2
10.9
11.8
13.1
12.2
12.5
0^2
Cone.
(PPM)
220
220
220
220
220
220
220
220
220
220
220
220
220
220
220
220
220
a/stem
Pressure
(cm H2O)
808
808
808
808
808
808
808
808
808
808
808
808
808
808
808
808
808
Mean
8. 2(4) 8. 2(4)
8.2(2) 1.70
1.70
1.69
100.3
12.1
220
808
Measurements on the XI scale; all others on X10 scale of Du Pont 400 Photometric Analyzer
-------
determined by use of calibrations of the DuPont 400 Photometer which were made
without use of the interface instrumentation.
The data for the thirty tests of the full system have been composited and
are shown in Figure 7. Each point represents a mean value obtained from five
tests. The fly-ash which was determined for each test, had a mean value of
12.1 g/m3 and a range from 8.7 to 15.4 g/m3. The sulfur dioxide was maintained
at 220 ppm; it was metered into the system by use of a rotameter.
Comparisons of the mercury levels obtained in tests with the complete
system with calibrations made without use of the interface instrumentation
show no significant differences. The curves are essentially identical. It
may be concluded that the technical objectives of the project have been achiev-
ed, at least when tested under laboratory conditions: (1) High levels of par-
ti culates and sulfur dioxide have been reduced by the interface instruments
to insignificant or zero levels; no interferences are obtained in the photo-
metric analysis of mercury vapor. (2) There is no evidence of hold-up of
mercury vapor by the interface instrumentation. (3) Within the precision of
the measurements, the analyses of mercury vapor by the system are identical
whether the original sample stream contains S02 and fly-ash or none of these
potential interferences.
Tests with Organic Mercury—
A comparison of peak heights obtained by introducing determined amounts
of dimethyl mercury and elemental mercury was used to establish the efficiency
of the prefilter and conditioner in converting dimethyl mercury to elemental
mercury.
With the furnace and prefilter* at 400°C, a calibration curve was obtained
with elemental mercury which was introduced at the prefilter inlet by means of
a syringe. These mercury samples collected by bubbler after passage through
the system. A KMn04/HNOs bubbler system was employed. The contents of each
bubbler was analyzed by reduction with 20% SnCl2 in 6 N HC1 and by determina-
tion of the resulting peak signal on a Perkin-Elmer atomic absorption
spectrophotometer.
For determinations of dimethyl mercury (DMM), a standard solution of DMM
in xylene was prepared volumetrtcally using a micropipet. Known amounts of
this solution were pipetted into the entry to the prefilter where it rapidly
evaporated. After correcting for the solvent blank, amounts of elemental
mercury obtained in the bubblers were determined.
The following data compare the experimentally determined amounts of mer-
cury with the calculated amounts introduced volumetrically as DMM in xylene.
* Rrefliter contains 60 g of oyrex wool.
17
-------
8.0
Mercury Concentration
(mg/m3)
Figure 7. Full System Performance.
Mercury Analysis in presence of Fly-Ash and Sulfur Dioxide
18
-------
Amount of Hg Experimental Determination
introduced as (CHsb Hg of Hg after Conversion
21 nanograms 21 nanograms
21 " 24
21 " 24
54 " 52
/
54 " 56
54 " 61
107 " 102
107 " 107
107 " 109 "
Each bubbler contained 0.5 ml of 0.25 KMnO^ 2.5 nl of 1:1 HNOs and
sufficient distilled water to make the volume up to 20 ml. Air flow was 2.5
1/min.
The data demonstrate that mercury introduced as dimethyl mercury is fully
recovered in the KMnO^yHNOg bubbler. Conversion of DMM to elemental mercury
appears complete. No loss of mercury in either form was noted.
Response Times--
Response time times were run by connecting and disconnecting the mercury
source stream immediately ahead of the pre-filter and determining the time
to response and clean-up which resulted. The average responses were as fol-
lows:
Time to Initial Response - 10 ± 5 seconds
Time to Peak Response - 50 + 9 seconds
Time to 90% Response - 46+6 seconds
Time to Clean-up from Peak Response (Fall Time) -66 + 6 seconds
Time to 90% Clean-uo -42 + 5 seconds
Detection Limits of System—
ExDermimental tests of the detection limit of the entire system were run
utilizing the mercury vapor generator. Initially, a steady low level output
of mercury was established with the generator. By adjusting the physical
parameters controlling the generator the signal was reduced until it was just
19
-------
detectable on the XI scale
to 1/5 of a chart division
disconnected from the prefi
baseline. Reconnection of
level mercury concentration
the KMn04/HN03 bubbler. R<
the XI scale show 0.4 ug/m
the photometer 4 ug/m3 was
Precision—
of the DuPont 400 Photometer. This is equivalent
on the recorder. For confirmation, the source was
Her allowing the signal to subside to the zero
the source established the detectability of the low
This mercury level was then determined by use of
plicate determinations of the detectable limit on
as the experimental limit. On the X10 scale of
detectable.
The precision of mercury vapor determinations may be estimated on the
basis of the standard deviations for each set of data included in Table 1.
Summarized these are:
Test
Nos.
16-20
1-5
21-25
6-10
11-15
26-30
Mean Mercury Cone. Std. Deviation
(mg/m3) (mg/m3)
0.038(4)
0.082(4)
0.23(8)
0.380(4)
0.959(0)
1.70
0.002(1)
0.004(3)
0.01(3)
0.015(1)
0.043(9)
0.16(1)
Mean C.V.
C.V.
(%)
5.5
5.2
5.5
4.0
4.6
9.5
5.7%
Thus, the estimated mean coefficient of variation (C.V.) for the six sets of
data is 5.7%.
Accuracy, Stability and Reliability—
The system calibration is based on comparison of the DuPont Analyzer
signals with standard bubbler determinations of mercury vapor, as quantitated
by the Hatch and Ott procedure. Accuracy estimates are therefore based on the
test values (Column 6 in Table 1) as compared with the mean values obtained
from the determinations before and after each test as tabulated in Columns 5
and 7. These latter values were determined by use of the by-pass around the
scrubber-diluter hardware.
For all thirty tests, determinations utilizing the full system agreed
with those obtained by use of the by-pass within ± 1.6%. That is, recovery
of mercury utilizing the interface was 100.3 ± 1.6% of that obtained without
its use. Since these data were obtained during a period of three weeks, the
stability and reliability of the interface system appears satisfactory.
20
-------
Tests of the Diluter
4
A series of tests of the diluter were made with system inputs identical
to the full system tests. That is, input particulate levels were 10-14 g/m3
and the sulfur dioxide concentration was maintained at 220 ppm. Mercury
vapor levels were varied from 0.94 to 1.87 mg/m3. A series of dilutions
ranging from 1/21 to 1/101 were tested. The data are shown in Table 2.
Comparison of the four right hand columns of Table 2 shows that the dilu-
ter yields the calculated concentrations of mercury vapor within a precision
varying from ± 2 to + 4%. Comparison of the measured mercury concentrations
in Column 6 with the calculated concentrations in Column 8 shows good agree-
ment. Calculated concentrations are obtained from input concentrations by
application of the dilution factor obtained from the diluter settings. Measur-
ed output concentrations are obtained from the recorder reading in millivolts
(Column 5) by use of a standard calibration.
It may be concluded that the diluter performs successfully and does not
serve to trap mercury vapor.
Tests of the Conditioner Subsystem
Two series of tests were utilized to determine the preferred configuration
of the conditioner: (1) the standard module consisting of the thermal conver-
ter and scrubber or (2) the same device to which a prefilter was added. At
high particle loadings, the latter configuration improves the operating perfor-
mance. Either modification may be used, depending on the properties of the
source under test.
Tests of the Standard Module—
The sample conditioner module was tested with coal fly-ash utilizing the
assembly shown schematically in Figure 6. The fly-ash was prepared by burning
commerically-available anthracite (10.2% solids as ash) and sieving the ash
through 200 mesh screens. About 250g of the sieved ash was placed in a two
gallon carboy where it was agitated with a 2 inch propeller rotating at 3500
rpm. Air drawn from the top of the carboy at 2-2.5 1pm contained 5-15 g/mj
of fly-ash. Although these concentrations are higher than the average exhaust
of most power plants, they serve as an adequate test condition in the labora-
tory.
Table 3 shows the results of a series of fly-ash removal efficiency tests.
The weight of ash collected on Filter No. 1 (Figure 9) during short periods
of operation with the 3-way valve diverting the stream through the filter,
served to quantitate the source level of particulates in the gas. The residue
deposited on Filter No. 2 established values for particulates which are not
removed by the process.
The procedure for determination of the ash on a filter included: (1) Each
Gelman Versapore filter was weighed and inserted in a 1 inch (diameter) holder.
(2) After completion of the test, the filters were dried for 20 min. at 110'C
and reweighed. The tare weight of clean filters was-'O.OGOg.
21
-------
TABLE 2. TEST OF DILUTER
Mercury Input
Concentration
Mean
R eading
(mvXIO)
6.8
6.8
7.0
7. 1
8.0
8. 1
8.2
8.3
8.5
6.8
6.8
7. 1
8.0
8. 1
8.2
8.3
8.5
6.8
6.8
7. 1
8.3
Cone.
(mg/m3)
0.94
0.94
1.03
1.07
1.54
1.60
1.67
1.74
1.87
0.94
0.94
1.07
1.54
1.60
1.67
1.74
1.87
0.94
0.94
1.07
1.74
Dilution
Volume Factor
Vol Sample
Vol Total
100/2100
M
"
"
tt
"
M
Tl
It
80/2080
II
It
II
tl
fl
1 1
"
60/2060
"
II
11
0. 0476
If
M
II
II
II
1 (
11
M
0. 0385
II
II
M
II
II
II
II
0.0291
II
11
II
Measurements
Mean
R eading
(mv,Xl)
Cone.
(mg/m3)
4.3
4.4
4.6
4.9
7.2
7.7
7.8
8.2
8.6
3.5
3.4
3.9
5.7
5.9
6.2
6.6
6.8
2.6
2.7
3.0
5.0
0.045
0.046
0.049
0.051
0.074(5)
0.079
0.080
0.085
0.089
0.036
0.036
0.041
0.059
0.061
0.064
0.068(5)
0.070(5)
0.027
0.028
0.031
0.052
Calculated
Mean
Reading
(mv,Xl)
4.3
4.3
4.6
4.8
7.0(5)
7.4
7.9
8.0
8.6
3.4(5)
3.4(5)
3.9(5)
5.7
5.9(5)
6.2
6.4(5)
6.9
2.6
2.6
3.0
4.9
Cone
(mg/m3)
0.0447
0.0447
0.0490
0.0509
0.0733
0.0762
0.0795
0.0828
0.0890
0.0362
0.0362
0.0412
0.0593
0.0616
0.0643
0.0670
0.0720
0.0274
0.0274
0.0311
0.0506
(continued)
22
-------
TABLE 2 (continued)
Mercury Input
Concentration
Dilution
Volume Factor
Out]
Measurements
Mean
Reading
(mvX 10)
8.0
8. 1
8.2
8.3
8.5
6.8
6.8
7. 1
8.0
8. 1
8.2
8.3
8.5
Cone.
(mg/m3)
1.54
1.60
1.67
1.74
1.87
0.94
0.94
1.07
1.54
1.60
1.67
1.74
1.87
Vol Sample
Vol Total
50/2050 0.0244
II 1)
11 11
t 1 II
11 II
40/2040 0.0196
II II
,, ,,
20/2020 0.00990
"
n ii
n ti
n n
Mean
R eading
(mv.Xl)
3.8
3.8
4.0
4.2
4.4
1.8
1.8
1.9
1.5
1.5
1.5
1. 7
1.9
Cone.
(mg/m3)
0.039(5)
0.039(5)
0.041(5)
0.044
0.046
0.019
0.019
0.020
0. 016
0.016
0.016
0.018
0.020
JUt
Calculated
Mean
Reading
(mv.Xl)
3.6(5)
3.8
3.9(5)
4. 1
4.4
1.8
1.8
2.0
1.4
1.5
1.5(5)
1.6(5)
1.7(5)
Cone.
(mg/m3)
0.0376
0.0390
0. 0407
0. 0425
0.0456
0.0184
0.0184
0.0210
0.0152
0.0158
0.0165
0.0172
0.0185
23
-------
TABLE 3. PARTICULATE COLLECTION
ro
Test
No.
1
2
.3
4
5
6
7
8
9
10
11
12
Collection Time
@ 2 1/m
No. 1
Filter
3 tmin
4
3
4
3
3
2
3
3
3
3
3
No. 2
Filter
Jjnin
15
15
15
15
15
18.75
141
15
12
15
30
Ash Collection
(R)
No. 1
Filter
0.0*54g
0.0419
0.0588
0.0973
0. 0307
0.0341
0.0285
0.0791
0. 0542
0.0951
0. 0437
0. 0285
No. 2
Filter
0. 0037g
0.0049
0.0051
0. 0066
0.0070
0.0072
0.0053
0. 0747
0.0111
0.0061
0. 0047
0.0053
Concentration
of participates
(g/m^)
Inlet
7.6g/m3
5.2
9.8
12.2
5.1
5.7
7.1
13.2
9.0
15.9
7.3
7.2
Exhaust
0. 21g/m3
0. 16
0.34
0.66
0.23
0.24
0. 14
0.26
0.37
0.25
0. 16
0. 15
Scrubber
Efficiency
97. 3%
97.0
96.9
94.9
95.7
96.0
98.1
98.1
96.1
98.5
97.9
98.0
Scrubber Operation
4. 4 ml/min of 0. 2M NaHCOs con-
taining 1 ml/1 of 10% Triton X100
5. 2 ml/min
6.0 ml/min
6. 0 ml/min
6. 5 ml/min
8. 5 ml/min
6. 5 ml/min
6. 5 ml/min
8. 5 ml/min
8. 5 ml/min 0. 2M NaHCO, +
3. 3 ml/I Triton X- 100
8. 5 ml/min 0. 2M NaHCO, +
7. 3 ml/1 Triton X- 100
8. 5 ml/min 0. 2M NaHCO, +
2 ml/1 Triton X-100 + 5.
-------
rsi
en
2.0 1/m
filter I
I Pump |
3-way Valve
Stirred
Participate
Source
&
No. 1
Filter
—a-
3-way Valve
Hg
Flask
Pump
0. 5 1/m
2.5 1/m
0.5 1/m
Figure 8. Paniculate test assembly.
-------
Air
ro
en
I
[Filter I
ILL_...__--"
[ Pump I ^^Three-way Valve
Stirred
Particulat
Source
Filter j
f
Gelman 5 fa.
Filter
Gelman 5
Filter
Figure 9. Particulate test assembly.
-------
x Tests 1-6 (Table 3) were conducted with the axis of the scrubber coil
positioned vertically. The tests averaged 96.3% removal of the ash. In
tests with the axis of the coil ti 1 ted ^20 * from the vertical, removal effici-
ency increased to 97.8%. In the latter position, the gas and liquid passed
as slugs through the coil. In tests 10 and 11 the detergent level caused
the formation of a froth in the separator. Thus some of the liquid was entra-
ined. The addition of salt together with a somewhat lowered level of Triton
X-100 resulted in improved mechanical performance in the separator but the
removal efficiency was not improved over Tests 7 and 8.
Since even the passage of~2% of the fly-ash would prove a considerable
operating problem during continuous operation, the effects of a filter between
the conditioner and the diluter were examined.
Tests of the Conditioner-Prefliter Subsystem--
Introduction of a prefliter immediately prior to the sample conditioner
simplifies the operation of the conditioner unit, which at high particle load-
ing levels, tends to become clogged. A series of heated prefilter configura-
tions were examined including: a ten inch section of cast iron pipe, an eight
inch glass trap and a vertical stainless tube designed for easy removal of
accumlated fly-ash. The latter design was selected. In use, it is hard
plumbed to the conditioner unit and the major portion of collected fly-ash may
be evacuated by removal of the plug attached at the bottom of the prefilter.
It is sized so that at loadings in the range of 10-15 g/m3 it must be emptied
once in each 24 hours of operation.
The test assembly used in examining the effects of particulates during
continuous mercury vapor analyses is shown schematically in Figure 8. In oper-
ation, the air containing mercury vapor and coal fly-ash from the stirred
particles generator was passed directly through the prefilter at 2.5 liters/
minute. The prefilter was maintained at 400°C by use of an external heating
tape. For determinations of particle levels, a by-pass containing a Gelman
5 ii filter was added around the prefilter.
All tests were conducted to establish possible losses of mercury in the
fly-ash. In general it was established that the prefilter temperature was
critical in avoidance of mercury losses. At prefilter temperatures below 300°C
some losses could be detected. At 400"C or higher, no measured losses were
observed. The operational technique involved establishment of a constant
mercury level through the system prior to introduction of fly-ash. This was
compared with the results obtained when fly-ash was added to'the system. Fly-
ash levels were determined by use of Gelman (5 u) 45 mm filters in front of
the prefilter and behind the sample conditioner to establish the particulate
removal efficiency. Deposits were quantitated by weight.
Data from a series of tests utilizing the system with prefilter are shown
in Table 4. The prefilter and conditioner temperatures were established at
400°C in all tests. The scrubber used 5 ml/min of 0.2M NaHCOs to which 1 ml/
liter of 10% Triton X-100 was added. In determinations of the quantities in
the table: Mercury levels are estimated to have a coefficient of variation
27
-------
TABLE 4. PARTICULATE REMOVAL DURING CONDITIONER/PREFILTER TESTS
ro
00
Test
No.
1
2
3
4
5
6
*
(a)
(b)
(a)
(b)
(a)
(b)
(a)
(b)
(a)
(b)
(a)
(b)
cv-2%
*
Mercury
Concentration
(wg/m3)
220
220
90
90
160
160
45
45
770
770
430
430
Particulate
Input
(g/m3)
7.3
0
6.5
0
14.8
0
10.6
0
9.6
0
12.8
0
Particulate
at Exit
(g/m3)
< 0.01
0
< 0.01
0
< 0.01
0
< 0.01
0
< 0.01
0
< 0.01
0
Particulate System
Removal Pressure
Efficiency (cm H2O)
100% 800
800
100% 800
800
100% 800
800
100% 800
14 800
100% 800
800
100% 800
800
Run
Duration
(min)
25
--
33
--
25
--
30
--
30
--
30
--
-------
(CV) of 2%. Participate concentrations have a CV of 6-101. However, an
absolute quantity of O.OOOZg of fly-ash is detectible on the filter behind
the sample conditioner. Thus, particle removal is essentially quantitative.
The data in Table 4 show highly satisfactory performance by the condi-
tioner-prefilter system in removal of particulates from coal fly-ash at high
loading levels.
Sulfur dioxide removal tests—The system with the prefilter was tested
for S02 removal and for mercury passage using the assembly shown schematically
in Figure 1-0. Initially, tests were run with the prefilter temperature at high
levels (700-800°C); in this case, a non-condensible vapor resembling H2S04
(probably $03) was generated. This vapor passed through the system to the
DuPont 400 Photometer where it interfered with the determination of mercury.
When the prefilter temperature was lowered to 400 C, the vapor disappeared as
did the interference with the mercury analysis. At temperature<300"C, some
mercury appeared to be lost in the prefilter.
Quantisation of sulfur dioxide removal was achieved by use of sodium
carbonate bubblers positioned on a by-pass around the prefilter and behind
the conditioner. Sulfur dioxide levels in these tests were achieved by use
of a tank source of S02 (Matheson) which was metered into the system by use
of a needle valve and rotameter. These are not shown in Figure 9.
Sulfur dioxide concentrations from 220-2175 ppm were tested against mer-
cury levels from 39 to 750 ;ug/m3. Table 5 shows that sulfur dioxide removal
was quantitative with the prefliter-conditioner system at all levels tested.
Removal of Particulates and Sulfur Dioxide by Prefliter-Conditioner—
Using a combination of the two test systems previously described, tests with
particulates and sulfur dioxide were carried out to determine whether mercury
vapor could be transmitted without losses. The schematic of the test assembly
is shown in Figure 11.
Input levels of fly-ash particulates and S02 were determined by use of a
by-pass. Downstream levels behind the sample conditioner were similarly placed
on a by-pass; no particles or S02 survived to this point in the test apparatus.
Tests were started by establishing and measuring each steady-state mercury
vapor level with the DuPont Photometer. Fly-ash and SO? were added to the gas
stream and Hg levels were redetermined. In Table 6, determination (b) for each
test shows the steady level of mercury vapor before addition of particulates
and S02, In no test did a variation outside the limits of the normal coeffici-
ent of variation of the mercury determination occur when particulates and S02
were added to the stream. In no case was S02 or fly-ash detected downstream
from the conditioner.
The conditioner-prefilter configuration appeared to be most successful in
these laboratory tests.
29
-------
Air
o
|Pump
| Pump ]
{Filter I
t
Air
Na,CO3
Bubbler.
Hj
Soul
;
rce
SO
Sou
Z
re
r
Thr<
Val
i
V_fc
rafilte
r 1—
Sample
Condition
Na2CO3
Bubbler
4-
DiluWr
Du Pont
400
Analyzer
aatt
Figure 10. Sulfur Dioxide test assembly.
P: esaur«
"1 iontrol
T
-------
TABLE 5. SULFUR DIOXIDE REMOVAL DURING CONDITIONER TESTS
Test
No. '
1
2
3
4
5
6
7
8
9
10
Mercury
Concentration
43
49
81
39
41
205
225
195
750
750
S02
Cone, at
Inlet
(ppm)
435
220
760
2175
750
410
720
2050
460
I960
SO,
Cone, behind Removal
Conditioner Efficiency *
(ppm) (%)
none detected 99. 99
n ii
"
II II
II II
II II
,r
II M
II II
II II
Duration
of
Test
(min)
57.8
60
42
25
46
50
44
25
45
30
System
Pressure
(cm H20)
760
755
810
810
810
810
810
810
810
810
Estimated coefficient of variation 2%
Detection limit is 10 jug of SO4/20 ml of solution in bubbler
-------
TABLE 6. REMOVAL OF SULFUR DIOXIDE AND PARTICULATES BY PREFILTER CONDITIONER
CO
ro
Test
No.
1 (a)
(b)
2 (a)
(b)
3 (a)
(b)
4 (a)
(b)
5 (a)
(fc)
6 (a)
(b)
7 (a)
(b)
Mercury
Level *
Oug/m3)
81
81
91
91
83
83
79
79
443
443
790
790
950
950
Part. Cone.
at Entry
(g/m3)
7.6
--
12.7
--
12.2
--
11.6
--
12.9
--
9.8
—
12.3
--
Part. Cone.
behind Cond.
(g/m3)
0
--
0
--
0
—
0
--
0
--
0
--
0
--
SO^Conc.
at Entry
(ppm)
110
--
210
--
1040
--
410
."
420
--
480
--
230
--
SO2Conc. #ltt
behind Cond.
(ppm)
0
--
0
--
0
--
0
--
0
..
0
--
0
--
Part.
Removal
Efficiency(%)
100
--
100
--
100
-_
100
..
100
-.
100
--
100
--
S02
Removal
Efficiency(%)
100
--
100
--
100
_-
100
--
100
-.
100
-.
100
--
Duration
of Test
(min)
22
--
30
--
25
-_
24
--
25
.-
25
..
25
--
System
Pressure
(cmH20)
800
II
800
tf
800
II
800
II
800
n
800
If
800
M
Coefficient of variation estimate 2%.
**
SO2 is undetectable; removal <99.99%.
-------
co
CO
Na2CC>3
Bubbler
Stirred
Particle
Source
Downstream
Filter and
Bubbler
Sample
Conditione
3
, Dilator
Du Pont
400
Analyzer
Waste
Pressu
Control
~Exha
ust
Air
Figure 11. Combined Sulfur Dioxide and Particulate test assembly.
-------
FIELD TESTS
During April, 1976 the interface system was tested in conjunction with
determinations of mercury in the exhaust gases of the No. 2 Unit of the J.E.
Amos Power Generating Plant of the American Electric Service Corporation at
St. Albans, West Virginia.
Description of Test Site
Unit No. 2 of the John E. Amos Power Generation Plan is a coal-fired
steam generating station with ia maximum output of 800 megawatts. The coal
used contains*; 1% sulfur and is burned at a maximum rate of 350 tons per hour.
Testing at ports in the 900 foot stack was not possible because of the
amount of equipment involved, the access difficulties to stack ports and the
limited catwalk space at such ports. A port located on the top of a horizon-
tal duct measuring 6.1 x 6.1 meters (20 x 20) feeding into the stack was
selected. The duct contained eight sample ports, 76.2 cm (30 inches) high
made of 12.7 cm (5 inches) i.d. pipe. These were installed at a point 18.9
meters (62 feet) from the stack but were located only 3.6 meters (12 feet) from
the junction of the duct from those two precipitators. Flow anomalies caused
by this location were not considered to be of importance in these mercury vapor
measurements. The port utilized was 1.9 meters (6-1/4 feet) from the duct wall.
Figure 12 shows the location of the sampling port relative to the stack and
precipitators. Figure 13 dimensions the probe assembly at the port.
The general operating conditions within the duct were:
Temperature: 280-320 F
Air Velocity: 26.8-35.8 m/sec
or: 1,610-2, 150 m/min
Gas Flow Rate: 1000-1300 m3/sec
or: 60-80,000 m3/mfn
Isokinetic Sampling
Rate: 114-150 Itters/mfn
(with 3/8" probe tip dfam.)
Estimated Gas
Components: NOX: to 600 ppm
S02: to 400-600 ppm
34
-------
OJ
C71
EL^SCTT ^O «5>T ATT \C
Figure 12. John E. Amos Plant, Unit #2, 800 Megawatt output.
-------
PUVSNIOOD
PL-VVs>OOD SPACER
ChA
TIP
V0.7. CM
Figure 13. Probe alignment.
36
-------
Estimated Paniculate
Level: 0.5-1.0 g/m3
Test Assembly
To obtain representative samples, a six foot, heated, glass lined, stain-
less steel probe was used. The probe extended 41.9 cm (16.5 inches) into the
duct and was fitted to the sample port by means of plywood spacers. The probe
was heated using a 10 amp (120 V) varfac set at 40-50 divisions which main-
tained a temperature of approximately 300 F at the probe wall. The duct gas
temperature was 280-320 F.
Gas was drawn through the probe at 22 1/min which avoided probe clogging
caused by fly-ash. At the top of the probe a glass tee was used to divide
the air flow. Twenty liters per mfnute were exhausted through a Thomas pump
operating through two 500 ml Greenberg-Smith impingers. The impingers were
partially filled with saturated sodium bicarbonate to protect the pump from
S02 and condensate. A volume flow; rate of two liters per minute was diverted
to the interface for analysis.
After passing through 76.2 on (30«) of 0.48 cm (3/16") I.D, amber latex
tubing and a horizontal 500 ml Greenberg-Smith impinger to remove water, the
sample gas, at a rate of 2 liters/minute, entered the conditioning unit.
Figure 14 shows the assembly schematically. In this operation, the thermal
decomposition furnace was not required. The gas was passed from the probe
through the 500 ml impinger to the scrubber coil which consisted of 3.66 m
(12 feet) of coiled (5 mm I.D., 7 mm O.D.) pyrex tubing. In the coil the gas
contacted a solution consisting of 80% 0.2 M sodium bicarbonate and 20% 0.2 M
ammonium bicarbonate. The solution was stored in a 2 gal. container and pumped
into the scrubber coll at 5 ml/mln. by the conditioner pump. After passing
through the coll, a portion of the gas (0.5 1/min) plus the waste liquid was
drawn off to a 2 gallon container. The remaining 1.5 1/min of gas went through
a condenser, cooled by the scrubber solution, to remove excess water in the
gas.
Two heated 25 mm Gelman in-line filter holdars containing Whatman #41 paper
removed any residual fly-ash before the rotameter. The gas was directed by
means of a three way valve to the DuPont 400 spectrophotometer or through 10%
silver alumina tablets and thence to the photometer. The DuPont 400 used dir-
ectly gave the measurement of the mercury yielding a base line or zero for
comparison when mercury-free gas was needed for passage through the photometer.
The difference between the two photometer readings was calibrated by putting
a known amount of mercury through the system.
The gas was exhausted through the pump unit. This unit served to keep the
pressure constant in the photometer cell by means of a pressure regulator. Also,
it pulled 0.5 1/min. of gas through the waste container.
To check the results against the reference method, measured volumes of gas
were drawn through midget impingers connected downstream of the probe. After
removing the S02 in the gas by passing it through saturated sodium bicarbonate
37
-------
COS1OE.KI S»^f^
NOTES;
Scrubber Fluid -
0. 2M NaHCO3 & 20% (0. 2M NH4CO3)
Pumping Rate: 4. B ml/min
Lines from probe to scrubber are amber latex rubber tubing
Gelman In-Line Filter Holders contained Whatman #41 Z5 mm Filter Paper
Probe Heater - Variac setting 40
Figure 14. Detailed schematic of measurement system.
38
-------
in the first midget impinger, mercury was collected in a ni'tric acid-potas-
sium permangate solution. The amount of mercury was determined by using
the photometer section of a GEOMET Model 103 Mercury Monitor to measure the
peak obtained from a modified Hatch and Ott procedure.
Operation of System
Maintenance--
Each morning the lines and impinger between the probe and scrubber were
rinsed with distilled water. Fly-ash was removed from the top of the probe,
and the 20 liter per minute exhaust line and impingers were cleaned. Every
two hours new filter paper was placed in the in-line filters. The first fil-
ter after the condenser restricted the air flow (because of collected fly-
ash) if left for long periods. Periodic air flow checks at the probe were
also necessary since the Thomas pump on the exhaust may cause the rotameter
in the measurement side of the system to give erroneous readings when varia-
tions from the fnttal pressure drop ratio take place.
Test Qperations--
In th.e tests conducted at the duct of the J.E. Amos Power Generation
Plant the results obtained by use of the instrumentation comprising the inter-
face system and the DuPont 400 Photometer Analyzer were compared with deter-
minations obtained by use of a simple absorption train together with conven
tional reduction and measurement procedures. Details of the procedures have
been included in Appendix B.
The mercury level as determined by the instrumentation was zeroed
against the signal obtained by diverting the air flow through 10% silver-alu-
mina tablets. The signal difference as measured by the DuPont Photometer was
calibrated by introduction of known quantities: of mercury vapor into the sys-
tem at the probe. This internal calibration technique was also utilized to
check the reproducibility of the operation from one test to another.
The stack mercury concentration was determined separately by hooking a
permanganate bubbler (1 ml 0.25 M KMn04, 2.5 ml 1:1 HNOa diluted to 15 ml)
to a saturated NaHCOs bubbler and connecting them directly to the probe. Sam-
pling intervals of 2.5 minutes were found to be about the best length of time
for these independent sample collections.
For comparison, a permanganate/acid bubbler was also used to collect Var-
ious volumes of mercury vapor introduced into the line behind the probe. This
was used to quantitate calibration peaks.
To analyse the bubblers, the contents were reduced (3 ml 20% Wt./Vol.
SnCl2 in 6N HC1) and caught in a second permanganate bubbler (0.3 ml 0.25 H
KMn04, 2 ml 1:1 HNOs diluted to 15 ml). The second bubbler was placed in ser-
ies with an NaBfy bubbler (-<-l g/15 ml), and empty bubbler and the Geomet ana-
lyzer, on reduction with 2 ml of SnCl2 the vapors were tranferred utilizing
an air flow of 2 1/min to the Geomet Model 103. The amount of mercury was
determined using peak heights from prepared calibration curves.
39
-------
Results
The results obtained in thirty-seven comparisons between measurements
made with the DuPont Photometric Analyzer and a direct bubbler measurement
of mercury in the stack (duct) are tabulated in Table 7. The three right
hand columns summarize the comparisons. Table 8 is a summary of the calibra-
tion test data (indicated by the designation P) and bubbler tests on the stack
gases (designated by S). All Run # designations correspond with the samples
on Table 7. While samples were not collected simultaneously with test runs
they were collected immediately before or after the determination by the
photometer.
Table 9 defines the terms used in Tables 7 and 8.
Discussion of Results
The results obtained and listed in Table 7 show that a good calibration
between the data obtained with the automated system and the bubbler technique
has been obtained. In the majority of the test runs, the two methods agreed
very well with respect to the amount of mercury in the stack gas. In a few
cases a significant difference appeared.
Photometric measurements with the interface and the DuPont Analyzer were
made considerably more easily than those with the bubbler. By this technique,
continuous monitoring of the effluent stream was clearly possible.
In general, the concentrations of mercury vapor in the exhaust gases were
relatively low, ranging from 1.74 to 6.96 ug/m3 in determinations made by the
instrumented system. The range was 1.60 to 7.25 in measurements made by the
bubbler techniques. The mean levels were 4.23 + 1.28 ug/nr (instrumental) and
4.66 + 1.42 ug/m3 (bubbler). The mean values represent the entire set of data
collected from 4/21 through 4/30.
These tests were conducted in the presence of relatively high levels of
potential interferences: 0.5 - 1.0 g/m3 of particulates,f400-500 ppm S02 and
-^600 ppm NOX. The levels of $03 were not defined. The presence of $03 neces-
sitated the changes in the experimental assembly involving the silver-alumina
tablets and the zero baseline determination. However, after incorporating
these modifications, the field performance of the interface-photometer system
was approximately equal to that of the laboratory-tested configuration discus-
sed in above. Response and fall times remained within the ranges previously
measured. The detection limit was etimated at 0.4-0.5 ug/m3, as previously.
Detailed replication of the field tests were not conducted; no reason for a
changed precision estimate was apparent. A coefficient of variation in the
5-6% range seems predictable.
The performance stability and reliability of the system were exceptional,
particularly when compared with manual methods.
Accuracy as determined by comparison of results obtained with the instru-
mentation and with the manual method appears more than adequate. However,
40
-------
TABLE 7. TESTS OF AUTOMATED INTERFACE INSTRUMENTATION
Ron Date
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
4/21
4/22
4/22
4/22
4/23
4/23
4/24
4/24
4/24
4/26
4/26
4/27
4/27
4/27
4/27
4/27
Time
15:35
10:25
15:03
17:23
10:30
17:10
10:15
14:45
16:47
15:05
16:06
10:37
12:39
15:19
16:45
17:50
Peak
Area in
Squares
719
857.5
2907. 5
1381
2802. 5
2277. 5
1705
2222. 5
2017.5
2482. 5
2712.5
928
869
766
1429
1247
1402
1241
Calc.
Stack
Hg Peak
Height in
Square*
•IPm-BBMMI—M—^V^H
35.95
42.88
72.69
27.62
62.28
56.94
37.89
55.56
40.35
62.06
67.81
23.2
21.73
19.15
35.73
31. 18
35.05
31.03
DuPont
Flow Content of Peak
Rate Bubbler Calib.
1/min in ng in tig /I
1.5
1.5
1.5
1.3
1.3
1.3
1.3
1.5
1.5
1.7
1.5
1.25
1.5
1.5
1.5
1.5
1.5
1.5
52.5
34.0
42.5
16.5
39.5
36.5
28.5
48.0
34.5
39.0
36.0
14.0
12.0
14.0
11.0
12.5
13.5
14.0
94.5
61.2
38.25
17.13
36.46
37.90
26.31
43.20
24.84
30.97
32.40
14.54
10.80
12.60
9.90
11.25
12.15
12.60
Photometric Ratio
Calibration Stack Hg Bubbler Hg Cone Bubbler to
Signal Hg Cone of stack Photometric
Factor in^ Heightin of Btack calculated Cone, of
ng-f-sq Squares in jig/m3 injig/mj Hg in stack
2.63
1.43
0.526
0.620
0.585
0.666
0.694
0.778
0.616
0.499
0.478
0.627
0.497
0.658
0.277
0.361
0.347
0.406
1
--
3.3
0.0
3.0
3.3
6.5
4.6
7.3
7.0
4.0
7.0
7.0
5.0
10.0
10.0
8.5
9.6
1.90
—
1.60
0.00
1.97
2.50
5.29
4.42
4.17
3.01
2.65
3.99
4.22
3.35
6.50
3.55
5.40
5.00
2.63
--
1.74
0.00
1.76
2.20
4.51
3.58
4.50
3.49
1.91
4.39
3.48
3.29
2.77
3.61
2.95
3.90
0.722
—
0.920
o. OOP;
1.12
1.14
1.17
1.23
0.927
0.862
1.39
0.909
1.21
1.02
2.35
1.54
1.83
1.28
(continued)
-------
TABLE 7 (continued)
ro
Run Date
19 4/28
20 4/28
21 4/28
22 4/28
23 4/28
24 4/28
25 4/28
26 4/29
27 4/29
28 4/29
29 4/29
30 4/29
31 4/29
32 4/30
33 4/30
34 4/30
35 4/30
36 4/30
37 4/30
Time
08:00
09:08
10:00
11:24
14:25
15:09
17:53
09:50
11:30
13:03
15:10
17:20
18:16
08:30
10:05.
12:50
14:35
16:03
17:03
Peak
Area in
Squares
754
760
1710
1033
866
791
1127
895
1262
1654
1095
2140
2167.5
1522
1098
1895
1540
1870
2562. 5
Calc.
.Stack
Hg Peak
Height in
.Squares
18.85
19.00
42.75 '-
25.83
21.65
19.78
28.18
22.38
31.55
41.35
27.38
53.50
61.93
33.82
24.40
47.38
44.00
53.43
73.21
DuPont
Flow Content
Rate Bubbler
1/min in ng
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
11.5
11.5
21.5
11.5
12.0
11.0
15.5
13.0
19.75
20.5
25.0
30.5 N
29.0
20.0
14.5
28.75
33.50
38.50
43.50
of Peak
Calib.
inng/1
10.35
10.35
19.35
10.35
10.80
9.90
13.95
11.70
17.78
18.45
22.50
27. 45
26. 10
16.00
11.60
25.88
34.47
39.62
44.76
Photometric Ratio
Calibration Stack Hg Bubbler Hg Cone. Bubbler to
Factor in Signal Hg Cone of stack Photometric
. . Height in of stack calculated Cone, of
ng-l~-»q Squares injig/m3 inpg/m3 Hg in stack
0.549
0.545
0.453
0.401
0.499
0.501
0,495
0.523
0. 564
0.446
0.822
0.513
0.421
0.473
0.475
0.546
0.783
0.742
0.611
9.3
8.5
11.3
12.5
9.3
8.5
8.6
13.3
11J6
11.3
5.75
8.3
11.5
10.66
11.0
9.5
8.0
6.8
8.5
5.00
4.75
4.75
4.50
4.95
4.20
3.80
7.25
6.50
5.05
5.61
4.85
6.75
5.05
5.95
5.30
6.90
5.50
5. 15
5.11
4.63
5.12
5.01
4.64
4.26
4.26
6.96
6.54
5.04
4.73
4.26
4.84
5.04
5.23
5.19
6.26
5.05
5.19
0.978
1.03
0.928
0.898
1.07
0.986
0.892
1.04
0.994
1.002
1.19
1.14
1.39
1.002
1.14
1.02
1.10
1.09
0.992
-------
TABLE 8. CALIBRATION AND DIRECT BUBBLER MEASUREMENTS
CO
Run
No.
1-A
1-B
1-C
2-A
2-B
3 -A
3-B
4-A
4-B
5-A
5-B
5-C
6-A
6-B
6-C
7-A
7-B
8 -A
8-B
Date
4/21
4/21
4/22
4/22
4/22
4/22
4/22
4/22
4/22
4/22
4/23
4/23
4/23
4/23
Type*
P
S
s
S
P
P
s
s
P
P
s
s
P
s
s
P
s
P
s
Bubbler
DVM
Reading
135
135
30
25
90
111
27
0
47
103
24
32
'96
32
30
76
58
124
58
Sample
Flow
Rate
1/min
2
2
2
2
1.6
1.6
1.6
1.6
1.6
1.6
1.9
DuPont
Flow
Rate
1/min
1.5
1.5
1.5
1.3
1.3
1.3
1.3
1.5
ng
Hg
52.
52.
9.
7.
34.
42.
8.
0.
16.
39.
7.
10.
36.
10.
9.
28.
21.
48.
21.
Concentration
Hg ng/1
5
5
5
5
0
5
0
0
5
5
0
5
5
5
5
5
0
0
0
1.
1.
1.
0.
1.
2,
2.
2.
5.
4.
90
50
60
00
75
18
62
38
29
42
Time
minutes
2.5
2.5
2.5
2.5
2.5
2.5
3.0
2.5
2. 5
2.5
2.5
P= Calibration tests by injection of vapor which passed thru as peak. S* Bubbler Stack Samples
(continued)
-------
TABLE 8 (continued)
Run
No.
9-A
9-B
9-.C
10 -A
10-B
10-C
11-A
11-B
11-C
12-A
12-B
12-C
13-A
13 -B
13-C
14-A
14-B
14-C
15-A
Date
4/24
4/24
4/24
4/24
4/24
4/24
4/24
4/24
4/24
4/26
4/26
4/26
4/26
4/26
4/26
4/27
4/27
4/27
4/27
Type*
P
S
S
P
S
S
P
S
S
S
S
P
S
S
P
P
S
S
P
Bubbler
DVM
Reading
91
57
56
102
45
50
95
37
38
42
57
41
53
78
36
41
71
73
34
Sample
Flow
Rate
1/min
1.9
2.0
2.2
2.2
1.8
1.9
1.7
1.5
1.8
1.8
2.0
2.0
DuPont
Flow
Rate
i/min
1.5
1.7
1,5
1.3
1.5
1.5
1.5
ng
Hg
34.
20.
20,
39.
15.
17.
36.
12.
12.
14.
20.
14.
19.
29.
12.
14.
16.
17.
JJ.
Concentration
Hg ng/1
5
5
0
0
5
S
0
0
5
5
5
0
0
0
0
0
5
0
0
4.
4.
Z.
3.
2.
2.
3.
4.
4.
6.
3.
3.
33
00
82
19
67
63
41
56
22
44
30
40
Time
minutes
2,
2.
2.
2.
2.
2.
2.
3.
2.
2.
2.
2.
5
5
5
5
5
5
5
0
5
5
5
5
P = Calibration tests by injection of vapor which passed thru as peak. S= Bubbler Stack Samples
(continued)
-------
TABLE 8 (continued)
en
Run
No.
15-B
15-C
16-A
16-B
16-C
17-A
17-B
17-C
18-A
18-B
18-C
19-A
19-B
19-C
20-A
20-B
20-C
21-A
21-B
•
Date
4/27
4/27
4/27
4/27
4/27
4/27
4/27
4/27
4/27
4/27
4/27
4/28
4/28
4/28
4/28
4/28
4/28
4/28
4/28
Type*
S
S
P
S
S
p
S
S
'p
S
S
p
S
S
p
S
S
p
S
Bubbler
DVM
R eading
88
83
37
76
72
40
74
71
41
69
67
35
68
68
35
53
76
59
62
Sample
Flow
Rate
1/min
2.
2.
2.
2.
2.
2.
2.
2.
2.
2.
2.
2.
2.
0
0
0
0
0
0
0
0
0
0
0
0
0
DuPont
Flow
Rate ng
1/min Hg
33.
31.
1.5 12.
28.
27.
1.5 13.
27.
26.
1.5 14.
25.
24.
1.5 11.
25.
25.
1.5 11.
19.
28.
1.5 21.
22.
Concentration
Hg ng/1
5
5
5
5
0
5
5
5
0
5
5
5
0
0
5
0
5
5
5
6.
6.
5.
5.
5.
5.
5.
4.
5.
5.
3.
5.
4.
70
30
70
40
50
30
10
90
00
00
80
70
50
Time
minutes
2.
2.
2.
2.
2.
2.
2.
2.
2.
2.
2.
2.
2.
5
5
5
5
5
5
5
5
5
5
5
5
5
P= Calibration tests by injection of vapor which passed thru as peak. S= Bubbler Stack Samples
(continued)
-------
TABLE 8 (continued)
Run
No.
21-C
22-A
22-B
22-C
23 -A
23 -B
23-C
£ 24-A
24-B
24- C
25 -A
25-B
25-C
26 -A
26 -B
26-C
27-A
27-B
27-C
Date
4/28
4/28
4/28
4/28
4/28
4/28
4/28
4/28
4/28
4/28
4/28
4/28
4/28
4/29
4/29
4/29
4/29
4/29
4/29
Type*
S
P
S
S
P
S
S
P
S
S
P
S
S
P
S
S
P
S
S
Bubbler
DVM
R eading
68
35
54
69
36
65
69
34
59
57
45
53
53
39
96
95
55
83
89
Sample
Flow
Rate
1/min
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
DuPont
Flow
Rate
1/min
1.5
1.5
1.5
1.5
1.5
1.5
ng
Hg
25.
11.
19.
25.
12.
24.
25.
11.
21.
20.
15.
19.
19.
13.
36.
36.
19.
31.
33.
Concentration
Hg ng/1
0
5
5
5
0
0
5
0
5
5
5
0
0
0
5
0
75
5
5
5.
3.
5.
4.
5.
4.
4.
3.
3.
7.
7.
6.
6.
00
90
10
80
10
30
10
80
80
30
20
30
70
Time
minutes
2.
2.
2.
2.
2.
2.
2.
2.
2.
2.
2.
2.
2.
5
5
5
5
5
5
5
5
B
5
5
5
5
P= Calibration tests by injection of vapor which passed thru as peak. S=Bubbler Stack Samples
(continued)
-------
TABLE 8 (continued)
Run
No.
28 -A
28-B
28-C
29-A
29-B
29-C
30 -A
30-B
30-C
31-A
31-B
31-C
32-A
32-B
32-C
33 -A
33-B
33-C
34-A
#_
Date
4/29
4/29
4/29
4/29
4/29
4/29
4/29
4/29
4/29
4/29
4/29
4/29
4/30
4/30
4/30
4/30
4/30
4/30
4/30
4** 1 " 1 __ _- LS .
Type*
P
S
S
P
S
S
P
S
S
P
S
S
P
S
S
P
S
S
P
A A. _ I
Bubbler
DVM
Reading
57
64
73
68
44
93
81
64
68
78
89
90
56
66
71
42
81
78
77
* • A.' f
Sample
Flow
Rate
1/min
2.0
2.0
1.6
1.9
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
.
DuPont
Flow
Rate
1/min
1.5
1.5
1.5
1.5
1.5
1.5
1.5
Hg
20.5
23.5
27.0
25.0
15.0
35.5
30.5
23.5
25.0
29.0
33.5
34.0
20.0
24.0
26.5
14.5
30.5
29.0
28.75
fe fm .« — A I.. C*
Concentration
Hg ng/1
4.70
5.40
3.75
7.47
4.70
5.00
6.70
6.80
4.80
5.30
6. 10
5.80
T5._i-L1 __ &A. 1_ f_— 1
Time
minutes
2.5
2.5
2.5
2.5
2.5
2.5
2.5
2.5
2.5
2.5
2.5
2.5
i _ __
p=
(continued)
-------
TABLE 8 (continued)
Run
No.
34-B
34-C
35-A
35-B
35-C
36 -A
36-B
36-C
37-A
37-B
37-C
Date
4/30
4/30
4/30
4/30
4/30
4/30
4/30
4/30
4/30
4/30
4/30
Type*
S
S
P
S
S
P
S
S
P
S
S
Bubbler
DVM
R eading
62
81
88
91
191
101
75
72
113
68
71
Sample
Flow
Rate
1/min
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
DuPonf
Flow
Rate ng
1 /rain Hg
22.5
30.5
1.5 33.5
34.5
75.5
1.5 38.5
28.0
27.0
1. 5 43. 5
25.0
26.5
Concentration
Hg ng/1
4.50
6. 10
6.90
15. 10
5.60
5.40
5.00
5.30
Time
minutes
2.5
2.5
2.5
2.5
2.5
2.5
2.5
2.5
£
P= Calibration tests by injection of vapor which passed thru as peak. S=Bubbler Stack Samples
-------
TABLE 9. DEFINITION OF TERMS.
Chart Paper
Hg Shot Peak
Stack Hg Signal
1
iHfeigjhtbfl
fcfectlngie 1
!/ * \l
4 >
Time in seconds
I- '
Baseline
Peak Area-
Stack Hg Peak-
DVM Reading -
Hg Content of
Bubbler -
Hg Concentration
of Peak-
Number of squares under the peak resulting from a
mercury shot.
Calculated by dividing PEAK AREA by time in seconds
where time corresponds to the base of the peak and the
length of a rectangle possessing the same area as the
peak. The quotient is the width of the rectangle. Thia
width would be the same as the height of a stack mercury
signal resulting from the same level of mercury which
produced the peak.
Width = Area i- length
Reading on the GEOMET 103 produced by a 1 cc shot of
Hg vapor collected in a KMnO4 bubbler. This corresponds
to the mercury level used to give the peak.
Amount of mercury in ng corresponding to the DVM read-
ing on the DVM - ng Hg calibration curve.
Concentration of mercury which produced the Hg SHOT
PEAK. Found by dividing Hg CONTENT OF BUBBLER
by the time in seconds which corresponds to the base of
the peak then multiplying by the inverse of the gas flow
rate and a correction factor, 0. 9, for the error in the
chart speed.
[(ng Hg/sec) (I/sec)"1 0.9]
(continued)
49
-------
Calibration Factor-
Stack Hg Signal
Height-
Hg Concentration
of the stack by
Bubbler-
Hg Concentration
of Stack Cal-
culated-
Table 9-Continued
DEFINITION OF TERMS
Calculated by dividing Hg CONCENTRATION OF THE PEAK
by STACK Hg PEAK.
Found by taking the difference between the signals/
produced by the stack gas with and without mercury
pills in the line.
Concentration found by collecting mercury from the
stack gas at the probe in a KMn04 bubbler proceeded
by a bicarbonate bubbler.
Found by multiplying STACK Hg SIGNAL HEIGHT
by the CALIBRATOR FACTOR.
50
-------
determinations made by the bubbler-manual precedures may not be appropriate
criteria for such comparisons.
In summary, the interface instrumentation appears to perform as required.
It will permit the continuous monitoring of stack gases for mercury. The
flexible nature of its design will permit its use in a wide variety of appli-
cations.
51
-------
Appendtx A
OPERATING INSTRUCTIONS
SECTION Al
CONTROLS AND CONNECTIONS
This appendix gives in detail the use of controls and connections to be
made in operating the Interface System. A detailed description will be given
of the Conditioning Module, the Diluter Module and the Pump Module.
The Conditioner Module
There are five operating controls on th.e conditioning module. Four of
these on the front panel are shown in Figure A-l. They are: the power switch,
the heater switch, the fluid switch, and the pyrometer set point knob. The
fifth is the fluid pump stroke adjustment knob which is located inside the
unit.
In addition there are 4 line connections on the unit. These are shown
in Figure A-2.
The Power Switch—
Power to the unit is controlled by the red toggle switch on the upper
right hand side of the front panel. In the "up" position, power is supplied
to the solid state electronic power supply. This permits the other switches
to function, and it activates the Rotron cooling fan in the side of the unit.
The light above the switch indicates that the power supply has been
activated.
The Heater Switch--
Power is supplied to the unit by connecting the 3 wire electrical cord
to the A.C. Connector in the lower left hand corner of the panel. 60 Hz,
110 VAC power is required. Power to the furnace heater is controlled by the
red toggle switch which is labelled "Heater". In the "up" position the power
is supplied to the heater.
The light above the switch indicates that power is being supplied to the
heater.
The Fluid Switch-
Power to the fluid pump is controlled by the white toggle switch. In
the "up" position the pump is activated and pumps fluid into the scrubber
coil. The light above the switch indicates that the pump is on.
52
-------
Right H&ttd
Side
Set Point
Knob
Powe
Switcl
Heater
Switch
A, C.
Conaector
Figure A-l. Front Face of Conditioner Module
53
-------
Scrubber
Inlet
i n
i
Figure A-2. Left Side View of Conditioner Showing
-------
The Pyrometer Set Point Knob--
In the upper left corner of the front panel, there is a single set point
pyrometer. This indicates the temperature of the furnace. The set point
knob located in the bottom right corner of the pyrometer controls the temper-
ature of the furnace. The furnace temperature is obtained by turning the knob
until the red set point needle is at the desired temperature. The temperature
will be maintained at that level.
The Fluid Pump Stroke Adjustment Knob--
The fluid pump stroke adjustment knob is located inside the case of the
conditioner in the lower front left hand corner of the instrument, immediately
behind the power terminal. It is reached by removing the cover of the unit.
Release of the four trunk latches is required.to remove the cover. This knob
varies the stroke of the fluid pump which allows for different scrubber solu-
tion flow rates.
Plumbed Connections--
All of the plumbing connections to the conditioner are shown in Figure
A-2. The Sample inlet for use when the furnace function is required is shown
on the right side of the photograph. It is located on the back face of the
unit. The cooling tube which passes the incoming gas from the furnace section
to the scrubber is shown in the upper center. The left hand end of the cool-
ing tube is connected to the lower inlet at the top of the glass scrubber and
its coil. When thermal decomposition of the source sample is not required,
the sample line from the probe or inlet should be connected to the scrubber
inlet port. The hard-plumbed connection attached to the sample inlet (rear)
may be moved to the scrubber inlet for this application.
At the lower left of Figure A-2 is the fluid inlet ("fluid in") connector.
Inside the case, the fluid inlet is connected to the bottom side of the pump
head. From the upper side of the pump head, the fluid line is connected to
the upper most inlet of the glass scrubber coil by means of 4 mm (i.d.) tef-
lon tubing. This connection is made with teflon tubing which is covered with
teflon tape. Outside the case, the fluid inlet line passes to a 2 gallon
heavy-duty polyethylene reservoir, which in operation will contain the scrub-
ber solution. The connecting polyethylene tubing should reach to the bottom
of the reservoir. It is sealed into a white plastic cap. This cap has only
one tubing lead attached to it.
Next to the fluid inlet are two plumbed connectors, labelled "sample
out" and "waste out".
From the "sample out" connector the scrubbed sample gas is removed from
the conditioner unit. It passes to the diluter under usual operating condi-
tions. A condenser and trap are connected to the "sample out" terminal to
cool the gas stream prior to its entry into the diluter. Excess moisture is
removed from the gas stream and collected in the glass trap by this techni-
que. Figure A-3 shows the arrangement. Inside the module the sample out
terminal is connected to the upper of the two takeoffs on the gas/1iquid
separator or trap located at the bottom of the scrubber coil. The connection
is made with teflon tubing covered with teflon tape.
55
-------
Condenser
to Diluter
Conditioner
Figure A-3. Arrangement of condenser between
conditioner and diluter modules.
56
-------
Liquid waste from the scrubber leaves the unit from the "waste out"
connector. Outside the case it connects to a two gallon container by means
of 4 mm (i.d.) teflon tubing which is sealed into the top of a white plastic
cap. The cap for the waste container has a second tubing lead extending from
it which will be connected to the pump module. Within the module, the liquid
waste line is connected to the bottom outlet from the trap (liquid/gas separa-
tor) located under the scrubber coil.
The Diluter Module
<*
The Diluter Module has seven operating controls and three plumbed connec-
tions. Six of the controls shown in Figure A-4 are: the power switch, the
dilution air switch, the low flow valve switch, the heater switch and two mass
flowmeter set point knobs. The seventh control is the adjustable flowmeter
which is shown in Figure A-5, a rear view of the diluter module. The three
rear connections are the sample inlet, sample outlet, and the excess sample
exhaust from the flowmeter.
The Power Switch--
Power to the unit is controlled by the red toggle switch on the right
side of the front panel. In the "up" position, power is supplied to the
solid state power supply and to the Rotron cooling fan. The light above the
switch indicate when the power supply has been activated.
The Dilution Air Switch-
In the "up" position this white toggle switch initiates the sample dilu-
tion process. When activated the three way valve (shown in Figure A-3) closes
off the sample line and allows the same flow rate of mercury free air to enter
the unit through the filter of silver-alumina absorbent. This also activates
both automatic servo valves for flow regulation.
In the "middle" position power is cut off to the automatic servo valves.
If desired the valves may then be adjusted manually. This requires removal
of the cover.
In the "down" position power is supplied to the high flow automatic ser-
vo valve. This position is used when the module is being used for flow regu-
lation only.
The Low Flow Valve Switch—
This white toggle switch opens the two way valve in the sample line used
during dilution (Figure A-3). It also activates the pump which maintains the
sample gas flow through the lines and exhausts the excess gas. Both this
switch and the dilution air switch must be "up" to have the sample gas dilu-
ted.
In the "down" position this switch, shuts off the pump and closes the two
way valve. This is the position used when no dilution of the sample gas is ^
necessary. Both this switch and the dilution air switch must be in the down
position when the unit is being used only for flow regulation and not dilution.
57
-------
01
! O
High Flow Mass
ter
V alvc :-- 'it< h
ov/ F J ow lv: ass
eter
Figure A-4. Front Panel, Diluter Module
-------
01
<0
Exhaust
Flowmeter
Flowmeter
,4,M. Adjustment Knob
Figure A-5. Rear View, Diluter Module
-------
The Heater Switch--
The white toggle switch which is labelled "heater" controls power to the
cartridge heater on the back of the unit. In the "up" position the heater
heats the incoming gas to prevent condensation in the unit.
The Mass Flowmeter Adjustment Knobs—
The 0-5 1/min. mass flowmeter is located in the upper left hand corner
of the front panel. By adjusting the set point knob, the desired flow rate
of sample gas or mercury air is obtained.
The 0-100 cc mass flowmeter is located in the lower left hand corner.
When the gas sample is being diluted, this adjustment knob controls the amount
of sample gas allowed to go through th.e unit.
By adjustment of the two mass flowmeters dilutions from 1/19 to 1/5000
may be readily achieved.
The Adjustable Flowmeter—
As shown in Figure A-4, there is an 0-4 1/min. adjustable flowmeter on
the back panel of the unit. When the sample gas is being diluted this con-
trols the rate at which sample gas is exhausted.
The flow rate through this exhaust flowmeter plus the sample gas flow
rate which is being diluted (shown on the low flowmeter) is the total flow
rate coming through the intake line. The flowmeter knob should be adjusted
so as to achieve the desired rate of gas flow.
The Diluter Module Connections—
The sample inlet is located on the lower right hand corner of the unit
when facing the back panel. The cartridge heater fitting is connected into
the fitting. The sample outlet, whether diluted or not, is located in the
middle of the left side of the rear panel. In the upper right hand corner of
the rear panel, the excess sample port is located. The adjustable flowmeter
connects to this port.
The Pump Module
This module has three operating controls, 3 line connections, and a mag-
nehelic gauge. In Figure A-6 the power switch, the flowmeter adjustment
knob, and the raagnehelic gauge are shown. Figure A-7 shows the line connec-
tions and the pressure adjustment. Figure A-8 shows the method of connecting
it into the system. It operates between the waste container and the outlet
from the measuring photometer.
The Power Switch—
The power switch is a white toggle switch on the lower right hand corner
of the front panel. In the "up" position this gives power to the pump. This
switch is labelled "vacuum pump".
The Flowmeter Adjustment Knob--
In the middle of the front panel is a 0-0.5 1/min. flowmeter with an ad-
justment knob on the bottom. This regulates the gas flow rate through the
60
-------
Ch
IQ
C
-
-i •
i
-j
'
o
11
c
ro
VACUUM
: :
E i
—FLOW :
FLOW I-:R
ADJUt, ^:. INI i .,O,B:
WAitI US-
SIR HOW
;;E
,DER
f/ A f, .,-».,-, , „
'
-------
: '
Pressure
Regulator
Adjustment
Waste Line
Vacuum
Sample Line
Vaetrana
Figure A-7. Interior, Pump Module
-------
CTt
CO
CONTAINER 50UUTIOM
Ft-OVN :
03L../MVN.
Duposrr
4oo
CONDITION »NB
MODOUE
50O M\~.
PRO&E.
PORT
RE.CORDEV*.
Figure A-8. Automated Mercury Interface system schematic.
-------
waste line. Generally, the flow rate is maintained at 0.5 1/min.
The Pressure Regulator Adjustment-
Figure B-7 shows the interior arrangement of the pump module. The regul-
ator which establishes and maintains a constant pressure inside the spectro-
photometer cell is located inside this unit. It is in the lower front right
hand corner when the unit is viewed from the front. By turning the threaded
stem, the vacuum of the system may be varied. A cylinder filter is generally
used on the air intake of the regulator to avoid plugging its orifice.
The Magnehelic Gauge--
The 0-300 Magnehelic gauge which reads in cm of H20 indicates the vacuum
in the system. The observed reading is the pressure below one atmosphere
which is operational in the system. For example, at 0°C, 1 atm= 1033 cm of
water. A reading of 100 cm on this gauge means that the pressure in the sys-
tem is 933 cm ^0. Any adjustment of the pressure regulator can be observed
on this gauge.
The pump module is connected to the waste container (waste line vacuum)
and the photometer (sample line vacuum) outlet. The connection to the waste
line is made by inserting the 1/8" line from the waste container to a 1/4"
polyethylene busing at the connector at the rear of the unit. Tightening
the teflon ferrule on the 1/4" bushing seals the connection.
The line from the photometer (DuPonfr 400) should be 1/4" polyethylene
tubing which is sealed by use of the ferrule at the sample line fitting
(right hand connector).
In operation, good liquid removal from the trap at the bottom of the
scrubber was achieved with the vacuum gauge reading 225 inches of water. This
included use of the diluter as a flow controlling monitor at 2.0 1/min. and
the DuPont Photometer in the line.
There are 3 line connections of the unit. When viewed from the rear,
two connections are on the lower right hand corner. The outlet line from the
spectrophotometer cell is connected to the "sample" fitting. The line from
the waste container is connected to the other. Inside the unit on top of the
pump, the pump exhaust fitting is located. A 1/4" tygon tube may be connected
to the exhaust if desired. Normally it is operated "as is".
64
-------
SECTION A2
REAGENTS AND INSTRUMENT PREPARATION
Once the units have been brought to a test site, they must be readied for
use. This entails preparing the reagents, installing glassware in the condi-
tioning module, and making the appropriate connections for the gas and liquid
flow systems. Figure A-8, a schematic diagram shows the interface system
connected and ready to use.
Reagents
For the operation and calibration of the system, five solutions are need-
ed. These are the scrubber solution, the stannous chloride solution, and the
sodium borohyride solution.
The Scrubber Solution--
The scrubber solution is 20% Q.2M ammonium bicarbonate (NfyHCOs), 80%
0.2M sodium bicarbonate CNaHCOs), to which 2 ml of 10% Triton X100 per liter
has been added. To make a two gallon supply, 20 g of reagent grade ammonium
bicarbonate, 102 g of commercial sodium bicarbonate, and 15 ml of 10% Triton
X100 are added to two gallons of distilled water. After thoroughly mixing
the solution, it is ready for use.
The Saturated Sodium Bicarbonate Solution—
The two impingers shown in Figure A-8 are partially filled with this sol-
ution. It is made by placing 30 g of sodium bicarbonate in each impinger :.ith
100 ml of water. This produces a saturated solution with excess solid remain-
ing to be dissolved as the sodium bicarbonate reacts with sulfur dioxide.
The Acid-Permanganate Solution--
This solution is used in some of the calibration procedures to collect
mercury. It is generally prepared in 15 ml amounts which is the amount held
by a bubbler. To make 15 ml, 1.0 ml of 0.25M potassium permanganate and 2.5
ml of 1:1 nitric acid are diluted to 15 ml with distilled water. This is u^ed
when the gas stream contains oxidants. When transferring collected mercury
from one bubbler to another, only 0.3 nl of permanganate are used with the
acid and water.
The Stannous Chloride Solution—
The stannous chloride solution is used to reduce the acid-permanganate
solution during analysis. The solution is 20% W/V stannous chloride in 6N
hydrochloric acid. To 100 ml of 6N HC1, 20 g of reagent grade stannous chlor-
ide is added. After thoroughly mixing the solution, it is ready to use.
65
-------
The Sodium Borohydride Solution--
This Is used in conjunction with the analysis of the mercury collection
bubblers. It is used in a bubbler which is connected to the bubbler to be
analyzed. For a full day of analyses, only 15 ml or a convenient level in
the bubbler is needed. It is prepared by adding 1 g of sodium borohydride
(NaBH4) to 15 ml of distilled water.
The Conditioning Module-Installation of Glassware
The Conditioning Module is prepared for use by the installation of the
glassware and by making the appropriate connections. The unit comes with two
pieces of glassware which are to be fastened in with ball and socket clamps.
Inside on the side of the unit there are three 12/5 socket joints. The waste
separator is clipped to the bottom two. The bottom ball joint of the scrubber
coil is clipped to the waste separator. The top socket on the unit wall is
clipped to the ball joint of the scrubber coil. The other ball joint on the
top of the coil is connected to the socket joint on the solution line from
the pump. Figure A-9 shows the unit with the glassware installed.
Gas Flow System
As shown in Figure A-8, the interface apparatus is connected to the probe
with a glass tee. The tee has portions of three ball joints: a 28/15 ball, a
28/15 socket, and a 12/5 ball. It is used to divide the gas flow between the
interface sample line and an exhaust line. To obtain a reliable sample, more
gas than is used by the interface system is drawn through the probe, and the
surplus is exhausted. For simultaneous comparisons, the exhaust system is
composed of two Greenburg-Smith impingers, a 36 1/min. flowmeter, and a -^30
1/min. Thomas Pump. These are connected by 1/2" I.D. rubber tubing and ground
glass joints.
The sample intake line to the interface instrumentation which has a 12/5
socket joint to connect to the glass tee is 30" of 3/16" I.D. amber latex
tubing which connects to an empty Greenburg-Smith impinger. This in turn is
connected to the sample inlet of the conditioning module by another 30" of
amber latex tubing.
In the conditioning module the gas flow is split. Most of the sample
exits the "sample out" connection through a short section of teflon tubing to
the condenser and thence to the diluter and photometer. However, approximate-
ly 0.5 1/min. of it travels through the "waste out" connection to the waste
container along with the waste scrubber fluid. The waste and fluid reservoir
containers are fitted with teflon tubing (1/8" O.D., 1/16" I.D.). From the
waste container this 0.5 1/min. gas flow is drawn to the pump unit where the
flow is regulated and then exhausted.
The condenser between the conditioner and diluter modules is connected
by means of ground joints to condensate collection flask, (Figure A-3).
The gas passes through the side arm of the flask to the two heated 25 mm
Gelman filter holders which are connected by short sections of amber latex
tubing. A heating tape connected to a variac heats the filter holders and
eliminate any problems with water on the Whatman #41 filter paper. These
66
-------
Scrubber
Inlet
c-i,
I
Waste
Separator
Scrubber ti;
Coil
Variable Stroke
Adjustment Knob
Check
Valve
Figure A-9. Top View, Conditioner Module-Cover Removed
-------
filters guard against passage of any residual fly-ash, in the event the scrub-
ber does not remove the last traces. ^SO^ mist ($03 + F^O) is removed at
these filters also.
Teflon tubing (0.25" O.D., 0.19" I.D.) which connects the remainder of
the apparatus conduces the sample gas from the filter holder to the diluter
module. After exiting from the diluter, the gas is directed either through
the 10% silver-alumina pellets or to the DuPont 400 photometer by a glass
three way valve. A polypropylene Y serves to connect the line from the pellets
back to the line to the photometer.
The'sample line is fastened to the teflon cell using polypropylene fer-
rule connections. With teflon tape on the threads, a leak free fitting is
obtained. After exiting from the monitor cell, the gas is exhausted through
the pump unit.
Liquid Flow System
The scrubber solution is held in a 2 gallon container. A teflon line
(1/8" O.D., 1/16" I.D.) connects it to the bottom of the condenser. If destr-
ed, part of this line may be placed in an ice bath so as to improve the
efficiency of the condenser. The fluid then travels out of the top of the
condenser to the "fluid in" connection of the conditioning module. The liquid
passes through the pump and a check valve before entering the scrubber coil.
After passing through the coil, the fluid goes into the waste separator.
Here it is drawn off from the bottom connection along with some gas through
the "waste out" connection to the waste container. The liquid is deposited
here while the gas is exhausted through the pump unit.
68
-------
SECTION A3
INSTRUMENT OPERATION
The Automated Mercury Interface hardware is flexible and can be used in
a variety of ways. With the modular composition of the apparatus, it can be
adapted to meet various testing situations. The conditioning module which
prepares the gas sample for photometric measurements and the diluter module
which regulates and or dilutes the sample may be interchanged or in some
cases omitted to meet different testing requirements. Descriptions will be
given of each of the operating modes.
Non-Flammable Sources
In testing sources such as power plants, there is no danger of sampling
flammable gases. The following sections will indicate different methods of
operating the interface system with non-flammable sources.
Complete Interface System—
When testing a source with non-flammable gas, the system may be set up
to operate as in Figure A-8. After connecting the apparatus, it is necessary
to establish the proper gas flow rates.
The middle two switches of the diluter module are initially kept in the
down position. With the sample line unhooked from the probe tee, the 0-5
1/min. mass flowmeter is adjusted to 1.5 1/min. or to. any other desired set-
ting. The flowmeter on the pump module is adjusted to draw 0.5 1/min. through
the waste line. This makes the flowrate 2.0 1/min. through the sample line
at the probe. With 100 ml of saturated sodium bicarbonate in the impingers,
the flowrate of the exhausted excess gas drawn through the probe is adjusted
to ~20 1/min.
The scrubber solution container is then filled, and a check is made of
the solution flowrate into the scrubber. This is done by unhooking the line
from the top of the scrubber coil and placing it in a graduated cylinder. The
flowrate is determined from the volume of collected liquid, and the length of
time during which the collection occured. A flowrate of 5 ml/min. is suffi-
cient for most applications. If the flowrate is unacceptable, it may be
adjusted using'the "Fluid Pump Stoke Adjustment Knob". Once the flowrate is
correct, it should not need to be changed during the test.
A 25 mm circle of Whatman #41 filter paper is placed in each of the filter
holders. These need to be replaced about every two hours, when sources havino
high paritcle loadings are'monitored. However, they last much longer depend^
ing on the amount of flyash present. The filter holder are kept warm to
69
-------
the touch by the heating tape which is controlled by a variac. If the variac
is set at approximately 5 divisions, the heating tape should provide enough
heat.
The temperature of the conditioning module furnace is generally set at
300'C. This may be varied depending on the testing circumstances. However,
300''C was found to be sufficient for most applications.
With the sample inlet line to the conditioner still unhooked at the probe
tee, a base line is obtained on the 10 mv scale of the recorder. After check-
ing to be sure that all gas and liquid flows are proper, the inlet line is
clipped to the probe tee. Once a steady signal has been achieved, the three
way valve is turned to divert the gas through the silver-alumina tablets. The
signal difference indicates the mercury contribution to the signal.
If the signal goes back to the base line after insertion of the tablets,
the entire signal can be attributed to mercury. However, if it does not, it
is an indication that a nonmercury interference is also giving a signal. The
calibration of these two cases will be discussed in a later section.
It may occur that a very small signal or no signal will be apparent on the
10 mv scale. In this case the 2 mv scale should be used. The contribution of
the mercury to the signal may be determined as before.
If a very high signal is obtained, dilution of the sample will be requir-
ed. The middle two switches on the diluter module front panel are placed in
the "up" position. The set point knob of the low flow mass flowmeter is then
turned to yield a flowrate which gives an acceptable signal. The rotameter
on the back panel of the diluter module must be adjusted to give approximately
the same flowrate through the sample line as before.
The apparatus between the probe and the conditioning module should be
cleaned daily. This involves rinsing the lines and impingers with distilled
water. The probe should be checked periodically and cleaned if necessary.
The pressure regulator in the pump unit should have a filter on the air intake.
If a 4" long, 1" I.D. cylindrical filter is used, it will last over a week in
normal operation.*
Operation Without the Diluter Module--
It may be convenient on occassion to remove the Diluter Module from the
system. The resulting interface apparatus will consist of the Conditioner in
such operation. In these situations, the Diluter Module is replaced by a
flowmeter and a glass valve. With this arrangement the operational procedure
indicated previously is followed with the following exception:
A periodic check of the air flow of the sample line at the probe tee
should be made. This is necessary because the vacuum of the system cause an
A cartridge filter such as the Gelman, Pleated Membrane Capsule, Cat. No.
12104 may be connected by use of a short section of rubber tubing.
70
-------
error in the flowmeter which replaces the diluter. This flowmeter and the
valve also may be heated to avoid condensation if the water content of the
sample if high.
Bypassing the Furnace--
It is possibe that the furnace of the conditioning module may not be nec-
essary in some tests. If mercury compounds are not present or are not of
interest, the furnace may be bypassed. The sample gas then goes directly
to the scrubber coil.
This is done by first removing the line from the sample inlet of the
module. The fitted sample inlet tube may be connected at the side port entry
to the scrubber, marked scrubber inlet in Figure A-9. Then, the sample from
the probe bypasses the furnace and enters the scrubber directly. The furnace
heater switch is kept in the "down" position when operating in this mode.
Otherwise, the operating procedures remain the same.
Flammable Sources
In many plants the gas streams contain hydrogen or other flammable sub-
stances. To test these sources requires that the sample be diluted prior to
entering the furnace. This mode of operation is shown schematically in Fi^jre
A-10. Operating procedures remain essentially the same with an important
exception.
Flammable sources require that the gas be diluted below the flammability
threshold. This is accomplished by changing the order of use of the modules:
The diluter precedes the conditioner. A filter will probably be necessary
in front of the diluter module to avoid clogging the mass flowmeter transducer.
If the mercury level of the source is high, this mode of operation pre-
sents no difficulties. However, with low mercury levels, the diluted sample
may not contain a measurable level of mercury. In this situation it may be
useful to operate as in the preceding section where the furnace is bypassed.
71
-------
ro
D\Y_UTIOH
SNAXSTE.
MODUUE.
EXCESS
VACUUM
Figure A-10. Use of Interface System with flammable substances.
-------
SECTION A4
CALIBRATION
Calibration of the instrument is partly contingent on the type of signal
which is obtained. Two methods of calibration will be detailed which should
cover all situations. The calculations necessary to obtain the mercury con-
centration in the gas stream will also be given.
Calibration of the Spectrophotometer in the Laboratory
Before going into the field, a calibration curve for the photometer in-
stalled In the operating system should be obtained in the laboratory. If it
is found to be valid in the test situation, it is the easiest method of deter-
mining the mercury content of the stack gas. This first method requires mer-
cury levels over/~»10 yg/m3 with little or no signal interferences.
To obtain a calibration curve similar to that of Figure A-ll, it is nec-
essary to have a system for generating various steady levels of mercury vapor.
Such a system is shown in Figure A-1.2. The photometer response to a constant
mercury level is obtained, and this level is then quantitated with acid-per-
manganate bubblers which are prepared as described in Section 2.
To quantitate the mercury level, a calibration curve for the acid-perman-
ganate bubblers must be obtained first. Known amounts of mercury are obtained
from a source such as in Figure A-13, which is used in conjunction with the
data in Table A-l. Various volumetric vapor samples are collected in bubblers,
and each bubbler is then analyzed to give a point on the calibration curve.
To analyze the bubbler, an atomic absorption Spectrophotometer is usually
used. A pump is used to draw 2 1/min. of clean air through the Spectropho-
tometer cell and a sodium borohydride bubbler (described in Section A-2), the
contents of the mercury-containing bubbler is then reduced with 2.0 ml of
stannous chloride solution (Section A-2), shaken vigorously and connected to
the borohydride bubbler. The resulting peak height is measured, and a cali-
bration curve is plotted. This allows quantisation of the mercury level which
in turn established the calibration of the Spectrophotometer.
If the photometer calibration is not carried out using the full Interface
system, it is necessary to correct for the vacuum in the photometer cell.
This correction can be built into the curve if the system is always run at the
same vacuum, or it can be calculated. The magnehelic gauge on the Pump Module
serves to indicate the cell vacuum.
If there are no signal interferences during testing, the method described
above is the easiest for determining the mercury content of the sample gas.
73
-------
10 TO THC CKNTtMCTCK
U. • I*MH CO. ••MM«*4
46 1512
0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.02.2 2
Mercury Concentration (mg/m^)
4 2. 6 2. 8 3.0
Figure A-ll. Du Pont 400 calibration curve.
-------
AIR
CHARCOAL. ^ I O o/o
•5>\V_V&R. A
PEU-ETS
cn
Figure A-12. Mercury vapor generating system.
-------
<3>V=tOON)C3 .Jot NTT
Figure A-13. Mercury reservoir flask for withdrawing mercury vapor samples.
76
-------
TABLE A-l MERCURY CONTENT OF SATURATED VAPOR '
Temperature - Volume Relationships of Mercury Vapor
Temp(°C)
19.0
19.1
19.2
19.3
19.4
19.5
19..6
19.7
19.8
19.9
20.0
20.1
20.2
20.3
20.4
20.5
20.6
20.7
20.8
20.9
21.0
21.1
21.2
(Continued)
ng/cc
12.0
12. 1
12.2
12.3
12.4
12.5
12.6
12.7
12.8
12.9
13.0
13.2
13.3
13.4
13.5
13.6
13.7
13.9
14.0
14. 1
14.2
14.3
14.5
cc/lOOng
8.34
8.26
8.20
8.13
8.06
8.00
7.94
7.87
7.81
7.75
7.70
7.58
7.52
7.46
7.40
7.35
7.30
7.20
7.15
7.10
7.05
7.00
6.90
Temp(°C)
23.0
23.1
23.2
23.3
23.4
23.5
23.6
23.7
23.8
23.9
24.0
24.1
24.2
24.3
24.4
24.5
24.6
24.7
24.8
24.9
25.0
25.1
25.2
ng/cc
16.8
17.0
17.1
17.2
17.4
17.5
17.7
17.9
18.0
18.2
18.3
18.5
18.6
18.8
18.9
19. 1
19.2
19.5
19.7
19.8
20.0
20.1
20.3
cc/lOOng
5.96
5.88
5.85
5.82
5.75
5.72
5.65
5.59
5.56
5.50
5.46
5.40
5.37
5.32
5.30
5.24
5.21
5.13
5.07
5.05
5.00
4.98
4.93
77
-------
TABLE A1 (continued)
Temperature - Volume Relationships of Mercury Vapor
Temp(°C)
21.3
21.4
21.5
21.6
21.7
21.8
21.9
22.0
22.1
22.2
22.3
22.4
22.5
22.6
22.7
22.8
22.9
tig fee
14.6
14.7
14.8
15.0
15.1
15.2
15.3
15.5
15.6
15.7
15.8
16.0
16.1
16.3
16.4
16.5
16.6
cc/lOOng
6.85
6.80
6.75
6.67
6.63
6.59
6.54
6.45
6.41
6.37
6.33
6.25
6.21
6.14
6.10
6.06
6.02
Temp(°C)
25.3
25.4
25.5
25.6
25.7
25.8
25.9
26.0
26.1
26.2
26.3
26.4
26.5
26.6
26.7
26.8
26.9
ng/cc c
20.5
20.6
20.8
21.0
21.2
21.4
21.5
21.7
21.9
22.0
22.2
22.3
22.5
22.8
23.0
23.2
23.4
c/lOOm
4.88
4.85
4.81
4.76
4.71
4.67
4.65
4.61
4.57
4.54
4.50
4.48
4.44
4.39
4.35
4.31
4.27
78
-------
The recorded signal height from the gas can be converted directly into a mer-
cury concentration using the spectrophotometer calibration curve. However if
interferences are present, this method probably will not be useful Despite
the use of the optical filter system in the photometer, small interferences
occasionally appear. In the presence of low levels of mercury vapor these
may be significant.
Dynamic Calibration
If interferences cause the spectrophotometer curve to be unuseable, cali-
bration in the field may be readily carried out.
When an interference is present, the 10% silver pellets may be used to
establish the mercury levels. The pellets quantitatively absorb mercury, and
when the sample gas passes through them, the photometer gives a signal which
defines the nonmercury interference. The difference between this signal and
the signal of the mercury plus the interference is the mercury signal. The
calibration of this mercury signal is then necessary.
This is accomplished using a mercury source such as in Figure A-13. A
known amount of mercury vapor (see Table A-l) is injected into the sample line
at the probe tee, and with a chart speed of 0.1"/sec., a large peak is obta-i'n-
ed. This peak is then used to quantitate the value of the mercury content
of the sample gas.
The area of the peak is calculated in chart paper squares (or by use of
an integrator or by weighing) and a rectangle represents a constant level of
mercury over the period of time that it took to form the peak. The mercury
concentration represented by this level is then calculated from the amount of
mercury which formed the peak and the volume of gas which carried the peak
through the system. The volume is based on the flow calibration established
previously.
The average value of the mercury concentration in the sample gas is ob-
tained as a result. Since the signal area is directly proportional to the
mercury content, comparisons of signals from known mercury levels obtained
in this fashion with signals from test levels will yield the value of unknown
mercury concentrations.
To obtain the best results with this mercury source, it should be placed
in a water bath or at least insulated from rapid temperature change. The
thermometer used in the source usually has a limited range. This may present
problems if the area where the equipment is set up is subject to rapid temper-
ature change. A constant temperature water bath would be most useful if this
problem is anticipated.
79
-------
SECTION A5
MAINTENANCE
Some maintenance other than the routine cleaning of lines and impingers
is required on each of the three modules. The maintencance required is not
difficult and does not require much time.
The Conditioning Module
Maintenance of this module concerns cleaning of the furnace and scrubber
coils. Both coils can be removed from the unit to be cleaned.
To remove the furnace coil, the top layers of insulation are first remov-
ed to expose the top of the coil. Both ends of the coil are then pushed back
through the instrument wall. The coil is then lifted out. A water rinse
followed by blowing out with an air hose is generally sufficient to clean the
coil.
Depending on the particulate level, it is anticipated that after a week
of operation, the scrubber coil may require cleaning. To do this the coil is
removed from the unit. Pass some of the stannous chloride hydrochloric acid
solution through the coil. This solution has proven to be a useful cleaning
agent in the field. In the event that deposits are difficult to remove, a
variety of other cleaning or oxidizing agents may be applied. Rinse carefully
prior to reinstallation in the Conditioner.
Diluter Module
Water in the transducer of the mass flowmeters is an occasional mainten-
ance problem with the diluter module. The symptoms of the problem are erratic
or zero air flow indications by the meter when gas is passing through the unit.
The problem is solved by removal and drying of the transducer.
First, disconnect the power supply from -the unit. To remove the trans-
ducer, the cover must be taken off the diluter instrument. This exposes the
inside of the unit which is shown in Figure A-14. The spacer bar is removed,
and the wires are disconnected from the circuit board. After disconnecting
the cable from the transducer, the module is effectively divided into front
and back halves. The front half may then be removed by removing the four
screws at the bottom of the front panel. This will expose the rear panel
as shown in Figure A-15. The transducer is then removed by disconnecting the
fittings on both ends and by removing the two large screws which hold it to
the back panel.
Once it is removed, the transducer should be rinsed with acetone. The
80
-------
o
Spreader
Bar
High Flow
Transducer
*»
*»
•
Low Flow
Transducer
Transducer
Cable
Figure A-14. Top View, Diluter Module
-------
\ o
Cool j 0g
Fan
Terminal
Low Flow
Servo Valve
Transducer
3 way
Solenoid
Valve
h Flow
Transducer
Figure A-15. Inside View, Diluter Module-Front Section Removed
-------
acetone must then be removed by pumping air through the transducer for about
15 minutes. The transducer may then be reinstalled and used immediately.
The Pump Module
The Pressure regulator is the maintenance concern of the pump module. If
a filter is used on the air intake of the regulator, it is only necessary to
change the filter when it gets dirty. However, if no filter is used and the
regulator gets dirt inside, it is then necessary to take the regulator apart
and clean it. The indication that the regulator is dirty is that the regula-
tor adjustment has no effect on the pressure reading of the magnehelic gauge.
Again, disconnect the power, first. Then the regulator is removed by
disconnecting the gas line and removing the two screws on the bottom of the
unit. A vise is needed to hold the regulator while the top is removed with
a wrench. Care must be taken when doing this as there are two springs inside.
Once apart the orifice is cleaned, and the regulator is then put back together
and replaced in the pump unit.
83
-------
APPENDIX B
SAMPLING PROCEDURES
For the field test, comparison of the instrumental technique with a
simple bubbler collection and Hatch and Ott analytical technique was carried
out.
The wet chemical technique utilized standard midget impingers which were
connected by glass ball and socket joints. The first bubbler contained 15 ml
of saturated sodium bicarbonate which served to remove sulfur dioxide from the
gas stream. The second contained 1.0 ml of 0.25 M potassium permanganate
and 2.5 ml of 1:1 nitric acid diluted to 15 ml. The sodium bicarbonate impin-
ger was used because of sulfur dioxide interference.
A vacuum pump and flowmeter were used to obtain a flow rate of 2.0 liters
per minute through the impingers. They were connected to the glass tee at the
end of the probe and then removed after 2.5 minutes. A sampling time of 2,5
minutes was found to be the optimum with the amount of potassium permanganate
used. After taking a duplicate sample, the samples were analyzed.
With the instrumental methods, a check was made to see if the proper oper-
ating conditions existed. The gas flows were set at 1.5 1/mtn through the
photometer and 0.5 1/mtn through the waste line. A visual check of the scrub-
ber was sufficient to determine that the scrubber fluid flow was appropriate.
The temperatureoof the probe was maintained approximately at the duct temper-
ature (280°-320°F) while the Gelman filter holders were kept warm to the touch.
New filter paper was put in the filter holders every 2 hours.
With the recorder on the 2 millivolt scale and a chart speed of 0,5 inches
per minute, a signal was obtained from the stack gas. The three-way valve was
then turned directing the gas through the 10% stiver-alumina tablets. With
the mercury removed, the gas was directed through the photometer again. After
about one minute, the three-way valve was returned to the ortgtnal position.
The quantity of mercury represented by the difference of the two stgnals was
then determined. From a mercury-containing flask, 1.0 to 2.0 cubic centime-
ters of mercury vapor was drawn into a syringe and injected into the sample
line near the probe. With a chart speed of 1 inch per second, a large peak
was obtained which was used to quantitate the unknown stack gas mercury level
(if the peak was too large for the 2 millivolt recorder scale, the 5 millivolt
scale was used). The amount of mercury was determined by injecting a similar
amount into a midget impinger containing the same acid permanganate solution
84
-------
" ^ PreVl°US procedure'* The conte"ts of the impinger were then
analyzed
Normally the amount of mercury could be determined by the vapor temperature,
however, during the test the thermometer available (19 -27 ) was off scale
most of the time.
85
-------
ANALYTICAL PROCEDURE
A modified Hatch and Ott procedure was used to determine the mercury con-
tent of the various impingers. To avoid possible interferences, the contents
of the impingers were reduced with 3.0 milliliters of 20 percent weight/volume
stannous chloride in 6 normal hydrochloric acid. The released mercury was
recaptured in a midget impinger containing 0.3 ml of 0.25 molar potassium
permanganate and 2.5 ml of 1:1 nitric acid diluted to 15 ml. With a vacuum
pump, air was pulled through the couoled impinqers at 2.0 liters per minute'for
3 minutes.
The photometer section of a GEOMET Model 103 Mercury Monitor was then
prepared to read the peak height obtained from reducing this impinger, An
empty midget impinger was connected to the photometer intake line which has an
air flow of 2.0 1/min. To the empty impinger, another impinger was connected
containing 1 gram of sodium borohydride in 15 ml of water. The contents of
the mercury-containing impinger was then reduced with 2.0 ml of the stannous
chloride solution. After 10 seconds of vigorous shaking, it was hooked to the
sodium boroyhydride impinger, and the peak height measurement was gtven by the
differential voltmeter of the mercury monitor. The amount of mercury present
in the impingers was determined from the calibration curve of the mercury
monitor.
86
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
. REPORT NO. 2. 1
EPA-600/2-78-178
\. TITLE AND SUBTITLE
AUTOMATIC INTERFACING SYSTEM FOR SAMPLING TOTAL
MERCURY IN STATIONARY SOURCE EMISSIONS
7. AUTHOR(S)
D. J. Sibbett and T. R. Quinn
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Geomet, Incorporated
281 4-A Metropolitan Place
Pomona, California 91767
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, N.C. 27711
3. RECIPIENT'S ACCESSION NO.
5. REPORT DATE
August 1978
6. PERFORMING ORGANIZATION CODE
8. PERFORMING ORGANIZATION REPORT NO.
10. PROGRAM ELEMENT NO.
1AD712 BA-H(FY-76)
11. CONTRACT/GRANT NO.
68-02-1789
13. TYPE OF REPORT AND PERIOD COVERED
Final 6/75 - in/7fi
14. SPONSORING AGENCY CODE
EPA/600/09
15. SUPPLEMENTARY NOTES
116. ABSTRACT
I An interfacing system to sample total mercury emissions in source streams and
suitably condition, dilute, and transport the sample to a mercury measuring instrument
was desicmed. fabricated, and tested. The svstem consists nf three romnonents : a
conditioner, a diluter, and a pump module. The conditioner contains a furnace to
thermally decompose compounds at temperatures up to 1000°C and a liquid scrubbing
system to remove particulates and interfering gases, such as sulfur dioxide and
nitrogen dioxide. The diluter module is used at sources where mercury levels are
above the calibration range of the measuring instrument. The pump module draws the
sample through the system, exhausts the waste scrubber liquid, and maintains a con-
stant pressure in the analyzer.
Field tests were conducted at a coal-fired power plant. Good correlation was
obtained between the interface/photomatic analyzer and a reference manual procedure.
Instrumental results ranged from 1.74 to 6.96 yg/m3, with mean value of 4.23 yg/m-3;
reference method results varied from 1.60 to 7.25 yg/m3, with a mean value of 4.66
pg/m3.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b. IDENTIFIERS/OPEN ENDED TERMS
COS AT I Held/Group
* Air pollution
* Mercury (metal)
* Metal vapors
* Hazardous materials
* Sampling
* Chemical analysis
138
07B
11F
lit.
14B
07D
118. DISTRIBUTION STATEMENT
19. SECURITY CLASS (ThisReport)
UNCLASSIFIED
95
RELEASE TO PUBLIC
•^••MM^H^M^HHMMM^MI^^M
EPA Form 2220-1 (Rev. 4-77) PREVIOUS EDITION is OBSOLETE
20. SECURITY CLASS (Thispage/
22. PRICE
87
------- |