&ER&
United States
Environmental Protection
Agency
Industrial Environmental Research
Laboratory
Research Triangle Park NC 27711
EPA-600/7-79-153
July 1979
Development of an
Automatic
H2SO4 Monitor
Interagency
Energy/Environment
R&D Program Report
-------
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 INTERAGENCY ENERGY-ENVIRONMENT
RESEARCH AND DEVELOPMENT series. Reports in this series result from the
effort funded under the 17-agency Federal Energy/Environment Research and
Development Program. These studies relate to EPA's mission to protect the public
health and welfare from adverse effects of pollutants associated with energy sys-
tems. The goal of the Program is to assure the rapid development of domestic
energy supplies in an environmentally-compatible manner by providing the nec-
essary environmental data and control technology Investigations include analy-
ses of the transport of energy-related pollutants and their health and ecological
effects; assessments of, and development of, control technologies for energy
systems; and integrated assessments of a wide range of energy-related environ-
mental issues.
EPA REVIEW NOTICE
This report has been reviewed by the participating Federal Agencies, and approved
for publication. Approval does not signify that the contents necessarily reflect
the views and policies of the Government, nor does mention of trade names or
commercial products constitute endorsement or recommendation for use.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/7-79-153
July 1979
Development of an Automatic
H2SO4 Monitor
by
B. A. Knight, E. F. Brooks, and R. F. Maddalone
TRW Defense and Space Systems Group
One Space Park
Redondo Beach, California 90278
Contract No. 68-02-2165
Task No. 105
Program Element No. INE624
EPA Project Officer: Frank E. Briden
Industrial Environmental Research Laboratory
Office of Energy, Minerals, and Industry
Research Triangle Park, NC 27711
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Washington, DC 20460
-------
CONTENTS
Page
Acknowledgement iv
List of Figures v
List of Tables vi
Sections
I Introduction 1
II Monitor Description 3
2.1 General Operation 3
2.2 Component Enclosures 3
2.3 The Sampling Train 10
2.4 The Sequencer 15
2.5 Electrical Control Panel 17
2.6 Temperature and Flow Control 25
2.7 Conductivity Instrumentation 30
2.8 Additional Equipment 34
2.9 Umbilical Connections 38
2.10 General Timing Sequence 39
III Laboratory Tests 43
3.1 Sequencer Timer Settings 43
3.2 Calibration of Conductivity Cell 44
3.3 Coil Rinse Efficiency 46
3.4 Acid Injection Tests 47
3.5 S02 Tests 48
3.6 System Collection Efficiency 49
3.7 Endurance Test 49
3.8 Results of Laboratory Tests 49
IV Field Test 51
4.1 The Test Facility 51
4.2 Test Description 51
4.2.1 Equipment Set-up 51
4.2.2 Initial Testing 52
4.2.3 Endurance Test 57
4.3 Test Results and Conclusions 59
ii
-------
Section Page
V Recommendations 67
5.1 Size Reduction 67
5.2 Impinger System 67
5.3 Gas Flow Measurement 67
5.4 Organic Removal 68
5.5 Filter Redesign 68
-------
ACKNOWLEDGEMENT
This document describes the apparatus developed on Task 47,
"Development of a Continuous $03 Monitor," on EPA Contract No.
68-02-2165, "Sampling and Analysis of Reduced and Oxidized Species in
Process Streams-" The Energy Technology Department, Applied Tech-
nology Division was responsible for the work performed on the task.
The work was conducted under EPA Project Officer Mr. Frank E. Briden
of the Process Measurement Branch of the Industrial Environmental
Research Laboratory at Research Triangle Park, North Carolina.
Dr. R. F. Maddalone was the Program Manager and the Task Manager
was Mr. E. F. Brooks.
I wish to thank Mr. Maynard D. Cole for his support during the
laboratory and field test phases of the program. The considerable
support from Mr. Steve Newton, Mr. Scotty Walthen, and Mr. Thomas
Augustyn of the TVA during the field test was greatly appreciated.
-------
LIST OF FIGURES
Page
Number
1 Sulfuric Acid Monitor Showing the Probe Unit (A) and Control 4
Unit (B) Enclosures
2 Control Unit Enclosure, Front Doors Open 5
3 Control Unit Front Panels 7
4 Control Unit, Back Doors Open 8
5 Umbilical Connections 9
6 Glass Sampling Train 13
7 Sequencer 14
8 Electrical Control Panel 19
9 Terminal Strip Connections 20
10 Electrical Circuitry Schematic 22
11 Function Circuitry Schematic 23
12 Temperature and Flow Control Panel 27
13 Air Flow Meter Calibration 28
14 Conductivity Instrumentation 31
15 Conductivity Cell Calibration, XI Scale 32
16 Conductivity Cell Calibration, XI0 Scale 33
17 Control Unit Pump Compartments 35
18 Filter Compartment 37
19 Schematic of S03 Monitor 41
20 Probe Unit at Scrubber Inlet 53
21 Probe Connections at Sampling Port 54
22 Filter Element After Test 56
23 Conductivity Recorder Output 60
-------
LIST OF TABLES
Number Page
1 Function Designations 16
2 Sequencer Input/Output Connections 18
3 Fuse Sizes and Circuits 24
4 Thermocouple Locations 29
5 General Timing Sequence 40
6 Sequencer Timer Settings 45
7 Temperature Control Data 58
8 Field Test Results 64
9 Analysis of Shawnee Power Plant Coal 65
10 Particulate Loadings During Field Test 66
VI
-------
SECTION 1
INTRODUCTION
This report has been prepared for the Industrial and Environmental
Research Laboratory of the Environmental Protection Agency, Research
Triangle Park, North Carolina as part of Task 47 of Contract No. 68-02-
2165, "Sampling and Analysis of Reduced and Oxidized Species in Process
Streams." The technical objective of this task was to develop a device
capable of automatic measurement of the mass emission rate of sulfur
trioxide (H?SO. vapor) within a precision of +20%. This document sum-
marizes the results of the development of a single prototype unit,
including a description of the monitor as well as the results of labor-
atory and field tests of the unit.
Currently only manual methods are available for the purpose of
monitoring sulfuric acid emissions: EPA Standard Method 8 and the Con-
trolled Condensation System (CCS). In the latter system, sulfuric acid
is selectively condensed out of a sample gas stream by cooling the gas
in a water-jacketed coil to a temperature below the dewpoint of H?SO..
The condensed acid is then titrated and the acid concentration determined
through a wet-chemistry procedure. The disadvantages of the manual
systems are that they require extensive manpower, they cannot provide
continuous measurements, and there are long delays associated with sample
analysis.
The automatic monitor described herein was designed to eliminate
these problems. The prototype device which was constructed and tested
is an automated Controlled Condensation System, with acid concentrations
being determined by measurement of the electrical conductivity of a
sulfuric acid solution. The monitor is capable of continuous unattended
operation for a 24-hour period in streams of moderate (5 g/m ) particu-
late loadings. Readings of solution conductivity are recorded contin-
uously, and new samples of the gas stream being studied are obtained
every 10 minutes. Sulfuric acid concentration can be determined from
the instrument and associated calibration curves within 5 minutes of
sample acquisition; determination requires only reading recorder output
1
-------
and sample gas volume, obtaining values from calibration curves, and
inserting these values into expressions for ppm concentration in the
gas stream.
The prototype device constructed is capable of detecting sulfuric
acid concentrations in the range from 0.5 ppm to 500 ppm, at temperatures
up to 300°C (527°F) with 3000 ppm S09, 8 to 16 percent H90, and up to
3
9 g/m of particulate matter in the gas stream. Field tests of the unit
indicate a system accuracy of ±7% at 10 ppm concentration under high
(11 g/m ) mass loadings and high (4000 ppm) SO- concentration.
Because of the success of the present prototype and the demand for
a system capable of this type of operation, continued development of
the monitor is highly recommended. Optimization of packaging and filtra-
tion system design are the two major areas where modifications are needed.
The potential demand for the device, coupled with the successful operation
of the present prototype, make continued development very desirable.
-------
SECTION 2
MONITOR DESCRIPTION
2.1 GENERAL OPERATION
The automatic sulfuric acid mist monitor consists essentially of a
heated Vycor probe, a cyclone and filter, a modified Graham condenser
(condensation coil), a conductivity measuring cell, three pumps, and an
orifice meter. In operation, the system condenses sulfuric acid out of
a sample gas stream by cooling the gas in a water jacketed coil which
is maintained at a temperature below the dewpoint of H-SO* but above the
H^O dewpoint. The coil is then rinsed with deionized f^O into a conduc-
tivity cell which measures the conductance of the HpO-H^SO^ solution and
outputs the conductance on a 0-10 mv scale. To automate the procedure,
a sequencer controls the operation of the various components and steps
in the sampling process. The entire sampling train temperature is con-
trolled to prevent condensation of the H-SO, aerosol on the component
walls prior to the condensation coil. The gas sample volume flowrate
can be controlled with a maximum sampling rate of 28 1/min provided by
a 4-stage diaphragm pump. Provision is made for the washing and drying
of the glass sampling train, so that measurements of sulfuric acid emis-
sions can be made on a semi-continuous basis unattended for a period of
time determined primarily by the particulate concentration encountered in
the gas stream and its effect upon the monitor's filtration system.
2.2 COMPONENT ENCLOSURES
The components of the monitor are housed in two environmentally
resistant all-steel cabinets, which are shown in Figures 1 and 2. The
probe unit shown in Figure 1 is a Nema Type 4 enclosure manufactured by
Hoffman Engineering, Anoka, Minnesota (Model A-36H30DLP). The enclosure
size is 91 x 76 x 31 cm (36 x 30 x 12 in.) and its weight is 41 kg
(90 Ibs). This unit contains the glass sampling train and probe connec-
tions, and in operation must be positioned at the stack sampling port.
-------
/. Sulfuric acid monitor showing the probe unit (A) and control unit (B)
-------
Figure 2. Control unit enclosure, front doors open
-------
Lifting lugs are provided at the top of the cabinet for ur,e with a hoist
for applications where the sampling port is located on a platform above
ground level. Handles are provided on two sides to aid in manipulation
of the unit. Openings through the cabinet are provided for the probe
inlet, probe heater, probe thermocouples, air inlet and exhaust, sampling
train thermocouples, solution flows, and gas sample flow.
Figure 2 shows the control unit cabinet which is a Nema Type 12
free standing two door dual-access enclosure. This enclosure was custom
manufactured by Hoffman Engineering of Anoka, Minnesota and is based
upon the model number A-724824FSDAD. It is 183 x 122 x 61 cm (42-1/16 x
48-1/16 x 24-1/16 in.) in overall dimensions and weighs 136 kg (300 Ib).
This cabinet contains all the control systems for the operation of the
monitor, fluid reservoirs, pumps, filters, and instrumentation. It is
a two door dual access enclosure; one set of doors provides access to
the control panels, the sequencer, and the conductivity instrumentation,
and the other set of doors provides access to the pumps and fluid reservoirs
Figure 3 shows the interior of the control unit with the front doors open.
The left panel contains the temperature and flow controllers and temperature
readout. The right side contains the sequencer, the main electrical control
panel, the conductivity meter, and the conductivity recorder. This arrange-
ment provides access to all controls from one position.
Figure 4 shows the interior of the control unit with the rear doors
open. The rear of the cabinet is divided into 4 main compartments. The
right compartment houses the deionization column, the deionized water
reservoir, air filter, air drier, and an orifice meter. The left side
contains 3 compartments; the first houses the purge air pump and 24 VDC
power supply, the second houses a peristaltic pump for transfer of the de-
ionized water, and the third houses the main sampling pump.
During normal operation, the probe unit will operate remotely from
the control unit. Connections between the cabinets are therefore provided
for gas flow, solution flow, waste flow, power, and thermocouples. Figure 5
shows the connections which are mounted on the bottom of the probe unit and
the side of the control unit. The tube fittings are all quick release
double-end shut-off fittings to facilitate removal without loss of fluid.
-------
Figure 3. Control unit front panels.
-------
00
Figure 4. Control unit, back doors open
-------
o
Figure 5. Connections between enclosures: (A) Sample and Fluid Flow Connections,
(B) Thermocouple Connections, and (C) Power Connections
-------
The power connections are provided by four 6-conductor shielded cables
which connect to multi-pin environmentally-resistant connectors. The pin
designations are discussed in the Operations and Maintenance Manual, under
separate cover.
This arrangement, separating the probe unit which must be maneuvered
into the sampling port location from the control unit which can be located
remotely, facilitates field handling. Once the two units are in position,
they can be connected easily by plugging the appropriate connector into
each fitting.
2.3 THE SAMPLING TRAIN
The sampling train used to collect the gas and measure the resultant
solution conductivity is composed of (1) a Vycor sampling probe, (2) a
pyrex cyclone, (3) a quartz filter and support, (4) two glass check valves,
(5) a modified Graham condenser, (6) a conductivity cell and probe, (7) a
measuring vessel, and (8) eight solenoid valves. Additionally there is a
constant temperature bath and circulator, and a magnetic stirrer which
function in conjunction with the glass sampling train. All of these items
are contained in the probe unit and are shown in Figure 6 with the excep-
tion of the sampling probe.
The probe, cyclone, filter and glass tube containing the check valves
are all maintained at a skin temperature of 300°C to prevent condensation
of the aerosol prior to the condensation coil. The gas sample is drawn
through the probe into the cyclone where particulate matter greater than
10y in diameter are separated out. The cyclone (B in Figure 6) is manu-
factured by Joy Manufacturing Company, Western Precipitation Division,
Los Angeles, CA (Part No. A-2070). The gas then enters the quartz filter
holder (C in Figure 6). This holder supports a 113 cm Tissuequartz filter
which removes particulate matter greater than 0.2u in diameter. This filter
holder was custom manufactured; sketches of it and all other sampling train
components are included in the Operations and Maintenance Manual. Following
the filter the gas enters a tube which contains two all-glass check valves
which prevent fluid from the condenser from entering the filter (D in
Figure 6).
10
-------
The controlled condensation coil (E in Figure 6) consists of a
Graham condenser modified to accept a medium quartz frit. The coil is
water jacketed and the water is maintained at 60°C (140°F) by a heater/
recirculator. This temperature is adequate to reduce the flue gas to
below the dewpoint of H2$04 but above that of water. The sulfuric acid
in the gas stream is selectively condensed on the inner surfaces of the
coils during sampling; sampling continues for a specified period of time
as discussed in Section 3.1.
Following sampling, the measuring vessel (F in Figure 6) which is
also water jacketed at 60°C (140°F), is filled with deionized water.
This fixed volume of water is then rinsed through the condenser into the
conductivity cell immediately after the coils are purged with air to remove
any S02 present. The conductivity cell (G in Figure 6) is also water
jacketed to maintain the temperature of the solution at 60°C (140°F).
This is important since electrical conductance of a solution is strongly
temperature dependent. The cell is equipped with a magnetic stirrer
(I in Figure 6, distributed by VWR Scientific) and a conductivity probe
(H in Figure 6, manufactured by Beckman Instruments, number K-l). In con-
junction with a conductivity bridge located in the control unit, the probe
measures the conductance of the H20-H2S04 solution. The stirrer keeps the
solution well mixed while the measurement is being taken.
After the conductivity measurement is taken, the condenser and cell
are washed with deionized water for several minutes and then dried with
60°C (140°F) air which has been filtered and dried. The actions of sampling,
purging, rinsing, washing, and drying are automated through the use of
8 solenoid valves as shown in Figure 6. These valves are of all-Teflon
construction to avoid the corrosive effects of sulfuric acid, and operate
on 24 VDC. The operation of these valves is controlled by a sequencer,
as described in Section 2.4 below.
The water bath used to maintain the 60°C (140°F) temperature of some
of the glassware is manufactured by Haake (Model E52). This unit provides
temperature control within ±1°C with a 1000 watt maximum capacity heater.
11
-------
Figure 6. Glass sampling train
LEGEND:
(A) PROBE INLET
® CYCLONE AND MANTLE
© FILTER SUPPORT AND MANTLE
(D) GLASS TUBE WITH CHECK-VALVES
© CONDENSATION COIL
© MEASURING VESSEL
© CONDUCTIVITY CELL
® CONDUCTIVITY PROBE
© MAGNETIC STIRRER
© SV-1
© SV-2
© SV-3
® SV-4
® SV-5
© SV-6
© SV-7
© SV-8
© WATER BATH CONTROLLER
© WATER BATH COILS
(T) GLASSWARE MOUNTING SUPPORT
12
-------
13
-------
FUNCTION I
SEQUENCER
rUNCTlON 2
FUNCTION J
FUNCTION S
STOP
Figure 7. Sequencer
-------
The heating mantles used to heat the cyclone and filter were custom manu-
factured by Glas-Col Apparatus Company, Terre Haute, Indiana. They
provide uniform heating with a minimum amount of electrical power input.
The glass components of the sampling train are mounted in the probe
unit with fixtures which were designed to reduce mechanical shock to the
components. Each mount is insulated from the glass component it supports
with a layer of asbestos insulation. The height of each component above
the mounting plate can be adjusted to provide alignment between adjacent
components.
2.4 THE SEQUENCER
To automate the operation of the sampling train, a sequencer was
designed to time and control the operation of key components in the sam-
pling process. Shown in Figure 7, the sequencer has 11 output channels
which supply power to each of 11 components which are termed "Functions".
The control panel of the sequencer contains 33 3-digit set point units
which control the time in a sampling cycle in which a function is turned
ON or OFF. When a function is in the "ON" mode, power from a 24 VDC power
supply is provided at the output terminal associated with that particular
function. This power switching is provided by a series of solid-state
relays which are incorporated within the sequencer itself.
Powered by a 24 VDC supply, the sequencer contains an oscillator
which serves as a timing base for the individual function timers. Initia-
tion of a sampling cycle is accomplished by activating a "Clock" circuit
which starts the timers. De-activation of the clock circuit stops the
cycle. A reset circuit resets all timers to 000 when activated. In
operation, the various components of the system are turned on or off by
the function timers which activate or deactivate the associated relay
according to the time (in seconds) selected on the set-point unit. The
time indicated on the set-point unit is the time an event is to occur
after initiation of a cycle. A "Function No. 1 START" time of 100 seconds
means that 100 seconds after initiation of a cycle, power will be supplied
at the Function 1 output terminal. The power will stay on until such time
as is indicated by the "Function No. 1 STOP" time. This type of operation
continues until a period of time determined by the "Sequence Duration"
15
-------
Table 1. FUNCTION DESIGNATION
FUNCTION NUMBER
COMPONENT CONTROLLED
1
2
3
4
5
6
7
8
9
10
11
Peristaltic (water) pump
Purge air pump
Magnetic stirrer
SV-1
SV-2
SV-3
SV-4
SV-5
SV-6
SV-7
SV-8
16
-------
timer. Whenever the cycle duration reaches this time limit, the internal
clock automatically resets to 000 and a new cycle begins. In this manner,
once the "Clock Enable" switch is activated with the sequencer power on,
the eleven functions controlled by the sequencer will operate in the
selected timing sequence continuously until such time as the "Clock Enable"
switch is turned off. Once the "Clock Enable" switch is turned off, the
"Reset" button must be pushed to stop the operation of any functions which
were activated whenever the "Clock Enable" switch was originally turned off.
Table 1 shows the components which are controlled by each of the
function timers. Because the sequencer operates on 24 VDC power, the first
three functions which control the peristaltic pump, purge air pump, and
the magnetic stirrer, operate separate relays which switch 115 VAC power
to those components. Examination of Figure 7 shows that several of the
functions have two sets of START and STOP timers. This is necessitated
by the requirement that during a complete test cycle, several of the com-
ponents must operate more than once. The timers are grouped under each
function according to their normal sequence of operation; the first set
of timers controls the operation of a function, and then the second set
of timers controls the function. This dual-control operation occurs only
for function numbers 1, 2, 4, 5 and 9. The remaining functions only
operate once during each cycle and therefore only have one set of START
and STOP timers.
The input and output connections to the sequencer are contained in
a 9-pin and a 25-pin rectangular connector, respectively. The pin
designations are shown in Table 2. Both the positive and neutral power
leads are switched through the relays in the sequencer. The output con-
nections from the sequencer are then directed to the main control panel
where distribution to the various functions occurs.
2.5 ELECTRICAL CONTROL PANEL
All of the electrically operated components of the monitor receive
power from the electrical control panel, shown in Figures 8 and 9. The
panel is essentially a switching and distribution point for both the
17
-------
Table 2. SEQUENCER INPUT/OUTPUT CONNECTIONS
INPUT: 9-PIN CONNECTOR
• PIN No.
1
2
8
9
FUNCTION
Clock start (+24 VDC)
Clock reset (+24 VDC)
Sequencer power (+24 VDC)
24 VDC neutral
OUTPUT: 25-PIN CONNECTOR
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
FUNCTION #1 (+24 VDC)
1 (neutral)
2 (+24 VDC)
2 (neutral)
3 (+24 VDC)
3 (neutral)
4 (+24 VDC)
4 (neutral)
5 (+24 VDC)
5 (neutral)
6 (+24 VDC)
6 (neutral)
7 (+24 VDC)
7 (neutral)
8 (+24 VDC)
8 (neutral)
9 (+24 VDC)
9 (neutral)
10 (+24 VDC)
10 (neutral)
11 (24 VDC)
11 (neutral)
18
-------
3 O • •
9 • • •
I
Figure 8. Electrical Control Panel
-------
ro
Figure 9. Terminal Strip Connections
-------
115 VAC and 24 VDC power circuits. The front panel contains fuses for all
the 115 VAC components, and switches to operate each of the 11 functions
as described in Table 1 as well as operate the sequencer and the main sam-
pling pump. Power is supplied to the unit through the main power switch,
and indicator lights indicate whether the power is connected (unit plugged
in), and if the power is turned on to the fuses. Figure 10 shows the gen-
eral schematic of the 115 VAC circuit, and the fuse values and components
controlled are listed also in Table 3.
The output from each of the sequencer channels is connected to the
respective component through a 3-position toggle switch as shown in
Figures 8 and 11. In the middle position, the component is isolated from
power. In the "ON" position, 24 VDC power is supplied directly to the
component, bypassing the sequencer. This option is convenient for checking
various component operation without having to use the sequencer. In the
"AUTO" position, the component is controlled by the sequencer.
The "Sequencer Standby" switch supplies power to the sequencer.
No functions will operate in the AUTO mode unless the sequencer circuit is
activated, as indicated by the "power on" indicator on the sequencer front
panel. The "Clock start/stop" switch will supply power to the sequencer
internal timer; activation of this circuit will be indicated at the "Clock
Enable" light on the sequencer front panel. The clock reset button, a
momentary pushbutton, will reset the timers as explained previously;
activation of this circuit is also indicated on the sequencer front panel.
To facilitate connection of the components to the control panel, all
connections are made via five terminal strips, illustrated in Figure 9.
All the 115 VAC neutral terminals are connected together and are connected
to one pole of the main power switch. All the 115 VAC ground terminals
are jumpered together and are connected to the control unit enclosure,
which is connected to the ground terminal of the input power lead. The
"hot leads" from each fuse are also connected to a terminal strip. Each
component requiring 115 VAC power is thus connected to 3 terminals; one for
the "hot" lead, the second for the neutral lead, and the third to ground.
-------
PWR
CONN
ro
Figure 10. Electrical circuitry schematic.
-------
24 Voc
FROM
ELECTR/C/U
P/1NETL
SE"OOE WtCR
AUTOC
TTF
OFF
1("
—
II f
)OM
'LACES
|
2
4
~
,5
x
7
e
0
/o
/ f
/
f
^
2.4 V
PELAY
2T V
24 V
ftrLfl y
h. ./•
'
^ ./•
hx y-
\
V
~
-
H»O PUMP
^/« PUMf>
S T 1 R REft
I
sv-/
1
5 V-^
S V-3
Ii
sv-4
1
S V-S"
1 1
5V/-6
1 I
^ \/~ 7
S V-8
FUNCT/OM SWfTCH
Figure 11. Function circuitry schematic.
23
-------
Table 3. FUSE SIZES AND CIRCUITS
FUSE NO.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
VALUE (A)
5
1**
1^
3
6
1
1%
3
5
i,
'2
h
3
1
1»5
h
COMPONENT CONTROLLED
main sampling pump
purge air pump
peristaltic (hUO) pump
glass tube heater
recirculator, heater
magnetic stirrer
main pump heater
cyclone heater
filter heater
conductivity meter
digital temperature readout
probe heater
spare
24 VDC power supply
conductivity recorder
24
-------
The output of the sequencer is also connected through terminal strips.
To facilitate troubleshooting, the components attached to each terminal
are listed in the Operations and Maintenance Manual.
2.6 TEMPERATURE AND FLOW CONTROL
To prevent condensation of H^SO, vapor in the sampling train prior
to the condensation coil, the temperature of the components must be main-
tained at a temperature above 288°C. As described previously, the
cyclone and filter are heated with heating mantles, and the probe, check
valves, and orifice are heated with heating tapes which are controlled by
a series of autotransformers, shown in Figure 12. Each autotransformer
has a maximum output of 5 amperes at 115 VAC which is more than sufficient
to operate the heaters.
The temperature of the sampling train is monitored at 9 separate
locations with Type 0 (Iron-Constantan) thermocouples. The thermocouple
locations are listed in Table 4. A rotating switch connects these thermo-
couples to the digital readout, manufactured by Omega Engineering, Stanford
Conn. (Model 250-J). No provision is made for recording of these tempera-
tures; they are to be set prior to initiation of sampling and will be
maintained by the power from the Variacs.
Since the measurement of sulfuric acid concentration requires an
accurate measurement of the sampled gas volume, a calibrated orifice meter
was manufactured and installed, shown in Figure 12. The differential
pressure across the orifice is measured and displayed by a magnehelic gauge
on a 0 to 4 in. H20 scale. The orifice is of all stainless steel construc-
tion to reduce corrosion due to acid in the inlet gas. It is located within
the control unit enclosure, immediately prior to the flow regulation valve.
The orifice has been calibrated in SCFM, the results of which are shown in
Figure 13. To reduce the error in meter indications due to temperature
variation, the orifice is wrapped with a heating tape to maintain a gas
inlet temperature of 60°C in cases where the unit is used in low ambient
temperature. In this manner, the volume of sampled gas in standard condi-
tions can be obtained from the differential pressure reading, and
used to calculate sulfuric acid concentration as described in Section 4, once
corrections for temperature and pressure are made.
25
-------
Figure 12. Temperature and flow control panel
LEGEND:
(A) HEATER CONTROL (AUTOTRANSFORMER)
(?) THERMOCOUPLE SELECTOR SWITCH
© DIGITAL TEMPERATURE READOUT
(O) DIFFERENTIAL PRESSURE GAUGE
(T) GAS FLOW CONTROL VALVE
26
-------
. rum
• (LASS TMC
. CCC IH£I
. 1UTEII J«ttT
>• KUK OO.OUK
27
-------
ro
CD
.40 .80 1.2 1.6 2.0 2.4
DIFFERENTIAL PRESSURE - IN.
2.8
3.2
3.6
28.0
- 24.0
ORIFICE METER CALIBRATION
AIR 14.65 PSIA 21°C
- 4.0
4.0
I
co
I—
T3
Figure 13. Air flow meter calibration
-------
Table 4. THERMOCOUPLE LOCATIONS
THERMOCOUPLE LOCATION
NUMBER LULMllUlN
1 Inside stack at probe inlet
2 Probe temperature
3 Cyclone Outlet
4 Filter Outlet
5 Glass tube temperature
6 Inlet to condensation coil
7 Water jacket temperature
8 Probe unit interior temperature
9 Orifice temperature
29
-------
2.7 CONDUCTIVITY INSTRUMENTATION
The output of the monitor is in two forms; the first being differen-
tial pressure from an orifice meter, and the second being millivolt output
from a conductivity meter. From calibration curves, these outputs can be
used to determine the mass of sulfuric acid contained in the sample gas.
As outlined in Section 2.1, the basic operation of the monitor
involves the condensation of sulfuric acid from the sample gas onto the
walls of the condensation coil. This coil is then rinsed with deionized
water and the solution directed into a conductivity cell (G in Figure 6).
Here the electrical conductivity of the acid-rinse solution is measured
with a conductivity bridge and recorded on a strip chart recorder.
The conductivity bridge and recorder are shown in Figure 14. The output
of the meter is then compared with calibration data of conductivity
(micromhos/cm) versus mass of sulfuric acid as shown in Figures 15 and 16.
Thus the mass of acid in the sampled gas can be obtained.
The conductivity cell contains a conductivity probe which houses two
platimum-coated electrodes which are part of the conductivity bridge circuit.
The conductivity probe maintains the electrodes at a fixed separation which
is essential for accurate conductivity measurement. The conductivity probe
is oriented at a 45° angle (see Figure 6) and is continually immersed in
the rinse/wash solution. Vent holes in the probe tip allow for the escape
of any trapped air so that the electrodes are completely immersed in the
solution. The probe is manufactured by Beckman Instruments, Inc.,
number K-l with a cell constant of 1.00/cm.
Since a wide range of acid concentrations can be expected to be
encountered, the conductivity meter used has a dual range capacity.
The meter operates on either a 0-500 micromhos/cm or a 0-5000 micromhos/cm
scale. This is sufficient to cover the range of 0.5-40 ppm at a sampling
rate of 12 Lpm for 7 minutes. A maximum concentration of 500 ppm can be
sampled by reducing the sampling time to approximately 1 minute. The
bridge is also manufactured by Beckman Instruments, Model RA5-X14-B-S8-T8.
30
-------
Figure 14. Conductivity Instrumentation
-------
ro
14
12 -
OO
a:
10
o
(/>
(NJ
a:
QC
o
u.
o
8
CONDUCTIVITY CELL CALIBRATION
RANGE XI
700
200 300
CONDUCTIVITY - MICROMHOS/CM
400
500
Figure 15. Conductivity cell calibration, XI scale
-------
CO
CO
300
280
240
200
s:
o
ir>
C\J
160
u_ 120
o
80
40
CONDUCTIVITY CELL CALIBRATION
RANGE XI0
500 1000 1500
2000 2500 3000 3500
CONDUCTIVITY - MICROMHOS/CM
4000 4500 5000
Figure 16. Conductivity cell calibration, X10 scale
-------
To provide a continuous record of the conductivity meter output, a
single-pen strip chart recorder is used. The conductivity meter output
is 0-10 mV DC in the XI range and 0-1 mV DC in the X10 range. This output
is supplied to the recorder which then gives a continuous trace of the
meter output. The recorder is a Hewlett-Packard model 680-015 with an
electric writing option. It requires the use of electrosensitive paper
(Hewlett-Packard number 9280-0136) which eliminates the problems associated
with ink pens when operated at low speeds, i.e., clogging, leaking, etc.
During normal automatic operation with a 15 minute total cycle duration,
a chart speed of 8 in/hr (20.3 cm/hr) is used. However, for shorter cycle
times, on the order of 1 or 2 minutes, a chart speed of 1 in/min (2.5 cm/min)
is desirable to provide sufficient time resolution between cycles. The
manufacturers manuals on these instruments are included in the Operations
and Maintenance Manual.
2.8 ADDITIONAL EQUIPMENT
There are several other components in the monitor which serve to
support the operation of the components discussed above. Most notable
is the main sampling pump which is a 4-stage diaphragm type air pump.
Manufactured by Air Dimensions, Inc., of Kulpsville, Pa., the model 299
used has a capacity of 28 Lpm at 458 mm of mercury vacuum when the
4 stages are connected in parallel. The heads are of 316 stainless and
are Teflon coated; the diaphragms are of Teflon and neoprene. Replace-
ment diaphragms are provided and can be easily interchanged in the field.
The pump can be seen in Figure 17.
To transfer the deionized water from the reservoir to the sampling
train, a peristaltic pump, shown in Figure 17, is used. A peristaltic pump
is used because the water only comes in contact with the transfer tubing
and not with any pump components, thereby preventing contamination. The
pump is manufactured by The Barnant Corp, Barrington, 111., and is their
model 7541-80 with pump head number 7015-20. The pump operates at 80 rpm
with a capacity of 134 ml/min.
Since the unit is to operate for extended periods unattended, provision
is made for the recycling of the water used to wash the sampling train.
A deionization column, shown in Figure 18 with the reservoir, is used to
34
-------
en
* 6
Figure 17. Control unit pump compartment: (A) VDC Power Supply, (B) Sequencher,
(C) Purge Air Pump, (D) Peristaltic Pump, (E) Electrical Control Panel
and (F) Main Sampling Pump
-
-------
Figure 18. Control unit filter compartment
LEGEND:
(A) DISTILLED WATER RESERVOIR
@ CALIBRATED ORIFICE AND HEATER
© DEIONIZATION COLUMN
@ PURGE AIR DRYER
© PURGE AIR FILTER
© PURGE AIR SOLENOID
36
-------
37
-------
purify the water from the rinse cycle. It is a Barnstead model D-8902
mixed resin cartridge with a total ion exchange capacity of 80 grams
as Nad.
The purge air system utilizes four components: a small capacity
pump (#7530-40 from The Barnant Corp., Barrington, 111.), a dryer and
filter (Gelman Instrument Co., Ann Arbor, Mich.) and a 115 VAC solenoid
valve. The filter housing contains a 0.2 micron filter cartridge, and
the dryer housing contains type 13X molecular sieve. The solenoid valve
is needed to prevent air from being pulled through the purge air pump
during sampling. It opens whenever the purge pump is operated, delivering
air to SV-1 (see Figure 6).
Power for the sequencer, the solenoid valves and relays is provided
from a 24 VDC Elexon power supply. The voltage output is quite stable in
order to provide a clean power input to the sequencer, and the unit sup-
plies 4 A at 24 VDC.
2.9 UMBILICAL CONNECTIONS
Since all of the pumps and controllers are located in the control
unit, while the solenoid valves and heaters are located in the probe unit,
connections are provided for easy assembly and hook-up in the field.
These connections are of three types: (1) electrical power leads to the
electrical components, (2) tube connections for gas and fluid transfer,
and (3) thermocouple connections. The electrical power connections are
made through four 6-conductor environmentally resistant connectors and
cables. The cables and connectors are numbered 1 through 4 on both units
and it is imperative that the numbers are matched during connection. The
pin allocations for each connector are listed in the Operations and
Maintenance Manual.
Provision for the deionized water supply and return, purge air supply
and conductivity cell drain are provided through 1/4" quick disconnect
fittings and tygon tubing. The designation of the fittings are as follows:
Connector #1 - Deionized water from pump
Connector #2 — Air from purge air pump
Connector #3 — Water overflow from SV-3
Connector #4 - Conductivity cell drain
38
-------
In addition, a 3/8" fitting provides connection for the main sampling line
through which the gas sample is drawn. This line is of nylon tubing to
prevent collapse under high vacuum levels. All of these fittings are
double-ended shut-off types which prevent loss of fluid from the tubing
whenever the umbilical is disconnected.
Panel jacks are provided for the thermocouple connections between the
two units. Type J (Iron-Constantan) connectors and extension wire are used.
The male connectors are numbered 1 through 8 and should be plugged into the
corresponding panel jack.
All the umbilical connections and cables are stored in the control unit
under the air filter housing. Access is gained through the rear cabinet
doors.
2.10 GENERAL TIMING SEQUENCE
In order to provide for the automatic operation of the sampling and
measurement systems of the monitor, a general sequence of operations was
established. From this sequence, the operation of the 11 function channels
could be determined which led to the dual operation of function numbers 1,
2, 4, 5 and 9 as described in Section 2.4. As part of the monitor design
criterion, it was determined that all operations involved in data acquisi-
tion would occur in 999 seconds or less for a given cycle. Thus the
sequencer timers were of 3-digit construction providing timing to a
1-second differential.
The general timing sequence for automatic sampling is shown in Table 5
which is keyed to the schematic of the monitor illustrated in Figure 19.
This sequence is the maximum sequence length used under automatic control.
Sample gas is drawn through the condensation coil for 400 seconds, fol-
lowed by the operations of rinsing, data acquisition, washing, and drying
as described in Section 2.3. The times listed in Table 5 are general
approximations as to the time in a cycle when a certain operation will
occur. The actual function timer settings were determined by laboratory
tests. By examination of Table 5 and Figure 19, one can see how the
systems operate during a sampling cycle to collect the sulfuric acid,
39
-------
Table 5. TIMING SEQUENCE FOR S03 MONITOR - 400 SECOND SAMPLING TIME
t - 0 sec.
300
320
400
410
430
500
700
990
SV-5 in 1-2 position (off). All other
valves off. Main pump begins to draw
sample.
SV-2 Open.
SV-3 Open.
SV-2 Closed.
SV-3 Closed.
SV-5 switched to 1-3 position (on).
SV-1 in 1-3 position (off).
SV-4 in 2-3 position (off).
SV-8 switched to 2-3 position (on).
SV-1 switched to 1-2 position (on).
SV-4 " " 1-2 " (on).
SV-8 " " 1-3 " (off)
SV-6 Open.
SV-1 switched to 1-3 position (off)
SV-6 Closed.
SV-2 Open.
SV-7 Open.
SV-1 switched to 1-2 position (on)
SV-6 Open.
SV-7 Closed.
SV-2 Closed.
SV-1 switched to 1-3 position (off)
SV-4 " " 2-3 " (off)
1 1-2 " (off)
SV-4
SV-5 "
SV-6 Closed.
Peristaltic pump on;
measuring vessel
filled.
Peristaltic pump off.
Purge air pump on;
low volume SOp purge.
D.I. H20 rinse of coil
Purge air pump off.
Magnetic stir on; conduc-
tivity measurement
taken.
Peristaltic pump on.
Magnetic stir off; coil
and cell rinsed.
Peristaltic pump off.
Purge air pump on;
coil and cell dried.
Purge air pump off.
Main pump begins to
draw sample.
Timer resets to 0; cycle repeats
40
-------
TUBING
sv~'2-
CALIBRATED
PERISTALTIC
PUMP
Figure 19. Schematic of Sulfuric Acid Monitor
41
-------
measure the acid content, and prepare the system for the next sampling
cycle. The solenoid valves are primarily responsible for isolating the
rinsing and drying systems from the main sampling system, and then allow-
ing the rinse solution or dry air to pass through the appropriate compo-
nents during the appropriate time in the cycle.
42
-------
SECTION 3
LABORATORY TESTS
Once the monitor construction was complete, an extensive series of
laboratory tests were performed to characterize the performance of the
systems and to calibrate the output of the instrumentation. The purpose
of these tests was to determine the system accuracy and to identify any
problem areas before performing an actual field test of the unit.
3.1 SEQUENCER TIMER SETTINGS
The timing sequence outlined in Table 5 is a general sequence only;
it does not provide for the overlap required in the operation of the
11 functions to account for component response times. Here the versatility
of the 3-digit set point timers becomes most apparent, for the adjustment
of the START or STOP time of a component can be performed by rotating the
proper digit to the desired setting to allow a component to operate at
precisely the right moment in a cycle.
Using manual operation of the 11 functions, it was determined by
observation of the condensation coil and fritted disk within the coil that
a drying time of 280 seconds would ensure that all moisture in the sampling
system would be removed following a wash cycle. Next, by placing a 1 molar
sulfuric acid solution in the conductivity cell in a sufficient volume to
obtain a reading of 5000 micromhos/cm (the maximum concentration readable
with the conductivity meter), the time necessary to wash out the sampling
system was obtained. A wash cycle of 200 seconds was found to be sufficient
to wash out the highest acid concentration expected. Thus 480 seconds of
a cycle must be devoted to washing and drying the sampling system.
A period of 85 seconds was allotted for taking the conductivity mea-
surement, a time which allowed for complete mixing of the HpO-H^SO, solution
in the conductivity cell. Of the 999 seconds in a cycle time, approximately
400 seconds were available for actual sampling of the gas. This figure was
used as the basis for setting the function timers for the maximum sampling
time permissible. By observing the operation of each of the components,
the START and STOP times on each function channel were adjusted until the
monitor operated satisfactorily under automatic control.
43
-------
As an example, the timer settings for a 400 second sampling time
(999 second cycle time) are shown in Table 6. It must be noted that the
time allocated for reading the solution conductivity, washing the system,
and drying the system, are fixed. However, the sampling time can be
reduced from 400 seconds to 60 seconds depending upon the concentration
of acid in the sample gas stream. At a high concentration (300-500 ppm)
a short sampling time is used to keep the conductivity meter reading
within range. At a low concentration (0.5-40 ppm) the maximum sampling
time is used to ensure accurate readings of conductivity. The Operation
and Maintenance Manual contains a table of suggested sampling times and
sample gas flow rates for various acid concentrations. The timer settings
are all reduced the same number of seconds according to the sampling time
variation from the standard 400 seconds. Thus, if a sample time of
300 seconds is desired, all the timer settings are reduced 100 seconds
from those shown in Table 6.
3.2 CALIBRATION OF CONDUCTIVITY CELL
The key component of the monitor is the conductivity cell; here is
where the determination of the mass of sulfuric acid collected is made.
Because of its importance, considerable care was taken in calibration of
the cell. A 1.000 molar solution of H2$04 was prepared and a precision
glass syringe obtained. The conductivity cell and other glassware were
maintained at 60°C, and various volumes of the H2SO. solution were
injected directly into the cell via the cell vent (outlet to SV-6 in
Figure 6). At each time when an acid solution volume was added, the
measuring vessel was filled and the deionized water was blown through
the condensation coil into the conductivity cell. After 85 seconds, the
reading was taken with the stirrer operating. It was noted that the opera-
tion of the stirrer did not affect the reading; however, the meter needle
deflected whenever the stirrer was first turned ON or OFF. The conduc-
tivity instrumentation therefore only responded to the transient magnetic
field created during startup or shutdown of the stirrer.
In this manner the conductivity cell was calibrated under conditions
exactly duplicating those encountered during normal operation of the
monitor. The water volume in the cell, the cell temperature, and the
44
-------
Table 6. SEQUENCER TIMER SETTINGS
400 SECOND
Function
Number
1
2
3
4
5
6
7
8
9
10
11
SAMPLING TIME; 990 SECOND
H20 pump
purge air pump
magnetic stirrer
SV-1
SV-2
SV-3
SV-4
SV-5
SV-6
SV-7
SV-8
CYCLE
Time
On
300
500
406
710
415
415
705
300
500
300
415
400
415
730
499
405
DURATION
Time
Off
320
700
435
990
655
436
990
330
700
330
990
990
436
990
732
415
45
-------
stirrer operation were identical to that of a normal operating condition.
Calibration curves were generated for the two conductivity meter ranges,
taking 3 sets of measurements at different times for each volume of
sulfuric acid solution injected. These results are shown in Figures 15
and 16. It can be seen that the relationship between the volume of sul-
furic acid injected into the conductivity cell to the conductivity meter
output is linear. From these calibration curves, the mass of acid col-
lected during a test can be obtained because the mass of sulfuric acid
in a 1.000 molar solution is fixed. A representative calculation is
shown in Section 4.3.
3.3 COIL RINSE EFFICIENCY
The operation of data acquisition involves the rinsing of a fixed
volume of the deionized water through the condensation coil, thereby
washing the sulfuric acid which has condensed on the walls of the coils into
the conductivity cell. The efficiency of the rinse was examined in the
following manner.
Sulfuric acid (150 ul of 1.000 molar solution) was injected into the
conductivity cell and 1 measuring vessel volume of deionized water was
added. The conductivity meter reading was recorded. This process was
repeated with 2 and then 3 measuring vessel volumes of water added to
the conductivity cell. This established a base line for the rinse
efficiency check.
The same volume of 1.000 molar H2S04 was then injected at the inlet
to the condensation coil and the measuring vessel emptied through the
coils into the conductivity cell. The conductivity meter reading agreed
with the base line reading within ± 0.7%. The process was repeated with
2 and 3 measuring vessel volumes of water rinsing through the coils.
These readings agreed with the baseline readings within ± 1.1%. Three
sets of readings were taken at each test which provided repeatability to
the tests. From these results it was concluded that the present rinse
volume was sufficient to rinse the coils completely.
46
-------
3.4 ACID INJECTION TESTS
A sulfuric acid aerosol generator consisting of a calibrated syringe
pump, a heated quartz injection manifold, a precision compound gauge,
five flowmeters, and three thermocouples was prepared for further testing
of the monitor. The injection manifold was attached to the probe inlet
connection of the cyclone, so that the manifold simulated the probe
inserted in a sulfuric acid ladden gas stream for the tests. Oxygen,
nitrogen, and S02 supply bottles were connected to the flowmeters and
the meters calibrated. These gases would represent a typical stack gas
composition for the tests. The compound gauge was used to monitor the
manifold pressure, and the main pump adjusted so that a slight negative
pressure (25 mm Hg vacuum) existed at a flow rate of 8 Lpm through the
system.
The heaters and mantles were checked for operation and proved to work
quite well. The glassware prior to the condensation coil was heated to
300°C by using only 50% of the capacity of the autotransformers. With
the probe unit door closed, the internal temperature of the unit reached
45°C after 8 hours and stabilized at that point.
The system was then leak checked prior to each sampling test; if any
flow registered on the flowmeters with the injection manifold sealed, the
system was dismantled until the leak was found. No unusual difficulty was
encountered in achieving a leak-free system.
Gas consisting of 8% 02 and 92% N2 was sampled at 8 Lpm through
the system. The entire sampling system prior to the condensation coil,
Including the acid injection manifold, was maintained at 290-300°C, and
the manifold pressure was maintained at 25 mm Hg vacuum. A total of
49 tests were performed using the system under automatic control.
For each test, the syringe pump was operated for a 3.75 minute
period, injecting a 1.000 molar H2S04 solution into the injection manifold
at rates varying from 0.0042 mL/min to 0.059mL/min, depending upon the
concentration of sulfuric acid desired for the test. The conductivity
meter reading was recorded on the conductivity recorder. These tests
covered the range from 1 to 100 ppm, and were repeated 7 times. The
results are summarized in Section 3.7.
47
-------
As a further check of the system operation, several tests were
performed with the acid being injected for 3.75 minutes but the gas
being sampled for 16.6 minutes, an additional 12.8 minutes of sampling
after the acid injection stopped. This was to determine if acid was
being held up in the system during the normal 400 second sampling time,
and if this acid could be recovered by additional gas sampling. The
measurements showed that almost no difference occurred with these read-
ings over those taken with a 400 second sampling time; the readings
agreed within t 2%.
From these tests the operating limits of the monitor could be
established. At the sampling rate of 12 Lpm, an acid concentration of
0.5 ppm will provide a conductivity reading which will be 20% of full
scale in the low range of the conductivity meter. This is the recommended
limit of system operation; readings lower than this may be too erratic due
to instrumentation limits. Converesely, by sampling for only 60 seconds
at a rate of 8 Lpm, a concentration of 500 ppm can be accommodated by the
upper range of the instrumentation.
3.5 S02 TESTS
At the reduced temperatures encountered in the condensation
the presence of water vapor in the gas stream can cause the S02 pr
to be oxidized to S03. The critical factor in preventing this is mat
taining the temperature above 60°C. In order to ensure that any
present in the gas sampled does not affect the SO, measurement, se\
tests were performed to determine the S02 effects upon the system.
The syringe pump was filled with distilled water and operated
same manner as in the acid injection tests described above. An S02
tration of 4000 ppm in the inlet gas was used with an oxygen flow rate!ef
0.64 Lpm (8% 02), a nitrogen flow rate of 6.75 Lpm, (82% N2) and a water flow
rate of 0.48 Lpm (6% H20). The S02 flow rate was 0.125 Lpmi. After per-
forming 9 tests using the water injection system at 4000 ppm S0~ concen-
tration, it was demonstrated that the S02 effect could barely be detected
at the most sensitive range of the conductivity meter. Readings on the
order of 6 to 8 micromhos/cm were obtained.
48
-------
3.6 SYSTEM COLLECTION EFFICIENCY
Finally, the collection efficiency of the glass sampling system was
investigated. The syringe pump used had a very steady feed rate; the
conductivity readings were fairly constant between tests at a given
syringe injection rate. However, the absolute value of the feed rate was
not necessarily that as advertised by the manufacturer. To calibrate the
pump, a 1.000 molar hUSO. solution was pumped into sample bottles using
feed rates ranging from 0.0042mL/min to 0.059mL/min for a period of
3.75 minutes. This duplicated the conditions used in the acid injection
tests.
The sample bottles were then titrated to determine the volume of
1.000 molar H,>S04 each contained. As an example, the pump operating for
3.75 minutes at a feed rate of 0.059 mL/min pumped 201.3 yL. By using
this accurate volume of acid with the results of the acid injection tests,
it was found that 92% of the acid injected at the probe inlet was recovered
at the conductivity cell.
3.7 ENDURANCE TEST
Since the monitor is to operate for at least 24 hours on a continuous
basis, an endurance test was performed demonstrating that capability. The
system was operated under automatic control with the glassware heated for
a 24 hour period. Every four hours, a specific quantity of 1.000 molar
HpSO, was added at the top of the condensation coil, and the reading taken
after the coil was rinsed and the solution added to the conductivity cell.
These readings were compared with the conductivity cell calibration curve.
During the entire period, no degradation in the system performance could
be detected. It was shown that the length of operation of the monitor on
a continuous basis will be determined primarily by the particulate loading
in the gas stream and its effect upon the filter.
3.8 RESULTS OF LABORATORY TESTS
From the outcome of the laboratory tests, it was demonstrated that
the monitor was capable of automatic operation to detect sulfuric acid
49
-------
concentrations in the range of 0.5 ppm to 500 ppm. It was also dem-
onstrated that the system was unaffected by 4000 ppm SOp in a gas
stream containing 6% H20,
The response of the system to H2SO^ concentration has been shown
to be linear, with an overall system accuracy of ± 9% at 0.5 ppm acid
concentration, and t 7% at 10 ppm acid concentration. The system col-
lection efficiency was shown to be 92%.
Setup and operation of the monitor was shown to be quite easy.
Adjustment of sampling times was a simple task, facilitated by the thumb-
wheel units used in the sequencer construction. Manipulation of the
probe unit was not difficult with two operators, and so the present
design was deemed quite satisfactory to continue in the test program.
50
-------
SECTION 4
FIELD TEST
Upon demonstration of the monitor's performance in the laboratory,
it was desired to demonstrate its performance in an actual field environ-
ment. In particular, the performance of the system's filtration system
needed to be tested as well as the duration cycle of the monitor under
field conditions. Hence, the monitor was prepared in its final deliv-
erable condition with all components securely mounted in the two enclo-
sures, and a field test scheduled at the Shawnee Steam Plant in Paducah,
Kentucky.
4.1 THE TEST FACILITY
Located 29 miles west of Paducah, the Shawnee Steam Plant is part
of the Tennessee Valley Authority's system of power plants. The unit
houses 10 generators with a total maximum power output of 1000 megawatts.
Associated with the power plant is a pilot scale scrubber operation being
managed by Bechtel Corporation. This venturi-type scrubber is being used
to reduce the S02 levels of the effluent gas of 1 out of the 10 coal-fired
boilers of the facility.
The automatic SO., monitor was transported to the Shawnee Power Plant
and the unit was hoisted to the third story level of the wet scrubber
facility by the use of a small hoist. This level contained the sampling
ports located at the inlet to the scrubber. The ports were 7.6 cm (3 in.)
in diameter, located at 90° intervals along the outer diameter of the
102 cm (40 in.) I.D. scrubber inlet. Two inlets were located at the same
level corresponding to two separate scrubbers, only one of which operated
at a given time. Both were tied into the same duct leading from the same
boiler.
4.2 TEST DESCRIPTION
4.2.1 Equipment Set-up
The glassware components which were shipped separately from the main
units were inspected upon arrival. Following the procedure as outlined
in the Operations and Maintenance Manual, the glass sampling train was
51
-------
assembled into the probe unit. The umbilical connections between the two
enclosures were made and the operation of the system components were
checked.
The sampling probe and its replacement, however, were severely
damaged during shipment. The probe was dismantled and a Teflon Swagelok
union used to repair the fractured vycor inner element of the probe.
Replacement vycor tubes were manufactured at TRW and sent to the test
site. In the interim, the test program was continued using the repaired
probe.
The probe unit was suspended on a rail at the sampling port at the
scrubber inlet. A set of turnbuckles held the unit and provided a height
adjustment for the probe. The enclosure was free to move along the rail
since the turnbuckles were held by a series of rollers. Because of space
limitations, it was not possible to mount the probe into the unit before
inserting the probe into the duct. Consequently, the probe was inserted
to the centerline of the duct (68.6 cm from the port flange face) and the
probe unit rolled into position so that the probe joint seated into the
cyclone inlet. Figures 20 and 21 show the probe unit in position at the
scrubber inlet prior to initiation of a test. Manipulation of the probe
unit proved to be fairly easy because of its relative light weight and
through the use of the roller suspension unit provided at the sampling
port.
4.2.2 Initial Testing
To ensure that the calibration of the conductivity probe had not
been altered during shipment, the conductivity cell and probe were
recalibrated. A 0.996 molar solution of H2S04 was prepared and injected
directly into the conductivity cell via the outlet to SV-6. The results
of this calibration provided conductivity readings nearly identical to
those obtained in the laboratory calibrations. It was determined that
the calibration curves obtained initially were valid and the unit was
readied for sampling.
52
-------
Figure 20. Probe unit at scrubber inlet
53
-------
Figure 21. Probe connections at sampling port
54
-------
With the probe inserted in the duct with the nozzle facing down-
stream the components were slowly heated to their prescribed temperatures.
Because the duct gas temperature was 150°C, a lower power setting on the
probe heater was required to heat the probe to 300°C.
Automatic sampling was initiated and the system performed exactly
as prescribed using the repaired probe. After one hour of sampling,
the manual Goksoyr-Ross system was prepared and the probe inserted
53.4 cm into the duct from a port located 90° from the automatic system.
A manual sample was taken for a 20 minute period, and the readings
recorded from the automatic system. The results are discussed in
Section 4.3.
After 2 hours of sampling, it appeared that water vapor from the
sampled gas was condensing in the vacuum lines and in particular in the
magnehelic gauge. A water trap and absorber were prepared and, during
the data acquisition and washing part of a cycle, were added to the
vacuum line prior to the magnehelic gauge. A rotameter and dry test
meter were also added to provide back-up data on the gas flow; it was
not known what effect the condensed water would have on the magnehelic
gauge.
At 3 hours and 45 minutes after initiation of automatic sampling,
the flow rate through the system began to drop off rapidly. Examination
of the medium frit in the condensation coil showed a covering of a dark
brown residue of unknown constitution. The system was shut down and
the probe removed from the sampling port.
The condensation coil was removed and cleaned for 2 hours in hot
chromic acid to remove the substance trapped in the frit. The filter was
removed and photographed, as shown in Figure 22. The Tissuequartz filter
pad was impregnated with a very fine brown dust. The filter was weighed
and it was determined that 0.26 grams of material had been collected in
the 3.75 hour period sampling at 4.16 Lpm (standard conditions).
55
-------
in
CTV
Figure 22. Filter element after test
-------
4.2.3 Endurance Test
Upon receipt of the replacement probe core, a new probe was
constructed to replace the repaired probe used in the initial phase of
testing. A new filter element was placed in the filter holder and the
now clean condensation coil was placed into the probe unit.
Officials at the power plant indicated that at approximately
2.5 hours into the initial testing run, the first scrubber shut down
momentarily, causing a large volume of particulates to be vibrated from
the walls of the duct. The momentary shutdown was caused by equipment
failure so a decision was made to stop operation of that scrubber.
Because of this, the monitor had to be moved to the second scrubber
inlet and the probe unit set up for testing once again.
The system was leak checked and the joints tightened until no flow
registered on the rotameter at 625 mm Hg vacuum. The glassware was then
heated to 300°C with the probe outside of the duct. Table 7 shows the
thermocouple readings before and during a test, and the corresponding
autotransformer settings. With the pump turned ON and the flow adjusted
for 5.58 Lpm (standard conditions), the probe unit was rolled into posi-
tion with the probe inserted to the centerline of the second duct
(68.6 cm from the flange face).
Approximately 1.5 and 6 hours after initiation of automatic sampling,
a sample was again taken with the manual Goksoyr-Ross system with the probe
inserted 53.4 cm into a port located 90° from the automatic system. The
results are discussed in Section 4.3.
Nine hours after initiation of the endurance test, it became apparent
that a brown oily substance was being collected on the frit in the conden-
sation coil. The air flow rate through the system was fairly unaffected,
but the pressure drop through the system had risen from 227 mm Hg to
375 mm Hg. Problems arose during the rinse cycle of sampling. The
substance on the frit seemed to impede the flow of water throught the coil.
At first, the conductivity readings began to decrease, indicating that not
all of the acid in the coils was being washed into the conductivity cell.
57
-------
Table 7. TEMPERATURE CONTROL DATA - TYPICAL DATA RUN
THERMOCOUPLE
NUMBER
1
2
3
4
5
6
7
8
9
LOCATION
Inside Stack
Probe
Cyclone Outlet
Filter Outlet
Glass Tube
Condensation Coil Inlet
Water Jacket
Probe Unit Interior
Orifice
TEMPERATURE °C
BEFORE TEST
152
316
336
310
310
40
60
40
30
DURING TEST
FLOW = 6.1 1pm
152
300
310
308
304
159
60
48
33
VARIAC
SETTING
-
82
53
48
78
-
-
-
0
en
00
-------
Then the water used in the wash cycle began to back up the coils and
into the purge air line (between SV-1 and SV-2). Because the water took
longer to be blown out of the system, the timing of SV-6 and SV-7 was
invalidated. The water level begain to rise in the conductivity cell and
so the test was halted.
A second condensation coil was installed and the system leak checked.
The glassware was then heated to 300°C again with the same filter element
in the filter holder. With the probe reinserted into the scrubber inlet
and an identical flow rate used, the pressure drop through the system
returned to its original 227 mm Hg level.
The system operated for 2.5 more hours until the same problem occurred
with the second frit. Water flow through the frit was again restricted.
During the time when the measuring vessel was rinsed through the coils,
the water appeared to bead up on the firt rather than pass smoothly
through it, as was the case in the laboratory tests. Because the timing
of SV-6 and SV-7 could not be adjusted long enough to accommodate the
erratic water flow, the test was stopped. The probe was removed from the
duct and the monitor turned off.
The filter was weighed after it had cooled and it was determined
that 0.46 grams of particulates had been collected during the 11.5 hours
of testing. No particulates could be seen in the region downstream from
the filter, indicating that the filter seal had been effective.
4.3 TEST RESULTS AND CONCLUSIONS
A total of 53 tests were performed during the field demonstration
covering a one week period. The sulfuric acid concentration ranged from
7 ppm to 20 ppm at the inlet to the scrubbers, with each scrubber
receiving a fairly constant concentration of acid with respect to time.
The conductivity recorder maintained a constant record of the con-
ductivity meter output. A typical trace of conductivity output is shown
in Figure 23. It can be seen that the system output, and hence the
sulfuric acid concentration, remained fairly constant between tests.
59
-------
Figure 23. Conductivity recorder output
-------
The trace for each test can be seen to remain at a steady baseline during
the sampling phase of the test cycle. This corresponds to the conductivity
of the wash-cycle water in the conductivity cell which is maintained at
the height of SV-7. With the addition of the acid contained in the rinse
solution from the measuring vessel, the conductivity meter output can be
seen to rise suddenly to a new level, determined by the conductivity
meter range and the recorder span. With the meter range in the XI posi-
tion and the recorder span at 10 mV DC, full scale on the recorder (100%)
corresponds to a conductivity of 500 micromhos/cm. With the meter range
in the X10 position and the recorder span at 5 mV DC, a scale reading of
20% on the recorder corresponds to a conductivity of 5000 micromhos/cm.
The response of the system to conductivity is linear, so that intermediate
readings correspond to a fixed percentage of the maximum scale reading.
The recorder trace can be seen to remain at the new level for approx-
imately 1.5 minutes at which point the trace begins to decay exponentially
to the baseline level during the washing phase of the cycle. The baseline
measurement will continue through the drying phase of the cycle and the
sampling phase of the next cycle.
By observing the baseline for the curves, the condition of the wash
water and hence the deionization column can be monitored. A steady climb
of the baseline between successive, measurements indicates that the column
is not deionizing the water sufficiently and should be replaced. This is
covered in the Operations and Maintenance Manual.
During the three periods when a sample with the manual Goksoyr-Ross
system was being taken, the dry test meter readings were recorded. The
total volume of gas sampled during each cycle was recorded along with the
meter gas temperature and pressure. The gas flow through the system was
not steady; an initial period of high mass flow occurred at the beginning
of a cycle, lasting for several seconds. This was due to the fact that
during the washing and drying periods the sampling train is valved-off
from the vacuum pump by SV-5, while the vacuum pump remained running
evacuating the lines and meters betwen the probe unit and the control
unit. Whenever SV-5 opened again, the pressure difference in the system
resulted in a high flow condition until steady state was reached.
61
-------
The gas volume sampled in each case was converted to standard
conditions using the following relationship:
where
M.V. = meter gas volume reading
TM = meter gas temperature (°F)
PBAR = amkient barometric pressure (inches Hg)
PJ^J = meter gas vacuum (inches Hg)
To calibrate the ppm concentration of sulfuric acid on a volumetric
basis, the acid volume recovered was converted to a gaseous volume at
standard conditions. Following the procedure as outlined in the
Operations and Maintenance Manual, the steps necessary were to
1. Obtain the volume of 1 molar FUSO. collected using the
conductivity meter output and the calibration curves
presented in Figures 15 and 16.
2. Obtain the mass of HpSO, in milligrams by using
AM = [volume of acid in microliters] [0.09808 mg/yl]
3. Using the ideal gas law, obtain the volume of gaseous
HUSO* at 294°K and 1 atmosphere pressure. The volume
should be in milliliters, obtained by using
A.V. = [mass of acid in milligrams] [0.24610 ml/mg] (2)
Since 1 ppm on a volumetric basis corresponds to 1 ml of acid per cubic
meter of sampled gas, the gas volume sampled was converted to cubic meters.
It was further assumed that the sample gas contained 9% HLO by volume
which was added to the dry gas meter readings to obtain a total gas volume.
The acid concentration was then
C-A.V/6.V. (3)
62
-------
Table 8 shows the comparative results of the automatic system versus
the manual system. The samples were taken at nearly identical locations
and times. It can be seen that the first measurement with the automatic
system using a repaired probe gave a reading 7% lower than the manual
system. Since the Teflon union used to repair the probe was unheated,
it is probable that sulfuric acid was condensing on the relatively cool
surface and therefore was not recovered.
Subsequent readings with the new probe provided much better agree-
ment between the systems, with a standard deviation of 4.9%. It can be
seen that the readings with the automatic system were consistently below
that of the manual system. This may be due to the transient behavior
of the gas flow through the gas meter, resulting in a higher gas volume
reading than what was actually sampled.
To provide reference information on the conditions encountered at
the scrubber inlet, an analysis of the coal fired during the testing
program was obtained. Table 9 shows the analysis of the coal as received
and as fired. In addition, samples were taken at the inlet to the second
scrubber using an Aerotherm High Volume Stack Sampler to determine par-
ticulate loadings and particle size distribution. These samples were
taken at the same level as the sulfuric acid monitor probe, so that an
accurate measurement of the particle sizes encountered by the monitor
were obtained. Table 10 shows the analysis of the particle size samples.
Examination of these results indicate that the monitor can be
expected to operate for periods greater than 12 hours in length in areas
where the particle concentrations are less than those encountered in the
field test. Because the particles were of such small size, a very small
mass was collected in the cyclone collection flask. Particles greater
thanlOy would be collected at the cyclone in streams where particles of
this size were encountered. In streams with a low grain loading (5 g/m3)
of particles less thanlOu in diameter, the monitor can be expected to
operate for periods much longer than 24 hours in length. In regions down-
stream of a scrubber, where particle loadings and organic compounds are
minimal, the monitor could operate for quite extensive periods, possibly
several days in duration.
63
-------
Table 8. FIELD TEST RESULTS
TEST
NUMBER
5
18
51
LOCATION
North Scrubber Inlet
South Scrubber Inlet
South Scrubber Inlet
PROBE USED
Repaired
New
New
SAMPLE AIR FLOW
(1pm)
5.8
6.1
9.9
ACID CONCENTRATION - ppm
MANUAL
SYSTEM
19.51
7.39
8.82
AUTOMATIC
SYSTEM
17.07
6.79
8.75
-------
er>
in
Table 9. SHAWNEE FACILITY NUMBER 10 BOILER COAL ANALYSES
Sample date: 8/2/78
ANALYSIS
PERFORMED
AS
RECEIVED
AS
BURNED
PERCENT BY WEIGHT
MOISTURE
12.6
12.6
VOLATILE
MATTER
38.6
33.7
FIXED
CARBON
43.8
38.3
ASH
17.6
15.4
SULFUR
3.5
3.2
HEATING
VALUE
J/KG
2.689 X 107
2.475 X 107
WEIGHT
PER
MONTH
KG
39,100
-------
Table 10. SHAWNEE PARTICULATE LOADINGS
Date: July 24, 1978
Location: Uet Scrubber
GRAIN LOADING
% MOISTURE
% ISOKINETIC
INLET
12.104
8.83
1.45
MASS LOADING
OUTLET
0.0854 G/SCM
14.13
11.30
(DRY)
BRINK SIZE DISTRIBUTION
CYCLONE
STAGE 1
STAGE 2
STAGE 3
STAGE 4
STAGE 5
FILTER
D 50
4.60
2.76
1.92
1.05
0.70
STGWTMG STG UT % STG CUMUL %
% ISOKINETIC 7.43
GRAIN LOADING G/SCM 5.1758
IMPACTOR FLOW RATE (Q) 0.0019
VOL SAMPLED CM 0.015
47.96
6.33
2.77
0.87
0.31
0.06
0.19
82.0
10.8
4.7
1.5
0.5
0.1
0.3
18.0
7.2
2.4
1.0
0.4
0.3
STACK TEMP C 125
STACK PRESS ABS 738.9 mm
DEL PC mm HG 38.35
STAGE
STAGE
STAGE
STAGE
STAGE
STAGE
STAGE
FILTER
% ISOKINETIC
GRAIN LOADING
METER TEMP DEG C
MRI SIZE DISTRIBUTION
D 50 STG WT MG STG NT % STG CUMUL %
36.11
17.40
,53
.89
,61
6.
2.
1
0.70
0.45
0.35
0.100
31.4
16.99
0.72
11
1.
6.
15
12.21
4.99
2.21
5.48
34.
1.
2.2
12.3
24.5
10.0
4.4
11.0
STACK PRESS
VOL METER
65.
64,
62,
49.
25,
15,
11.0
721.1
0.537
66
-------
SECTION 5
RECOMMENDATIONS
As a result of the intensive laboratory and field test program, it
is felt that the present prototype performed satisfactorily, providing
reasonably accurate measurements of sulfuric acid concentration on a
semi-continuous basis. However, several suggestions can be offered for
the improvement of the unit for future development.
5.1 SIZE REDUCTION
Since the present unit was a prototype, size and weight considera-
tions were not primary; development of a workable accurate system was
the goal. Now that the design has been proven, steps should be made to
reduce the size and weight to make a more field-worthy unit.
The control unit enclosure could be reduced to one-half its present
size by repositioning the components. By using solid-state triacs
instead of variacs, considerable size and weight savings could be made
in the temperature control panel. Constructing the cabinet of aluminum
would further reduce the weight so that a control unit weight of approx-
imately 150 pounds might be attained.
Through redesign of the filtration system, the probe unit could be
reduced in size by 20%. Constructing the outer case out of aluminum
would result in a substantial weight savings.
5.2 IMPINGER SYSTEM
The field test demonstrated that water from the sampled gas will
condense in the vacuum lines causing problems with the flow measurement
devices. A water trap and absorber should be added to the sampling
system to eliminate this problem.
5.3 GAS FLOW MEASUREMENT
Since the measurement of the gas volume sampled is critical, the
use of a dry test meter is advised. This type of meter would provide
the total volume of gas sampled during a test cycle, accounting for
67
-------
variations in flow which may occur during initiation of the sampling.
A thermocouple and vacuum gauge should be used in conjunction with the
meter to correct the flow readings to standard conditions.
If it is desired that a magnehelic differential pressure meter be
used, a series of calibration curves at various pressures and tempera-
tures should be developed. The advantage of the orifice meter is one
of weight; however it can only provide gas flow readings at a steady--
state condition.
5.4 ORGANIC REMOVAL
In situations where the monitor will be sampling gas streams con-
taining heavy organic compounds, it appears as a result of the field
tests that provision must be made for the removal of these compounds.
Otherwise, depending upon the condensation temperature of the com-
pounds, they may condense in the condensation coil and present problems
with the operation of the system. Operation in gas streams free of
these organics present no problems; however for universal use the
problem must be resolved.
5.5 FILTER REDESIGN
Although the filter designed and used in the prototype performed
well, handling of the component proved difficult. Care had to be taken
when wrapping the filter element with the Tissuequartz filter pad.
Sealing the pad against the filter support proved to be a tenuous
process - applying too much pressure to the sealing surface would tear
the pad whereas applying too little pressure would not result in a
proper seal. Additionally, because of the filter size the probe unit
enclosure was necessarily large.
It is suggested that a small program be initiated to evaluate filter
designs and develop a new filter. By reducing the filter housing size,
the probe unit may be made smaller and, additionally the S-shaped tube
housing the check-valves (D in Figure 6) might be eliminated.
68
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
^REPORT NO.
EPA-600/7-79-153
3. RECIPIENT'S ACCESSION NO.
„. TITLE AND SUBTITLE
Development of an Automatic H2SO4 Monitor
5. REPORT DATE
July 1979
6. PERFORMING ORGANIZATION CODE
AUTHOR
Task Final; 9/77 - 10/78
14. SPONSORING AGENCY CODE
EPA/600/13
919/541-2557.
NOTESIERL.RTp project officer is Frank E. Briden, Mail Drop 64,
16 ABSTRACT The report describes the development, construction, and testing of a pro-
totype automatic H2SO4 vapor and aerosol monitor. The device was based on the con-
trolled condensation (Goksoyr/Ross) approach to H2SO4 measurement. In this appro-
ach, H2SO4 is condensed out of a filtered gas stream (at 250 C) using a water -
jacketed coil maintained at a temperature (62 C) below the dewpoint of H2SO4. The
H2SO4 collected in the coil is recovered automatically and its electrical conductivity
is correlated with H2SO4 concentrations. The monitor is capable of continuous unat-
tended operation for a 24-hour period in streams of moderate (5 g/cu m) particulate
loadings. Readings of solution conductivity are recorded continuously, and new sam-
ples of the gas stream for analysis are obtained every 10 minutes. H2SO4 concentra-
tion can be determined from the Instrument and associated calibration curves within
5 minutes of sample acquisition; determination requires only reading recorder out-
put and sample gas volume , obtaining values from calibration curves , and inserting
these values into expressions for ppm concentration in the gas stream. The proto-
type can detect H2SO4 concentrations in the range of 0. 5 to 500 ppm, at tempera-
tures up to 300 C, with 3000 ppm SO2, 8-16% H2O, and up to 9 g/cu m of particulate
m atter in the gas stream. _
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
i.IDENTIFIERS/OPEN ENDED TERMS
c. COS AT I Field/Group
pollution
Sulfuric Acid
Sulfur Trioxide
Vapors
Aerosols
Monitors
Automation
Condensing
Pollution Control
Stationary Sources
Goksoyr/Ross Method
13B
07B
07D
14B
13H
Release to Public
Unclassified
i (This Report)
21. NO. OF PAGES
69
20. SECURITY CLASS (This page)
Unclassified
22. PRICE
69
------- |