&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

-------