EPA-650/2-75-045
May 1975 Environmental Protection Technology Series
JDRECOVE
0V/
UJ
C3
-------
EPA-650/2-75-045
STONE & WEBSTER/IONICS
SO2 REMOVAL AND RECOVERY PROCESS
PHASE I, FINAL REPORT
by
Wisconsin Electric Power Company
231 West Michigan
Milwaukee, Wisconsin 53201
Contract No. 68-02-029?
ROAPNo. 21ACX-082
Program Element No. 1AB013
EPA Project Officer: Michael A. Maxwell
Control Systems Laboratory
National Environmental Research Center
Research Triangle Park, N. C. 27711
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
OFFICE OF RESEARCH AND DEVELOPMENT
WASHINGTON, D. C. 20460
May 1975
-------
EPA REVIEW NOTICE
This report has been reviewed by the National Environmental Research
Center - Research Triangle Park • Office of Research and Development,
EPA. and approved for publication. Approval does not signify that the
contents necessarily reflect the views and policies of the Environmental
Protection Agency, nor does mention of trade names or commercial
products constitute endorsement or recommendation for use.
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environ-
mental Protection Agency, have been grouped into series. These broad
categories were established to facilitate further development and applica-
tion of environmental technology. Elimination of traditional grouping was
consciously planned to foster technology transfer and maximum interface
in related fields. These 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
9. MISCELLANEOUS
This report has been assigned to the ENVIRONMENTAL PROTECTION
TECHNOLOGY series. This series describes research performed to
develop and demonstrate instrumentation, equipment and methodology
to repair or prevent environmental degradation from point and non-
point sources of pollution. This work provides the new or improved
technology required for the control and treatment of pollution sources
to meet environmental quality standards.
This document is available to the public for sale through the National
Technical Information Service, Springfield, Virginia 22161.
Publication No. EPA-650/2-75-045
ii
-------
ABSTRACT
The report covers Phase I of a potential
three—phase program to evaluate the Stone & Webster/lonics
process at 1 MW pilot plant scale with the option to scale
up and denKrnstrate process viability at the 75 MW prototype
level. The report cites the objectives, approach, results,
and conclusions, and discusses a program that incluaed: the
design, construction, and operation of, and completion of a
test program for, the pilot plant; the design, construction,
and testing of prototype—size electrolytic regeneration
cells; the design, engineering, and estimation of
construction and operating costs of the 75 MW prototype; and
preparation of a test program and operating schedule for the
prototype. An executive summary includes the background and
objectives of the overall program and pilot—scale effort,
and highlights significant results and conclusions.
Although technical feasibility was demonstrated at the pilot
scale, the economics of a 75 MW prototype plant at the site
of the pilot plant do not appear favorable. There are no
current plans to continue into Phase II (detailed design,
procurement, and installation of the 75 MW prototype) or
Phase III (12—month start—up and operational test period for
the 75 MW prototype).
iii
-------
ACKNOWLEDGEMENT
The contributing parties uld like to acknowledge personnel
from Stone & Webster Engineering Corporation, lonics Inc.,
the Environmental Protection Agency, and Wisconsin Electric
Power Ccinpany who contributed to the overall project effort
and the preparation of this report.
iv
-------
TABLE OF CONTENTS
Section Pac7e
EXECUTIVE SUMMARY xiii
PHASE 1A — PILOT PLANT OPERATION
1 • ABSTRA - PHASE 1A 1
2. CONCLUSIONS 3
3. RECOMMENDATIONS 5
4. INTRODUCTION 7
5 • TEST PROGRAM OPERATING RESULTS 11
5 • 1 Absorption—Stripping Section 11
5.2 Electrolytic Cell Section 12
5.3 Overall Conclusions 13
6. DISCUSSION OF PROCESS RESULTS 15
6.1 Absorption is
6.1.1 302 Re val 15
6.1.2 Mass Transfer Coefficients 15
6.1.3 Caustic Utilization 18
6.1.4 E ntrainment 20
6.1.5 Pressure Drop 21
6.1.6 Particulate Removal, NOX and 503 27
6.2 Oxidation 29
6.2.1 Literature Survey 29
6.2.2 Data Reduction 30
6.2.3 Results 33
6.2.4 Discussion of Results 34
6.3 Stripping 54
6.3.1 Stripping Steam Rate 54
6.3.2 Acid Addition 56
6.4 Electrolytic Regeneration 61
6.4.1 General Description of the Cell System 61
6.4.2 Cell System 61
6.4.3 Cell System Operations 66
6.4.4 Initial Cell System and Process
Debugging Period 66
6.4.5 Cell System Evaluation Period 68
6.4.6 A.dvanced Component Evaluation Period 72
6.5 Process Material Balances 77
b 5.1 Process Losses 77
6.1.2 Loss Analysis Run Results 77
6.1.3 Discussion of Results 78
6.6 Cell Feed Liquor Processing 82
6.6.1 Contaminant Removal 83
6.6.2 Major Stream Analysis 88
V
-------
TA 1LE OF c N’PENTS (CONT’D)
Section Page
6.6.3 Sources of Metal Contaminants 89
6.6.4 Cell Feed Liquor Conditioning in
Future Installations 92
7 • DISCUSSION OF OPERATING RESULTS 93
7.1 Process Consistency 93
7.2 Process Control 95
7.2.1 Absorption 95
7.2.2 Stripping 97
7.2.3 Feed Liquor Preparation 98
7.2.4 Cell Room 98
7.3 Materials of Construction 98
7.4 Equipment Operating Experience 98
7.4.1 S02 Analyzer 98
7.4.2 Flowineters 99
7.4.3 pH Meters 99
7.4.4 Temperature and Pressure Indicators 100
7.4.5 Level Control 100
1.4.6 Forced Dratt Fan 100
7.4.7 Pumps 100
7.4.8 Absorption Tower 101
7.4.9 Stripper Column 101
7.4.10 Cell Room 101
8 • DISCUSS ION OF A1’1ALTYICAL PROBLEMS 103
8.1 Basic Analysis Problems 103
8.1.1 NaHSO3/Na2SO3,’NaHCO3/Na2CO3/NaOH
Analyses 103
8.1.2 Na2SO4 111
8.2 Iron Analysis 113
8.3 Other Analyses 116
9. REFERENCES 119
10. APPENDICES 121
A. Limnetics, Inc. Report on Foreign Materials
Analysis 123
B. Chronological History of Pilot Plant Opera-
tions 133
C. Oxidation Calculations 143
D. Sulfur Rai val Calculations 147
E. British to Metric Conversion Table 151
F. Tabular Summaries and Measurements 155
G. IGT Trace Iron Analysis 169
VI
-------
Page
PHASE lB - PROTOTYPE CELL DEVELOPMENT (DESIGN,
FABRICATION, AND PERFORMANCE) 175
1. ABSTRACT 177
2. CONCLUSIONS 179
3. RECOMMENDATIONS 181
4. INTRODUCTION 183
5 • DESIGN AND FABRICATION OF THE PROTOTYPE CELL
COMPONE NTS 185
6. PROTOTYPE CELL TEST FACILITY $89
7. PROTOTYPE CELL PL.RFORMANCE 195
7.1 A” Cell Performance 195
7.1.1 Current Efficiency 195
7.1.2 Energy Consumption 195
7.1.3 Cell Voltage 198
7.2 B Cell Performance 198
7.2.1 Cell Voltage 198
7.2.2 Current Efficiency and Energy 201
Consumption
PHASE 1C - PROTOTYPE PLANT COST ESTIMATh. PROJECTED
OPERATING COSTS, AND DESIGN AND
CONSTRUCTION SCHEDUL}. 203
1. GENERAL DESCRiPTiON OF PROCESS WORK 205
2. COST ESTIMATE FOR INCLUSION IN WEPCO
75MW POWI.R PLAN’I 207
2.1 Design Basis Cost Suimnary Sheet
(4250 lb S02/Ifr) 207
2.2 Design Basis Cost Sununary Sheet
(3100 lb S02/Hr) 207
3. PROJECTED MONTHLY OPEICATING COSTS 211
4. DESIGN, ENGINEERING AND CONSTRUCTION SCHL DULE 215
VII
-------
TABLE OF CONTENTS (CONT’D)
Section Page
PHASE 1D - TEST PROGRAM AND OPERATING SCHEDULE FOR
75MW PROTOTYPE PLANT 217
1. INTRODUCTION 219
2. TEST PROGRAM O] SJJ ’ICPWES 221
2.1 Process Performance 221
2.2 Absorption Section 221
2.3 Stripping Section 223
2.4 Feed Liquor Treatment 223
2.5 Data Collection and Evaluation 223
2.5.1 Technical Program 223
2.5.2 Process Maintenance 224
2.5.3 Process Operating Costs 2214
VIII
-------
ILLUSTRATIONS
Fiqure
PhASE 1A Page
I Stone & Webster/lonics S02 Recovery and Removal XV1
System - Simplified Block Diagram
II Stone & Webster/lonics S02 Removal Pilot Plant - xvii
WEPCO, Valley Station. Milwaukee
III Pilot Plant Absorber xviii
IV Pilot Plant Stripper xix
V Electrolytic Cell Module in Cell Room at lonics, xxi
Inc.
4.1 Pilot Plant Flow Diagram 9
6.1 Height of a Transfer Unit 17
6.2 Overall Gas Phase Mass Transfer Coefficient
versus Superficial Gas Velocity 19
6.3 Entrainment from Stage 1 as a Function of
Gas Rate 22
6.4 Entrainment from Stage 1 as a Function of
Liquid Recirculation Rate 23
6.5 Entrainment from Quench Section as a
Function of Gas Rate 24
6.6 Entrainment from Quench Section as a
Function of I-late Water Flow 25
6.7 Pressure Drop through Absorption Tower 26
6.8 The ffect of Flue Gas Flow on Oxidation 35
6.9 The Effect of S02 Inlet Concentration on
Oxidation 36
6.10 The Effect of Total S02 Removed on Oxidation 37
6.11 The Effect of Flue Gas Flowrate on the Net
Draw Sodium Bisulf ate Concentration 38
6.12 The ± .itect of S02 Inlet Concentration
on the Net Draw Sodium 39
6.13 The Effect of Total S02 Removed on the Net
Draw Sodium Bisuif ate Concentration 1 40
6.14 The Effect ot Absorber Net Draw Sodium
Bisulfite Concentration on Oxidation 142
6.15 The Effect or S02 Absorption Profile on
Oxidation 43
6.lo The Effect ot S02 Absorption Profile on
Oxidation (Total) 44
6.17 Mole Percent Sodium Bisulfite versus
S/C Ratio 46
6.18 The Effect of S02 Absorption Profile on
oxidation 47
Ix
-------
ILLUSTRATIONS (CONT ‘ D)
Figure Page
6.19 The Effect ox S02 Absorption Prof ile on
Oxidation (Total) 48
6.20 S02 Removal in the Top Stage Versus
Sodium Bisulfite Concentration in
Top Stage Effluent 49
6.21 The Effect of Total Salt Concentration
in Net Absorber Net Draw on Oxidation 50
6.22 The Effect of Gas—Liquid Contacting on
Oxidation 51
6.23 The Effect of Heavy Metals Contamination
on Oxidation 53
6 • 214 Limiting Stripping Steam as a Function
of S02 Removal 55
6.25 Excess Acid Residual 502 in Stripper
Bottoms as a Function of pH 57
6.26 Optimum p1:1 for Stripper Bottoms
Operation 59
6.27 Titration Curve - Sodium Sulfate
Solution versus Water 60
6.28 Schematic Diagram of Electrolytic Cells 62
6 • 29 Design Point Material Balance for 56 Type
NAN SULFOMAT Cells 64
6.30 Design Point Material Balance for 32 Type
B SULFOMAT Cells 65
6.31 Design Cell Voltage — Current Density
Relationships for Types NAN and “B
SULFOMAT Cells 67
6.32 Absorption Tower Response Testing
7 • 1 Absorption Tower Response Testing 96
8.1 Distribution of Aqueous Sulfite Species
as a Function of pH 104
PHASE iFs
5.1 Cell Amode I ssembly 188
6.1 Assembled Prototype Cell Stack 190
6.2 Design Point Material Balance
for 16 Prototype NA Cells 192
6.3 Design Point Material. Balance for
16 Prototype li Cells 193
6.4 Design Cell Voltage - Current Density
Relationships for Types NAN and B
SULFOMAT Cells 194
7.1 Current—Voltage Curve for Prototype •A M Cells 199
7.2 Current-Voltage Curve for Prototype B Cells 200
PHASE 1C
4 • 1 Proposed Construction Schedule for
Prototype Plant 216
x
-------
TABLES
Number Page
PHASE 1A
6.1 Particulates NOX and S02 Levels on
Absorber Inlet and Outlet 28
6.2 Bad Case Operating Conditions for
Stripping Study 58
6.3 Selected Design Values of WEPCO
Pilot Plant 63
6.4 Summary of WEPCO Pilot Plant Current
Efficiency Measurements 69
6.5 Summary of Voltage-Current Data and
Specific energy Consumption Results
in WEPCO Pilot Plant Cell System 73
6.6 Summary of A and •B” Cell Current-
Voltage Data 77
6 • 7 Comparative Operating Conditions and
Measured Losses 79
6.8 Cell Feed Specification 83
6.9 Profiles ot S02 and Fe in Feed Liquor
between Stripper Drum and Cell Stacks
during Initial Running with Heavy”
H202 Addition 86
6.10 S02 and Fe Profiles in Major Streams of
WEPCO Pilot Plant 87
6.11 Typical Fly Ash Analysis of
WEPCO Pilot Plant 90
6.12 Elemental Breakdown of Seven Typical
Fly Ashes 91
7.1 Pilot Plant Operation and Availability 93
7.2 Absorption Tower Response Testing 97
7.3 Comparative S02 Measurements 99
8.1 Sulfite—Bisulfite Determination 105
8.2 Bicarbonate in Absorber Liquid 112
8.3 Sodium Sulfate Analysis Graviinetric
Determination 114
8.4 Iron Analyses in 3 N Sodium
Sulfate Solutions 115
8.5 Total iodium Analysis 117
PHASE lB
5.1 Drawing List of Prototype Cell
Components 186
6.1 Selected Design Values of Prototype
Cell Test Facility 191
7.1 Summary of Performance Tests of
Prototype “A Cells 196
XI
-------
TABLES (CONT’D)
Number Page
7 • 2 Summary of Prototype A” Cell Energy
Consumption Data 197
7.3 Summary of Performance Tests of
Prototype RBU Cells 202
PHASE 1C
2.1 Cost Summary Sheet for 4250 lb S02/Hr
Removal 208
2.2 Cost Summary Sheet for 3100 lb S02/Hr
Removal 209
3.1 Estimate of Operating and Maintenance
Costs — 75 MW Demonstration Plant for
3100 lb S02JHr Removal 212
PHASE 1D
2.1 Test Program Objectives 217
X II
-------
EXECUTIVE SUMMARY
Stone and Webster E.ngineering Corporation (SWEC) in conjunction
with lonics, Incorporated, has designed and engineered a process
for the removal and recovery of sulfur dioxide from stack gases.
In July 1972, EPA and Wisconsin Electric Power Company (WEPCO)
initiated a potential three phase program to evaluate the
SWEC/Ionics process at pilot plant (1MW) scale with the option to
scale—up and demonstrate process viability at the prototype
(75MW) level. The flue gas S02 removal and recovery process
employs electrolytic regeneration of the working fluids. It
considerably reduces solids handling and disposal problems
inherent in lime/limestone slurry systems and offers operational
advantages for utilities over other regenerable systems. This
report covers Phase I of the contract.
The overall goal of Phase I work was to provide information for
the design, construction, and operation of a 75 MW Prototype
Plant using the Stone & Webster/lonics S02 Removal Process. This
goal was accanplished by means of the following steps:
Phase 1A — Design, construct, and test operate a Pilot Plant
sized at approximately 1 NW. The results are
summarized as follows:
1. The plant was built and operated successfully,
demonstrating feasibility of the process.
Information was obtained on the mechanical
operability and reliability of equipuent.
2. The performance of the electrolytic cells was
demonstrated successfully in an industrial
environment.
3. The effects that process variables have on
process performance were determined.
Phase lB — Design, construct, and test commercial size
electrolytic regeneration cells which would be
suitable for use in a 75 MW Prototype Plant. The
results of this effort were:
1. Cells were designed, constructed, and tested,
demonstrating a five—fold increase in capacity
over the Pilot Plant cells at the same or a
lower energy factor.
2. Mechanical integrity of the larger cell units
was demonstrated.
3. Component designs were developed that
simplified field assembly of the electrolytic
cells.
xiii
-------
Phase 1C - Design a 75 MW Prototype Plant to be installed at
Wisc sin Electric Power Company’s Valley Station
site and establish estimates of capital and
operating costs. The results were:
1. The design of the plant at the specified
locations was completed.
2. Cost estimates for construction of this plant
were established for two design capacities.
3. Operating cost for a one year operating and
testing program following plant start—up was
estimated.
Phase 1D — Develop a test program and operating schedule for
the 75 MW Prototype Plant. The results of this
effort were:
1 • An orderly operating schedule was established
for start—up and testing of the 75 MW plant.
2. A test program was established designed to
supply data which could not be obtained from
the Pilot Plant operation.
Each of these, steps is discussed in correspondina sections of
this report. This Executive Summary gives an overview of the
Phase I results.
xiv
-------
PHASE IA, DESIGN, CONSTRUCTION, AND OP] .RATION
OF THE PILOT PLANT
Operation of the Pilot Plant provided data to form a basis of
work for the other parts of Phase I. The Pilot Plant was
constructed and installed at WEPCO’s Valley Station in Milwaukee
(Figures I II). The plant was designed to remove more than
90 percent of the 502 from flue gas flowing at 2200 (actual)
cubic feet per minute.
Pilot Plant Desian Criteria
Flue Gas Flow
2200 Ft 1Mm
Flue Gas S02 In
2000 PPM
Flue Gas S02 Out
200 PPM
This would be roughly equivalent to the flue gas from a boiler
with a capacity of about 0.75 MW. Design of the Pilot Plant
progressed satisfactorily; there was no design procedure which
was beyond proven practice. Construction encountered the
equi znent delivery and cost inflation problems experienced by the
entire industry during 1973. The Pilot Plant was commissioned in
June 1973. After overcoming some typical shakedown difficulties,
the test program was carried out and the technical feasibility of
the process was established. An important part of the
establishment of feasibility was demonstrating the operability of
individual process steps. These results are reviewed below.
The absorber, shown in Figure III, demonstrated S02 removal
efficiencies of 90 percent or more at SO2 inlet concentrations
of up to 3600 p zn. This was accomplished in two ten—foot packed
stages. An average of 13 percent of the absorbed S02 was
oxidized in the Pilot Plant (“oxidation” raises both the capital
and operating costs of most regenerable processes and should
generally be minimized). The effects of several parameters on
oxidation in the absorber were studied as part of the test
program. It was found that oxidation was lowered by shortened
gas/caustic solution contact times and by increased sodium
bisulfite concentrations in the absorbing solution. Early
problems of overhead entrainment and quench water carryover in
the absorber were also resolved.
The stripper, shown in Figure IV, satisfactorily removed S02 from
the absorbing solution with steam after that solution was
acidified to pH 2.5 to 3.0 with recycled acid sulfate. The
stripper’s internal steam requirements were found to be 1.5 to
2.0 pounds per pound of S02, depending upon the precise condition
of the stripper bottoms and the efficiency of the packing. Some
unanticipated operating difficulty was experienced that was
particular to the installed stripper equipment. Future designs
should eliminate this difficulty.
xv
-------
SULFUR DIOXIDE
RECYCLE CAUSTIC SODA
D ES UL F U R IZ E D
GAS
ABSORPTION
SECTION
FLUE GAS FROM I
DILUTE SULFURIC
PRECIPITATOR
ACID PURGE
FIGURE I. STONE WEBSTER/IONICS SO 2 REMOVAL PROCESS
xvi
-------
FIGURE II .
STONE a WEBSTER/IONICS SO 2 REMOVAL PILOT PLANT—
WEPCO, VALLEY STATIONS MILWAUKEE
xv ii
:
.0
-------
øk
‘u_I
\4 “\i
1
/i.
QES ULFURIZ FLUE GAS
;Y
I
CAUSTIC SOLUTION
CIRCULAT
DEMISTING STAGEJ
\\
.1
‘:•
STAG \ :1.
ABSORPTION
V
\ V
QUENCH INLET
FLUE GAS INLET—_.
_____— ---j
FIGURE III. PILOT PLANT ABSORBER
xv i ii
-------
SULFUR DIOXIDE STREAM
PILOT PLANT STRIPPER
. . -fi
T
L.-
STRIPPER REFLUX
FIGURE IV.
X IX
-------
The Pilot Plant’s cell feed was kept sufficiently free of
contaminants (heavy metal cations and chloride ions) by a series
of treatment steps. Oxidation by hydrogen peroxide addition
followed by pH ad]ustment to 8—9 precipitated most contaminants
and allowed their removal in a filtration step. The dissolved
iron was reduced to less than 0.1 mg/l in this manner - The
oxidation step was at first attempted with a manganese zeolite
bed which failed due to the higher than anticipated concentration
of oxidizable material in the stripper bottoms. This system
operated consistently well, however, after the beds were replaced
by the peroxide addition. Simple aeration may give the same
results as peroxide addition, but that requires further study.
The electrolytic cells (shown in Figure V) regenerated the
caustic and acid solutions used in the process. This was
accomplished by applying an electric charge across a stack of
cell compartments, containing the cell feed liquor, separated by
membranes which selectively passed ions of positive or negative
charge. The membranes were arranged to regenerate acid or
caustic solutions in alternating compartments down the length of
the stack. Oxidation in the absorber and other parts of the
process produced nonregenerable sulfate which was purged by
special cells, with an extra membrane, as dilute, about
10 percent. sulfuric acid. The rate of production of this acid
was directly proportional to the rate of oxidation. The pilot
plant cells had shakedown operating difficulties that were
related to inadequate feed treatment and the use of “off—spec”
caustic soda, which contained chloride, for sodium replacement.
It was also found that the sodium sulfate concentration in the
cell feed should be less than 3.5N (Noi:mal) to prevent
crystallization in the cells. Once these difficulties were
resolved and the limiting operating conditions found, the cells
operated reliably and exceeded design efficiency in normal
operating conditions.
It was also found that use of “inert” anodes rather than those of
lead alloy would reduce electric power consumption costs, an
observation that was used in estimating the Prototype Plant costs
in Phase IC. The energy required for operation of cells using
inert anodes in the 75 MW plant was scaled from results obtained
with the Pilot Plant. Scale—up requirements were verified during
testing of the Prototype Cells in Phase lB.
The estimated operating costs for the 75 MW Plant were
$60,000/Year lower than for an equivalent plant with lead anodes.
The estimated initial investment was about $100,000 higher with
inert anodes, but the reduced operating costs could repay this
difference rapidly.
Calculations predicted that 4500 kw (6.0 percent of 75 MW
capacity) would be required to operate the Prototype Plant cells
and 11433 kw (1.9 percent of 75 MW capacity) would be needed for
other process equipment. The total energy requirement for this
installation at the Valley Plant of WEPCO was estimated to be
7.9percent of the 75 MW capacity.
xx
-------
DIAPHRAGM STACK
IN POSITION
FIGURE V. ELECTROLYTIC CELL MODULE IN CELL ROOM
AT IONICS, INC. BEFORE PILOT PLANT
INSTALLATION
x xi
-------
The Pilot Plant achieved closed loop operation. This was
indicated by closure of the sodium balance around the system and
demonstrated the feasibility of operation in the regenerable
node. During the course of the Pilot Plant’s operation, specific
studies were also made of the overall material balance.
4 P RIA L BALANCE STJt4MARY FOR MALTOR PROCESS SPECIES. %*
Species
Present
in Process
Streams
Accounted for
in Mechanical
or Handling Losses
Unaccounted
for
Na+
97.8%
2.2%
0.0%
S03
97.9%
2.1%
0.0%
H20
94.3%
1.3%
‘4.4%
Water was lost in the Pilot Plant operation, but test procedures
determined at which points it was being lost. A large amount was
apparently used to saturate the flue gas in the absorber; hence,
a more effective quench water system upstream of the absorber
would be used in the Prototype Plant. Losses can be more easily
monitored in the Prototype Plant, once having been identified and
located in the Pilot Plant work.
In addition to the process related experience, mechanical
requirements for the process equipment were determined. Based
upon Pilot Plant experience, equipment of improved design and
materials of construction has been specified in Phase IC work.
Rotating equipment specifications and operation procedures which
increase operating reliability have particulary been determined.
Sample metal coupons were placed in several environments in the
process and examined upon completion of the Pilot Plant’s testing
schedule. Material of construction specifications based upon
coupon examination have been used in the design and engineering
work of Phase IC .
The Pilot Plant denonstrated the overall operability and
technical feasibility o the SWEC/lonics S02 process. Further
Pilot Plant work is not required. Process experience can now
advance only with a larger unit. The Pilot Plant woric has
identified the following specific objectives for the Prototype
Test Program:
1. Utility-type mechanical reliability should be demon-
strated for all pieces of equipment.
2. A quantitative water balance should be made.
3. Operation of the stripper should be more precisely
characterized.
4. The use of large-scale equipment in the Cell Feed Treat-
ment system slould be demonstrated and air oxidation
should be studied for use in place of the peroxide
addit3.on.
•Percent of species in daily cell teed
xxii
-------
5 • Parameters affecting oxidation in the absorber which
stxuld be studied are packing type and vol mie, and flue
gas oxygen concentration.
xxiii
-------
PHASE IB,_DEVELOPMENT OF PRO )TYPE
EL TROLYTIC CELL SYSTEM
Prototype cell developn nt was carried out in the laboratories of
lonics, Incorporated. Cell assemblies suitable for use in the
Prototype Plant were designed, fabricated, and tested under
simulated process conditions. The Prototype cells were scaled up
to a production capacity per cell in excess of five times the
capacity of the Pilot Plant cells. nphasis was placed on
develoilnent of a prefabricated •cell packages for easy assembly
and disassembly.
Particular problems of scale-up which were resolved within this
context included achieving adequate flow distribution in the
cells, obtaining a recirculation rate for the ele ctrode
compartment high enough to allow disengagement of the gas
produced, and construction of a unit that avoids exposure of the
electrodes to the process solution in the cell manifolds. One
type of cell assembly experienced some leakage externally and
internally early in the program, but this was corrected by minor
modifications which uld be incorporated into existing
production molds for cclnponents in the future.
The Prototype cell performance met the design criteria of the
contract. There was no loss of current efficiency with the
increased size • This demonstrates that the cell operation is
understood well enough to be scaled-up confidently and indicates
that the cells are ready to be denonstrated in the Prototype
Plant. These cells should be installed and operated as part of
an actual process application as the next step in their
develo ent.
xxv
-------
PHASE IC, PRELIMINARY PROTOT!PE PLANT
ENGINEERING DES IGN
Work in Phase IC included the design, engineering, and
preparation of a cost estimate for a Prototype Plant to be
installed at WLPcO’s Valley Station. This plant would remove S02
from flue gas exiting Boilers 3 or 1$ which have individual
capacities of about 75 MW. Projected costs for operating the
unit and carrying out the test program outlined in Phase ID were
tabulated for each of the 12 operation and test months.
Two cost estimates were prepared for design and construction of
the Prototype Plant. These estimates are detailed in Tables 2.1
and 2.2 in Phase IC, pages 214 and 215 herein. One was based on
removal of 4250 lbs S02/hr for operation of the cells on “off
peak” hours only and the other on 3100 lbs S02/hr for continuous
cell operation. The practical minimum design basis is 3100 lbs
S02/hr. This represents the highest average S02 absorption over
a one-month period (February) during 1973, the latest year from
which complete coal consumption and sulfur content data for the
Valley Plant were available. This design capacity must be
considered as minimum even though it is more than enough for the
remaining months. Short—lived increases of S02 removal rates
above the average rate for the entire year can be handled by the
system’s storage capacity, but it would be impractical to supply
the large storage capacity needed to spread regeneration
necessitated by a month—long surge over other months. The
overall costs tabulated were $14.475 MM for the 4250 lb S02/hr
plant with an automated cell room, and $ 12.229 MM for the 3100 lb
S02/hr plant with a semiautomatic cell room. The estimates are
higher than those generated at the inception of this contract due
in part to the rapid escalation in materials, equipment, and
labor costs in the intervening period.
The Phase IC costs are specifically related to the Valley Station
site and reflect severe space and retrofit limitations which
require intricate arrangement of equipment. It is anticipated
that the costs for this unit at a more adaptable or new site
would be lower. Also reflected in these estimates is the expense
of including the equi inent flexibility and instrumentation
necessary to conduct the Test Program (Phase ID). An operating
plant would not require such cost-raising extras, but with them,
this Prototype Plant could provide data for design optimization
and accurate prediction of large scale equi mient costs.
The operating and maintenance costs similarly reflect the fact
that the Prototype Plant will be a test facility for most of the
time covered by the program. (Refer to Table 3.1, Phase IC) page
220. A large technical support effort will be required for
start—up and operation during the test program. This period is
expected to require spending about $156,000 per month (wages and
prices February, 1975, no credit for by—products) in operating
costs. However, when normal operation ensues, the monthly costs
are expected to be about $102,000 per month. If credits for
xxvii
-------
condensate returned and by—products are taken, the costs could be
as low as $97,000 per month during the test program and
$43,000,ktonth during normal operation if the current market for
S02 hold (February 1975 basis).
It should be pointed out that the indicated favorable economics
of by—product credits uld probably not hold true for a large
scale installation. conversely, capital costs per kw capacity
would decrease for a larger plant.
xxviii
-------
PHASE ID, TESP PROGRAM AND OPERATING SCHEDULE
FOR 75 MW PROTOTYPE PLANT
Phase ID included planning the operation of the S02 Removal
Prototype Plant and specifying test objectives. The S02 Removal
Plant has no polluting effluents and therefore, inherently meets
all operating codes and legal restraints. Marketing surveys by
SWEC have shown that there is a market for the S02 produced.
Dilute acid produced by the process could be used to regenerate
the demineralizer water units of the WEPCO power plant. There is
a relatively small amount of filter cake from the cell feed
liquor treatment which could be disposed of with the ash from the
boilers. The sulfur dioxide uld be sold as a process
by—product. Although the disposition of by—products appears
favorable for the 75 MW Prototype Plant, this condition cannot be
readily forecast for a full scale installation, complete
mechanical testing is planned before process operation is begun.
The test program will extend study in several areas which could
not be completely studied in the Pilot Plant Program. Water loss
by saturation of the flue gas and/or entrainment overhead will be
studied in the absorption section. The effect of NOx, heavy
metal cations, S03, and flue gas 02 concentrations on oxidation
will be studied, especially in the absorber. The minimum
stripping steam for residual S02 concentration in the stripper
bottoms effluent will be determined. The economic optimum S02
concentration In the solution from the stripper bottom will then
be known in terms of stripping steam economy. Cell feed liquor
treatment using peroxide addition and filtration will also be
tested. All data will be evaluated to provide equilibrium and
equipment correlations for scale—up use. Accurate records of
process maintenance will be kept to determine the proper
application of estimating techniques. Process operating costs
will be kept in detail to provide a basis front which to project
costs to units larger than 75 MW.
xxix
-------
SWEC and Ionics believe that the results of the Phase I work
justify evaluation of this process on a coimnercial scale. WEPCO
has decided that the difficulties and costs of retrofitting a
75 MW prototype insta]. lation at the site of the Pilot Plant
cannot be justified at this time. The Pilot Plant results
indicate the SWEC/Ionics S02 Ren val Process is technically
feasible. Accurate data for prediction of capital and operating
costs for a 500 MW plant can only be obtained by the construction
and operation of a Prototype Plant.
x x xi
-------
PHASE 1A
Prwr PLAZ4’P OPERATION
-------
I
ABSTRACT-PHASE 1A
The first phase of the EPA-W PCO sponsored program to
evaluate the Stone & Webster/lonics closed cycle S02 removal
system has been successfully completed. The technical
feasibility of the process was demonstrated on a 2200 ACFM
pilot plant by meeting emissions and operating requirements
over a representative range of conditions. The average S02
removal was 85—95 percent at exit concentrations of
200—300 p n 502. Oxidation of S02 in the absorber averaged
3•14 lb/hr* or 13 percent of the S02 removed. The data and
experience obtained will allow for the design and operation
of a reliable prototype unit. The test program also has
defined areas requiring further study beyond the scope of
this work, which will permit additional design optimization
when their effects are fully determined.
*Metric conversions may be found in Appendix E,
.. ritis to-4 eric Conversion Table.
1
-------
II
CONCLUSIONS
The following conclusions were reached from the operation of
a 2200 AC 4 Stone & Webster/lonics S02 removal system pilot
plant.
1. The technical feasibility of the process was
demonstrated.
2. S02 removal efficiencies of 90 percent or more (for
S02 inlet concentrations up to 3600 ppm) were
attained with two 10 foot packed stages.
3. The process was operated at oxidation levels less
than 15 percent.
4. Minimum oxidation was favored by minimizing gas-
liquid contact residence time and maximizing sodium
bisulfite concentration in the absorbing liquor.
5. High sodium bisulfite concentration was favored by
high inlet S02 concentration.
6. Gas rate ane total salt concentration had little
effect on oxidation.
7. The overall energy consumption per mole of caustic
produced by the electrolytic cells was lower than
the desicrn objective.
8. Cell production capacity exceeded the design
objective.
9. Closed loop operation was achieved as indicated by
closure of the sodium material balance.
10. The vapor pressure relationships for the aqueous
sodium sulfite-bisulfite-sulfate system agreed with
the data of Johnstone*. Sodium sulfate
concentration did not affect 302 vapor pressure
over the solution.
11. The pH of the stripper bottoms effluent should be
controlled to between 3.2 and 3.5 to optimize
recycle caustic consumption and S02 stripping.
12. Stripping steam reauirements in the pilot plant
were between 1.5 and 2.0 pounds per pound of S02
removed, exclusive of process preheat which was
*See Section 9, References.
3
-------
estimated to be 4.25 pounds per pound of S02 for
the pilot plant.
13. The maximum concentration of sodium sulfate in the
cell feed liquor should be 3.5 normal in order to
avoid crystallization.
14. The heavy metal cation specifications for cell feed
liquor were met using a conbination of hydrogen
peroxide addition and filtration.
15. Process water losses can be expected; they are
contingent on process design and maintenance. In
the pilot plant water losses varied from 60 to
200 gallons per day attributable to electrolysis,
humidification, and operating losses.
16. Operating problems with the pilot plant were mainly
n chanical. Improved equipment and materials
selection will minimize operating difficulties in
future plants.
-------
III
RECOMMENDATIONS
In order to establish commercial feasibility and costs, the
process should be demonstrated on a prototype plant scale of
at least 75 to 150 MW. Items to be studied more closely
during that period should include:
1. Maintenance and Operability
The most important facet that requires
demonstration is the mechanical operability of the
process. Process oriented problems are minor
compared to the need to demonstrate utility-type
reliability from rotating equipment, commercial
sized cell components, and other items of
equipment.
2. System Water Balance
The water balance for the pilot plant was
inconclusive. Evaporative losses from the cells,
the absorber, stripper, and various tanks should be
monitored in order to establish specific rates.
3. Stripping Section
The lack of reliable flówmeters around the pilot
plant stripper prevented the measurement of
stripping rates for the S02-Na2SO4-H20 system. The
stripping steam rate, mass transfer coefficients,
and pH relationships within the tower require
additional study.
14• Feed Liquor Preparation
The demonstration of large scale feed liquor
treatment is needed including the use of precoat
filters and their effectiveness in removing small
quantities of iron and aluminum hydroxide.
Additionally, the use of air oxidation in place of
peroxide addition should be investigated.
5. Oxidation
Areas not investigated during the test program, but
rthy of study with the objective of reducing
oxidation, are the effects of packing type and
volume and the effect of oxygen concentration on
oxidation.
6. Trace Elements
A Study of the baildup of trace elements and means
for their removal should be made.
5
-------
Iv
INTRODUCTION
The present annual emission rate of sulfur dioxide into the
atmosphere in the United States is estimated at nearly
40 million tons. Approximately 60 percent of this is
emitted from the stacks of power stations which burn fossil
fuels as primary energy sources. Roughly two—thirds of
this, or approximately 40 percent of the total, is from
stations which burn coal. Removal of S02 from stack gases
is a feasible method which may be used to abate these
emissions.
Stone & Webster Engineering Corporation, in conjunction with
lonics, Inc., has designed and engineered a process for the
removal and re very of sulfur dioxide f rain stack gases
based on the absorption of sulfur dioxide in an alkaline
solution which then is regenerated - The key component of
this process is the lonics electrochemical cell which
regenerates the spent absorbent.
The Stone & Webster/lonics S02 removal process is applicable
to gaseous effluents from stationary power plants burning
fossil fuels containing sulfur and to tail gases from sulfur
recovery plants, smelters, and sulfuric acid plants.
The principal advantages of this process are:
1. Very small amounts of makeup chemicals are required
and no secondary waste products are generated which
represent additional contamination to the
environment. A relatively small dilute sulfuric
acid purge stream is pure and may or may not
represent a penalty to the process.
2. Flue gas S02 content can be reduced to a minimum
level since NaOH is the absorbent and S02 has
negligible vapor pressure over a caustic solution.
3. The process chemistry is simple and well known.
14. The components in the regenerating electrolytic cell
system are readily mass produced and can be
improved through evolutionary engineering
development even after installation.
5. The modular cell operation enables cell regeneration
to fluctuate with boiler operation without loss of
efficiency.
In order to evaluate this process in a realistic manner, a
pilot plant was installed at Wisconsin Electric’s Valley
7
-------
Plant in Milwaukee. to process 2200 acfm of flue gas at an
inlet S02 concentration of 3600 ppn. This is approximately
equivalent to flue gas from a power plant of 0.75 Mw
capacity. The concentrated S02 stream was returned to the
stack since there was no need to demonstrate S02 recovery
technology.
The primary goals of the test program were to:
1. Demonstrate the system operability and reliability
from both a process and mechanical standpoint.
2. Measure cell performance in an industrial
environment.
3. Determine the effects of process variables on
oxidation.
Oxidation has a direct effect on the power requirements for
the system as well as the initial capital investuent,
because it increases the amount of caustic to be produced by
regeneration and the number of “B cells (more expensive)
required to purge sulfate ions.
This report summarizes the pilot plant operating experience
and results obtained during the period from July 1973 to
June 1974.
The Stone & Webster/lonics S02 removal process pilot plant
is shown in Figure 4.1. It is a closed loop system for
removing sulfur oxides from gas streams. The acid gases are
initially absorbed by sodium hydroxide in a packed tower.
The resultant sodium sulfite-sodium bisulfite solution is
acidified with sulfuric acid. Sodium sulfate is formed and
the released S02 is stripped from the process liquor for
recovery. The acid and caustic feed chemicals are
regenerated by splitting the sodium sulfate product in an
electrolytic cell system.
The process tan be summarized by the following reaction
sequence:
1. ABSORPTION
2 NaOH + S02 Na2SO3 + H20
Na2SO3 + S02 + 1120 2NaHSO3
2. ACIDIFICATION
Na2SO3 + 112S04 Na2SOU + H20 + S02
2NaHSO3 + 112S04 Na2SOL4 + 21120 + S02
8
-------
FIGURE 4 I-SCHEMATIC FLOW DIAGRAM OF
FINAL VALLEY STATION PILOT PLANT
CAUSTIC
RECYCLE
FLUE
GAS
F ROM
9
-------
3. REGENERATION
Na2SO4 + 2H20 + E. 2NaOH + R2SOL$
The primary side reaction is the absorption of sulfur oxides
as S03 either by direct absorption or by the oxidation of
sulfites and/or bisulfites. This will form sulfate directly
and the sulfur oxides will not be released upon
acidification. thereby resulting in a net sulfate increase
within the system. Special electrolytic cells are used to
purge this excess sulfate as dilute sulfuric acid.
The electrolytic cell system consists of two types-
designated as types A and RBM. A cells consist of two
diaphragms and a cation membrane. The sodium sulfate feed
is split into mixed caustic (NaOH and Na2SO4) and mixed acid
(H2S04 and Na2SO4) streams for use directly in the process.
Conversions are kept low to increase cell efficiency. •B”
cells are similar to “A cells except that they contain an
anion membrane in addition to the other components. In
addition to a mixed acid and caustic stream, a pure sulfuric
acid stream is produced, thus maintaining the sulfate
balance in the system. This stream could be sent to a
sulfuric acid plant for use as water makeup.
10
-------
V
TEST PROGRA ? ! OPERATING RESULTS
5.1 ABSORPTION—STRIPPING SECTION
The major results of the test program in the absorption—
stripping section of the pilot plant operation were:
1. S02 removal efficiencies of 90 to 95 percent were
attained with inlet S02 concentrations up to 3800 ppm. This
was accomplished with two 10 foot absorption stages at
2000 lb/sq ft/hr gas flow and a pressure drop of 10 to 15
inches of water.
2. The process was operated at an average oxidation
level of about 13 percent at an overall caustic utilization
in the absorber of 80 percent or greater.
3. Maximizing the concentration of sodium bisulfite in
the absorption tower minimizes oxidation and maximizes
caustic utilization. It appeared that sodium sulfite was
the oxidized species.
14• High sodium bisulfite concentrations were favored
by lowered pH (approximately 5.0) and high inlet S02
concentrations.
5. Gas rate and sodium sulfate concentration appeared
to have little effect on oxidation.
6. Increased gas—liquid contact residence time in the
tower increased oxidation. Oxidation is apparently
controlled by the mass transfer of oxygen into the absorbing
solution.
7. Stripper steam requirements (including preheat and
losses) were close to 6 lb/lb of 502 removed. This total
included 4.25 lb steam/lb S02 for preheat with the balance
goinq to stripping and heat losses.
8. Closed icop operation was achieved except for
sodium losses which were accounted for by mechanical
leakage. Process liquor carryover in the towers can be
controlled using demisters..
9. Heavy metal cation specifications for the cell feed
liquor were obtained through appropriate treatment.
11
-------
10. The operational problems with the pilot plant were
mainly mechanical. Proper equipment and materials of
construction selection will minimize operating difficulties
in future plants.
5.2 ELECTROLYTIC CELL SECTION
The major results obtained in the electrolytic cell section
of the pilot plant were:
1. The current efficiency for total caustic production
by the “A” and “B” cells combined averaged 90—91% except
when crystallization occurred in the cells. The high
efficiency exceeds the design basis by 5—6 percentage
points.
2. The maximum sodium sulfate feed concentration
should be 3.5 N Na2SO4. Above this level, crystallization
of the catholyte inside the cells occurs damaging the cells.
3. Overall “A” cell current efficiency averaged 90—93%
(except when crystallization occurred in the cells),
exceeding the design basis by 5—8 percentage points.
4. Overall “B” cell current efficiency averaged 96% in
three measurements free of external and internal leakage.
The current efficiency for purged H2S04 (“B” anolyte)
averaged 45%, exceeding the design basis by five percentage
points.
5. Current efficiency appears to be independent of
current density, at least in the range covered (i.e., 70% to
100% of design point).
6. The overall energy consumption per pound of caustic
produced in t] e pilot plant was less than or equal to the
performance objectives at each production level, except when
crystallization occurred in the cells • The measured energy
factors “bettered” the performance objectives by as much as
9% at the 100% production level.
7. Reduction of dissolved iron in the sodium sulfate
cell feed stream to less than 0.1 mg/i was achieved using
hydrogen peroxide to oxidize the ferrous ion to ferric
before filtratior.
8. Cell system availability during the last nine
months ot the twelve—month test program was over 90%. Cell
maintenance time during the nine months was 1408 hours.
12
-------
9 • Substantial reductions
consequently, energy consumption
alloy anodes were replaced with
can reach over 25% at the higher
5 • 3 OVERALL CONCL1JS IONS
in cell voltage and,
were realized when lead
“inert” anodes. Reductions
current densities.
The process is technically feasible with proper
materials of construction and equipment. The features of no
secondary waste disposal other than dilute sulfuric acid and
on—site chemical regeneration are advantages, and the sulfur
dioxide is recoverable as a product.
13
-------
VI
DISCUSSION OF PROCESS RESULTS
6.1 ABSORPTION
The absorption of S02 in caustic soda follows the reaction
sequence:
2NaOH + C02 ‘ - Na2CO3 + H20 (1)
Na2CO3 + 502 Na2SO3 + C02 (2)
2Na2CO3 + S02 + H20 2NaHCO3 + Na2SO3 (2a)
NaBCO3 + S02 NaHSO3 + C02 (2b)
Na2SO3 + S02 + H20 2NaHSO3 (3)
In order to minimize the caustic required for a given S02
removal, a high sodium bisulfite to sodium sulfite ratio is
required. This ratio (S/C) is conveniently expressed as the
ratio of total dissolved S02 to total sodium as sulfite and
bisulfite (all expressed in gm-ions/liter). This concept
was first used by Johnstone, et al., ( ) in correlating S02
vapor pressure over sodium sulfite—sodium bisulfite
solutions. Operating S/C ratios for the absorber liquid
effluent varied from 0.85 to 0.95 during the test program.
The absorption tower contained a quench section for flue gas
cooling and fly ash removal and three packed absorption
sections. Each packed section contained ten feet of two
inch Tellerettes. The t lower packed stages were used as
absorbing stages and the upper stage as a demister. This
avoided the high liquid carry-over experienced during the
initial shakedown runs when all three stages were used as
absorbing stages.
6.1.1 S02 Removal
With two absorption stages in operation, 90 to 95 percent
S02 removal was achieved with inlet S02 concentration as
high as 3800 ppn. These data are presented in Table F—2 in
Appendix F. The system had enough capacity at the higher
S02 inlet concentrations to handle normal fluctuations in
the inlet concentration. The absorber was never overloaded
during the test program.
6.1.2 Mass Transfer Coefficients
Whitney, Hans, and Davis( 8 ) measured mass transfer
coefficients for the absorption of S02 into caustic media.
They found that for very alkaline systems the gas phase
resistance was controlling. However, as the absorption
increased and sulfite and bisulfite became the predominant
species, the liquid phase resistance became limiting. This
conclusion was based on lower overall gas phase mass
transfer coefficients as conversion increased.
15
-------
Johnstone and Singh(’ 1 ) studied the absorption of S02 in
alkaline media and found the mass transfer coefficients
varied with the gas rate raised to the 0.8 power. Their
system however was more dilute and more alkaline than
concentrations studied in the present work. Rein, Phillips,
and Young(’°) found that gas rate had no effect on the
absorption of S02 with ammonia solutions. The pH of their
absorbing liquor varied from 5.6 to 6.8, indicative of high
ainmonium bisulfite concentrations.
For the magnesium sulfite—bisulfite—sulfate system,
Pinaev C ) found that the presence of the sulfate increased
the sulfur dioxide vapor pressure at a given
bisuif ite—suif ite concentration.
It was found that S/C ratios in the draw liquor of 0.90 and
above could be easily obtained at concentrations of sodium
sulf ate up to 24 percent (wt). This verified the Johnstone
vapor pressure data. Thus, it can be concluded that sodium
suif ate does not affect the S02 vapor pressure for the
sodium system at concentrations up to 24 percent (wt).
Mass transfer coefficients were calculated for the present
system using interstage data collected during the program.
The equilibrium S02 concentration in the gas phase was found
using the vapor pressure data of Johnstone, et al. C )
Overall gas phase mass transfer coefficients were calculated
for both the top and bottaa stages and are presented in
terms of height of a transfer unit (Hog) using a packing
height of 10 ft per stage.. The results are suxranarized in
Table F—3 in Appendix F.
The absorption was found to be limited by the S/C ratio of
the absorbing liquor. Below an S/C ratio of 0.925, Hog is
constant at 3.5 ft. However when the S/C ratio reaches
0.925. Hog rises dramatically. This is shown in Figure 6.1.
At S/C ratios much above 0.90, the S02 vapor pressure curve
increases rapidly. In calculating the number of gas phase
transfer units for each packing stage, it was assumed that
operating and equilibrium lines were straight. For stage 2,
this gave conservative answers since the S/C ratio of the
interstage liquor was usually 0.85 or less. The vapor
pressure of S02 starts to become significant at S/C of 0.80
and above.
For stage 1, the assumption of a straight equilibrium line
is not as appropriate because of vapor pressure
nonlinearity. Also, experimental accuracy for measuring
inlet S02 concentration and liquid phase compositions in
this region was such that small errors caused a large
variation in the calculation of Hog. This factor is
16
-------
FIGURE 6.1 HEIGHT OF A TRANSFER UNIT, Hog VS.S/C
I-
LU
U i
WI
0
5000
45.00
4000
3500
30.00
25.00
20.00
15.00
10.00
2.50
d ’ 0
—-a-
0
.__ 0-
)
0
—
0
. —&c - — — —.
0
, 0
0.70
080
s/c
0.90
1.00
17
-------
probably responsible in large measure for the almost
vertical rise.
Figure 6.2 presents the overall gas phase mass transfer
coefficient (Kga) as a function of superficial gas velocity
(Gm). The data points shown are only those where the S/C
ratio is below 0.90. The overall gas phase mass transfer
coefficient is proportional to G 0.84. This is consistent
with the data of Johnstone and SinghC 1 1) for the absorption
of sulfur dioxide in alkaline solutions. At low levels of
S02 removal (60 to 70 percent), tbe mass transfer
coefficient was proportional to G 0.8. In view of the
errors involved in measurement and the assumption of
straight line equilibrium, this agreement is quite good.
Insufficient data were taken to determine the controlling
resistance. However, at low S02 ren vals, the gas phase
resistance appears to predominate. The gas phase
contribution will decrease as conversion increases as a
result of the increased S02 vapor pressure above the
solution. Thus, the overall transfer coefficient is
determined by gas phase or liquid phase resistances
dependent upon the region of operating conditions.
In the majority of the test runs, the bottom stage was
ineffective for S02 absorption.* Furthermore, as shown in
Table F—2. it contributed very little to the overall
oxidation. However, the bottom stage did increase overall
sodium utilization by increasing the S/C ratio. An increase
in the S/C ratio from 0.80 to 0.90 at 10 percent oxidation
decreased caustic requirements by approximately 9 percent
for a given S02 removal. This is very significant to the
overall process economics as 90 percent removal could not be
achieved in a one stage operation with recirculation. The
packing height in this stage was not varied during the test
program but results indicate that a lower height would have
been as effective.
6.1.3 Caustic Utilization
Caustic utilization can be defined as “the ratio of the
theoretical minimum caustic required for a given S02 removal
to that actually used.” Process limitations that increase
the requirements above the theoretical minimum are
oxidation, pH adjustment required in feed liquor
preparation, and completeness of the absorption reaction
(S/C ratio) -
The theoretical minimum caustic requirement is defined as
the stoichiometric equivalent of the S02 removed from the
gas. This assumes that:
1. No oxidation occurs,
2. No pH adjustment of feed liquor is required, and
3. All S02 is absorbed as bisulfite.
*See Table F-2, Appendix F.
18
-------
20
‘5
I0
7
0 84
K 9 9 - AGm
FIGURE 6.2 OVERALL GAS PHASE
MASS TRANSFER COEFFICIENT
vs. SUPERFICIAL GAS VELOCITY
C
0
E
.0
—a
I I
a
a ’
VELOCITY Gm(Ib mole /ft 2 lir)
19
-------
The first assumption is obvious, for any oxidation will
require 2 moles of caustic per mole of sulfur. The second
assumes stoichiometric acid addition to the net draw liquor
from the absorber and 100 percent effectiveness in releasing
the 502. This is not the case as will be seen in Section
6.3. The minimum recycle caustic required was 3.2 percent
of the caustic feed to the absorber • This is based upon the
measured stripping effectiveness of the pilot plant tower.
With more stages the required amount of recycle could be
reduced. The third assumption disregards 502 vapor pressure
over the absorbing liquor. At the absorption temperature
encountered in the pilot plant (120 F), the limiting S/C
ratio according to the Johnstone data is about 0.925. This
fact is verified by the rapid increase in the height of a
theoretical transfer unit at a S/C ratio of 0.925. Thus,
the minimum caustic requirement is not equal to the S02
removed but is 11.3 percent greater than the S02 removed.
This corresponds to a caustic utilization of 0.898. The
base value of 1.0 was chosen because the caustic requir nent
for a given S02 removal is the parameter of interest.
It is important to realize that overall caustic utilization
is a function of all the factors mentioned and not just the
level of oxidation. If one assumes a base case, as shown on
Table F—i in Appendix F, caustic utilization in the absorber
is 83 percent. Furthermore, decreasing the S/C ratio from
0.90 to 0.85 will increase caustic consumption by
4.9 percent, and increasing the amount of oxidation from
10.0 to 15.0 percent will increase caustic consumption by
3.7 percent. Additionally, incomplete stripping ot S02 from
neutralized cell feed liquor will increase the caustic
consumption. A five percent excess recycle acid flow will
increase caustic consumption by 4.2 percent and 1000 mg/i of
residual S02 in the stripper bottoms will increase caustic
consumption by 1.9 percent.
Caustic utilization averaged about 0.78 for the test
program. This can be translated into an oxidation of
approximately 13 percent, an S/C ratio of 0.925 and a
caustic recycle of five percent.
6.1.4 Entrainment
During the early part of the test program, significant
chemical losses were experienced. These losses were
attributed to the entrainment of process liquor overhead
from the absorption tower. Considerable dilution also
occurred as a result of quench water carry-over into the
bottom stage. In addition to the penalizing effects of the
dilution, the high calcium and magnesium content of the
entrained quench water was detrimental to the performance
and operation of the cell system.
20
-------
The process fluid entrainment loss was traced to the
recirculation liquor feed distribution nozzles. These were
high—pressure drop Bete nozzles which produced a fine
atomizing spray. Figures 6.3 and 6.4 present process fluid
entrainment as a function of gas rate and liquid
recirculation rate, respectively. The combination of high
gas velocities through the chimney trays and the atomized
spray produced by the Bete nozzles was responsible for the
entrainment. When these nozzles were replaced with a splash
plate type distributor, the entrainment dropped drastically.
The top packed section was also used as a de!uisting section
to further limit process losses. The third stage of
Tellerettes was not effective for removing the condensed
water caused by cooling of the flue gas. This water loss
was due to the humidity profile of the flue gas across the
tower and cannot be attributable to inefficient deinisting by
the Tellerettes. P11 overhead chemical losses were
eliminated using the above—mentioned modifications.
Figures 6.5 and 6.6 show the entrainment from the quench
section as a tunction of gas rate and impingement tray water
flow, respectively.
With no demister present, carry—over was large over the
range of gas rates and plate water rates studied.
Apparently, the high gas velocities through the chimney
trays were responsible for the entrainment. A polypropylene
demister pad was installed between the quench section and
the bottom stage. This eliminated carry-over below gas
rates of 1700 lb/hr/fta; measured values were in milliliters
of water per hour.
Due to tower geometry, the demister had to be installed with
an insufficient clearance between the pad and the bottom
chimney tray. Therefore, as the gas rate was increased, the
effective area of the pad was reduced as the gas passed
through the smaller chimney tray. This resulted in demister
flooding at gas rates above 1700 lb/hr/ft2.
6.1.5 Pressure Drop
Figure 6.7 shows tne pressure drop through the entire
absorption tower, including quench section, as a function of
gas rate. The tower internals account for most of the drop
since the unit was operating below flooding conditions when
the highest pressure drops were measured.
The relative contributions of tower internals and
circulating liquor to the pressure drop can be found by
examining a specific case. The pressure drop for a typical
run was 7.6 in. of water for the entire tower including the
21
-------
z
S..
-J
w
>
0
>-
U
1=5 GPM BETE NOZZLE
LIQUID RECIRCULATION CONSTANT 1 :25 GPM SPARGE NOZZLE
0.8
0.6
0.4
0.2
0
78 13.0 18.2 23.3
GAS VELOCITY(FT/SEC CHIMNEY)
FIGURE6.3 ENTRAINMENT FROM
STAGE I AS A FUNCTION OF GAS RATE
I I
2.2 3.6
SI
GAS VELOCITY (FT/SEC TCWER
I I
6.5
600 600 1000 1200 1400 1600 1800 2000
GAS RATE (LB/FT HR)
22
-------
0.20
LIQUID FLOW TO STAGE (GAL/MIN)
FIGURE 6.4
OF LIQUID
ENTRAINMENT FROM STAGE 1 AS A FUNCTION
RECIRCULATION RATE-GAS RATE CONSTANT
@ 1200 LB / FT 2 HR.
z
-J
4
C,
U
>
0
4
C .)
.
0.10
0
.
WITH BETE NOZZLE
X
WITH SPARGE TYPE (MODIFIED SPLASHPLATE)
x
I
I
I
I
I
0
5.0
10.0 15.0 20.0
23
-------
2.9
TOWER GAS VELOCITY(FT/SEC
10.4 15.6 23.4
CHIMNEY TRAY GAS VELOCITY (FT/SEC)
FIGURE 6.5 ENTRAINMENT FROM QUENCH
SECTION AS A FUNCTION OF GAS RATE-BAFFLE FLOW CONSTANT
z
-J
4
C D
U
>
0
5
P.O
0.9
0.8
0.7
0.6
0.5
0.4
03
02
0I
0
• PLATE WATER FLOWIOGPM
X PLATE WATER FLOW5GPM
/
/
A___.__
-
-,
NOTE. NO DEMISTER PRESENT
I I I. I I
600 eoo 1000 1200 p400
GAS RATE LB/FT
600 1800 2000
2 -HR
6.5
24
-------
FIGURE 6.6 ENTRAINMENT FROM QUENCH SECTION
AS A FUNCTION OF PLATE WATER FLOW-
GAS RATE CONSTANT 12OO lb/ft 2 hr
z
-J
4
Lu
>
0
>.
4
U
PLATE WATER FLOW (GAL/MIN)
25
-------
GAS RATE
FIGURE 6.7
PRESSURE DROP THROUGH ABSORPTION TOWER
(Including Quench Section)
20.0
10.0
5.0
0
oJ
I
3-
0
U i
C l )
U ,.
U i
a-
500 000 3000
(LB/ FT 2 - HR)
26
-------
quench section and demister. The drop across the top stage
including chimney tray (used as a demister and hence
essentially dry) was 1.9 in. of water. Thus, the drop
across the three packed stages was 5.7 in. of water or 75
percent of the total. The operating pressure drop across
the impingement tray and demister was 1.6 in. of water or 21
percent of the total drop. The circulating liquor,
therefore, contributed only 0.3 in. of water (1$ percent) to
the pressure drop across the tower • This level is well
below flooding conditions.
6.1.6 Particulate Removal, NOx and S03
During the latter part of the test program, EPA performed a
series of measurements to determine the amount of
particulates, NOx and S03, at the absorption tower inlet and
their removal from the flue gas. The removal results were
inconclusive, but the absolute level of these components was
reasonably established.
The measured outlet concentrations for all three variables
were greater than the inlet concentrations. The inlet and
outlet measurements were taken on different days and at
different times of day, so varying boiler load conditions
may account for some discrepancy. Table 6.1 presents the
EPA results.
NOx levels were also measured using the WEPCO “theta”
sensor. Both inlet and outlet concentrations were measured
for constant boiler load conditions. The NOx levels by this
analysis were reduced from 385 ppm to 231 ppm in the
absorption tower. The indicated removal is consistent with
published data on NOx removal by caustic scrubbing systems.
27
-------
TABLE 6.1
PAR ICUL TES NOx AND S03 LEVELS
ON ABSORBER INLE T AND OUThET
(MEASUREMENTS BY EPA EXCEPT AS NOTED)
Inlet Outlet
Total
Particulates (gr/SCF) 0.006—0.007 0.04 ± 0.014
Particulates as
Sulfate (gr/SCF) 0.0014
503 (ppm) 9 — 29 30 — 100
NOx (ppm) 134/220 329/287
167/1145 375/417
385* 231*
Water (% moisture) 7.0
1. No oxidation occurs,
2. No pH adjustment of feed liquor is required, and
3. All S02 is absorbed as bisulfite.
*Measurements by WEPCO ‘theta” sensor
28
-------
6.2 OXIDATION
Oxidation was defined as the amount of S02 removed that
appears as sulfate in the adsorber net draw. It is
expressed as percent of S02 removed. The determination of
the amount of oxidation that occurred in the absorber and of
those parameters affecting oxidation was a primary goal of
the test program. Increased oxidation directly affects
power and capital costs by increasing the number of “B”
cells relative to “A” cells as well as increasing the total
caustic required for S02 removal. Furthermore, the quantity
of S02 product will be reduced with increased oxidation -
Thus, operation with a minimum amount of oxidation is the
most economical for a given sodium sulfite-sodium bisulfite
ratio in the absorber.
Anitoxidants were considered as a possible means of
minimizing oxidation in the absorber. It is necessary, of
course, that any antioxidant used must not adversely affect
the performance of the electrolytic cells. Therefore, the
effect of several candidate phenolic—based antioxidants was
studied using small, bench—scale, Type “A” cells. All four
candidates tested (hydroquinone, para—aminophenol,
paraphenylerte diantine and catechol) resulted in fouling of
the cells from decomposition products. No further testing
was performed.
6.2.1 Literature Survey
There is very little data available in the literature on the
oxidation of a mixed sodium sulfite-bisulfite salt solution.
However, several authors (1,2,3, 4) have published
information on sulfite oxidation relating to the calibration
of biochemical reactor systems. Sulfite oxidation is used
for measuring oxygen absorption into the liquid phase.
Fuller and Cristc 1) found that the oxidation of sodium
sulfite is first order with respect to sulfite concentration
below a concentration of 0.0 15 mole per liter. Above this
concentration the reaction is limited by oxygen transfer
into the liquid phase and is independent of sulfite
concentration. They also noted a marked catalytic effect by
heavy metals such as copper at concentration levels as low
as 10_a moles per liter. The amount of oxygen absorption
was decreased by the addition of acid to the system. Acid
addition will convert the sulfite to bisulfite. Since
oxygen absorption was used as a measure of oxidation, Fuller
and crist report that bisulfite solutions are not oxidizable
and the decrease in oxygen absorption is dependent only on
the sulfite ion present. The rate of oxygen absorption,
however, does not change, absorption terminating upon the
reaction of all available sulfite.
Srivastava, et al. (2) confirm the Fuller and Crist results
by reporting the oxidation ox sodium sulfite is zero order
29
-------
with respect to sulfite concentrations above 0.04
moles/liter and first order with respect to oxygen
concentration. They also noted a catalytic effect of cobalt
addition at levels of 30 to 40 ppm. Similar results are
reported by Phillips and JohnsonC 5 ), and Yagi and InoveC3).
ChertkovC 4 ) studied the oxidation of magnesium—sulfite—
bisulfite sulfate solutions. He noted that in dilute
solutions (less than 12—wt percent salts) the absorbed S02
is almost completely oxidized. A lower oxidation was
observed in solutions with magnesium sulfate concentrations
above 12 percent. Lowered oxidation was attributed to an
increased liquid phase transfer resistance for oxygen
absorption resulting from the increased liquid viscosity and
density. Oxidation was found to be greater at a lower pH,
contrary to the Fuller and Crist results. This apparent
discrepancy can be explained if one examines the study by
Potts, et al.C6). Using a proposed free radical mechanism
for sulfite oxidation and the derived rate expression as
follows:
d(S03 ) = g(H303 )
dt 4I I
Potts, et al, calculatea the hydrogen and bisulfite ion
activities for a calcium salt system. The ratio (HSO3—) Hi
was shown to increase or decrease with pH depending upon
whether or not the system is saturated with CaSO3 1/2H20.
Decreasing pH inhibited oxidation for unsaturated solutions
and the opposite was true for saturated solutions. Chertkov
also noted the effect of gas rate on oxidation. As the gas
rate was decreased over a considerable range, the amount of
oxidation decreased. No explanation was offered for this
occurrence.
In general, the extent of oxidation should be influenced by
the concentration of oxygen in the flue gas, the
characteristics of the absorbing liquor, and the design of
the absorption .tower (incorporating such variables as gas
rate, gas-liquor contact time, aM type and volume of
packing). This test program studied the effect of operating
variables only. These included the characteristics of the
absorbing liquor and gas and liquid rates within the tower.
Oxygen concentration, packing volume and physical system
characteristics were not studied.
6.2.2 Data Reduction
During a test run, samples of caustic feed to the absorber,
interstage circulating liquor, and absorber net draw were
taken and analyzed for sodium sulfite, bisulfite,
bicarbonate and sulfate. Caustic flow rate, gas flow rate,
and S02 inlet and outlet concentrations were monitored
continually and averaged hourly. Material balances around
the absorber were used to determine the reliability of each
30
-------
data point.. Typically, there were five data points
generated per day for a three to five day run at specified
conditions.
The S02 disappearance from the gas stream was calculated
from the outlet gas flow rate and inlet and outlet S02
concentrations. Gas flow measurements, based on the
pressure drop over a straight run of pipe, calibrated using
a standard pitot tube, could be determined to 1360
pounds per 1 ur (2 percent of full scale). S02
concentrations, determined using an Intertech analyzer
calibrated by a standard span gas, could be determined to
±50 ppm. This would result in an error of ±0.3 gin-mole/hr,
or less than one percent of the average removal. Inlet and
outlet 502 levels were measured as dry gas eliminating any
variances due to water content. S02 measurements agreed
quite well with those determined by EPA using wet chemical
analysis (see Section 8).
The S02 appearance in the liquid stream was also determined.
The total sodium fed to the absorber was determined fran the
caustic flow rate and the chemical analysis. Free sodium
was determined by titration with hydrochloric acid and
sodium sulfate was determined gravimetrically by barium
chloride precipitation. The total sodium determination was
accurate to ±5 percent and the caustic flow rate to
approximately 2 percent. Measurements of the total sodium
leaving the absorber and the total liquid phase sulfur
appearance were hindered by the lack of a reliable flowmeter
on the net draw.
A total sodium balance agreeing to within 15 percent was
used as one criteria for data reliability. Based upon the
system water balance, some concentration of the process
liquor is expected (See Section 6.5). However, the absorber
net draw should have shown no more than a 15 percent
concentration. Total sodium analyses for the net draw and
caustic feed were compared directly (assuming equal mass
flow rates) and a data point was, eliminated if the ratio
deviated by more than 15 percent. This ratio also includes
analytical errors which could be isolated from the
concentration effect. The net draw flow rate was calculated
by assuming a sodium balance across the absorber.
The total liquid phase sulfur appearance was calculated by
two methods, one being a slight variation of the other. In
the first method, sulfur as S02 was found from the sodium
sulfite and bisulfite concentrations and the calculated net
draw flow rate. The sulfur as sodium sulfate was determined
by the difference between the inlet free sodium and sodium
as sulfate—bisulfite and bicarbonate analyzed in the draw
stream.
In the second method, the sulfur appearance as S02 was
calculated as above. However, sulfur as sodium sulfate was
31
-------
found from the difference between caustic and net draw
sulfate analyses. Liquid phase sulfur appearance to within
10 percent of gas phase sulfur disappearance was used as the
second criteria of data reliability. Sample calculations
have been presented in Appendix C.
Oxidation was primarily calculated by two methods. A sodium
balance method was based on sodium and sulfur material
balances. Sulfur appearance as sodium sulfate (determined
by sodium difference as above) was compared to the total
liquid phase sulfur appearance (sodium sulfite, bisulfite,
and sulfate). This method is subject to the errors in
absolute chemical analysis and net draw flow rate mentioned
above. A ratio method compares the total sodium to sulfate
ratios of the caustic and net draw streams • Given an
analytically determined S/C ratio, there is only one
oxidation level that can satify both the sodium to sulfate
ratio and the inlet caustic conditions. This method is
independent of S02 removed and net draw flow rate and has,
therefore, been used in all the data correlation.
A third method, used only as a check, is based on the acid
flow to the stripper. Given the acid molar flow rate to the
stripper and the molar flow rate of caustic required to
neutralize excess acid in the pH buffering system, one can
calculate the amount of S02 released. Since S02 absorbed as
sodium sulfate will not be released, the gas phase
disappearance of S02 through the absorber will be greater
than the ai unt of S02 released in the stripper. The
difference is the amount of oxidation that occurred in the
absorber. However, due to the unreliability of the acid
flow meter and the batch—wise measurement of buffering
caustic requirements, this method was applied for conditions
averaged over a whole run and has a very low confidence
level. Examples of these calculation methods are presented
in Appendix C.
The data correlations presented in the following section are
based upon a least squares analysis of all the 100 plus
acceptable data points (or 16 acceptable points for
interstage data.) The variance about the straight line was
calculated and a band of one standard deviation (J) has been
shown on all plots. The actual data points shown on the
graphs are individual run averages and are used as
representative of all the data points.
The wide scatter in the data is probably a result of several
factors. The imprecision of the gravimetric sulfate
analysis (see Section 8), variations in S02 inlet
concentrations due to boiler load changes, normal process
fluctuations, and instrumentation errors all contributed to
the scatter. The samples used were spot samples and no
attempt was made to obtain averaged ones. In view of the
magnitude of other errors and fluctuations involved, such an
32
-------
attempt uld result in a negligible increase in overall
accurancy -
6.2.3 Results
Table F—i in Appendix F summarizes the oxidation results of
the pilot plant operation. The data are composed of 22 runs
based on varying operating conditions. Run 1 to run 6 were
initial shakedown runs and have significant material balance
errors. They were therefore omitted from the final
tabulation. Runs 8.9,16, 17, 19, and 33 were omitted due to
material balance errors, significant process dilution, or
unsteady state operation. The results cover S02 inlet gas
concentrations from 1000 to 3700 ppm, gas rates from 4500 to
8700 lb/hr (4-7 FPS), and cell feed liquor concentrations
from 2.5 N to 4.0 N sodium sulfate. S02 removal varied from
83 to 97 percent. Each run consists of three to fifteen
distinct data points after screening for material balance
reliability. The majority of those points contained
absorber feed and net draw analyses only, however, several
included interstage analyses for the determination of
absorption and oxidation profiles. These runs are presented
in Table F—2 in Appendix F.
The major results of pilot plant oxidation studies were:
1. Oxidation levels in the absorber varied from 7
to 25 percent with an average of 13.3 percent.
2. Oxidation decreased as the amount of S02
removed increased from 125 to 300 gm—mole per
hour.
3. The effect of the amount of S02 removed (i.e.
the combination of inlet S02 concentration and
gas rate) on oxidation can be related to
increased sodium bisulfite concentrations.
The higher the sodium bisulfite concentration,
the lower the oxidation. Thus sodium sulfite
is the oxidized species.
4. The most oxidation occurred in the top stage,
the stage with the highest sodium sulfite
concentration.
5. Oxidation is a function of the solution pH as
determined by the concentrations of sodium
sulfite and sodium bisulfite. At a given
level of dissolved S02, oxidation decreased as
the sodium bisulfite concentration increased.
6. Oxidation decreased as the absorption profile
shifted toward the top stage. As the fraction
of the total SO2 removal in the top stage
increased, the sodium bisulfite in the top
33
-------
stage increased. This minimized the time that
sodium sulfite existed in the tower.
7. Splitting the caustic flow between the top and
bottom stages increased oxidation at constant
S/C in the draw liquor by shifting the
absorption profile toward the bottom stage.
8. Increasing the amount of heavy metals in the
absorber feed increased oxidation. Low
contaminant levels (less than 0.1 mg/i) can be
achieved through proper materials of
construction -
9. The sodium sulfate concentration in the
circulating liquor had little effect on
oxidation.
10. Increased gas—liquor residence time increased
oxidation. Oxidation appears to be controlled
by the liquid phase mass transfer resistance
to oxygen absorption.
6.2.4 Discussion of Results
There were t K primary goals for the pilot plant oxidation
studies. The first was to define where oxidation was
occurring, what species were being oxidized, and what
chemical parameters affected the amount of oxidation • The
second was to define the influence of operating parameters,
such as gas rate, on oxidation and determine what operating
region minimized oxidation.
The effect of gas rate on the amount of oxidation is shown
in Figure 6.8. As the gas rate through the absorption tower
was increased by a factor of two, oxidation decreased by
approximately 30 percent. This fi.nding is consistent with
Chertkov’s results for the magnesium salt system.
Figure 6.9 presents the effect of S02 inlet concentration on
oxidation. As the inlet concentration was increased by a
factor of two, oxidation decreased by approximately
30 percent. The product of gas rate and S02 inlet
concentration, at a specific percent removal, is the total
S02 removed in the absorber. Figure 6.10 shows that as the
removal increased by a factor of two, oxidation decreased by
50 percent. The amount of S02 removal can be related to the
system loading.
System loading can be defined as “the t al amount of
dissolved S02 in the absorbing liquor.” As the loading
increases, for a constant caustic feed rate, the absorption
reaction is driven more toward completion and sodium
bisulfite becomes the predominant species. Figures 6.11 -
6.13 show the influence of gas rate, S02 inlet
311
-------
25
-
—
o
c x
—
o
—
-
—
+
—
p
-
0
0
0
6 ’
—
°
— — —
—
(.
—
—
—
0
0
—
4000 5000 6000 7000 8000 9000
STANDARD DEVIATION
FLUE GAS RATE (LBIHR)
FIGURE 6.8
THE EFFECT OF FLUE GAS FLOW ON OXIDATION
(SO 2 INLET CONCENTRATION VARIED)
20
I5
z
0
I-
10
x
0
5
0
U,
-------
25
SO 2 INLET CONCENTRATION (PPM)
FIGURE 6.9 THE EFFECT OF SO 2 INLET CONCENTRATION
ON OXIDATION (FLUE GAS FLOWRATE VARIED)
20
a
0
15
z
0
I-
0
5
0
0
0
—.---
0
w
2
C
-.......--.--
oc
0
cr•
.o 0
.—.
r - ---- -
-..u-..--..
...II...-
O
+o.
)
—
800 1200 1600 2000 2400 2800
3200
3600
-------
25
- 15
z
0
I0
0
FIGURE 610 TIlE EFFECT OF TOTAL SO 2
REMOVED ON OXIDATION
20
C
.
0
l...0.._. .
0
.
.
o
L
0
p
—
8
0
— -
+ 0•
—0
5
0
100
I SO
200 250
TOTAL SO 2 REMOVED (GMOL/HR)
300
-------
FLUE GAS RATE (LB/HR)
FIGURE 6.11 THE EFFECT OF FLUE GAS FLOWRATE
ON THE NET DRAW SODIUM BISULFITE CONCENTRATION
(so 2 IN LET CONCENTRATI ON VARIED)
180
w
I.-
-J
4
I6O
0
4
140
I-
U i
z
z
120
I—
-J
U)
100
0
0
(I )
L)
0
•o
0
.
— -
— -
— — —
a
0•.
—
0
-4-o_
-
0
d
—
)
P
.—
—
——
0
.
0
———
0
—-0--——-
—O-
)
0
.
80
4500 5000 5500 6000 6500 7000 7500 8000 8500 9000
-------
-J
C,
4
I-
w
z
z
LU
I-
U-
-J
C l )
0
Cl)
200
180
160
140
120
100
80
INLET SO 2 CONCENTRATION (PPM)
FIGURE 6.12 THE EFFECT OF SO 2 INLET CONCENTRATION ON
THE NET DRAW SODIUM BISLJLFITE CONCENTRATION
1000 1500 2000 2500 3000 3500 4000
39
-------
TOTAL $02 REMOVED GMOL/HR
FIGURE 6.13 THE EFFECT OF TOTAL SO 2 REMOVED ON
THE NET DRAW SODIUM BISULFITE CONCENTRATION
U i
I-
-j
4
4
0
I-
U i
z
z
I i i
I-
LL
-J
D
U)
0
0
U)
200
175
150
125
I00
75
C
100
125 150 175 200 225 250 275 300 325
-------
concentration, and total S02 removal on the absorber net
draw bisulfite concentration. The S02 inlet concentration
strongly influenced the ultimate bisulfite concentration in
the net draw liquor. This was expected since the high S02
partial pressures drove the absorption reactions to
completion. The net draw bisulfite concentration is plotted
against oxidation in Figure 6.14 and it can be seen that as
the bisuif ite concentration increased, oxidation decreased.
Figure 8.1 (Section 8) sbows the distribution of sulfate and
bisulfite as a fraction of pH. The lower the pH, the more
sodium bisulfite present in the system and the lower the
oxidation. This result agrees with the work done by Fuller
and Crist(’), and is consistent with the work of Potts, et
al.C’) for unsaturated solutions. Typically, the Pilot
Plant absorber net draw pH varied between 5.5 and 6.0.
The effect of alkalinity on the amount of oxidation is made
clear from the basic oxidation reaction mechanism.
Na2SO3 + 1/202 Na2SOL& (1)
NaHSO3 + NaOH - Na2SO3 4 H20 (2)
NaHSO3 + 1/202 + I aOH NA2SO4 + H20 (3)
The oxidation reactions (1) and (3) are influenced by the
sulfurous acid equilibria of reaction (2) and oxidation can
only occur to the extent that sodium sulfite exists in the
system. This is consistent with the findings of the pilot
plant. Sodium sulfite is the oxidized compound and
minimizing the sul1 ite will minimize oxidation. This also
confirms the Fuller and Crist conclusion that bisulfite is
not oxidizable.
The fact that sodium sulfite was the oxidized compound is
reconfirmed by the analyses of the absorber interstage data.
Figures 6.15 and 6.16 show that when S02 removal in the top
stage increased, both in terms of absolute S02 removal and
fraction of total S02 removal, oxidation decreased. This
shift minimized the region of sodium sulfite within the
absorber and therefore minimized oxidation.
Shifting the absorption profile toward the top stage is
equivalent to operating with a lower absorber net draw pH.
This would result from a decrease in the amount of free
sodium fed to the absorber at a certain S02 removal. The
penalty in this type of operation is a lower S02 removal
efficiency.
The effect of alkalinity also explains the increased
oxidation observed when splitting the caustic flow to the
absorption stages. By feeding caustic directly to the
bottom stage, the overall alkalinity of the tower increased
(in terms of length) and there was more sulfite in the
bottom stage. Consequently the amount of oxidation
41
-------
25
NaHSO IN NET DRAW (GRAM/LITER)
FIGURE 6.14 THE EFFECT OF ABSORBER NET DRAW
SODIUM BISULFITE CONCENTRATION ON OXIDATION
20
I5
0
I-
4
0
0
5
0
90
00 110 120 130 140 150 160 170
-------
250
SO 2 REMOVED IN TOP STAGE (GMOL/HR)
FIGURE 6.15 THE EFFECT OF SO 2 ABSORPTION PROFILE
ON OXIDATION (502 REMOVED IN THE TOP STAGE VS. OXIDATION)
20D
0
15.0
z
0
2 0.0
x
0
5.0
0
60 80 100 120 140 160 180 200 220 240
-------
30
a
. 20
2
0
I—
0
I5
0
FIGURE 6.16 THE EFFECT OF SO 2 ABSORPTION
PROFILE ON OXIDATION (PERCENT OF TOTAL SO 2
REMOVED IN THE TOP STAGE VS. OXIDATION)
25
A SPLIT
CAUSTIC RUNS
0
p
‘
—
10
5
30
40 50 60 70 80 90
PERCENT OF TOTAL SO 2 REMOVAL IN TOP STAGE
100
44
-------
increased. The amount of S02 removal that occurred in the
bottom stage can be related to the S/C ratio of the top
stage effluent. As the S/C ratio increases, by definition,
the bisulfite concentration increases (Figure 6.17) and the
available sodium for S02 removal decreases. Figure 6.18
presents the relationship between the total amount of
oxidation and the S/C ratio of the top stage effluent. As
the S/C ratio increased, the amount of S02 removal in the
bottom stage and the overall oxidation decreased.
The effect of the absorption profile on oxidation can be
demonstrated by considering the fraction of the total
oxidation that occurred in the top stage. Figure 6.19 shows
that as the absolute amount of S02 removal in the top stage
increased the fraction of the total S02 oxidized in the top
stage increased. Therefore, as the absorption profile
shifted toward the top stage, the amount of oxidation that
occurred in the whole tower decreased and that oxidation
occurred preferentially in the top stage (compare
Figures 6.15 and 6.19). In other words, shifting the
absorption profile toward the top stage maximized the sodium
bisulfite in the top stage (Figure 6.20) and thus minimized
oxidation.
In sununary, oxidation was minimized by operating the
absorption tower with maximum sodium bisulfite levels
consistent with a desired S02 removal.
The literature indicates that oxidation is controlled by the
liquid phase mass transfer resistance for oxygen absorption.
Thus, oxidation should be affected by those parameters that
affect the mass transfer of oxygen, namely the gas—liquid
contact time and the liquid phase viscosity, density, and
turbulence.
Oxidation was plotted against the total salt concentration
of the net draw, as shown in Figure 6.21. There appeared to
be no effect on oxidation as the sodium sulfate
concentration increased from 2.5 to 3.5 gin—mole per liter.
ChertkovC 4 noted an effect of salt concentration on
oxidation and related it to increased solution viscosity.
However Chertkov’s study involved solutions considerably
more dilute (10 to 15 weight percent salts versus 20 to
25 percent) ana it was also an Mg system.
Residence time however had a very noticeable effect on
oxidation. A hypothetical residence time was defined as the
stage circulation rate in gallons per minute divided by the
total free sodium feed rate to the absorber in gm-moles per
hour. While the units have no significance, this “time” can
be visualized as the number of times a given volume of
solution contacts the oxygen—laden flue gas before it is
changed over with fresh absorbing liquor. Figure 6.22 shows
that as this “time” increases, oxidation increases. The
ratio was multiplied by 3/2 for the split caustic runs to
45
-------
0.90
0.80
0.70
0.60
0.75
S/C RATIO
FIGURE 6 I7 MOLE PERCENT SODIUM BISULFITE
VS. S/C RATIO
LU
I-
I i -
-J
C l ,
0
0
U,
I-
z
LU
C-,
L i i
a-
LU
-J
0
0.80 085
0.90
46
-------
25
S/C OF TOP STAGE EFFLUENT
FIGURE 6 18
PROFILE ON
STAGE
THE EFFECT OF SO 2 ABSORPTION
OXIDATION (THE S/C OF TOP
EFFLUENT VS. OXIDATION)
20
— IS
z
0
0
XI
0
5
-4
0
0 70
075 080 085 090
-------
80
502 REMOVED IN TOP STAGE (GMOL/HR)
FIGURE 6.19 THE EFFECT OF SO 2 ABSORPTION PROFILE
ON OXIDATION (TOTAL SO 2 REMOVED IN TOP STAGE VS. FRACTION
OF TOTAL OXIDATION THAT OCCURRED IN THE TOP STAGE)
j TO
C!,
U)
a.
o 60
z
z
0
50
0
x
0
-J
40
I—
0
0
U
U-
30
20
60
80 100 120 140 160 180 200 220 240
-------
140
80 100 120 140 160 80 200 220 240
SO 2 REMOVED IN TOP STAGE(GMOL/HR)
FIGURE 6.20 SO 2 REMOVED IN TOP STAGE VS.
SODIUM BISULFITE CONCENTRATION IN TOP STt&GE EFFLUE NT
I-
z
L i i
-J
L i
U-
LU
LU
I-w
U,’-
o-J
1=’
I 30
120
110
tOO
90
80
•60
-------
25.0
TOTAL SALTS IN NET DRAW(GMOL/LITER)
FIGURE 6.21
THE EFFECT OF TOTAL SALT CONCENTRATION
IN ABSORBER NET DRAW ON OXIDATION
U i
C,
20 0
5.0
z
0
0
)( 100
0
5.0
0
2 50
275 3.00 325 350
-------
25.0
20.0
z
2 15.0
4
U i
0
10.0
5.0
0
0.020 0.040 0.060 0.080 0.100 0.120 0 140
STAGE RECIRCLJLATIONRATE(GPM)
FREE SODIUM FEED RATE (GMOL/ HR)
FIGURE 6.22 THE EFFECT OF GAS-LIQUID CONTACTING ON OXIDATION
-------
reflect the increased changeover time with only half the
caustic flow to the top section. Figll.re 6.22 also
incorporates the high liquid recirculation rate run (run
27). The effect or LLquld recirculation rate on oxidation
could not be isolated since only one run was made at high
liquid circulation rates. However, based upon the effect of
contact time ana the results ot the high circulation rate
run, oxidation should be minimized at the minimum
recirculation rate required to wet the pacKing. An
associated variable which should also affect oxidation is
the volume of a particular type of packing. Although not
specifically studied during the test program, gas-liquid
contact beyond that required to achieve the desired S02
removal would increase oxidation due to increased oxygen
absorption by the liquor.
Figure 6.23 shows the effect of iron concentration in the
caustic feed on oxidation. As the iron content increased,
the level of oxidation increased. This agrees with
literature results on the catalysis ot sulfite oxidation by
heavy metals. The iron levels in the caustic were not
monitored frequently and the values reported are based upon
a single measurement for an entire run. The source of the
iron contamination appears to be the steel cathodes in the
electrolytic cells. Cell membrane cross leaks, caused by
the crystallization discussed in Section 6_U, causec
corrosion in several compartments leading to the
contamination of the caustic feed to the absorber.
Fluctuation in these iron levels occurred as a result ot
cell operation, but qualitatively the effect of the iron
content on oxidation proved significant. It should be
noted, however, that for most runs, the iron levels in the
caustic were less than 0.5 mg/i indicative of a low degree
of contamination. Proper cell operation and materials of
construction should allow operation at. contamination levels
below 0.1 mg/i.
In suzmnation, oxidation appears to be significantly affected
by the characteristics of the absorbing liquor and the
extent of gas—liquid contact. Minimizing sodiui i sulfite
within the absorber, both in terms of concentratic n and
residence time, minimizes oxiaation.
Additionally, minimizing the amount of gas—liquid contacting
to only that necessary for good absorption should minimize
oxidation.. The effect of oxygen concentration in the riue
gas as well as the type and volume of absorber packing which
determine the contacting surface area were not studied as
part of this program but should affect oxidation as
indicated above.
52
-------
z
0
I.-
4
0
0
25 0
20 0
15 0
10 0
5
0
—
+O
0 0
-.------.--
.
- --
—
—
0 040
080 I 20 160 200 240
IRON LEVEL IN RECYCLE CAUSTIC (MG/L)
FIGURE6.23 THE EFFECT OF HEAVY METALS
CONTAMINATION ON OXIDATION
280 320
-------
6.3 STRIPPING
The net draw liquor from the absorption section was
acidified with recycle sulfuric acid releasing the S02 in
the sodium sulfite/sodium bisulfite mixture. The released
S02 is then stripped from the liquor using indirect steam.
The acidification reactions are as follows:
Na2 S03 + H2S04 Na2SO4 + S02 + H20 (4)
2NaHSO3 + H2S04 Na2SO4 + 2S02 + 2H20 (5)
Any S02 oxidized in the absorber was fixed as sodium sulfate
and could not be released. The total acid consumption will
be less than the caustic consumption by that amount. The
resultant increase in sulfate ions within the system makes a
I SBN cell acid purge necessary to maintain the overall acid—
base balance.
The required stripping steam rate and effectiveness of 0 cid
addition were studied during the test program. A more
detailed study was rDt possible because of a lack of
instrumentation. There were no liquid flowirteters on either
the feed or effluent flows and the overhead gas flowmeter
was plagued with problems throughout the test program.
6.3.1 Stripping Steam Rate
Figure 6.24 presents tI stripping steam rate (as a function
of S02 removed from the flue gas) versus the residual S02
levels in the net strirper bottoms.
The minimum steam required for the reboiler including heat
losses and feed preheat requirement was about 6 lb steam per
pound of S02 removed. Estimates of the preheat required to
raise the feed from 110 F to 235 F were made. Of the
measured steam flow of 6 lb per pound of S02, 4.25 pounds
can be accounted for as process preheat. The remainder is
process stripping steam and heat losses. If the steam
equivalent of this duty is subtracted from the total steam
fed to the reboiler, the resultant stripping steam required
is 1.0 to 2.0 lb per pound of S02 removed including heat
losses. The final design steam rate will depend upon the
heat exchange method cbosen. Based upon absorption data for
the S02 —H20 system, a minimum of about 0.5 —1.0 lb of steam
per pound of S02 removed is required for stripping. This is
exclusive of heat losses and the solubility depression of
sodium sulfate in the liquor.
Theoretical transfer unit heights could not be calculated
due to the lack of reliable flow rate data.
54
-------
STEAM (LB/HR)
SO 2 REMOVED(LB/HR)
FIGURE 6.24
LIMITING STRIPPING STEAM AS A FUNCTION OF SO 2 CONCENTRATION
IN STRIPPER BOTTOMS
800
700
600
500
400
300
200
100
0
w
I .-
-J
E
U)
0
I-
I —
0
U i
0
ct
I-
(I)
z
N
0
Cl)
U,
01
a- STRIPPING STEAM I
(WITHOUT PROCESS x
PREHEAT)
I
I
I
0
x
X
—
—TOTAL STEAM
(INCLUDES LOSSES AND PREHEAT)
(ACTuALMEASUREDSTEAM)
(
X
0
X
%
‘
‘
‘
‘I
‘
‘
\
.‘
‘
— eo
\
‘S
S
00 0
•%
— i,—-—gI -—
KK
K
I
I
I
I
I
I
0 2.0 4.0 6.0 8.0 10.0
-------
It should be noted that the stripping steam tests were
performed after the bottom eight—foot packed section had
dropped into the reboiler drum. The support tray for the
bottom stage may have tailed as a result of the rupture disc
relieving several times previously.
6.3.2 Acid Addition
The addition of the stoichiometric amount of acid to the net
draw liquor was an important part of the overall process
economics. Since cell feed liquor pH must be about 8.5 for
iron and aluminum removal, any excess acid addition to
stripper feed must be neutralized with recycle caustic.
This results in an increase in the number of electrolytic
cells required for a given S02 removal.
Conversely, if less than the required amount of acid is
added, unreleased S02 (above 200 ppm) will remain in the
stripper bottoms. The oxidation of ferrous iron to ferric
iron takes place in the feed liquor treatment section. Any
large excess S02 must also be oxidized to ensure the
oxidation of divalent to trivalent iron. This would result
in increased peroxide costs as well as additional “B” cells
for removal of the additional sulfate formed.
Based on operating conditions from Table 6.2, 400 milligrams
per liter of S02 in the stripper bottoms is the equivalent
of 1.0 percent oxidation. Additionally, an excess acid
level of 0.05 g—moles per liter i i the stripper bottoms is
equivalent to 16 percent excess acid flow in the absorber
net draw.
The acid addition to the net draw stream was controlled by
the pH of the stripper bottoms. The typical operating pH
range was from 2.5 to 3.0.
Figure 6.25 presents the residual S02 levels in the stripper
bottoms and the excess acid in the stripper bottoms
(expressed in grnoles sulfuric acid per liter of bottoms
effluent) as a function of pH. This curve is based on
samples taken over the entire test program and in all cases
the stripping steam was well above the required minimum.
Since caustic is needed for neutralization of excess sulfate
ions from S02 oxidation as well as the excess acid, one can
calculate the optimum pH for stripper bottoms operation as a
function of recycle caustic required. A base case with
operating conditions in accordance with Table 6.2 was
assumed. For several pH levels in the stripping bottoms,
the recycle caustic (as a percentage of caustic feed to the
absorber) was calculated from excess acid and residual S02
56
-------
pH OF STRIPPER BOTTOMS
FIGURE 6.25 EXCESS ACID, RESIDUAL SO 2
IN STRIPPER BOTTOMS AS A FUNCTION OF pH
012
w
I-
-j
-J
o 0 0
3
0
-J
U.
U)
0
I-
I .-
006
LU
a-
a-
0 04
(I )
z
002
U)
U)
U
U
x
LU
(I)
0
C))
-I
0
-u
rn
cu
0
-4
- I
0
(I)
6)
r
-I
r n
18 20 22 24 26 28 30 32
0
34
-------
values using Figure 6.26. The optimum pH was found to be
3.2 with a minimum caustic recycle of 3.2 percent.
This does not however represent the economic minimum since
peroxide costs and differential power costs ( BN cells for
oxidation versus “A” cells for caustic) were not included.
Incorporation of these cost figures would shift the minimum
to a lower pH. Since Figure 6.25 was based on actual pilot
plant data, the minimum recycle calculated is peculiar to
the pilot plant operation. In future installations,
additional stripping trays should be included. This t uld
lower the caustic recycle required.
The titration curve for a sample of feed liquor (20 percent
wt sodium sulfate) was determined and compared to one for
water (Figure 6.27). The buffering effect of the salt
solution is significant and allows pH control in the range
of interest without the complex instrumentation and
incremental acid or caustic addition that pure water
requires.
TABLE 6.2
BASE CA OPERATING CONDITIONS
FOR STRIPPING STUDY
Gas Rate 8000 lb/hr
S02 Inlet 2000 ppm
S02 Removal 90%
SIC in Net Dra’i 0.90
Caustic Normality 2.0
Acid Normality 1.0
Oxidation in Absorber 10%
58
-------
14
pH OF STRIPPER BOTTOMS
FIGURE 6 26 OPTIMUM pH FOR STRIPPER
BOTTOM OPERATION (EXCLUDING COST FACTORS)
IC
8
12
LU
0
C l)
4
0
I.-
U
LU
Li-
Li
0
0
I-
C /)
4
0
U
-J
0
>-
C)
U
I —
z
U
0
U
0
6
4
2
0
2.2
24 26 2.8 30 3.2 3.4
-------
12
ACID EQUIVALENT X tO — I—)m’ BASE EQUIVALENTX iO
FIGURE 627 TITRATION CURVE—SODIUM SULFATE SOLUTION VS WATER
SAMPLE SIZE5OmI
I0
8
9
7
I
0 .
6
4
6 5 4 3 2 I 0 I 2 3 4 5
GO
-------
6 • L L .CTkOLYTIC }
-------
1/202
D
No 2 SO 4
1
I
I
CMIO
H 2
2 No +
2OH
FIGURE 6.28A SCHEMATIC DIAGRAM OF “A” ELECTROLYTIC CELL
D
No SO 4
- 4j SO 4
D= POROUS DIAPHRAGM
CM= CATION SELECTIVE MEMBRANE
H 2 $04
No SO 4
H 2 SO
AM
+
— H
4 S0
D POROUS DIAPHRAGM
CM CATION SELECTIVE MEMBRANE
AM ANION SELECTIVE MEMBRANE
FIGURE 6.28B SCHEMATIC DIAGRAM OF “B” ELECTROLYTIC CELL
4
No 2 SO 4
No 2 SO 4
2e. ,..p 1/202
NoOH
No 2 SO 4
. 2e
2e
Na OH
CM
2No
—2.
I
62
-------
Table 6.3
SELECTED DESIGN VALUES OF WEECO PILOT PLANT
“ A”Cells “ 8”Cells Total
No. of CelLs 56 32 88
Active Electrode Area
(0.945 sq.ft.pcr cell) ,sq.ft. 52.9 30.2 83
No. of Stacks 2 1 3
Cells per Stack 32 & 24 32 —
Cells per Electric Module 8 8 —
Maximum Current Density, ASP 180 180 —
Design Current Density, ASP 142 127 —
Stack Current, amps 1074 960 —
Cell Voltage, volts 6.48 6.63 —
Total Power, ]C4DC 48.7 25.5 74.2
DC Energy per lb NaO}I,a kwhr/lbNaOH 2.32 2.37 2.34
Production Rates, lb-mole/hr
NaOH in mixed caustic 0.525 0.269 0.794
2 SO 4 in mixed acid 0.263 0.071 0.334
142504 as pure dilute acid 0 0.063 0.063
Make—up waterY” lb—mole/hr 0 3.56 3.56
GE 4 0 0.128 0.128
Ccoling Requirements Cc)
Anolyte, B1-u/hr 47,800 23,300 71,100
Catholyte, I3tu/hr 31,900 15,500 47,400
cooling Water Cd ) CPU 5.3 2.6 7.9
(a) Phase I rerfortance Objectives (3N ta SO 4 feed solution, 140 0 F,
85 NaOH curront efficiency) adjusted Etoatward for the above Design
Current Densities.
(1 ,) Consists of water fed to the anolyte of “fl” cells only.
C c) ucced on above cell voltages less 2 volts reversible voltage.
So’ l]un sulfate soluUon assunad fed at 1000 F and make—up water
at P U 0 F. Latent heat of v.ipor ration of water at electrodes
taken iiito accotnt b it leszes to environ-tent neglected. Total
cooling split 60— 0, ano] tc—catho] te.
(d) Based on 30° F tertpcrature rise of cooling water.
61
-------
FIGURE 6.29
De9ign Point I atorial lialance for 5 Type “1 SULFCMAT
Coll o EPCO SO 2 Removal Pilot Plant.
2 3]
— 5 6
7 8 9 1O
:OL
:oL
LB
7
HR
I OL
LB
MOL
21.0 .52.5
l7#.3 I. o6 5 ,35 .037w 6?.o . f6 £3 .0377 69.d •4/
ç 7 3I. 17.73 . ? zi IZ. 7 17.73 • gç 263
I.i’ 7 .0637
.6I .3o9
73 23.)
23.1
297
13 L
761
2 . 903
3r?6
MOL
5 36
16 1.3
133.9
25
5 ’lS
OL
.263
zg.6
.f7’I .03 9
‘1.94 .if4 ,
fl44
43
roai.
iil 7
MDL
4.40
47?
:IOL L9 0L
“hR 1R hR
c :i
3
377
, q
O377
)
L 1(b [
— —
— —
5 p. Cr.
1.7.7-7
/.127
1.27.7
1.7.7.7
I .3’/z()
I.3 ”/Z.
i.ns(”)
g .,_g5 ’
/00
/ 0—I’l0
,. .
/00
ioo-1 f 0
(. o.*
J’1O
130-19o
out’
I’jo
3c-aqo
Press, psia
—
17.0
17.0
c , 0P2
—
—
—
—
,.n7( .%)
A cells, o.1’
Feed Noz nality
Current Density
NaOH Current Efficiency
Purged I1 S0 4 Current Efficiency
Operating Temperature
Catholyto Caustic Norin.ility
i¼nolyte Acid )4orrnalily
Iiid—f .u1”t Aci ‘J si, I Ity
sq. ft. ea., 52. sq. ft.
4 N Na 2 SO 4
142 . 7 SF
- i 1/f3
= iqo ° ‘
- 2 . w
, I l
-—
— Ii
:: 04
1120 Liquid
H .,0 Vapor
02
Basis:
©
-------
.S32. O4
1-
H 2 0 Liquid
1120 Vapor
If-’
02
Totals
FIGURE 6 30 Do iqn roint ?:aterial r .nlance for 32 Type “B” SULFOMA? 4
Cc11 of C! O SO 2 Rer oval Pilot Plant
1
2 3 J
—— 5 6 7 0 — 9
) I i
L 3 :OL
d
L3 :CL. LU 0L LB ‘“CL LU DL LU 10L LB ;:OL LU i:OL
V V V V V Z zz V , V ‘
HR HR I HR fIR HR HR HR hR 1111 HR B un ‘HR HR
g74 . I7 2.7’? •CI 3 35.4 .2’19 2.7’? .0I ’ 13
290 i&ia 9.07 .ço’ ii&I v.07
iS2.2
,J 71;
.2.i i
0
1 10.3
3 ’ O
co4 I31. 7 qg I4J7
.517 .0326
.317 .,5 3
IV.7
2.9
6.z .0 3.3 4 I 4.56
61.0 3.39 j339 241
.2.93 .6/63
2.53 .677!
70.0 4779
64.1
64.1
3.c6
6S. 6
“97
2( 7
343
0L
• r7i I
GPI
.6/&
•ci ’7Z
.2.4w
.o’97..
.2t ,9(l 1 )
2.9w
.i27((. )
.szg
•c .o
. Gr.
f.Z7..7
/.LL7
,.2.Z7
f.z17
iYf2 . .( )
I.3’IZ
i.o :()
,.o I
/.o0
1.270
OF
—
/60
#60 -#110
, 1 O t
/60
leo -/40
I ..
#40
131-I’iO
, o.’t
/‘f C)
,jc-140
I
g
,‘,to
Prcz , psia
17.0
—
17.0
cr: , OPT
—
—
—
—
!iOu1 J l s)
—
.6o2.(r5)
11111 t
AM D CM
O Basis:
5
32 13 Cells,O.94&sq.
Food Normality
Current Density
HaD!! Current £fficiency
Purgcd u1 2 S0 4 Current Efficiency
Operating Tcnpcroture
Catl ’olytc Caustic Normality
Anolyte 1 cid Norrnality
flid-AnOlyt ’ i cid f zii.1lIty
ft. ea., 0.L sq. ft.
4 N Na 2 50 4
= 127 ASF
g5• .4
- 40 %
140 O .
= 2 . :1
= 2_N
= o.c3 N
-------
The change in cell voltage with current density, consistent
with contract pertormance objectives, is given in
Figure 6.31.
6.14.3 Cell System Operations
It is convenient in discussing the electrolytic cell system
to divide the pilot plant operating time into three periods:
initial cell system and process debugging period (three
months), cell system evaluation period (six months), and
advanced component evaluation period (three months). Each
period ended with disassembling of the cells for inspection
and rebuilding the cells with new membranes and diaphragms
for the next period.
6.4.4 Initial Cell System and Process Debugging Period
In this period, the fully integrated S02 removal pilot plant
was operated for the first time.. Process and system
deficiencies and refinements were identified and corrected,
and personnel became familiar with the operation of the
plant.
In the electrolytic cell area, the principal problem
encountered was the quality of the cell feed liquor. It was
intended that dissolved iron would be removed by a manganese
zeolite (Mn e) conditioner which oxidizes the ferrous ion to
ferric ion and also filters out the precipitate ferric
hydroxide under alkaline conditions.
The initial problems with the MnZe conditioner resulted from
pH excursion.s in the process steam. At low pH’s, iron
passes through the unit and at high pHs, aluminum was
leached out of the bed. Both metals in hydroxide form
precipitated out mainly on the cathode diaphragms of the
cells causing increasing pressures within the cells. A new
automatic pH control point was therefore installed between
the heat exchanger, (T—103), which is downstream of the
stripper, and the sultate surge tank, (M—1O1). Also, the
MnZe conditioner was relocated to the electrolytic cell
section of the pilot plant to permit operation at the
required pH level of 8 to 9.
Subsequently, temperature excursions of the feed liquor
stream occurred unaer the Nisolateu mode of operation of the
cell system and these led to degradation of the bed in the
MnZe unit. Also, bed regeneration was sometimes incomplete
and iron breakthrough occurred. Each upset resulted in
additional deposits in the cells, attecting the internal
pressure distribution. Removal of deposits was accomplished
to some e,ctent by acid-washing the catholyte diaphragms in
situ , but with some detrimental effects to the cation
membranes then in use.
A heat exchanger was installed to eliminate the temperature
excursions under the isolate” mode of operation. The bed
66
-------
FIGURE 6.31 DESIGN CELL VOLTAGE-CURRENT DENSITY
RELATIONSHIPS FOR TYPES “A”AND”B” SULFOMATTM CELLS
S
C’,
w
0
>
L u 5
-J
a
0
0
.J3
-j
L i i
0
2
0
0 20 40 60 80 I0 i 120 140 160 ISO
CURRENT DENSITY,AMPERES PER SQUARE FOOT (ASF)
200
67
-------
of the MnZe unit was replaced and regeneration water for the
unit was changed from city water to steam condensate. De—
ionizers were installed to treat the latter at the pilot
plant due to contamination from new pipes which brought the
condensate to the pilot plant area.
Because diaphragm-grade caustic, rather than rayon-grade
caustic, was used inadvertently to charge the pilot plant
initially, excessive chloride levels were present in all
process streams. Consequently, at the end of the debugging
period, process liquors were drained and new solutions were
prepared using rayon—grade caustic.
Also, because of the numerous upsets that had occurred which
led to deposits in the cells leading to the resulting need
for in—situ acid washes and the above—mentioned excessive
chloride levels throughout the period, the cells were
rebuilt at the end of the period with new membranes and
diaphragms - Anodes and cathodes were reused.
During the initial debugging period, the “A” cells were
operated 443.5 hours and the “B ” cells were operated
354 hours.
6.4.5 Cell System Evaluation Period
The bulk of the test program for the integrated pilot plant
was conducted during this period. Cell operating times were
2093.5 hrs and 2089 hrs for the A” and B” cells,
respectively. There were five continuous runs of ten days
or more during this period. The longest was 13 days.
Cell operating levels varied from 70 percent to 100 percent
of design levels, depending on the caustic feed rate to the
absorber and the relative liquid levels in the caustic,
mixed acid, and sulfate surge tanks (M—105, M—104, and
M—101, respectively).
6.4.5.1 Current bfficiency Results — Current efficiency
results for this period and for the final “Advanced
Components Evaluation Period” are suimnarized in Table 6.4.
The overall performance of all the cells in the pilot plant,
both AS and “B”s, is given by the current efficiency of
the “Combined Catholytes.” This measurement compares the
actual amount of NaOH produced by all the cells to the
theoretical amount that could be produced according to
Faraday’s law for the current applied to the cells.
Table 6.4 shows that, for the first three-and-a—half months
of the period (through C.E. No. 21), the current efficiency
of the combined catholytes was consistently from 86 percent
to 93 percent, and averaged 90 percent. This compares
favorably with a design value of 85 percent (see
Figures 6.29 and 6.30).
68
-------
TaAle 6.14
SUMMARY CF WEPCO PILOT PLANT CURRENT EFFICIENCY MEASUREMENTS
• re 19j’’’. t’ii’..itto, I”.’%
1.12
‘2 13/’/73 2.6’. fl 1.31
3 17/’. 2.44 756
11 I’/lt ?. 1 1.03
1’. I’’)O 3.1’l 156 I 03
16 15/6 3.111 91 % 1.04
17 17/1 ” 3.0’l 5 03
it 11/’S 3.lM 756
1’) 22/29 4.07 1 1074 1.01
1 ,i4 3.0% 915 0.59
‘1 In 3.19 912
0.91
22 1/23 4.29 715
0.99
23 2/4 3.0.1 710 1
24 2/’) 4.ON 909
0.96
25 2/22 3.811 756
0.95
24 2/13 3.’.’ 733
0.99
21 2/16 3.Pi 915
0.59
39 1/7 3.611 5174
1.00
2’ 3/6 3.711 1074
1.03
30 3/10 3.19 1071
1.00
71 3/16 3.211 171’
1.04
32 3,25 3.511 912
11. %dvarcd Co-rnent Evaluation 55 ciod
33 ‘n m’ 3.51 1.00
34 4/15 3.471 756 .97
35 4/30 7 .5’i 9] 1.00
R.’tlo of total equiv.7lants of 11a0’I amI 12 .24 irvo.ved in the C.E. ),asuresent.
(a) Acc-j—u lotlon tens n isu’c’5 ro;119 1’fle duo to 2n’ aai’pIe titration inconsistency.
lb s proIM?,1c.. alIhvJJ, urc.,rtair, rciult — die te Incons Istency nf o,e of three
ample tltr il Ions.
Ic) Slight e,tc:nat it t 1 ta7o of 2-Ancilyte occurred In these runs (‘se teat).
‘24 - cc:— 1 1—Cells
P el t a c tt r ,. 1 lt,% Current Ccsthinod Cathofl$5
c.r’.o. or. _ Ccr’t. . Cr..’rt D j cu r:nt llesitj inns C.E.
— .- , L1 ti 1 131 )11
A-CslIs 9 — Cells
9-Ano lyte 8-Mid Anolyte 9-Anolyta
litt le. C.f. llnrn. C.S. he m. CS.
112 504 112504
1.52 92% 1.40 95%
1.81 91% .92 91%
1. O 93% .91 99%
1.89 69% .96 93%(b)
2.27 91% 1.55 939 (b)
2.02 69% 1.25 90%
1.69 66% 1.31 99%(8)
2.02 91% 1.36 93%
1.97 68% 1.15 91%
1.94 ILL. 1.13 !!!.
90% Av.— 93%
9,. ,,
130.0
91 1.1 l
130.0
121.
‘39.2
i o n.0
152
1 ’ l. C’
1 :0.6
121.0
99.2
120.2
100.0
90.6
171.0
112
142
142.5
147.1
120.6
675 89.3
675 89.3
663 87.7
672 86.9
816 107.9
(69 66.5
672 86.9
963 127
919 108.3
813 107.5
939 124.2
666 89.1
819 108.3
72 89.9
669 88.5
831 109.9
960 127
813 107.5
916 107.9
960 127.0
610 107.1
___________ 9-Coil Consiatency Ratie
Tot.tcId C.E. ( ‘SitU ll, j _
.38 29% 1.96 — —
.51 40% 1.98 36% 7 6 %Ic)
.49 30% 1.85 — —
.54 32% 2.03 11% 735 (c )
.34 22% 2.16 58% 809 (c)
.50 41% 2.02 35% 7 69 1c)
.59 53% 1.89 21% 749 (c)
.74 54% 2.24 34% 98%
.46 19% 1.98 62% 91%
.64 36% 2.11 49% 95%
.71 41% 2.10 38% 79%
.46 17% 2.13 56% 75%
.44 19% 2.01 64% 83%
.50 24% 2.11 63% 87%
.49 29% 2.05 55% 84%
.45 29% 2.06 51% 80%
.57 37% 1.95 32% 69%
.67 43% 2.06 39% 82%
.69 39% 1.92 39% 77%
.56 27% 1.60 50% 77%
.60 42% 1.50 3’.t 77%
Tot.Au33% Tot.Au’ 41% ?ot.Av-79%
1.93
1.87
1.95
1.99
2.04
1.99
1.97
2.14
1.95
1.91
1.84
75%
91%
80%
81%
79%
64%
72%
73%
90%
74%
69%
Av.
.87 84%
1.02 66%
1.26 78%
1.05 83%
1.04 83%
1.11 90%
.87 75%
.78 69%
.98 79%
.75 72%
.75 62%
Tot .A85%
99.6 672 99.9 1.73 91% 1.35 89% .67 46% 2.00 47% 93%
1130.0 691 90.1 1.99 91% .99 93% .48 54% 1.99 43% 97%
139.3 792 104.8 1.95 91% 1.02 98% .45 53% 1.91 45% 99%
Tot Av. 91% 1,v — 00% Av.I1% Av .
-------
In the last two—and—a—half months of the period
(C.L. Nos. 22—32), the catholyte current efficiencies were
all significantly lower. The deterioration of the cells was
caused by crystallization as a result of too high a Na2SO4
feed concentration, i.e., in the vicinity ‘4N Na2SO4. This
high concentration was first used for an extended period
starting in nid—January. In the concentration ranges
encountered, the crystallization point of the a2SOL& —NaOH
catholyte mixture decreases with increasing Na 2S04
concentration. However, crystallization occurs in this case
when the temperature exceeds the crystallization point. As
a result, localized crystallization occurred withirh the
cells, specifically in the catholyte compartment, the mid—
catholyte compartment, and extending into the cation
membrane. The crystallized material was still present in
five of eight “B” cells that were disassembled in late
January to investigate unaccountably high voltages for that
electric module.
It is noteworthy, based on the results obtained before
crystallization occurred, that the current efficiency is
independent of the current density, or production level, at
least over the range covered, which is 70 percent,
85 percent, and 100 percent of design point.
As stated above, the combined catholyte current efficiency
gives the performance of both the “A” and B cells
together. The A ” anolyte current efficiency measures the
“A” cells alone.* Table 6.L4 indicates an average “A” anolyte
current efficiency of 93 percent tor the period before
crystallization occurred, significantly exceeding the design
basis of 85 percent. Current efficiencies were considerably
lower after crystallization had occurred, indicating
significant internal crossleaking.
The sulfuric acid produced in the “B” cells is divided
between the “B” anolyte stream, which is essentially free of
sothuin (Figure 6.30, Stream 7), and the mid-anolyte stream,
which contains sodium sulfate (Figure 6.30, Stream 10).
This mid—anolyte stream is combined with the “A” anolyte
product, which also contains sodium sulfate, and flows to
the “front end” of the pilot plant for recycle to the
stripper -
*Note: The independently measured catholyte and anolyte
current efficiencies must be equal for a given cell
provided there are no undesirable side reactions at
the electrodes and no external leakage of either
product. Our prior experience with these cells
indicates that the catholyte current efficiency may
run slightly higher than the anolyte current
efficiency.
70
-------
The current efficiencies of the “B” anolyte and mid—anolyte
are measured separately and are given in Table 6.14. The sum
of these t efficiencies gives the current erficiency of
the total acid proauced in the “B” cells, also given in
Table 6.4.
Of the results obtained before crystallization occurred,
Table 6.4 shows that the “B” cell total acid current
efficiencies are significantly below those of the combined
catholytes in C.E. Nos. 13, 15—18. This suggests leakage of
some acid from the system to the outside. The amount of
leakage can be estimated by assuming an 85 percent overall
current efficiency for the “B” cells. On this basis, the
indicated leak rate for the “ rst case” is
1.6 GPH (100 ccAnin) of “B” anolyte acid. Leakage in the
“B” anolyte recirculation pump (into the mechanical seal
cooling water) was discovered in early’ December; new seals
were installed by mid-month. C.E. No. 19 data were taken
after the pump was rebuilt, and the “Is” cell total acid
current efficiency of 88 percent (Table 6.14 is consistent
with the current efficiencies of the combined catholytes and
“A” anolyte for that run, indicating no further external
leakage. (Note: The design point throughput of the anolyte
recirculation pump is 9.01 gpm so that even at 70 percent of
aesign point, the 1.6 GPH leak rate was about 0.14 percent of
the pump throughput.)
The t columns of current efficiencies for the “B” anolyte
and mid-anolyte in Table 6.4 show wide scatter. However,
where the mid—anolyte current efficiencies are high
(C.E. Nos. 18 and 19). the “B” anolyte current efficiencies
are low; and vice versa (C.E. Nos. 16 and 20.) Furthezi ore,
where the mid—anolyte efficiency is high, the mid—anolyte
acid normality is high; and where the inid-anolyte efficiency
is low, the acid normality is low. These characteristics
indicate the presence of an internal crossleak between the
mid—anolyte and the “B” anolyte streams. The crossleak has
been in one direction in some cases (e.g., C.E. No. 16) and
the other in other cases (C.E. Nos. 18 and 19), depending on
the relative pressures.
For C.J .. Nos. 33—35, Table 6.4 shows little scatter for the
“B” anolyte and mid—anolyte current efficiencies. Average
values of 45 percent and 51 percent, respectively, were
obtained in the absence of internal crossleaks. These
exceed the design basis of 40 percent and 45 percent,
respectively.
71
-------
6.4.5.2 Energy Consumption Results - The total d—c energy
applied to the bus bars for both the “A” and “B” cells
divided by the total production rate of caustic by the two
types of cells gives the overall specific energy consumption
for caustic production. KWH/lb NaOH. Specific energy
consumption results for both the “Cell System Evaluation
Period” and the “Advanced Comporiei t i .valuation Period” are
summarized in Table 6.5.
The “Performance Objective” values in Table 6.4 for the
overall NaOH energy factors are based on the Phase I
performance objectives (3N Na2SO4 feed solution, 140 F
operating temperature, 85 percent NaOH current efficiency)
but adjusted downward to reflect tfle actual current
densities. The adjustment was made in accordance with the
cell voltage—current density relationship given in
Figure 6.31.
The “Measured” overall NaOH energy factors in Table 6.11 are
calculated from the measured caustic production rates from
both types of cells, the d—c bus bar current and the
“Average Cell Voltage Excluding Highest E.M.” (see
Table 6.5). The latter is derived by subtracting the
voltage of the electric module (eight cells in parallel
electrically per electric module, Table 6.3) having the
highest voltage from the bus bar voltages and dividing the
result by the number of remaining electric modules. This
was done to eliminate the effects of incomplete gas venting
from the center compartment and/or mid—catholyte compartment
of the pilot size cells. Table 6.5 contains both the
average cell voltage with all E.N.’s and the “Average Cell
Voltage Excluding Highest E.Zi. ” The latter are also given
in Table 6.6 arranged in accordance with the Na2SO4 feed
concentration and current density.
Table 6.5 shows that during the three-and-a—half months
prior to the occurrence of crystallization in the cells
(i.e., up to and including C.E. No. 21) • the overall NaOH
energy factors met the performance objectives at each
production level. At the 70 percent production level
(C.E. Nos. 12—15, 17. and 18), the overall NaOH energy
f actors matched performance objectives. For the 85 percent
production level case (C.1.. No. 16, 20, and 21), the
measured energy factors “bettered” the performance objective
by three percent. Finally, for the 100 percent production
level case (C.L. No. 19), the measured energy factor
“bettered” the performance objective by almost nine percent.
6.4.6 Advanced Component Evaluation Period
Because ot the deterioration of the cells due to crystal-
lization, the cells were rebuilt and some advanced
components were installed. In particular, fluorocarbon
based cation membranes were installed in most cells. Eight
ot these membranes had been installed as test specimens at
the end of the “Initial Cell System and Process Debugging
72
-------
Table 6.5
SUWMARY OF VOLTAGE-CURRENT DATA AND SPECIFIC ENERGY CONSUMPTION RESULTS IN WEPCO PILOT PLANT CELL SYSTEM
i.e. ct
C’!, ‘ Lrrg
C.!. A/9 re.,d Sn,
‘ !‘i e. , p .
3. C—Il S _ nt— C ‘5 ation P ri’I
12 ‘‘/,2 7,6, 140
II ‘/72 7_C IU
I I 1’/32 2.91 un
1 ’ ‘‘/ 7.’. ’ 5.1 .
1’ ‘/32 1.22 fl2
57 t/!3 3. ” 241
11 ‘/32 7.1:. 135
13 ‘‘/32 4 .‘75 141
:0 1.0/32 3.’fl 140
25 t’/72 Ia.’ 140
77 4’/72 4.21. 140
‘(/32 3 l 137
21 ‘ (fl2 4 I 347
Pt t’/ 2 3. ” 136
:s ‘r/.3 3, 92 l ’ s
77 t’/31 7,r1 339
t’/37 2.’’. 137
:i ‘(/32 3,r: 113
30 Or/J O 7.) 1 13’)
‘/32 3.? 119
;43/33 ),‘l 145
12, 7. ‘S;:— 2 r.---,- rt fl’5!’9 rnr,’v !
2’ t’/32 ‘. 1 37.’
74 : 1 .122 3.4) 177
‘9 4 ./74’ 3,”) 535
An. roll vo hrP Ir — . 9 .
Onsber All Eeclu.U,,g mother Current Bother All !,rlu.line
Voltage EM. Highest cM:o Current 06rnty Voltage !.M,s Unsent EM ,
volts volts volts Sep., 57.2’ ,ulL.. l1l I .alt..
41.5 5.93 5.90 675 09.3 23.5 5.07 5, 07
43.0 6.00 1.95 675 09.3 22,5 t.V7 ‘.41
42.0 6.00 6,00 653 07.7 34.0 6,20 .‘.lI
41.3 1.V0 673 60.9 34.0 6.00 5 ‘0
43.0 6.14 6.12 016 201,9 26,0 6.50 6.’O
42,4 6.06 669 00.5 26,7 6.67 6,4)
39.5 5.61 5 7 673 00.9 33,0 5,95 6,13
45.0 6.43 6.20 963 127.4 27,0 (‘.75 1.10
43.0 6.26 5.97 019 100.3 25.0 6.45 1.10
43.0 6.14 5.90 013 107.5 21.0 6.21 1.23
35.5 6.04.6.30 5.92.6.30 939 121.2 30.5 7,62 0.’)
79.9 5.70 5.62 666 60.1 23.9 5.97 5.5
42,4 6,06 5.95 019 100.3 25.0 6.4% 0,33
39.7 5.67 5.6 1 672 00.9 23.5 5.61 5..’)
39.5 5.64 5.50 669 OP,6 24.0 (‘.14’ ‘,‘7
42.0 6.11 6.03 631 109.9 36.4 6.60 (,‘7
475 5,79 6.59 960 527.0 20.0 7.00 0’. 7
47.5 6.79 6.53 013 107.5 26.0 6.50 6,47
47.2 6,74 6.47 016 107.9 26.0 6.00 6.4’
45.3 6.42 960 127.0 30.0 7.30 7..’
40.5 5,62.6.05 5.10.6.00 010 107.1 26.2 6.55 6.51
0 ereI1 i ’ M
1 .1 ) ‘41’ 1
‘ k.s. r’ 2 -sr I. — —
K. , 51 Ha.”
2.? )
5, •‘
5.4 ’ 1.’?
‘1
2.59
2.10 2.00
1.” 7.21
3.51 2.:.
3.11 2,14
2.73 2,19
2. 12 1 _ n
2. 5 ’ :. u
2.24 2.1 9
7, 5 1
:. 7o
2.14 2.24
2.79 7.35
2.11 2.2?
a. ASS ‘ 9 A—’—!. lr ? ’ , ot l,. ’j i. ,Irs.,ll.s55 ,
5. All 32 5—c’!llJ ‘fly ( 51. 1.7 Ir ,Itsill’alty.
r, tilts’s o: n e et-trlc —nO un hatirg lncrr anods..
d, C31’—u’at” rr .. ‘ rann’lr prw:,—tinn rat’i, oc, surbar current and average cell voltage
5,2.2117
5, Ps.a I Ft’f .ro,,’t : ‘ ‘ 0 ‘3 !‘‘ 704 ‘n . .) s .1.,’son, 140 0 F, 055 Si a N) current nflictr’n.-yS
but a1y. zte.2 ‘Sr”.-’ so ac;—J.n • with tI. ’ ccli voltage—current .Psnnlty relationship 9ivon to
rig. 4 tfl S’2Ouit for 129 5!.’.fl’ a.tuul C’Jtroflt C l sttien.
A — itLSS
B — LtS
Pus5ar
Cu’rnnt
4’
053
747
:
750
754
1034
915
915
‘to
“7
7 1.6
753
955
1074
1074
1077
‘ 974
912
Current
0 ”imtty
ASS
99.6
100.0
94.9
100.0
121.0
99.2
100.0
542.1
121.0
120,6
121,161
9s2
120.2
100.0
99.6
121.0
542,1
142.1
142.5
112.1
121,161
71.) 95,6 30,3 5 , 72 , 495 c 5.72.4.95
71.6 100.0 39.0 5 , 63 , 5 , 25 c 5.61,5,25
2053 139 3 36.0 6.33 6.33
672 00.9 23.4 G.lS.l. QS 650495
601 Vu.! lid u.lJ.Snkft.A’, .00
792 104.0 59.1 5.31 6.35
1.91 2.21
1. ’ J
2. 12 2?6
-------
Period.” Four of the eight were reinstalled after visual
examination. The other four were analyzed and found to be
in excellent condition despite some adverse conditions that
had occurred during the nearly 2100 operating hours.
In addition, one electric module of “A” cells and one
electric module of “B cells were outfitted with “inert”
anodes. These anodes have a coating of noble metal or noble
metal oxides on a titanium substrate. They offer the
potential of significantly lower voltage (see below), are
free of corrosion products, and are dimensionally stable.
Aside from these cation membranes and inert anodes, new
diaphragms and anion membranes 0± the same types as before
were installed. Most of the lead alloy anodes and the
cathodes were reused.
During the “Advanced Component Evaluation Period,” the “A”
and “B” cells operated for 885 hours and 889.5 hours,
respectively. Operation of the pilot plant during the
second half of this period was conducted with little
contractor supervision to determine the operability of the
S02 removal plant by power plant operators. A continuous
run of 412 bours, or 17.2 days, was made. This was not only
the last run but also the longest uninterrupted run of the
entire test program. The run had to be terminated due to
failure of the acid sulfate pinup, (P—10’5), which fed the
mixed acid to the stripper.
6.4.6.1 Current Efficiency Results — Three current eff i-
ciency measurements were made during the first month of the
period before the reduction in contractor personnel. The
results are included in Table 6.4. The catholyte current
efficiency, which gives the overall performance of the
cells, was 91 percent, exceeding the design value of
85 percent.
The “A” anolyte current efficiencies averaged 90 percent and
the “B” cell total acid current efficiency averaged
96 percent. both above design values. The B” cell total
acid current efficiency consists of the “B” mid—anolyte
current efficiency of 51 percent and the “B” anolyte current
efficiency of 145 percent (Table 6.3).
6.14.6.2 Energy Consumption Results — Energy consumption
results for the Advanced Component Evaluation Period appear
in Table 6.5, and cell voltages for the electric modules
with “inert” anodes are summarized at the bottom of
Table 6.6.
Table 6.5 shows that the overall NaOH energy factors for
C.E. Nos. 33—35 are well within the performance objectives
in each case.
Substantial voltage reductions with the “inert” anodes were
realized as shown in Table 6.6, and can be more than
714
-------
Table 6.6
SUMMARY O ’ A” AND”B”CELL CURREL T-V0LTAGE C1 TA
Na 2 SO 4 A t Cells ttBtt Cells
Anode Feed Current C .E • Cell Cuxrent C .E • Cell
Type Conc. Density No. Volt o Density No. Voltage 4
ASP volts
Pb alloy 2.4—2.6 N 100 (12) 5.901 89 (12) 5.87 5 8
— (13) 5.98 (13) 5.83
2.9—3.2 N 100 (14) 6.00 89 (14) 6.13 ..,
(15) 5.88 5.82 (15) 5.90 ( 6.ll
(17) 5.93 (17) 6.50
(18) 5.47 (18) 5.92
121 (16) 6.12 108 (16) 6.50
(20) 5.97 6.02 (20) 6.40
(21) 5.98 (21) 6.23 6.39
(30) 6.43
142 (30) 6.47 6 127 (31) 7.45
(31) 6.421
3.4—3.7N 100 (33) .72) 66 90 (33) 6.10 6 5
— (34) 5.611 (34) 6.20 .1
121 (32) 5.58 107 (29) 6.47?
(32) 6.53 6.45
(35) 6.35)
141 (28) 6.58) 127 (28) 6.97
(20) 6.53 6.48
(35) 6.33
161 (32) 6.80
3.8—4.2 N 100 (23) 5.62 89 (23) 5.05
— (25) 5.61 5.60 (25) 5.83 5.92
(26) 5.58 (26) 5.97
121 (22) 5.92 109 (24) 6.43 6 50
(24) 5.96 5.97 (27) 6.57
(27) 6.03
142 (19) 6.20 126 (19) 6.60 6 5
(22) 6.53 6
161 (22) 6.30
“1n ’rt’ 3.4—3.SN 100 (33) 4.95 90 (33) 4.95 4 97
(4/]2) 4.75 . (34) 5.00
(4/13) 4.C0
(34) 5.25
163 (4/26) 5.4
* ‘ rv c cefl :1cctr F 1c) ;olta e cx luchng or .c L .M. with
hig c’ t vol1ac ( co T i.c )IX ). fh’2 a cr c es, thorefore, ?LC
gcn ra1i . c r I .o1. ?. rcllz t: .‘l.
volt:-cs ‘r i ce1I . “ Li ’ ‘ iLa cells w3th “iricrt” nodcs
‘i•e rc,r a s l’ J’c c --
75
-------
25 percent at the higher current densities. At 160 ASP, the
cost of electrical, energy for the 75 MW prototype plant
(3100 lb of S02 absorbed/hr) is reduced by approximately
$60,000/yr (at $0.00781/kW) by using inert anodes. The
incremental capital. st of inert anodes is not a
significant part of the prototype plant costs.
76
-------
6.5 PROCESS MATERIAL BALANCES
6.5.1 Process Losses
The mech nicQl loss of water and process fluids, sometimes
totaling hundreds of gallons per day, occurred durinc the
life of the pilot plant pro ram and caused operatinq
limitations. These losses were related to mechanical and
handling problems. The sodium losses monitored in these
streams accounted for all the make—up sodium required;
therefore, make—up liquids were not required due to process
problems. In addition to operating losses, shutdowns of
iuore than a few hours required that the equipment be drained
completely in order to c void crystallization and pluaqincT.
These conditions represented a significant economic penalty
and emphasized the need for careful loss monitoring and
recovery facilities.
During the first several months of operation, several types
of losses were identified and corrected. They were:
1. Absorber entrainment loss — corrected by using the
third packing stage as a demister.
2. Entrainment from catholyte circulation tank—
caustic loss controlled by lowering liquid level in
separation tank.
3. Cell area pump seals - stopped by replacinq pump
seals with double mechanical seals and monitoring
flush water for seal leakage.
4. External cell leaks — corrected by collecting and
returning liquids to process.
6.5.2 Loss Lnalysis Run Results
During January, 1974, the absorption-stritpinq section and
i-he cell room were operated separately for a period of
22 hours in an attempt to identify and quantify liquor
losses frctn the system. This test was conducted after the
electrolytic cells had been damaaed by internal
crystallization due to operation at high Na2SO4
concentrations in the cell feed liquor (see Section 6.6).
The main results obtained from this test were:
1. Pump drit pinas, samples, and miscellaneous losses
in the absorption-stripping section amounted to
28.2 gal per day, representing a sodium loss of
580 gm—ion per day and a sulfate loss of 190 cm—ion
per day.
77
-------
2. The sodium inventory for the adsorption—stripping
section (front end) remained constant, indicating
no chemical losses other than that attributable to
1 above. SilTlultaneously, the sulfate inventory
increased at a rate equivalent to 16.8% oxidation
in the absorber. The observed sodium sulfate
increas€ in the cell feed tank and the calculated
sodium sulfate increase agreed within experimental
error.
3. The water loss for the front end was 65 gallons per
day or 5 1 40—pounds oer day. This increased salt
concentration in the cell liquor from 20.24 to
21.37 wt%. The maximum probable error was
approximately 200 pounds of water per day.
4. Miscellaneous losses in the cell area were
7.5 gallons per day, representing a sodium loss of
80 gm—ion per day and a sulfate loss of 147 gm-ion
per day.
5. The sodium and sulfate inventories for the cell
area remained constant during the test indicating
no chemical losses other than accounted for in
4 above.
6. The water disapPearance in cell area amounted to
175 gallons per day: 11.8 percent due to
electrolysis, 149•7 percent for humidification of
the effluent 02 and F2 streams, 4.1 percent due to
leakage, and the remaining 314•4 percent (or
60 aallons) could not be accounted for.
6.5.3 Discussior Cf Results
:ear the end of the test proqram, EPA carried out a series
of tests on the absorption tower to determine the water
content of the effluent gas and the entrainment of process
fluids was deternined. The flue qas rate and absorber
temperature profile were essentially identical to those
existing during the Loss of Analysis run. It was therefore
assumed that similar losses existed in both cases. The
comparative ooeratina conditions and the measured losses are
presented in Table 6.7.
6.5.3.1 Chemical Losses - Pump drippinas and miscellaneous
losses were the only chemical losses in the front end.
Assuming a cell feed rate of 2.0 GPM, these losses were
calculated to be less than 1% of the total fluid processed.
The closure of the sodium balance indicated no extraneous
leaks, such as continued absorber carryover. This was
confirmed by the EPA study. An entrainment of
0.004 grains/SCF as sulfate ion corresponded to
78
-------
TABLE 6.7
COMPARATIVE OPERATING CONDITIONS ND MEASURED LOSSES
( as S02
rate TEMPERATURES (F ) removed
( lb/hr) inlet quench outlet ( lb/hr )
Loss
Analysis 5310 280 120 110 2U.0
Run
EPA Study 5541$ 270 120 105 33.3
Front end
Stripper Flue gas Water Entrainment
Overhead outlet loss arains/SCF
temperature (F) % moisture ( gal/day) as SO1$
Loss
Analysis 83 NA 65 NA
Study
EPA
Study 110 7.0 NA 0.004
79
-------
0.014 aallons per hour of process fluids: this was only
0.05 percent of the net absorber flow and accounted for
virtually no chemical losses.
The chemical losses in this test run were, therefore, due to
the various leakages. They amounted to about 660 gm-ions of
sodium per day or the equivalent of 62 gallons of cell feed
licruor per day. Durina the test program substantial
chemical losses did exist. In addition to leakacie, other
sources of chemical loss were absorber carryover and sodium
lost via the “B” cell anolyte purge stream. The absorber
carryover was corrected by using the third packing stage as
a demister. Losses via the “B” cell anolyte were a result
of cross leaks through the anion membrane caused by liquor
crystallization and operational upsets. Measured values of
Na in this stream were as hiqh as 13,000 mg/i which at
design flow rates would represent a sodium loss of
308 am—ion per day. Replacement of the anion membranes late
in the test program corrected this loss.
Not included in the analyses above were substantict]. liquor
losses incurred by system drainage upon shutdown and losses
resulting from operational errors. The shutdown losses
amounted to as much as 100 gallons per shutdown when ambient
conditions necessitated complete drainage.
6.5.3.2 Water Balance - The water balance for the process
indicated a loss of 65 gallons per day from the front end
and 175 aallons per day for the cell area. Water could be
lost through carryover in the absorption tower, with the S02
in the stripper overhead flow, by electrolysis within the
cells and by humidification of the cell vent streams.
Based on the stripper overhead temperature and the S02
remov& during the Loss Analysis run, the water loss with
the S02 would be 3.2 gallons per day. The correspondina
equilibrium S02 concentration would be 984 wt%. A catch
pot was installed in the stripper overhead line, but no
substantial amount of water was collected. This indicated
that entrainment of condensed water was minimal.
The water balance around the absorber was more complicated.
The flue gas from the boiler, containinq some water, entered
the quench section at a temperature between 270 F and 310 F.
It was cooled by water evaporation to 120 F and passed
through a demister to the absorption section. The water
pickup in the quench section was a function of gas flow,
flue gas temperature, and inlet humidity. The gas passed
through the absorbing stages at near quench temperature, but
was cooled in the demisting (third) stage to about 105 F at
the absorber outlet. This cooling appeared to be a result
of heat losses to the atmostthere.
80
-------
The EPA study indicated that the outlet gas contained 7
moisture as entrained water. This corresponded to water
entrainment of 0.0017 lb H20/SCF dry gas or 305 gallons per
day for the test. The water condensed in the third stage,
assuming a temperature drop from 120 F to 105 F and
saturated gas leaving the absorber, was
0.0021 lb Ff20 SCF dry gas or 369 gallons per day. Thus it
appeared that the majority of the condensed water was
carried overhead. Depending upon the exact temperature
profile across the third stage, the effective height of
packing was reduced and demister efficiency decreased
correspondingly.
Based on the above, the process liquor inventory should have
shown a net water gain of about 61$ gallons per day if the
flue gas was saturated at 120 F as it left the quench
section. Fowever, with a water loss in the absorber of
61.8 gallons per day for the 22 hour Loss Analysis run, the
relative humidity of the flue gas leaving the quench section
was calculated to be 90 percent. This would result in a
water evaporation in the chemical section of
0.00071 lb H20/SCF dry gas or 124 gallons per day and close
the system water balance. If the holdup in the absorber and
the net throughput are assumed to be 60 gallons and
30 gallons per hour respectively, a loss of 61.8 gallons per
day from a 20 percent salt solution will concentrate the net
draw solution by 7%. This concentration increase is
comparable to that measured by sodium balances through the
tower. During the Loss Analyses run, the ratio of total Naf
in the caustic feed to total Na in the net draw was 1.06
to 1.08. This shows excellent agreement with the
concentrating effect predicted by assuming that the inlet
aas was not saturated.
The humidity of the inlet flue qas could not be measured
during the test program. However, a boiler manufacturer’s
efficiency figures indicated the water content of the inlet
flue gas might be 0.0504 lb Ff20/lb dry gas. This was based
on an unknown moisture and hydrogen content in the coal.
With that inlet humidity and a quench section effluent
humidity of 90%, the water pickup in the quench would be
0.00 193 lb H20/SCF dry gas. This estimate does not check
with the adiabatic saturation temperature of 120 F. The
heat balance will be satisfied only if there is
0.0372 lb H20/lb dry cyas in the inlet gas. Flue gas
humidity at this level can result fron’ the combustion of
coals with hydrogen and moisture contents of approximately
2.7 and 8.3 percent respectively. These values are not
typical and lower inlet humidity was apparently closer to
the correct value for this test run.
Water loss via electrolysis in the cell area was estimated
to he about 20 gallons of water per day. This loss was
-------
calculated from the cell amperaqe measured for the Loss
Analysis run. This value varied with particular operating
levels for the test program.
The water loss by humidification of the 02 and H2 vent
streams from the cells was calculated from the difference in
humidity between the vent streams and the ambient air used
f or dilution and the CFM ratings of the cell dilution air
fans. No attempt was made to measure the actual gas flow
rates of these vent streams. Water loss by this
humidification was estimated to be 87 gallons per day.
The balance of the water loss, 60 gallons per day, was
unaccounted for.
During the last months of the test program, water was added
to the cell feed liquor tank to maintain a constant density
feed to the cells. This water was added at a rate dependent
upon hourly hydrczneter readings and averaged about
60 gallons per day. This was lower than the expected
230 gallons per day measured during the Loss Analysis run.
Changes in inlet flue gas water content, ambient humidity,
as well as varying levels in the cell system surge tanks
could well account for these differences. Additionally,
slight density changes, undetectable with a hydrometer,
could have resulted in a difference of several hundred
gallons of water in the 4000 gallon tank. continuous
operation was not m intained long enough (the longest run
being nine days) to permit resolution of these differences.
6 • 6 CELL FEED TREATMENT
The feed to the electrolytic cells should be free of
substances which will cause operational difficulties.
Examples of such substances are calcium carbonate and iron,
aluminum and magnesium hydroxides which will themselves, or
after reaction in the cell environment, form materials which
will settle out in the cells or be filtered out by the
microporous diaphragms. The cell feed should also be
relatively free of substances which will deleteriously
interact with cell components. An example is chloride.
which at high concentrations reduces the useful life ot lead
alloy anodes. There is an economic trade—off between
keeping such substances out of the cell feed (by selection
of suitable materials of construction or by selection of
chemicals for teed liquor makeup, for example) and rer ving
the substances from the feed. At the beginning of the pilot
program, a tentative cell feed specification was formulated
based on experience with other electrochemical systems.
This specification is summarized in Table 6.8. The table
also lists the present specifications based on experience
with the pilot plant.
82
-------
6.b.1 Contaminant Removal
In the initial pilot plant design, the acidic stripper drum
effluent was cooled to about 130 F and sent to the cell feed
liquor tank. There was some pH buffering in the stripper
drum but the bulk pH control was by caustic adc.ition
directly to the feed tank. Poor mixing in the feed tank and
the lack of continuous pH monitoring made control very
dirt icult. This resulted in large pH fluctuations and
excessive caustic consumption. The recycle caustic
requirement for pH control has an effect on the plant
economics since it increases the overall NaOH/S02 ratio arid,
hence, the number of cells. The cell teed liquor was then
filtered for bulk iron removal as the hydrous metal oxide
pz ssed through a manganese zeolite filter for ferrous iron
rer ioval and sent to the cell system feed tank.
TABLE 6.8
CMJL EE.ED SPECIFICATION
Tentative Present
Feed Specification Recon nenaation
(PPM) (PP 1)
Fe < 0.1 < 0.5
Mn <0.5 <0.5
I i < 0.5 < 0.5
Cu <0.5 <0.5
Co <0.5 (0.5
A]. <0.5 <0.5
Ca ( 5.0 < 15
Mg ( 0.5 < 15
Si < 12.0 < 12
Cl < 100 < 200 ( 1 )
(1) with lead alloy anodes
83
-------
An intermittently regenerated manganese zeolite tilter
consists of a bed of microporous natural or synthetic
alumino.-silicate ion exchange granules treated with
manganous chloride to convert it to manganous zeolite -
Intermittent treatment with or a continuous body feed of
potassium permanganate results in tne precipitation ot
higher oxides of manganese on the zeolite grains. Oxidation
reactions occur between the ferrous iron content of the cell
feed and the manganese oxides on the grains. Regeneration
is required because the oxidizing capacity of the bed is
consumed in the process of iron removal. The bed also
functions as a filter media for the ferric hydroxide formed.
Backwashing the precipitated iron from the bed is necessary.
Filtration rate is about three gallons per square foot per
minute and backwash rate a minimum of eight gallons per
square foot per minute.
The pH of the feed tank was initially between 10 and 12 to
maintain the caustic environment necessary for iron removal.
Operation at a lower pH level was not controllable with the
initial system. However, the electrolytic cells fouled
fairly rapidly as a result of occasional upsets when the pH
dropped below 7. The cell area polishing filters also
reçuired frequent changing during this period. The iriaterial
on these filters was analyzed as the hydrous metal oxide of
aluminum.
It was suspected that the aluminum was being leached from
the aridsorb resin. Further investigation, in conjunction
with the resin supplier, determined that at a pH of 11.6 or
greater, aluminum was leached from the silica—alumina base.
It was recognized that the existing system could not effect
control in the desired pH range, 8—9, and modifications were
undertaken. A pH buffer tank with automatic pH control was
installed and the manganese zeolite filter moved to a point
between the cell feed tank and the cell area polishing
filter. The manganese zeolite filter was still required for
ferrous iron removal, but was moved downstream of both pH
adjustments to prevent any upsets in its operation. An
auxiliary caustic tank was also installed to provide an
accurate measure (batchwise) of the recycle caustic
rec uirements.
Over the next two months, the pilot plant was plagued with
problems resulting from the poor performance of the
manganese zeolite filter. The cause was traced to a
combination ot resiaual oxidizable materials in the cell
feed liquor and high inlet temperatures which resulted from
periodic operation of the cell system in isolation from the
front end. The MnZe unit was designed for a feea of several
parts per million ferrous iron. However, due to the several
hundred parts per million S03 which remained in the cell
84
-------
feed after stripping, the resin at times would saturate
allowing break through of iron and release of manganese. In
addition, operation of the unit with the teIr perature above
110 F broke down the resin. The situation was clouded for a
time as a result of erroneous iron analyses being received
due to interference from sodium sulfate with the atomic
absorption method used. A wet chemical method based on
orthophenanthroline was found to give satisfactory results
without interference and was used thereafter foz iron
determinations (see Appendix G).
zeolite filter was
these conditions;
the manganese zeolite
hydrogen peroxide.
Following laboratory tests demonstrating that 11202
satisfactorily oxidized Fe+2 to Fe+3 in strong sodium
sulfate solutions containing residual sulfite, the method
was applied to the pilot plant. Hydrogen peroxide was added
to the process stream at the pH control point upstream of
the large cell feed liquor tank.
For the initial evaluation in the pilot plant, H202 at c i
concentration of 23.5 percent was added at a conservative
rate of 15.4 ccflnin. This rate of addition is sufficient to
oxidize stoichiometrically an S02 concentration of 790
mg/i in the stripper bottoms stream at a flow rate of 1.8
gpia. Table 6.9 shows the effectiveness of the peroxide.
Column 4 shows that good iron removal was accomplished
throughout the period, even with two high excursions of S02
in the stripper bottoms.
The amount of H202 that was added for the initial evaluation
was clearly more than was necessary and was reduced in
subsequent tests. For the period from 12/16 tO 12/19,
Table 6.9 sI ws an average S02 concentration in the stripper
bottoms of 294 mg/i, or 89 mg/i if the two high excursions
on 12/18 are omitted. Therefore, the H202 addition rate
averaged more than 2.5 times the stoichiometric amount based
on 294 ing/i of S02, or 9 times based on 89 mg/i of S02.
Additional data for the first week in January is given in
Table 6.10. All subsequent running was done with peroxide
addition and with the MnZe unit removed from service.
Satistactorily low iron levels were consistently obtained in
the cell feed at the stacks.
The conservative minimum amount of peroxide to be added for
iron oxidation is that amount which will react
stoichiometrically with the S02 and Fe in the stripper
bottoms. Over a nearly three—month period starting in
mid—December, S02 and Fe concentrations in the stripper
It was decided that the manganese
impractical for iron removal under
therefore, the oxidative function of
was replaced with a continuous feed of
8f
-------
Table 6.9
PROFILES OF 502 AND Fe IN FEED I I’flJOR BETWEEN STRIPPER DPUM AND
CELL STACKS DURING I NITIAL RUN’It WIT)’ “HEAVY” 11202 ADDITION
1 2 3 U
Stripper Bottana Feed Tank Contents Cell Feed a 11K—S Cell Feed S Stacks
( 3 “—103 inlet l 43 R—103 suct ion l — ( llslve 5-21) (Valve 5—20 )
DATE TI ?! S02 Fe 502 Fe 902 Fe £02 Fe
1973 ni/i irg/l mg/i ma/i mg/i mg/i mg/i mg/i
12—16 1300 — — nil ‘0.01 — — — —
12—IS 1700 nil 6.00 nil 0.30 12.8 <0.01 nil (0.31
12—16 2000 nil 6.50 nil 0.80 nil <0.01 nil (0.01
12—17 0000 40.0 6.00 nil 1.50 — — — —
12—17 0400 12.8 0.45 nil 1.10 — — — —
12—17 0RD0 262.4 5.0 2 1.00 nil <0.01 nil (0.01
12—17 1330 28 5.0 nil 1.20 nil <0.01 nil (0.01
12—17 1600 (44.8 4.1 6.4 1.0 12.8 0.02 nil (0.01
1 —I7 :ooo 153.6 .1 nil 1.1 iii 0.02 nil (0.01
12—18 0000 51.2 4.4 nil 0.90 25.6 <0.01 nil (0.01
12—18 3430 217t, 5.0 12.8 0.70 25.6 <0.01 12.8 (0.01
12—18 0800 25.6 4.8 25.6 0.71 12.8 <0.01 nil C0.01
12—18 1200 1069 4.5 25.6 0.80 25.6 <0.01 nil (0.01
12—18 iaOO 38.4 4.4 12.8 1.0 iii — nil (0.01
12—15 2000 51.2 4.5 19.2 0.8 25.6 0.0 1 1 51.2 (0.01
12—19 0000 51. 4.3 51.2 1.3 IR.4 (0.01 19.2 (0.01
12—19 0400 121.b 4.4 12.8 1.3 12.8 <0.01 12.8 <0.01
12—14 1210 19.2 4.3 nii 1.3 — — — —
12—19 2000 — — — — — — 19.2 (0.01
12—23 2200 — — — — — — nil ‘O.Oi
12—28 0800 — — — — — — 45 ‘0.01
12—24 1200 21.6 4.8 15.0 4.7 — — — —
12—24 2200 — — — — — — 8.3 (0.01
IOTES: (1) Stripper l’otto’na flow 3 1.8 GPM. (p112) throughoul the period.
12) 11202 added between Cols. 1 and 2 at 15.4 cc/mm, 23.5% 11202.
(3) 0-103 filter between Cola. 2 and 3.
(8) CF 1/3 (or 2/4) filters between Cola. 3 and 4,
(5) MnEe Cone’it toner 6-105 by—passed throuqhout the period
(6) r.lnztrun detectable limit of “S02” analysis is 1 .9 sio/l.
(7) Fe analysis by orthophenanthro jine method, by W PCO lab.
86
-------
TABLE 6.10
S02 and Fe PROFILES IN MPJOR STREfYAS OF WEPCO PILOr PLAI T
1 2 3 14 5 6
Acid Sulfate
Cell Feed Caustic Recycle Recycle Drum Absorber
Stripper Bottoms Feed Tank Contents at Stacks Drum CM-lOS) (M-10Ll) Net Draw
( T—103 Inlet) ( P—103 Suction) (Valve S-20) ( P—lOS Suction) ( P—lOll Suction) ( *21—1037. Outlet)
Date Time S02 Fe 502 Fe S02 Fe Fe ____ Fe
19714 mg/i mg/i mg/i mg/i mg/i mg/i mg/i mg/i mg/i
1—2 1600 1.0 (0.01
1—2 2200 nil <0.01
1—3 0800 nil <0.01
1—3 1200 nil 14. 18 114.9 1.8 0.70
1—3 1600 0.36
1—3 2200 41.3 <0.01
1—4 0800 nil <0.01
1—4 1200 51 4.6 nil 0.10 1.0
1—4 1600 0.60 (0.01
i—a 2200 nil nil <0.01
1—5 osoo nil <0.01
1—5 1200 nil 1.1 nil 0.60 0.29
1—5 1600 0.19 (0.01
1—5 2200 nil (0.01
1—6 0800 nil (0.01
1—6 1200 nil 1.0 nil 1.7 0.31
1—6 1600 0.10 <0.01
1—6 2200 nil <0.01
1—7 0800 — <0.01
1—7 1200 nil 2.5 1.5 0.60
1—7 1600 0.15 <0.01
1—7 2200 38.4 (0.01
1—8 0800 nil (0.01
NOTES: (1) Stripper bottoms flow 1.3SGPN
(2) 41202 added between cols. 1 6 2 7.1 co/mm, 22.5%
(3) Filters G—103 and CF 1/3 (or2/4) between Cols. . afld 3.
( (4) MnZe Conditioner G-105 by—passed throughout the period.
(5) Minimum detectable limit of S02 analysis is 25 mg/i.
(6) Fe analysis by orthophenanthrOlifle method, by WEPCO lab.
-------
bottoms averaged 54 and 3.2 mg/i, respectively. The S02
average is based on 43 analyses; it excludes four other
analyses, three of which clearly indicate upset conditions
and the fourth is during a response test. The Fe average is
based on 35 analyses during the subject period.
6.6.2 Major Stream Analysis
Table F .l4 in Appendix F presents a summary of the cations
founa in the major plant streams on several different
occasions with the accompanying chemical analysis.
detailed discussion of one such set of analyses is presented
in ippendix A. Several general comments can be made,
however, concerning these analyses.
1. A comparison o sample 349 with sample 350 will
show Mn leaching from the manganese zeolite filter.
2. The drop in the iron levels during the test program
is indicative of the removal of corroding
equipnent, mainly in the stripper. The chromium
analyses also are indicative of this trend.
3. Ca and Mg levels in the net draw are indicative or
quench water carry—over into the chemical section.
Samples 45, 64, and 246 are prior to the
installation ot the demister.
4. Na in the •8 anolyte was a result of leaks in the
anion membrane and represents a process fluid loss.
In cells tree of cross leaks, sample 242, the
sodium losses are negligible.
5. Chloride levels over 200 mg/i are deleterious to
the anodes. The drop in the chloride levels
ref lect the use of mercury cell caustic as opposed
to diaphragm cell caustic for feed makeup. Anode
breakdown may be the cause of the Pb levels in the
anolyte samples. During pilot plant operation, the
chloride ion concentration did not buila up to
greater than 200 mg/i after mercury cell caustic
was used for solution makeup. Since no chloride
buildup was observed in the pilot plant, chloride
introducea from the fuel gas was inconsequential.
6. Silica levels are indicative of fly ash dissolution
in the process fluids.
7. Sodium can be used as a tracer for the relative
mixing volumes of the plant streams. For example,
•A anolyte (45,000 mg/i) mixes with the absorber
net draw (65,000 mg/i) to form cell feed liquor
88
-------
(51,000 mg/i). Thus the relative volumes are 1 net
draw plus 2.3 anolyte to form 3.3 feed liquor.
6.6.3 Sources of Metal Contaminants
The two primai-y sources of metals contamination are
equipment corrosion and fly ash dissolution.
There was a significant amount of corrosion in the pilot
plant especially during the earlier stages of operation.
The most severely attacked piece of equipment was the
stripper. The overhead spool piece on the outlet of the
overhead condenser and the stripper bottoms piping were
completely corroded and made a significant contribution to
the iron cxntaminants • The carbon steel spool piece
(installed by mistake) was later plastic coated and the
stripper bottoms piping problem (a field fabrication error)
was corrected.. The iron levels dropped significantly after
the repairs were completed. This can be seen in Table F—L4
in Appendix F.
The detailed contribution of dissolved fly ash to the metals
contaminants is not easily determined and can only be
inferred. The efficiency of the hydroclone was not
measured. While qualitatively there was more fly ash in the
hydroclone bottoms, the ratio of the flow of quench and
makeup water to the solids involved was so large that
efficiency measurements were meaningless. This fact
combined with insufficient data input on inlet fly ash
quantity makes any calculation of fly ash entry to the
chemical section of the absorber speculative. Furthermore,
there is little data on the solubility of metals contained
in the fly ash in the plant solutions.
At a precipitator efficiency of 98 percent, WEPCO estimates
the dust loading to the absorber to be 0 .04 grains per cubic
foot of gas. Table 6.12 is an elemental breakdown of seven
typical fly ash analyses.
At iron and aluminum levels based on sample 7 and the dust
loading as above, the expected iron and aluminum hydroxide
to be filtered from the cell feed liquor will be 1.31 x 10_6
and 1.91 x 10—’ lb/ft 3 gas. This assumes 100 percent of the
fly ash enters the ch nical section and the iron and
aluminum are 100 percent dissolvea. A typical boiler gas
flow of 250,000 C 4 should contribute no more than 1.96 and
2.86 lb/hr of iron and aluminum as hydroxides from we fly
ash.
A sample of typical fly ash was sent to Commercial Testing
and Engineering Company for analysis. Their results are
shown in Table 6.11. The most significant number is the
weight percent water soluble fraction. If the quench
89
-------
section renxwes 90 percent of the *sh and only 2.1 percent
of that left is soluble (the iron and aluminum are small
fractions of the total solubility). the hydroxide loadings
could be reduced by a factor of 500. Translating this into
unit concentrations (mg/i) at the base case of 2000 CFM gas
flow, an absorber net draw flow of 0.8 gpm, and the. factor
of 500 as deriveci above, the expected elemental iron and
aluminum concentrations in the net draw would be 0 • 4 and
0.11 ing/l, respectively. This agrees quite well with what
was actually measured in the net draw (see Table F —LI in
Appendix F).
ThBLE 6.11
TYPICAL FLY ASH ANALYSIS FROM
WEPCO VALLEY PLAL T*
Water Soluble Insoluble Residue
(2.12% of total, Dry Basis (97.887 . of total, Ignited
dry basis) ( Wt. %) dry basis) ( Wt. 7. )
Suif ate, 504 1.39 Phos. pentoxide, 0.15
P205
Calcium, Ca 0.66 Silica, SiO 45.00
Sodium, Na 0.04 Ferric oxide, 24.83
Fe 20 3
Silicon, Si 0.01 Alumina, Al203 19.91
Aluminum, A]. 0.01 Titania, Ti02 1.08
2.11 Lime, CaO 4.95
Magnesia, MgO 0.99
Sulfur trioxide, 0.21
S03
Potassium oxide K20 2.48
Sodium oxide, 0.36
Na20
Undetermined 0.04
100.00
*Carbonate carbon is 0.18% of total, dry basis;
noncarbonate carbon is 2.OS . of total, dry basis.
90
-------
TABLE 6.12
ELEMENTAL BREAKDOWN OF SEVEN TYPICAL FLY ASHES (WEPCO DATA)
(Units are parts per million, ppm)
ELEMENT 1 2 3 5 6 7
S ‘4600 7600 15000 5500 12200 12700 3800
Br 25 1 7.6 15 15 13 5.3
Cl 370 150 365 272 280 660 250
F 170 120 175 327 134 100 20
Al 86000 94000 72000 102000 73600 95500 116000
Sb 15 19 25 21 13.6 10.4 7.5
As 213 220 84 168 100 92 75
Ba 641 735 1900 637 657 644 877
Be 11.7 3.3 5.3 9.0 16.0 7.7 7.3
B 70 70 85 205 170 125 85
Cd 1.6 .80 1.0 2.0 1.6 1.25 .83
Ca 9900 12700 32800 16600 26000 43300 14340
Cr 219 178 191 1714 165 179 178
Co 108 71 87 94 87 101 69
Cu 115 157 110 100 110 170 110
Fe 126000 147000 146000 166000 149000 103000 120000
Pb 275 301 269 586 158 222 237
Mg 9300 7600 10300 7100 8050 11460 9500
Mn 2514 302 532 280 366 460 325
Hg (.06 (.05 .18 <.14 .16 <.2 <.2
Mo 90 ‘44 54 39 18 25 5
Ni 580 622 527 585 607 512 1485
K 17200 17500 22800 17800 18500 13400 15500
Se .24 5.0 .55 <.8 3.0 2.2 <.9
Ag <.4 (.25 1.2 1.3 <.3 1.8 <.8
Na 6380 7410 13600 5120 6900 18300 3600
Ti 23 35 29 50 29 21 50
Sn (9 51 26 56 92 (18 107
Ti 5700 i4450 4000 5140 5900 6630 6300
Zn 200 163 215 189 218 222 200
V 296 310 427 304 256 276 320
I <1.2 <3 <6 N.D. 26? 6 1.
91
-------
6.6.4 Cell Peed Liquor Conditioning in Future Installations
The cell feed liquor conditioning system in future plants
will be designed to carry out the following processes:
1. Oxidation ot rerrous iron at a level of 1 to 5
milligrams per liter of cell feed with hydrogen peroxide in
the presence of about 50 milligrams per liter of sulfite.
2. At least one hour of ferric hydroxide conditioning
after peroxide precipitation. This permits the ferric
hydroxide gel particles to grow and/or coalesce. The rate
of coalescence depends on the temperature, pH, concentration
of iron, concentration of other insoluble substances, and
presence of conditioning assistants. The rate of growth is
maximum at high temperatures and, in the absence of other
insoluble substances, at pH’s near 5.5. II the
concentration of iron is low (i.e., 1 milligram per liter or
less) it may be desirable to add one or more
milligrams/liter aluminum sulfate, sodium aluxninate, or
ferric sulfate or conditioning assistants to increase the
rate of conditioning and improve subsequent filtration.
3. Precipitation of other heavy metal oxiaes and
hydroxides (i.e., those of chromium, manganese, cobalt,
nickel, and copper) and the oxides and hydroxides of silicon
and•alujninum by adjusting the pH of the cell feed to about
8.5. Such pH adjustment will be virtually simultaneous with
the peroxide addition discussed above. Conditioning
requirements are essentially the same as those for ferric
hydroxide. The optimum pH for precipitation and
conditioning depends upon the relative concentrations of the
various insoluble metal oxides and hydroxides.
4. Precipitation of calcium as the carbonate. The
carbonate required (about twice the concentration of
calcium) is expected to be supplied trom carbon aioxide
(unavoidably) absorbed from the air into the cell feed
liquor drum (cell feed surge tank) and the caustic recycle
drum (effluent catholyte surge tank).
5. Renxwal of solids by precoat filtration using a
diatomaceous earth precoat and bony feed. Precoat
requirements are estimated to be about 0.2 pounds (dry
weight) per square foot of filter area. Body feed
requir nents are estimated to be about 0.3 pounds (dry
weight) per thousand gallons of cell feed liquor. The 75 MW
prototype plant would use 62 tons of diatomaceous earth
annually. This would generate 1214 tons of 50 percent wet
cake annually or about 2.5 tons per week. The ash disposal
rate fran the boiler to be serviced by the prototype plant
averages about 180 tons per weeic. The filter cake uld
increase the amount ot solids by about 1.4 percent.
92
-------
VII
DISCUSSION OF OPERATING RESULTS
7 • 1 PROCESS CONSISTENCY
In general, the majority of operational problems experienced
during the pilot plant operation were mechanical. Examples
of these problems were forced draft fan vibration and
recycle acid pump failure. Table 7.1 presents a summary of
the inteqrated pilot plant operation and availability. As
the various mechanical problems were corrected, the
operating frequency increased dramatically. High
availability and low operating frequency during the later
months was attributable to power plant boiler failure
resulting in loss of flue gas to the unit. Appendix B
presents a chronological history of all pilot plant
operations from initial conunissioning through later stages
of operation. The test program encompassed the period from
July 1973 to March 19711.
TABLE 7.1
PILOT PLANT OPERATION AND AVAILABILITY
Total Hrs. Hrs.
Hrs. Punning Avail .
July, 1973 7414 30 4.2 30 4.2
Aug., 1973 744 181 214.3 181 24.3
Sep., 1973 720 0 0 0 0
Oct., 1973 7 144 2149 33.5 469 63.0
Nov., 1973 720 248 34.4 2118 34.4
Dec., 1973 744 142 19.1 212 28.5
Jan., 1974 7414 1419 56.3 1463 62.3
Feb., 19714 672 484 72.1 588 87.9
March, 1974 7414 499 67.0 549 711.0
April, 1974 720 378 52.5 64.3 89.5
93
-------
Since the electrochemical cells are the only basically new
canponent in the overall S02 removal process, it is
worthwhile to break out the availcthility of the cells during
the test program.
During the first three months of operation of the integrated
pilot plant (“Initial Cell System and Process Debugging
Period”, see Sections 6.4.3 and 6.I$.4) the cells were
subjected to excessively abusive conditions stemming from
the MnZe conditioner, in situ acid washes and excessive
chloride levels. Durinq this time, several shutdowns were
made due .to the cells or to cell—related causes for in situ
acid washes, replacement of the bed in the MnZe unit, etc.
At the end of the period, the cells were rebuilt with new
membranes and diaphragms. Anodes and cathodes were reused.
After the cell rebuilding and for the remaining nine months
of the test program, the availability of the elctrochemica].
cells was over 90%, as the following figures show:
Period: 2 October 73 to 29 June 714
Period Time 6,504 hours
Cell System Operating Time 2,978 hours
Down Time
For cell maintenance 408 hours
For non-cell maintenance 3,0 118 hours
(i.e., FD fan vibra-
tion and bearings, pump
seals, and shafts, re—
charqing system chem—
ica].s, chemical system
revisions, causes other
than pilot plant, etc.)
Unaccounted f or 70 hours
Total Down-Time 3,526 hours
Overall Cell System Service Factor = 2,978 (100) = 1s5.8%
6,5014
Cell System Availability Factor = 2,978 + 3,048 (100) = 92.7%
6 ,504
94
-------
7.2 PROCESS CONTROL
The parameters of process control, including system response
times and instrumentation design basis, are discussed below
for the four major sections of the pilot plant: absorption,
stripping, feed liquor preparation, and electrolytic cell
operation.
7.2.1 Absorption
The caustic flow rate is the most important parameter in the
absorption section. Insufficient caustic will result in
lowered 302 removal while excessive caustic will increase
oxidation and decrease caustic efficiency. During the pilot
plant operation, the caustic flow was controlled by
monitoring the S02 outlet concentration and the pH of the
net draw. The net draw pH was a good indication of the
relative sodium sulf ite—sodiuin bisu].fite concentrations
(Figure 8.1). The system was normally operated at pH of
5.0 ± 0.2 which was easily maintained by hourly monitorir g.
In later stages of the test program, the response of the
absorber to step changes in S02 inlet level and gas rate was
measured. There were five specific tests performed. They
are shown in Table 7.2 with their specific objectives. The
results of the test are presented graphically as Figure 7.1.
When control of the caustic flow was feed forward, based on
S02 inlet conditions and gas rate, the specified S02 outlet
concentration was maintained over a wide range of inlet gas
conditions (Tests A, B, and E). When control of the caustic
flow was by the S02 outlet concentration, long response
times due to the large absorber holdup were experienced
(Tests C and D). Since such slow responses would result in
periods during which S02 outlet requirements could not be
met, the more reliable feed forward control would be
preferred for future installations.
Inaccuracies inherent in the measurement of large gas flows
can be overcome with instrument tuning. Minor adjustments
to attain a specific net draw pH or for slight variances in
the outlet S02 concentrations can be handled by the
operator.
95
-------
TEST A
TESTB TESTC TESTD
TEST E
I I
1000 — I
I I
:1 I I
-a
U
I—. I ii t I
: 6000- I
I. I I I I
5000
I I
100 — I
I I
700— I I
I I ‘,I
600 -
a
500 - I I I.
-a I I
400- I I
-o
300-
200- I I I
I 1 I I
3500— I I I I
•11
3000— 1 I I I
I I
I-
p1
-J
I II I I
2500— I f —._. p
I I I I
. 350— I I I I
6
I I
U
300 -
a ’
o i i II II
a’
0
I- _______
L____ I I i
0
ii Ii
1A.M. 9A.M. lOAM. 1 1A.M. 12A.M. P.M. ZP M. 3P.M. 4P.M.
TU.IE
Figure 7.1 Absorption Tower
Response Testin 9
96
-------
TABLE 7.2
ABSORPTION TOWER RESPONSE TESTING
TEST OPERATION OBSIECPIVE
A Cut S02 inlet and caustic flow Can system balance
simultaneously with gas flow be maintained? S02
constant. outlet on specifica-
tion?
B Raise S02 inlet and caustic flow As above.
simultaneously with gas flow
constant.
C Cut caustic flow with S02 How rapidly will S02
inlet and gas flow constant. outlet exceed speci-
fication level?
D Cut S02 to correspond to lower How rapidly will
caustic flow. system recover?
E Raise gas flow and caustic flow Can system balance
simultaneously with constant S02 be maintained? 502
inlet, outlet on specifica-
tion?
7.2.2 Stripping
The rate of acid add itio is a very important parameter in
the process economics, as discussed in Section 6.3. The
initial pilot plant pH control system required the
withdrawal and pH measurement of a small slipstream at the
base of the stripper. This control system failed as a
result of inadequate winterizing. The acid addition was
then controlled by bouriy readings of the stripper drum
outlet pH. Due to the large holdup time in the stripper
drum, control of the acid addition was poor, but not
impossible. However, as the stripper drum volume to feed
rate ratio decreases, controllability will increase. A pH
level of 4.0 should be attainable without elaborate
neutralization according to the titration curve for sodium
sulfate (Figure 6.27).
Steam flow to the reboiler is best set manually. Since S02
removal is a strong function of the acid addtion, the steam
rate is better set at a rate slightly above the anticipated
minimum required with minor adjustments for ambient
temperature added.
97
-------
7.2.3 Feed Liquor Preparation
The pH neutralization system for the cell feed liquor was
discussed in Section 6.6. The system as modified worked
quite well.
Control of the feed liquor density is also important.
Gradual salt concentration due to water loss will increase
feed liquor density. However, if one monitors the density
of the whole tank in order to control the water addition,
the large volume of solution will mask density changes. The
density should be monitored on the pH buffer tank effluent.
Continuous control is not required but a density indicator
for use with the water flowmeter is essential.
7.2.14 Cell Room
The cell room process control as described in previous
sections functioned adequately.
7.3 MATERIALS OF CONSTRUCTION
As part of the test program, sample metal test coupons were
placed in the various chemical environs of the process.
After exposure, these coupons were removed and analyzed.
The results of these analyses are presented in Table F—S in
Appendix F. The cell room materials of construction as
determined in previous sections herein performed adequately.
7.4. EQUIPMENT OP RATI?G EXPERIENCE
7.’Ll S02 Analyzer
The Intertech Uras analyzer performed quite well for the
entire test program. Table 7.3 compares measured S02 levels
as determined by the Intertech analyzer, wet chemistry
method (EPA), and the WEPCO Theta analyzer. The high end
values determinea by EPA and the Intertech agree quite
closely, but the Intertech appears to be reading high at low
S02 levels. This may be due to nonlinearity in the I/P
converter (the emf signal from the Intertech is coverted to
a pressure signal for a chart recorder) as opposed to a bias
in the Intertech. Periodic cleaning of the probes was
required. Fly ash deposition and ice buildup (due to low
ambient temperatures) restricted flow through the inlet and
98
-------
outlet probes, respectively. The probes can be cleaned on
line and restricted gas flow is easily detectable. The
plugging was apparently caused by the intermittent operation
of the unit.
TABLE 7.3
CONPARM’IVE 802 MEASURE MENTS
(All Values In Parts Per Million)
EP1i INTERTEcH W PCO
288 440
352 472
286 420
322 300
3130 3160
3102 3080
3160 3000
1275 11142
39 18
*1894 1870 1779
* Calibration Gas
7.4.2 Flowmeters
The stripping column ov erhead flowmeter was inoperative for
most of the test program due to obstructed impulse lines.
It is believed that S02 hydrate formation near the orifice
taps was responsible. Impulse lines should be heat—traced
and as short as possible.
The recycle acid flown ter required frequent recalibration.
This was necessary due to the attack of the acid sulfate
solution on the differential pressure cell internals. The
316 SS internals were r t satisfactory. The dp cell was
filled with barometer fluid to eliminate contact with acid
sulfate solution, but fluid losses from valve and fitting
leaks made frequent recalibration necessary.
7•L4.3 pH Meters
The stripper bottoms pH meter was inoperable for most of the
winter months due to inadequate insulation and cooling water
control. While operable, the Unioc 320 held calibration
and performed well, but the frequent sample line plugging
made continued operation inconvenient.
99
-------
7 .14.1k Temperature and Pressure Indicators
The plastic protective coating of the thermowells was
unsatisfactory. Nearly 90 percent of the thermowells failed
due to corrosion after the plastic coating had split.
Pressure indicators in stripper service also failed due to
corrosion. They should be of the diaphragm type with a non-
freezing liquid.
7.14.5 Level Control
The stripper bottoms level control failed due to corrosion.
The internal KEL-F coating was unsuitable for the 240 F acid
sulfate solution. The float chambers on all level controls
should be insulated, heat—traced, and have a provision for a
bottom drain to avoid crystallization.
7.14.6 Forced Draft Fan
Fan vibrations and subsequent bearing failure was a
continuous source of trouble. Most of these vibrations were
a result of acid sludge buildup on the blades causing
imbalance. Operation at fan inlet temperatures below 225 F
increases deposition on the blades. This occurred only when
the absorber overhead gas was recycled to control the 802
inlet concentration. This should not be a factor in future
work. Vibrations were reduced by water washing the fan
blades. The fan blades were washed only when high vibration
occurred causing flue gas flow interruption for only a few
minutes.
7.14.7 Pumps
Sethco pump shaft failures were a major source of trouble.
A different design should be used that avoids the hole in
the •shaft for the sleeve retainer. A condensate flush
should be used for all packed pumps. Pumpage should not be
used as seal flush since it does not remove salts from the
packing material.
The Dorr-Oliver acid and caustic recycle pumps failed during
the test program. Failure was traced to the radial seal
which permitted attack of internal metal parts. Total
operating time until failure was 30 days for the acid pump
and about three months for the caustic pump. The
construction of the Dorr-Oliver pumps is quite complex and
repairs were difficult. Packing gland leakage from the acid
100
-------
recycle pump caused severe corrosion to the pump baseplate
and pedestals.
7.4.8 Absorption Tower
No problems were experienced with the chemical section of
the absorption tower. The plastic lining is intact as are
all the internals. However, there was a problem in the
quench section. A 316 stainless steel pad demister was
installed between the impingement baffle tray and the lowest
chimney tray. After several weeks of operation, the
demister was completely corroded around the edges. Because
the demister was installed during operation, the clearance
between the pad and the chimney tray was only about one
foot. This low clearance resulted in gas channeling which
formed a stagnant ring around the edges of the demister as
the flue gas passed through the restricted flow area • The
acidic quench water (pH=2.5) in this ring attacked the
demister and failure resulted. A polypropylene pad was
installed and has operated satisfactorily for over six
months.
7.4.9 Stripper Column
The packing support grid for the lower stage broke and
dumped the packing into the reboiler. It is believed that
pressure fluctuations, as a result of the rupture disc
relieving on several occasions, contributed to the failure.
The column performance was not diminished by the packing
loss and operation continued using only one stage.
The carbon steel overhead reducer above the overhead
condenser (T—1O1) corroded and failed after short service.
A CPVC replacement performed quite well • However, CPVC has
an upper temperature limit of 185 F and care was required to
avoid overheating and sure adequate condenser performance.
The condenser tubes plugged periodically due to S02 hydrate
formation. The blockage was cleared by raising the
condenser outlet temperature.
The Kynar lined steel piping for the stripper bottans flow
failed on several occasions. These failures were
attributabe to improper field •flaring of the plastic liner
allowing the process fluid to attack the carbon steel.
7.14.10 Cell Room
Cell room equipment operation was adequate at the operating
levels reported in previous sections.
101
-------
VIII
DISCUSSION OF ANALYTICAL PROBLEMS
8.1 BASIC ANALYSIS PROBLEMS
There were several analytical problems encountered during
the test program. The following discussion details the
problems associated with the major analyses encountered
during operation and presents some general calibration data.
8.1.1 NaHSO3 / Na2SO3 / NaHCO3 / Na2CO3 / NaOH Analyses
The major analysis for the pilot plant was the determination
of the absorption system net draw liquor composition. There
are five possible species that can exist.
The quantitative determination of sodium bisulfite (NaHSO3)
by titration with caustic was found to be quite acceptable
for both pure and mixed bisulfite solutions. Titrations
were carried out using a phenolphthalein end point, as well
as a potentianetric end point. The potentiometric end point
occurred at pH of 8.8, as determined by the titration of
pure bisulfite solution. The results of the calibration
runs in terms of the original solution norinalties (1.0 N)
and expressed as grams per liter are presented in Table 8.1.
They do not reflect the final mixed concentrations of the
species. Figure 8.1 presents the titration curve for sodium
sulfite and sodium bisulfite as determined on site. The end
points show excellent agreement with the published sulfurous
acid equilibrium constants.
The quantitative determination of the sodium sul fite
(Na2SO3) proved to be difficult. Direct titration with
hydrochloric acid was found to yield results that were
initially low. These results are also shown in Table 8.1.
The acid titration s were carried out using both the methyl
orange and potentianetric end point (pH L4 .5). Furthermore,
when a mixed su].fite-bisulfite solution was analyzed, the
sulfite determination was even more inaccurate followinq the
bisulfite titration with caustic soda.
The majority of the sulfite inaccuracy was found to be
attributable to oxidation as a result of excessive exposure
to the atmosphere. The evidence for this was:
a. No change in analysis occurred when boiled water
was used for standard solution makeup (deoxygenated
water).
b . 2.L$ gm/i Na2SOL4 was found in a titrated Na2SO3
solution (qravimetric analysis).
103
-------
9
MOLE PERCENT
FIGURE 8 I DISTRIBUTION OF AQUEOUS SULFITE SPECIES AS A FUNCTION OF pH
x
B
7
6
5
4
c
1
0
HSO
0 20 30 40 50 60 70 80 90
I I I I I I I — I I
90 80 70 60 50 40 30 20 10 0
-------
SANPLE
ml NaJ SO3 ml a2SO3
95.8
106.3
106.0
106.0
TABLE 8.1
SULFITE - BISULFITE DETERJ’1I? ATION
RESULTS (gm/i) ‘THCD
? aHSO3 Ma2SO3
NOTFS
(Titration Metho’)
1
102.2
Ma OH
hand
1
106.9
raOH
M. chinp
1
105.7
MaOH
flac’une
1
105.7
)‘a0h
Mdchlne
1
1
1
2
2
115.4
115.4
119.0
119.0
111.8
MC i
MC i
IIC1
MC i
MCi
h’and
Machine
Vachine
! ‘ac in,
Mecl, ine
2
105.5
MaO t !
Math’ine
2
104.2
t aOM
!hachire
2
1
104.5
100.0
NaOhl/MC1
Iac h r.t
2
1
104.5
106.5
NaOH,’VC].
!1 chin,
2
2
123.0
120.0
MCi
PCi
J1achine
uzinq new
point ion
2
105.5
?4aOH
Machine
2
104.2
2
106.2
1
1
1
1
1
1
1
2
2
4
4
88.5
NaOP,’PC i
Machine
12 1.5
PCi
Machin
118.0
1 Cl
Mac’iine
112.0
ITC1/NaOH
Machine
97.5
NaOH/HC1
ai contrary
to
xnethod
126.0
N OM/D zct
Macfline for
MaOM
125.0
12
Hand tor £2
105
-------
TABLE 6.1 - (CONT’1 )
SVLFIT - BISULFI?R DETERIIINATION
SANFLE RESULTS (qin/l) r . ET R OD NOTES
ml NaHSO3 ml Na2SO3 ?7a11S03 Na2SO3 (Titration Nethcd
4
11
1
I
105.0
106.0
116.0
116.0
aOH Exceas
12 then
Machine for NaO’
Hand for Na2S2O3
a2S2O3
U
I
104.1
61.5
NaOH/HC 1
‘achine
4
1
106.0
45.4
NaOH/HC 1
of above Solution
0.5
111.9
excess 12
Hand (new standard)
2
1
104.0
65.8
Na0 1 1 /HC 1
Nachini’
2
1
104.0
65.8
Na OH/I!Cl
(with anothcr now
standarc
1
1
105.0
104.0
VaOH/ 12
Mach ne for UaOH
1
1
105.0
8.O
Direct
Hand for 12 -
1
1
125
130
12 Direct
Excees 12
Rand
1
1
104.0
126.0
NaOR/ 12
M cttine for NaOH
1
1
106.0
1211.0
Direct
Hand for f2
1
1
104.0
130.0
NaOH/Excess
Machine for NaOH
1
1
106.0
126.0
12 then
Na 2 5 203
I’and for Na2S2O3
1 126.1 RCI Hand
1 125.0 NC1 (Ir nedi8eely after
1 126.1 MCi Sample Prep)
1 1 104.0 126 HaOH/HC1 Hand for both
1 1 104.0 126 NoON/HC 1 (lnvnedi tely)
1 1 105.0 125 HaOH/HC1
106
-------
IAPLE 8.1 - (C ONT’D)
SUL} ITE - BTSU FITE T)ETEF.MINA ’ION
SAMPLE RESULTS (qm/l) METfiOD NOTES
ml NaHSO3 ml Ka2 O3 Mai!S03 ?.a2SO3 (Titration Method)
2
2
0_s 102.0 126 NaOU/HC 1
0.5 103.0 126 NaOtI/HC1
Hand for both
(Invnediately)
The following series of samples were timed to note
deqradation. I.ottles remained closed and titration
was irru ediate1y after removing sample. All titrations
were done by hand.
2
1 103 128 NaOfl/P.Cl
0 hrs. ti me
2
1 103 128 NaOH/1C 1
2
1 104 126 NaOH/Fxcess
0 hrs. time
2
1 103 132 12 then
Na2 5203
2
1 104 127 NaOH/HC1
3 hrs. time
2
1 104 126 NaO)./ fCl
.
.
2
1 104 129 MaOH/Exces s
3 hrs. time
2
1 10 13 128 12 then
Ma2 5203
2
1 104 124 aO1!/i Cl
20 hrs. time
2
1 104 125 NaOfl/IIC 1
2
1 104 126 NaO!!,’ xces
20 hrs, time
2
1 104 126 12 th ’
N. 2S 203
2
1 10 (3 126 NaOH/PC 1
0 hrs, time
2
1 10(3 126
2
1 101 126 NaOH/HC1
3 bra. time
2
1 101 124 N . 0I’fHCl
107
-------
TA L 8.1 - (CONT’D)
SL’LPITE — BXSULFXTE DETER}IXNATION
SA.’ LE RESULTS (an/i) METHOD
ml NaBSO3 I?l Na2SO3 VaWSO3 Na2SO3 (Titration Method)
2 1 100.5 125 NaONIPC 1 6 hrs. timP
2 1 100.0 126 NaOH/HC 1
The auto titrator was also used for the above series
of samples. While the results Qre not reported, they
were signifacar.tly low as in previous tests.
The following series of aamp].es were used to test the
effect of timed exposure to the air as well as the
use of riunnitol, n oxidation inhibitor.
(Samples titrated itmuediately)
Titrator with-
1 1 10 1 1 51.9 NaOH/HCl
Auto
out ?4annitol
1 1 100 82 NaOV/HC 1
Auto Titrdtor
MjinflitOl
1 1 1011.2 86 NaOH/}!Cl
?itr.stor with
1 1 101 .2 67.5 1.aOI1/1 Cl
Auto
MannitOl
,
(Samples exposed to air for two hours)
(I4annitOl added at start of exposure time)
1 1 1OS..1 56.9 NaOP/ 1IC1
Hand titration with-
out ? nnitOl
1 1 103.5 121 !‘ aOH/HCl
)‘and titration
rannitol
(Samples stirred and exposed to air for two hours. Mannitol
added
at start of
exposure)
Titr.itor with-
1 1 10 1 1.0 0 NaOH/ Cl
out V.innitol
with
1 1 1011.5 96 NaO1 ,’I Cl
Auto Titrator
Mannitol
-
(Samoles stirred and exposed to air for two hours. ‘annitol addd at start of
exposure.)
1 1 1011 2.5 NaO1 /HCl Pand titration with-
out ‘annitoE
1 1 105 125 NaOH/RC 1 Rand titration with
Mannitol
108
-------
TABLE 8 • 1 - (CONT’ )
SULPITE - BISULFITE DETER1 I ATION
SA!WLE PF.SULTS (qm/l) METHOD
ml Nai SO3 ml Na2503 NaHSO3 N32S03
1 1 1014 5 i aOH/I2
0 irt’ct
1 1 101 4 1211 l aOH/I2
Direct
(Test of formaldehyde coTplexinq method)
(Samples titrated lethately)
1 1 105
1 1 104
1 1 104
1 1 104
saiiole of net draw liquor was also titrated as a
Sample bottle was closed.
on/L
;a2SO3 Na2CO3 arCO3 MtT1 IOD
13.4 2.1 12.7 NaOH/HC 1/12
13.11 2.1 12.7 H iO1T/HCl/i2
13.3 2.1 12.7 1 aOH/ 1!Cl/I2
uOTr: (1) Both sulfite and bisulfite solutions used were 1.0 N slutions.
(2) All results ad)usted to reflect input normality.
NOTES
(Titration rethod)
)I nc1 fitr t1on with—
o :t P nr.nitol
Hand titration with
annitol
SB
NaOP/0 430”—
I?C1
. uto
titration
96
118
NaO)i/Cfl3011—
MC i
PCi
! aO1!/CI-3OI1-
Auto
Han
titration
titration
1214
NaO!!/CI’304 1—
PC ’
iTand
titration
fun. t.on of ti1fl( .
NO S
0 )‘ a ti-4e
3 ha. time
109
-------
c. Continued sample degradation upon exposure to
atmosphere.
d. Accuracy increased upon the addition of an
antioxidatant (mannitol).
The initial solutions were titrated using an automatic
titrator which, for a particular sample requiring both
caustic and acid titration, meant that the sample was
exposed to the atn sphere for at least ten to fifteen
minutes. These titrations yielded quite low results.
H ever, upon reverting to hand titrations and minimizing
exposure time, the test accuracy was increased.
The fact that sodium sulfite is the oxidized species, is
confirmed in Section 6.2. The accuracy of the acid
titrations directly reflects the overall analytical accuracy
in that carbonates and bicarbonates are determined by
difference. The iodctuetric determination of total dissolved
502, either by direct titration with iodine or by the back
titration of excess iodine with sodium thiosulfate (Na2S2O3)
also proved to be satisfactory when done rapidly. The
iodine titration follows the equation:
! a2S03 + 12 + H20 Na2SO1 + 2H 1 (1)
(with excess 12)
Na2SO3 + 12 + H20 - Na2SO4 + 2H 1 + 12 (2)
and the back titration
with Na2S2O3
12 + 2Na2S203 Na2S1406 + 2NaI (3)
The pH of the solution for equations (1) and (2) should be
less than 8.0 to prevent the formation of hypoi ous acid
13+011 T-TIO+21 (14)
which in turn, oxidizes S203 to S04 instead of
tetrathionate as given in equation (3).
The -acid/base titration was used for sulfite-bisulfite
determination. Iodine titration was used to determine the
total S02 in solution and the difference between the two
titrations used as a measure of carbonates, if there were
carbonates present. The test accuracy was approximately
one percent for the sulfite—bisulfite determination.
During the course of the calibration work on this system,
another method was tried for sulfite-bisulfite
determination. Formaldehyde addition complexes the sulfite
ion with the stoichiometric release of NaOH. Therefore, in
c sulfite—bisulfite titration, one would titrate with
caustic to the phenolphthalein end point, thereby convertina
110
-------
all bisulfite to sulfite, add formaldehyde, and titrate with
acid to the phenolphthalein end point again. This method
has the advantage of very sharp potentiometric end points,
but the use of formaldehyde in a confined field laboratory
without a hood precluded its use. This method did, however,
yield results comparable to the other methods, the
difference being in the sulfite concentration.
In order to limit sample oxidation during storage and
analysis, mannito]. was added to each sample. Mannitol is a
carbohydrate (C6H1406) anti-oxidant. Its use proved quite
successful in limiting sample oxidation as can be seen in
Table 8.1 by comparing the timed titration both with and
without mannitol addition. While one closed sample showed
only slight deqradation over a 214 hour period, rnannitol was,
nevertheless, added to each sample in order to prolona
sample shelf life.
During the later part of the test program, as the total S02
solution increased c ue to increased system loading, sodium
bicarbonate appeared more frequently and at significantly
higher concentrations. Several samples of solutions
containing high bicarbonate levels were analyzed for total
carbon at a commercial laboratory in order to determine if
the bicarbonate levels were real or a result of titration
errors. Table 8.2 presents the laboratory-determined
results. Obviously, the titrations were in error. In these
determinations, the iodine titration was used for total
sulfur as S02 and compared to that obtained from the acid
titration which includes the amount of bicarbonate present.
Thus, either the iodine titration was low or the acid
titration was high. It is believed that with more
controlled laboratory conditions, the overall test accuracy
can be increased.
8.1.2 Na2SOII
The analytical determination of sodium sulfate is extremely
important with respect to oxidation calculations. The
sulfate formation due to oxidation is usually a small.
percentage of the total sulfate in solution. For example,
i-f 100 liters/hr of caustic solution containing 200 gin,’l of
sodium sulfate were fed to the absorption tower to effect
the removal of 1.0 g-inole/hr of S02, absorber oxidation in
the amount of ten percent would add only 6.1$ grams/liter to
the sulfate already in the caustic. This is equivalent to a
3.2 percent increc&se in the total sulfate concentration.
:i ii
-------
TABLE 8.2
BICARBONATE IN ABSORBER LIQUID
HCO3 — Determined HCO3 — Determined by
On Site - Skinner & Sherman with a
Sample Titration Difference Total Carbon Analyzer
Number (Grains/Liter) (ppm)
1358 2.6 546
1368 3.11 714
1375 7.31 1426
1379 3.411 5116
1387 6.97 1010
1422 (4.12 524
1449 3.61 8114
1461 3.78 716
1533 10.25 626
1534 5.12 774
1537 10.92 906
15145 12.60 708
112
-------
The solution sulfate concentrations were determined
gravimetrically by precipitation with barium chloride. This
methodology has been used for sulfur determination for many
years. However, it is prone to errors, especially errors of
coprecipitation and occlusion. Furthermore, all sulfite and
bisulfites must be reu ved by acidification and stripping
prior to the barium chloride addition in order to avoid
precipitation of barium sulfite. Table 8.3 summarizes the
calibration samples and their analysis.
They exhibit an overall standard deviation of five percent,
which is well above that required for a definitive oxidation
calculation. This error was the major reason for the large
an unt of data scatter experienced. An alternate method for
sulfate determination should be considered such that the
overall accuracy is less than one percent of the total
sulfate present.
8 • 2 IRON P NMJYS IS
In the analytical scheme setup for the WEPCO S02 removal
pilot plant, iron determinations from the various plant
streams were assigned to a Milwaukee—based commercial
laboratory, Liinnetics, Inc.
Of the samples withdrawn for Fe analyses, cell feed liquor
samples were of primary importance. Iron levels exceedinq
0.1 ppm in the cell feed liquor were considered to be
detrimental to the cell operation. Analytical results
obtained from Limnetics, however, showed a systematic
tenfold increase over the permissible safe iron levels; yet,
no serious deterioration in the cell performance was
experienced. Reconciliation of the obviously conflicting
facts can only be explained (a) by a much higher than
anticipated tolerance of the cells for iron or (b) by
erroneous iron analyses. In checking the latter
possibility, comparison of Fe determinations were
simultaneously made on standard and test solutions in both
Limnetics and lonics laboratories. The iron determinations
were run in approximately 3 N aqueous sodium sulfate
solutions containing about 213 g Na2SO4 per liter.
Limnetics used the instrumental atomic absorption (AA)
analytical technique by direct aspiration. lonics applied
the phenanthroline colorimetric method for the iron
determination.
In Table 8.4, a summary of the results received from the two
sources is shown. Along with these data, results of a later
comparison study at the Institute of Gas Technology (IGT),
Chicago, were also entered.
113
-------
TABLE 8.3
SODIUM SULFATE ANALYSIS CRAVIMETRIC T)ETERMIrATION
STANDARD RESULTS
Na 2804 NaHSO3 Na2SO3 Na2SO I4 PERCENT
Sample (gm/i) (gm/i) (cm/i) (gm/i) ERROR
258 150 145/138 —3.31—8.0
259 200 196/188/189 —2.0/--6.0/--5.5
A 200 207 +3 .5
B 150 — 152 +1.3
C 200 80 — 205 +2.5
D 150 80 — 154 42.7
E 200 80 20 191 -11.5
F 150 80 20 162 +8.0
590 150 50 20 165 +10.0
620 142 154 48.5
I - ’ 621 213 221$ +5.2
622 28 1 1 — 282 —0.7
623 213 52 221 +3.8
6211 213 104 215 +0.9
625 213 130 229 +7.5
1079 100 — 96 —14.0
1080 100 105 103 +0.3
1129 150 100 — 157
1171 150 100 — 157 +4.7
1253 150 100 — 143 —4.7
12514 143 100 (+2N NaOH) 1Lz8 +3.9
1446 150 100 151/152 +1.0
1532 150 100 162 +8.0
Average = +1.3
Standard Deviation = +5.1
-------
TABLE 8.14
Source Stream
(Sainpl e
No.)
ppm
Atomic Absorption
(Direct Aspiration)
ppm
Phenanthroline
(Colorimetric)
(Extr ac-
tion)
(g/ml
Ferrozine
Colormetr ic)
5814
585
586
587
588
589
Upstream Mn Zeolite
Downstream Mn Zeolite
Upstream Mn Zeolite
Downstream Mn Zeolite
Stripper Bottoms
Cell Feed
0.8
0.8
0.9
0.9
42.0
1.3
11.4
7.5
0.11
0.02
0.14
0.02
1.8
1.5
0.0 14
0.01
0.10
0.01
35
1.14
3.3
0.148
0.514
0.58
0.614
0.74
146
1.9
3.9
1.0
IRON ANALYSES IN 3 N SODIUM SULFATE SOLUTIONS
Limnetics
lonics
ICT
Standard Samples
1 (3 ppm Fe)
3A(0..5 ppm Fe)
3.8
0.8
-------
Comparing the two sets of data in columns 1 and 2, the
generally much higher trend in the Limnetics results was
obvious. The prepared standard samples seemed to indicate,
after allowing for the amount of iron introduced as Na2SO4
impurity, that the results furnished by Liinnetics were in
serious error.
The analytical laboratory of the Institute of Gas Technoloay
(IGT) investigated the effect of the presence of a large
quantity of sodium sulfate on the iron determinations. The
same test and standard solutions, as used earlier by the
other two sources, were submitted for the analysis.
IGT conducted a systematic study using both AA and
colorimetric approaches. A technical report (Appendix G)
was prepared detailing the results of the investigation from
which the foflowinq primary conclusions were made:
1. Direct aspiration of a concentrated (3 N) sodium
sulfate solution into the AA unit for iron
determination is unworkable and was responsible for
the erratic Limnetics results.
2. Reducing the salt concentration by water dilution
renders the direct aspiration AA workable: however,
in case of low iron concentration, dilution can
depress the Fe level beyond its detectability.
3. Removing the iron from the aqueous electrolyte
medium into an organic phase by extraction and
using the extract for the determination is the most
promising and safest method for specific Fe
determination with AlL
8.3 OTHER ANMJY”SES
Some comments are in order on several other analytical
problems encountered during the test program.
The determination of the total sodium in the plant streams
should be done by atomic absorption. A sodium specific
electrode gave very erratic results due to the high
dielectric constant of the solutions; specific ion electrode
determinations are most applicable for dilute salt
solutions. Table 8.5 compares total sodium as analyzed by
atomic absorption versus total sodium in certain standard
samples.
Dilute S02 gas samples are quite sensitive to degradation
with time if taken in glass burettes for subsequent gas
chroinatographic analysis. Samples determined online by the
Intertech at several thousand ppm S02 degraded completely
within 214 hours in glass burettes. This was not true of the
concentrated S02 samples taken in the stripper overhead.
Absorber interstaqe gas samples should be analyzed
immediately, preferrably using the Intertech analyzer or a
similar online device.
116
-------
TABLE 8.5
TOTAL SODIUM ANALYSIS
ATOMIC
STANDARD ABSORPTION
Na RESULTS PERCENT
SAMPLE ( mg/i) ( mg/i) ERROR
A 9t4,200 93,600 -0.6
B 37,700 314,000 —9.8
C 18,800 19,000 +1.1
117
-------
Ix
REFERENCES
(1) Fuller, E.C. and R.H. Criat “The Rate of Oxidation of
Sulfite Ions by Oxygen’ Journal American Chem. Soc.
Vol. 63 , pp 16114—1650, June, 1941.
(2) Srivastava, R.D., F. McMillan; and 1.7. Harris, “The
Kinetics of Oxidation of Sodium Suiphite” The Canadian
Journal of Chemical Engineering, Vol. 46 , pp 181—1811,
June, 1948.
(3) ‘ragi, S., and H. move, “The Absorption of Oxygen Into
Sodium Suiphite Solution’ Chemical Engineering Science,
Vol. 17 , pp 411—421, 1962.
(4) Chertkov-, B.A., “Oxidation of Magnesium Sulfite and
Bisulfite During Extraction of S02 from Gases”
Zhurnal Prikladnoi Khimii, Vol. 33, No. 10 ,
pp 2165—2172, October, 1960.
(5) Phillips, D.H. and M.J. Johnson “Oxygen Transfer in
Agitated Vessels’ Industrial and Engineering Chemistry,
Vol. 51 , No. 1, pp 83—88, January, 1959.
(6) Potts, J.M., A.V. Slack, and J.D. Hatfield, “Removal
of Sulfur Dioxide from Stack Gases by Scrubbing with
Limestone Slurry: Small Scale Studies at TVA, New
Orleans, Louisiana”: Environmental Protection Agency,
Proc. Second Intern. Lime/Limestone Wet Scrubbina
Sym osium (APrD—1161J November 8—12, 1971.
(7) Pinaev, V.A. ‘S02 Pressure Over Magnesium Sulfite-
Bisulfite-Suif ate Solutions” 3. Appi. Chem. USSR 36
( 10) . 2049—53, 1963.
(8) Whitney R.P., S.T. Han and J.L. Davis “On the
Mechanism of Sulfur Dioxide Absorption in Aqueous
Media’ TAPPI, Vol. 36, No. 4 , pp 172—175, April, 1953.
(9) Johnstone, A.F., H.J. Read, and B.C. Blankmeyer,
“Recovery of Sulfur Dioxide From Waste Gases”
Industrial and Engineering Chemistry Vol. 30 No. 1,
pp 101—109, January, 1938.
(10) Hem, L.B., A.B. Phillips and R.D. Young, “Recovery of
Sulfur Dioxide from Coal Combustion Stack Gases”
Air Pollution , Chapter 15, F.S. Wallette, Editor,
Reinholt Pubi. ., 1955.
(11) Johnstone, H.F., and A.D. Singh, “Recovery of Sulfur
Dioxide from Waste Gases” Industrial and Engineering
Chemistry, Vol. 29 , No. 3, pp 286—297 March, 1937.
119
-------
x
APPENDICES
A. LIMNETICS, INC. REPORT ON FOREIGN MATERIAL ANALYSIS AS
SUBMrj-r v BY IONICS, INCORPORATED
B • WEPCO PILOT PLANT OPERATU LOG
C • OXIDATION CALCULATIONS
D • SULFUR RI2IOVAL CALCULATIONS
E • BRITISH TO METRIC CONVERSION TABLE
F • TABULAR SUMMARIES AND MEASUREMENTS
G. IGT TRACE IRON ANALYSIS
121
-------
APPENDIX A
LIMNETICS, INC., REPORT ON
FORE IGH MATERIAL ANALYSIS
AS SUBMITTED BY
IONICS INCORPORATED, WATERTOWN, MASS • 02172
123
-------
APPENDIX A
LINNETICS, INC., REPORT ON
FOREIGN MATERIALS ANALYSIS
Concentrations of various foreign materials in the
process streams were determined from time to time during the
pilot program. An example of such an analysis, made by
Limnetics, Inc., early in the pilot program, is given in
Table A—i. Such ana1yses are frequently very useful for
diagnostic purposes. The interpretation of the data in the
table is as follows:
1. Sodium : This element is conveniently used as a
tracer. The Cell Feed (Column E — 51,000 ppm) is sent as
feed to the Combined Catholytes (Column I)) and the center
compartments. Sodium ion is transferred through the cathode
cation exchange membranes into the catholyte so one expects
the Combined Catholytes (Column D — 90,000 ppm) to have a
substantially higher concentration of sodium. This is
indeed the case. One also expects the “A” Anolyte Effluent
(Column A - ‘45,000 ppm) and the “B” Mid—Anolyte Effluent
(Column C — ‘ 3,800 ppm) to have reduced concentrations of
so ium owing to transfer of sodium into the catholytes. The
apparent reduction in concentration is low, however, owing
to the expected electroendosmotic water transfer
accompanying sodium ions from the center compartment to the
catholytes. The “B” Nid—Anolyte (Column C) is separated
from the “B” Anolyte (Column B — 15.0 ppm) by an anion
exchange membrane which permits substantially only transfer
of sulfate ions from the “B” Mid—Anolyte to the “B” Anolyte
and of hydrogen ions from the “B” Anolyte to the “B”
Mid—Anolyte. The low concentration of sodium in the “B”
Anolyte (Column B - 15.0 ppm) is typical and completely
expected.
The Combined Catholytes (Column 1) — 90,000 ppm) are
fed to the Absorber (Column F). The Drawdown from the
latter (65,000 ppm) is much less than 90,000 ppm and
indicates a substantial liquid carry—over from the Quench
section of the Absorber. This conclusion was confirmed by
other results indicated below. Subsequent measurements on
the Absorber confirmed such a substantial carry—over from
the Quench section.
2. Potassium : This element is also conveniently used
as a tracer. The concentrations in the system are at least
an order of magnitude less than those which might cause any
problems. At such much higher concentrations, the first
effect would probably be the crystallization of potassium
pyrosulfite. The concentrations of potassium ion in the
125
-------
TABLE A-i
LIMNETICS, INC., ANALYSES OF VARIOUS WEPCO PROCESS STREAMS SAMPLED AUGUST 22, 1973
(PARTS PER MILLION)
A B C D E F G H
•A” Anolyte •B• Anolyte •B Mid-Anolyte Combined Cell Absorber Stripper Quench
Item Effluent Effluent Effluent Catholytes Feed Drawdown Bottoms Return
1. Sodium 45,000 15.0 43,800 90,000 51,000 65,000 50,000 30
2. Potassium 145 0.6 130 400 170 250 175 4.0
3. Calcium 5.2 0.05 5.6 1.3 5.6 5.3 6.3 52.0
14• Magnesium 2 (1 (1 <1 2 10 5 38
5. Aluminum <0.5 (0.5 (0.5 <0.5 <0.5 0.6 0.5 (0.5
6. Manganese 0.30 0.06 0.36 0.23 1.35 0.35 0.60 0.15
7. IronS 1.8 2.0 1.4 1.8 1.6 3.6 58.0 3.6
8. chromium 0.21 <0.02 0.10 0.15 0.12 0.63 0. 143 1.0
9. Nickel 1.9 0.4 1.6 2.2 1.8 2.1$ 2.1 0.3
10. Lead 5.6 3.8 1.5 1.7 1.4 2.0 3.5 (0.02
11. Zinc 0.63 0.28 0.56 0.51$ 0.58 0.68 0.62 <0.02
12. Nitrates (as N) 130 110 18 30 29 (0.01 13 0.8
13. Nitrites (as N) 0.004 0.004 0.004 <0.002 (0.002 <0.002 0.006 0.012
114. Chloride 152 38 698 161.5 741 29,000 950 152
I 1
0 ’
*These iron analyses by Li.innetics • Inc. • found to be incorrect, see Section 8 • 2.
-------
various process streams follow those of sodium and confirm
the probability of substantial carry—over of liquid from the
Quench section into the absorber.
3. Calcium : The highest value of calcium in the
various process streams is found in the Quench Return
(Column H — 52.0 ppm). The latter is the probable source of
calcium in the system through liquid carry—over into the
Absorber section. The difference between the value in the
Cell Feed (Column E - 5.6 ppm) and the Combined Catholyte
(Column D — 1 • 3 ppm) indicates that one should expect
precipitation of calcium compounds in the mid—catholyte and
center compartments. The precipitates found in these
compartments did contain a high concentration of calcium.
The low concentration of calcium in the “B” Anolyte
Effluent is as expected from the use of anion exchange
membranes separating the “B” Anolyte from the “B”
1’lid—Anolytes -
The difference in concentration between the
Combined Catholytes (Column D - 1.3 ppm, which is feed to
the Absorber) and the Absorber Drawdown (Column F - 5.3 ppm)
again confirms the carry—over of Quench Return into the
Absorber.
4. Magnesium : The analyses for this ion again
illustrates the carry—over of Quench Return (Column H -
38 ppm) into the Absorber Drawdown (Column F - 10 ppm). The
Combined Catholyte (Column D) fed to the Absorber contained
only (1.0 ppm. The difference between the Cell Feed (Column
B - 2 ppm) and the Combined Catholytes (Column D - <1 ppm)
again predicts magnesium precipitate in the center and
Zlid—catholyte compartments. Such magnesium bearing
precipitates were found in these compartments. The
difference in concentration between the Cell Feed (Column E
- 2 ppm) and the Stripper Bottoms (Column G - 5 ppm)
probably indicates the precipitation of magnesium hydroxide
at alkaline pH’s in the main cell feed storage tanks. The
difference in concentration of this ion between Stripper
Bottoms (Column G — 5 ppm) and Absorber Drawdown (Column F -
10 ppm) is caused by adding “A” Anolyte Effluent (Column A —
2. ppm) and “B” ?Iid—Anolyte Effluent (Column C - <1 ppm) to
the Absorber Drawdown to strip off sulfur dioxide in the
Stripper.
5. Aluminum : The aluminum levels are essentidlly
undifferentiated by the sensitivity of the analysis. It was
concluded, however, that the dissolution of aluminum
hydroxide from the manganese zeolite filter by high pH cell
feed experienced earlier in the pilot program had been
corrected. The aluminum levels are acceptable.
6 • Manganese : This ion catalyzes the oxidation of
sulfite by oxygen in the Absorber and should be kept at low
127
-------
levels. The highest concentration of manganese is found in
the Cell Feed (Column E — 1.35 ppm) indicating pick—up from
the manganese zeo].ite filter presumably from excessive
sulfite in the Cell Feed. The difference between the Cell
Feed (Column E - 1.35 ppm) and the Combined Cc tholytes
(Column D — 0.23 ppm), the “A” Anolyte Effluent (Column A -
0.30 ppm) and RBN 141d—Anolyte Effluent (Column C — 0.36 ppm)
indicates precipitation of manganese compounds in the
mid—catholyte and center compartment cells. Indeed,
moderate amounts of manganese were found in the precipitates
in these compartments, probably present as the carbonate.
7. Iron(see note): This ion also catalyzes the
oxidation of sulfite by oxygen in the Absorber and should be
kept at low levels. The highest concentration of iron is
found in the Stripper Bottoms (Column G - 58.0 ppm)
indicating corrosion of some component. (It was later found
that a steel spool piece in the top of the Stripper and an
inadvertently unlined length of steel pipe carrying bottoms
from the stripper had been corroding. Both were replaced
with plastic lined components.) The difference between the
Stripper Bottoms (Column G - 58.0 ppm) and the Cell Feed
(Column E — 1.6 ppm) indicates that the various filters
between the Stripper and the Cell are indeed removing iron
but not quite well enough. The undesirably high levels in
the Cell Feed are probably due to the sulfite loading
contributed by the Stripper. The comparatively high level
of iron in the Absorber Drawdown (Column F - 3.6 ppm)
indicates corrosion of some component in the Absorber (such
as the stainless steel rods on which corrosion test coupons
were mounted or perhaps the coupons themselves) or
carry—over from the Quench Return.
Note: Analyses for iron in concentrated Na2SO4 (3N) by
Limnetics, Inc. were later found to be in error.
Nevertheless, the relative amounts of i-ron in the
various streams are correct.
The high level in the “B Anolyte i riluent is
unexpected and either indicates corrosion of some component
in the “B” Anolyte circuit or contamination in the water fed
to the “B” Anolyte.*
8. Chromium : The highest level is found in the Quench
Return (Column H - 10. ppm), liquid carry-over of which into
the Absorber is probably responsible for much of the high
value in the Absorber Drawdown (Column F - 0.63 ppm), the
remainder coming from the Combined Catholyte (Column D -
0.15 ppm) and from corrosion of chromium containing coupons
and support rods in the Absorber. The concentration in the
*]t was later found that the latter was the principal case.
The water was subsequently treated to remove the iron by
ion—exchange demineralizers.
128
-------
Stripper Bottoms (Column G - 0.43 ppm) is adequately
explained by dilution of the Absorber Drawdown with “A”
Anolyte Effluent (Column A — 0.21 ppm) and “B” Mid—Anolyte
Effluent (Column C - 0.10 ppm). The difference in
concentration between the Stripper Bottoms (Column G -
0.43 ppm) and the Cell Feed (Column E — 0.12 ppm) indicates
a removal process, presumably precipitation and filtration
of the gray—green gelatinous, hydrous chromium sesquioxide.
The level in the “B” Anolyte Effluent is satisfactory and
probably indicates that the iron alloy (such as Hastelloy C)
corroding in the “B” Anolyte circuit contains only a low
amount of chromium. The level in the “A” Anolyte Effluent
(Column A — 0.21 ppm) is significantly higher than the Cell
Feed (Column E — 0.12 ppm) and probably indicates corrosion
of a stainless steel type alloy in the “A” Anolyte circuit.
9. Nickel : This ion also catalyzes the oxidation of
sulfite by oxygen in the Absorber and should be kept at low
levels • The highest level is found in the Absorber Drawdown
(Column F - 2.4 ppm) and indicates corrosion of some
stainless steel components, presumably the stainless steel
rods used to srpport the corrosion test coupons in the
Absorber. The concentration in the Stripper Bottoms (Column
G — 2.1 ppm) is adequately explained by dilution of the
Absorber Drawdown with “A” Anolyte Effluent (Column A -
1.9 ppm) and “B” Mid—Anolyte Effluent (Column C - 1.6 ppm).
The small difference between the Stripper Bottoms (Column G
— 2.1 ppm) and the Cell Feed (Column E — 1.8 ppm) indicates
that the Cell Feed conditioning system is not very effective
in removing this contaminant. The somewhat higher
concentration in the Combined Catholyte (Column D — 2 • 2 ppm)
compared to the Cell Feed indicates that much of the nickel
is soluble and is being transferred through the cathode
cation exchange membrane. The high value in the “B” Anolyte
Effluent (Column B — 0.4 ppm) indicates corrosion of an
iron—nickel alloy (such as Hastelloy C) containing a low
level of chromium in the “B” Anolyte circuit (see
paragraphs 7 and 8 above). The levels of nickel in the
various cell effluents compared with the Cell Feed indicate
that we should not find significant quantities of nickel in
any precipitates in the cells. This was found to be the
case.
10. Lead : The highest levels of lead are round in the
“A” Anolyte Effluent (Column h - 5.6 ppm) and “B” Anolyte
Effluent (Column B — 3.8 ppm). These levels are both higher
than expected probably due to the high chloride levels (line
14, columns A and B) in the two effluents (see paragraph 14
below). Both the lead and chloride levels in the “A”
Anolyte Effluent are higher than in the “B” Anolyte
i ffluent. The concentration of lead in thee Absorber
Drawdown (Column F - 2.0 ppm) cczupared with the Combined
Catholytes (Column D - 1.7 ppm) is surprisingly high
considering the dilution of Absorber Drawdown by Quench
Return indicated by other evidence. The most probable
129
-------
explanation of this vc lue is that the system is not at
steady—state with respect to chloride. Note that the
Stripper Bottoms (Column C, line 14) contain about 950 ppm
chloride compared to 152, 698. 162, and 741 in the “A”
Anolyte Effluent, “B” Mid—Anolyte Effluent, Combined
Catholytes, and Cell Feed, respectively. The lead level in
the Absorber Drawdown corresponds to that in a volume of
Combined Catholyte made several hours before. As will be
shown in Paragraph 14 below, the system rejects chloride and
hence, both the lead and chloride levels should decrease
with time.
The higher than expected levels of lead in the “A”
and “B” Anolyte Effluents indicate higher than expected, but
not catastrophic, attcuCk on the lead anodes. As will be
discussed in Paragraph 14, the chloride levels in the
process streams should be reduced.
The difference in levels between the Stripper
Bottoms (Column G - 3.5 ppm) and the Cell Feed (Column E —
1.4 ppm) indicates substantial removal of lead by the Cell
Feed conditioning subsystem.
11. Zinc : This ion is not expected to cause any
difficulties at the concentrations found in the various
process streams. However, the very low level found in the
Quench Return compared with the other process streams
indicates dissolution of some zinc containing component in
the latter process streams. The value found in the “B”
Anolyte Effluent (Column B - 0.28 ppm) is particularly
surprising. Possible sources are zinc compounds (zinc
oxide, zinc stearate) used in the molded rubber cell
components; as mold release agents for such components; as
heat and light stabilizers in CPVC or PVC components; as
mold release or extrusion lubricants for such components.
Less likely but also possible are Babbit metal, solder, or a
Drass or bronze component in the circuit.
12. Nitrates : The high levels of nitrate in the “A”
and “i ” Anolyte Effluents (Columns A and B - 130 and
110 ppm, respectively) indicate probably corrosion of the
nitrile rubber anode frames by oxidants in the anolytes. It
is suspected that the corrosion rate is aggravated by the
high level of chloride in the process stream. At the
anodes, such chloride should be oxidized to chlorine which
is known to attack nitrile rubber. The anode frames in the
conunercial scale plant will be filled polypropylene, known
to resist wet chlorine attack.
The low level of nitrate in the Absorber Drawdown
(Column F — <0.01 ppm) is surprising compared to the 30 ppm
in the Combined Catholytes (Column D), the feed to the
Absorber. The low level iiriay indicate catalytic reduction of
nitrate by bisulfite under acid conditions through the
transition element ions (Mn, Fe, and Ni) as intermediaries.
130
-------
One would also expect attack of ferrous alloys (such as the
stainless steel support rods for the corrosion test coupons
in the Absorber) by dilute nitrate under acid conditions.
The reported low level may also mean some interference by
the high levels of bisulfite or sulfite with the nitrate
analysis used.
The difference between nitrate in the Stripper
Bottoms (Column G — 13 ppm) and in the Cell Feed (Column £ -
29 ppm) is unexplainable. The difference between Cell Feed
and “B” Mid—Anolyte Effluent (Column C - 18 ppm) may be
explained as transfer of nitrate from the Mid—Ariolyte
compartment through the anode anion exchange membrane into
the B Anolyte.
13. Nitrite : The nitrite levels in all streams are
unremarkable and at very low levels.
ill. Chloride : Chloride levels in all process streams
(other than the Quench Return) are very high and generally
unsatisfactory. The 29,000 ppm reported for the Absorber
Drawdown appears to be in error and is otherwise
unexplainable. Comparing the level in the Stripper Bottoms
(Column G — 950 ppm) with that in the Quench Return (Column
H — 152 ppm) indicates that the high levels in the system
did not come from the Quench Return. It was later found
that the chloride was a contaminant in the initial charge of
diaphragm caustic soda into the system. Chloride is a
substantial contaminant of diaphragm grade caustic. Low
chloride, rayon grade caustic should be used as charge and
makeup. The differences between Cell Feed (Column E -
7LII ppm) and the Stripper Bottoms (Column G - 950 ppm) and
between the Cell Feed and the Combined Catholytes (Column D
— 162 ppm) cannot be satisfactorily explained. The Stripper
Bottoms levels (Column G - 950 ppm) would be expected to be
a weighted average of Combined Catholytes (Column D -
162 ppm), “A Anolyte Effluent (Column A - 152 ppm), and “B”
Mid—Anoltye Effluent (Column C — 698 ppm) but clearly is
not. The differences may perhaps be explained by
differences in the time at which the samples were collected
since the system is rejecting chloride, probably as gaseous
chlorine in the oxygen gas produced at the “A” and “B”
Anodes • For example, the concentration in the A Anolyte
Effluent is 152 ppm (Column A) compared to 741 ppm in the
Cell Feed (Column E) to that anode. One would expect
chloride to be transferred across the “B” cell anode anion
exchange membrane from the “B” Mid—Anolyte compartment
(Column C — 698 ppm) into the “B” Anolyte compartment
(Column B — 38 ppm) and part of that chloride converted to
gaseous chlorine.
131
-------
APPENDIX B
CBRONOLOGICAL HISTORY OF PILOT PLANT OPERATIONS
JUNE 1, 1973 — MAY 9, 1974
WEPCO PILOT P lANT
S W/IONICS PROCESS
MILWAUK , WISCONSIN
133
-------
APPENDIX B
CHRONOLOGICAL HISTORY OF PILOT PL7 NT OPERATIONS
JUNE 1, 1973 — MAY 9, 1974
June 1—June 19, 1973
Debugging and preliminary operation with flue gas and
process fluids.
June 19—June 29, 1973
Vendor repaired tank, 3-1014, which had failed.
June 29—July 6. 1973
Prepared new feed liquor charge and processed through the
cells.
July 6. 1973
WEPCO accepted plant and integrated plant operation was
started.
July 7, 1973
Shutdown after 214 hours due to low levels in caustic and
feed liquor tank and plugging of cell system. Reason for
low levels was excessive pump packing leakage, absorber
carryover, and cell venting. Plugging of cells was due
to metals released from water conditioner.
July 7-July 10. 1973
Diagnosed problems and flushed cell system.
July 10. 1973
Restarted cells.
July 11, 1973
Restarted integrated plant operation. Shutdown after six
hours because of plugging in cells.
July 11-August 3, 1973
Obtained analysis of various streams and contaminants in
order to diagnose problem. Identified problem and took
corrective action which included relocating the water
conditioner, and the installation of a new automatic
pH control system.
134
-------
August 3-August 5, 1973
Preliminary operation and debugging of new pH control system.
Cells restarted to process feed liquor.
August 5, 1973
Restarted integrated plant operations.
Shut down plant after five hours due to excessive vibration
of fan R—1O1.
August 5—August 7, 1973
WEPC0 balanced fan and a new batch of feed liquor was
prepared.
August 7. 1973
Restarted integrated plant operation.
Shut down plant after 13 hours operation due to low level in
caustic tank.
August 7-Auqust 9, 1973
Built up level in caustic tank with only the “A” cells able
to operate, since “B” cell rectifier had failed.
August 9, 1973
Restarted integrated plant operation.
Shut down after five hours operation due to excessive
vibration of fan R—1O1.
August 9-August 13, 1973
WEPC0 rebalanced the fan and lonics repaired the ‘B” cell
rectifier circuit.
August 13, 1973
Restarted integrated plant operation.
August 16, 1973
Shut down after 68—hour run due to packing failure of P-102A.
August 16—August 19, 1973
WEPCo repacked P-102A ten times.
135
-------
August 19, 1973
Restarted cells.
August 20, 1973
Restarted integrated plant operation.
August 21, 1973
Shut down after 26 hours due to failure ot weld in stripper
overhead spool piece. Cells continued to run.
WEPCo repaired spool piece.
Restarted integrated plant operation.
August 24, 1973
Shut down after 69 hours due to l level in caustic tank
which was caused by gradual fall-off of cell performance.
August 24-October 3, 1973
Cells were disassembled and some parts were returned to Watertown
for evaluation. Analyses were made of all major streams to
determine source of contaminants. Various modifications
requiring new materials were made to the plant to avoid
future problems. New cell components were manufactured and
installed. Calibration of instruments and entrainment tests
were conducted during the shutdown. Existing liquor inven-
tories were dumped, tanks washed and a new batch of feed
liquor was prepared.
October 3, 1973
Restarted integrated plant operation. Frontend: 1650,
Cells: 2100
October 6, 1973
Shutdown after three-day run due to failure of P—lOS motor
overload relay (1650).
October 6-October 9, 1973
WEPCo obtained a replacement relay for P—105, and cleaned
fan R—101.
October 9. 1973
Restarted integrated plant operation (1410).
136
-------
October 11, 1973
Shut down due to excessive vibration of fan R—101, caused by fly
ash deposition in fan (0830).
October 11-October 15. 1973
WEPC0 cleaned and balanced the fan. Also, it was necessary to
align the suction piping flange and repair the foundation grout
which had failed due to past vibration of the fan.
October 15. 1973
Restarted integrated plant operation.
October 16, 1973
Shut down after 24-hour run due to failure of the WEPCO #4 boiler
discharge main steam line.
October 16-October 26, 1973
During the shut down, the absorber quench section demister was
found to be severely corroded. A replacement polypropylene
demister was ordered and installed. An air purge was added
upstream of the stripper overhead pressure control to improve
pressure control. The hydrogen peroxide injection system was
readied. A broken pump shaft (P—102A) was replaced.
October 26, 1973
Restarted integrated plant operation.
November 6. 1973
Shut down after 11-day run • The cause was corroded spool piece on
stripper overhead and corroded FIC—15.
November 6-November 13, 1973
Replaced spool piece with CPVC piece, repiped stripper feed line
to reduce carryover to condenser which had been subject to
plugging, reversed cooling water flow to condenser, repaired
FIC—15 and filled with buffer fluid, replaced LC-34 with a spare
absorber IC, repaired solenoid in cell area.
November 13, 1973
Restarted integrated plant operation.
137
-------
November 17, 1973
Shut down due to high D I ’ in cells, and fan bearing failure.
November 17-November 24, 1973
Replaced bearing, backwashed cells; also repaired PlC-iS and LC—311
and recharged water conditioner with new “green sand”.
November 24. 1973
Restarted integrated plant operation.
Shut down due to broken shaft on absorber circulation pump.
November 24—December 5, 1973
Replaced shaft. Removed acid pump for repair in WEPC0 shop
because pump had corroded between baseplate and pump.
December 5. 1973
Restarted integrated plant operation (0800).
December 6. 1973
Shut down due to power failure caused by short in mixer motoi
(0230).
December 6-December 16. 1973
Noticed leaking mechanical seals in cell area pumps and leaking
shaft seal in acid pump. Ordered and replaced parts in pumps.
Christmas rush delayed air shipment of pump parts and when arriv-
ed one part ordered was missing. Missing part was hand carried to
site by Vendor. Also, impeller was stock size and had to be
turned down in WEPCo shop.
December 16. 1973
Restarted integrated plant operation (14.00/1000).
December 20, 1973
Shut down due to ground fault in cable to fan motor.
December 20-December 23. 1973
Repaired fault. Also noticed mysterious white gummy material in
lower circuit of absorber which had to be washed out of circuit.
Sent sample to downtown lab for analysis.
138
-------
December 23, 1973
Restarted integrated plant operation.
December 25, 1973—January 2, 1974
Shut down due to leaks from the lined pipe between the stripper
and T-103. Located replacement materials; fabricated and reinst-
alled pipe by December 28. Drips from pipe caused a small
electrical fire which destroyed a switch box which controls lab
power and electrical tracing circuits. WEPC0 completed electrical
repairs associated with the fire on December 31. During startup
on January 1, the WEPCo turbine serving No. 3 and No. 4 boilers
tripped, and the pilot plant was returned to shut-down status
pending stable operation of the WEPCo power house which was
achieved the evening of January 1, 19714.
January 2. 1974
Cold weather caused a slow startup, but integrated plant operation
resumed late on January 2.
January 8, 19714
Shutdown at 1600 hours due to cracking stripper bottoms drum out-
let nozzle. Crack caused by incorrect installation of pipe hanger
and clamp.
January 11, 1974
Plant restarted at 1100 hours. Cell startup delayed by
crystallization in the cell components.
January 12
Cell system started at 1100 hrs. Front end shutdown at 1200 hrs
due to fan vibration. All the anchor bolts on east side of frame
were snapped.
January 13
Cells shutdown at 0900 because product tanks M—104 and M—105 were
full and large cell teed tank M—101 was empty due to front end
shutdown for fan repair.
January 114
Plant restarted at 1100. No problems in front end or cell
system.
139
-------
January 15
A 22 hour loss analysis run was initiated with front end and cell
system operating independently.
January 2L$
Plant shutd n occurred at 1400 hrs. Low cell area throughput
and resultant low caustic inventory forced shutdown. Inspection
of M—106 indicated the bottom packed section of Pellerettes had
dropped into the drum.
January 28 - February 2, 19714
Repairs on pinnps and partial cell rebuilding completed.
Integrated plant operation started at 0930 on February 2.
February 6—7
Continuous operation interrupted for 17 hours due to shutdown of
boiler #4.
February 16
Plant operation shut down at 0930 due to low feed liquor
inventory. Installed line for continuous water addition to
M—1 01.
February 18, 1974
Integrated plant operation restarted at 1100 hrs.
February 21
WEPC0 boiler $14 shutdown caused shutdown of S02 plant at
2015 hrs.
February 25
Integrated plant operation restarted at 1300 hrs.
February 26 - 27
EPA testing of absorption system being completed.
March 9, 1974
Plant shut down at 2100 hrs due to leaks in stripper 01! piping.
L O
-------
March 12
Operation resumed at 10145 hrs after replacement of R—l01
bearings.
March 15
The last test was completed at 0800. Plant continues to operate.
March 25 — April 8
Plant shut down for planned rebuilding of cell system at 0830 on
March 25. Startup postponed from March 29 at WEPC0 request due
to boiler $ 14 shutdown and wet coal problems. Start of integrated
operation on April 8 at 0800.
April 17
Plant shut down at 1300 hrs to repair stripper OH piping leak.
Startup delayed by boiler *14 problems.
April 23
Integrated plant operation started at 1300 hrs.
April 26
April 30. 19714
Plant shutdown at 11400 hrs for planned repair and feed liquor
makeup.
Plant restarted at 1600 hrs. Delay caused by off—spec caustic
delivered and used in feed liquor makeup.
Shutdown
at 0200
caused
by
boiler
$4 shutdown.
Cell system
restarted
at 0500.
Front
end
startup
delayed by
split rubber
boot on
G—101.
Ready
for
startup
at 1000 hrs
but
delayed at
WEPCO request. Promt end
restarted at
1300 hrs.
U’
-------
APPENDIX C
OXIDATION CALCULATIONS
1L3
-------
APPENDIX C
OXIDATION CALCULATIONS
A. Sample Calculation of Oxidation as Determined by Sodium
Balance Method
Caustic Feed Rate to A—101 = 0.40 gpm
= 1.52 liters/mm
Caustic Concentration = 1.85 N
Reactable sodium to A— 101 = (1.52) (1.85) = 2.82 g—ion/Thin
Analysis of Draw Liquor:
NaHSO = 1.30 g-mole/liter
Na2 503 = 0.165 mole/liter
Reactable sodium in draw liquor in form of sulfite-bisulfite:
NaHSO = (1.30) (1.52) = 1.97 a—mole/mm
Na2SO3 = (0.165) (1.52) (2) = 0.50 q—mole/min
2.47
Sodium disappearance — 2.82—2.47 = 0.37 g-ion/inin
Sodium in form of sulfate = 0.37/2 = 0.185 g-ion/min
S02 absorbed in liquid phase:
(1.30) (1.52) + (0.165) (1.52) + 0.185 = 2.405 g—mole/fltin
Therefore, oxidation is
(0.185/2.405) (100) = 7.8%
B. Sample Calculation of Oxidation as Determined by RatiO Method
Data: Gas Flow = 9,000 lb/hr
S02 in 1,075 ppm
out = 50 ppm
S/C = 0.90
(Sodium/sultate) as determined by chemical analysis is
3.07
Caustic = 1.85 N sodium sulfate feed = 1.32 g—mole/
liter
S02 disappearance in gas phase:
(9,000 ].b/hr) (1,075—50) (14.9/lOb) = 137.5 g—mole/hr
145
-------
Assume Oxidation = 10%
I aHSO + Na2SO3 in draw = 0.90(137.5) = 123.5 g—mole/hr
Na2SO4 in draw = 137.5 — 123.5 = 14 g—mole/hr
Sodium with NaHSO3 - Na2SO3 = 123.5/0.90 = 137.5 g-mole/hr
Sodium with Na2SO4 14 (2) = 28.0 g-mole/hr
Total Sodium 165.5 g—mole/hr
Caustic flow to A—101 = 168.5/1.85 = 89.5 1/hr
Sodium Sulfate in = (89.5) (1.32) = 118 g—mole/hr
Sodium in Sodium Sulfate = 118(2) = 236 a-ion/hr
Sodium/Sulfate in Draw Liquor 236 + 165.5 = 3.014
118 + 1l4
Sodium/Sulfate as measur& = 3.07
Therefore, oxidation is slightly less than 10%.
C. Sample Calculation of Oxidation by Acid-Base Balance
Recycle Acid Feed Rate to Stripper = 0.90 qpm
Recycle Acid Concentration = 0.50 g—nole/l
pH Buffer Tank Feed Rate = 10.9 gallon/hr (average)
Let a = NaHSO3 = 1.30
b = Na2SO3 = 0.165
C = Na2SO4 = 3.07
S/C = a + b = 0.88
a + 2b
a/b = 7.88
Total acid addition =
(0.90) (3.78) (60) (0.50) = 102.1 q—mole/hr
ccess acid =
(10.9) (3.78) ( 1.85 ) = 38 q—mole/hr
2
S02 released =
( 7.88 ) (2) (64.1) + ( 614.1 ) = 121.0 g—mole/hr
(8.88) (8.88)
Total gas phase S02 disappearance =
137.5 g-niole/hr
Oxidation =
137.5 — 121.0 (100) = 12.0%
137 • 5
146
-------
APPENDIX D
SULFUR REMOVAL CALCULATIONS
147
-------
APPENDIX t
SULFUR REMOVAL CALCULATIONS
TOTAL GAS PHASE SULFUR DIS1 PPEARANCE
Gas flow = 8000 lb/hr
S02 in = 1600
out = 200
Therefore S02 removed = 214.0 lb/hr
= 170.0 q-mole/hr
SULFUR REMOVAL CALCULATIONS
1’IETEOD I
Caustic Flow rate = 0.50 qpm
Caustic Analysis c nn/i g—mole/l
t aOH 78 1.950
Na2SOLI 200 1.1408
Net Draw Analysis
NaHSO 110 1.058
Na2SO3 20 0.159
Na2SOI4 250 1.761
S/C = 0.884
TOtal Na in caustic = 4.766 g-mole/l
Total Na in net draw = 4.897 a-mole/i
Net draw flow rate = 0.50 (14.766/4.897) = 0.1487 gpm
Total sulfur appearance as S02 0.487(60) (3.78) (1.058 + 0.159)
134.3 g—mole/hr
Total Na with sulfur as S02 = 134.3/0.884 = 151.8 a-n’ole/hr
Total free Na in = 0.50 (3.78) (60) (1.95) 221.1 g-r oie/hr
Sulfur appearance as Na2SO4 = (221.1 = 151.8)/2
= 347 q-nioie/hr
Total liauid phase sulfur appearance = 34.7 + 134.3
= 169.0 c’-mole/hr
149
-------
METRO!) II
Total sulfur appearance as S02 (as above) = 1311.3 g-mole/hr
Sulfur appearance as Na2SO 1 4 ((0.1487) (1.761) — (0.50) (1.1108))
3.78(60) = 311.8 g—n le/hr
Total liquid phase sulfur appearance = 34.8 + 134.3
= 169.1 g—xnole/hr.
150
-------
APPENDIX E
BRITISH TO METRIC CONVERSION TABLE
151
-------
APPENDIX E
BRITISH TO METRIC CONVERSION
To Convert from To Multiply by
acfm mm 3 /hr 1.70
ft m 0.305
ft/sec rn/sec 0.305
g 4ncf l/ 3 0.134
gpm 1mm 3.79
gpm/ft 2 l,kiin,kt 40 • 8
gr/scf gr n 3 2.29
in. cm 2.54
in. H20 nun Hg 1.87
lb—moles gm-moles 4.54
lb-moles/hr gm-moles/mm 7.56
lb-mo les,4nin gm-moles/sec 7.56
Subtract 32 and
Divide by 1.8
153
-------
APPENDIX F
TABULAR SUMMARIES AND MEASUREMENTS
155
-------
TP.BLE F-i
SUMMABX OF TEST RESULTS
Quantity of’ SO2 Removed Caustic
Gas Phase Liquid Phase Na to Circ.
Test SO 2 Inlet 502 Removal. Gas Rate Die. App. Flow Na Conc. Absorber NaSO14 Rate
No. _______ _________ Lb/Hr g mol/hr g moi/hr g g mol/L g inoi/hr g niol/L G n
7 1,815 86.2 8,1465 200 203.9 0.611 1.822 252.8 1.12 7
8 ABORT
9 ABORT
10 2,265 88.5 8,6314 262 255.7 o.6 1i6 2.088 306.14 1.205 7
11 1,016 96.3 8,883 132 1140.8 0.389 2.117 187.0 1.283 7
12 1,925 92.1 8,790 235 217.1 0.5146 2.026 251.2 1.378 10
13 1,550 93.3 8,716 176.9 176.7 0.532 1.778 2114.8 1.317 7
114 1,975 93.7 7,375 189.9 1714.2 0.506 1.909 219.14 1.261 7
15 1,815 88.0 8,352 205.9 210.6 0.616 1.767 2147.2 1.652 10
16 ABORT
18 2,100 97.5 14,1475 136.7 129.6 0.1498 1.1485 168.0 1.378 10
19 ABORT
20 2,380 93.2 6,621 216.3 205.8 O.51e6 1.9140 2140.6 1.356 10
21 2,3143 89.2 5,336 167.5 165.14 0.1483 1.767 193.8 1.5146 11.5
21* 2,370 89.0 5,320 170.2 167.6 0.1490 1.781 198.2 1.536 11.5
22 2,098 87.5 14,1499 123.14 126.6 0.367 1.879 156.6 1.7149 11.5
23 2,1614 814.5 5,360 1145.9 151.3 0.1i06 1.981 182.7 1.798 10
214 2,026 89.2 7,886 213.5 219.2 0.638 1.812 262.6 1.561 11.5
25 2,020 90.0 7,930 2114.8 228.1 0.606 2.110 290.14 1.1498 11.5
26 1,693 88.14 7,666 170.8 171.5 0.1490 1.992 221.7 1.609 11.5
27 1,732 90.2 7,555 176.7 177.0 0.513 1.998 232.8 1.578 30
28 2,682 92.9 6,1e81 2141.7 250.6 0.6147 2.070 3014.2 1.681 10
29 2,328 91.0 6,3314 196.9 199.0 0.522 2.097 2148.6 1.5140 10
30 2,9214 90.14 6,373 2149.9 268.8 0.61414 2.1514 315.1 1.5314 10
31 3,071 91.14 5,8314 2314.9 252.1 0.627 2.079 296.1 1.1475 10
32 3,61 10 93.14 14,880 257.0 278.8 0.731 1.9914 331.1 1.1488 10
33 ABORT
314 2,068 89.3 8,093 226.9 235.7 0.716 1.778 289.1 1.258 7
-------
TABLE F—i (Continued)
SUMMARY OF TEST RESULTS
Absorber Oxidation
Net Draw Liquor Analysis Fe in Caustic Sodius Overall
Test in (BIlL NaOH Utilization Balance Ratio Acid Na Caustic
• Na su 3 BaR sb 3 Na2SO I# m g/L SO2ftla Method Method Method Na Net Draw Notes
7 11.2 1)7.8 196.8 0.9314 - 0.791 17.1 15.56 7•5* 0.988
8
9
10 11.9 l5I .2 1S.5.8 0.913 - 0.855 10.5 9.0 0.998
11 214.7 133.5 208.5 0.883 - 0.706 12.8 11.5 1.003
12 21.2 1143.1 203.14 0.905 - 0.935 7.6 15.8 1.0 1 49
13 18.4 120.6 205.6 0.899 - 0.823 10.14 10.2 1.011
20.0 123.6 212.0 0.885 - 0.866 114.3 114.0 0.977
15 18.6 1141.2 261.0 0.910 . 0.833 7.02 7.0 0.955
16
18 23.5 95.3 2140.5 0.851 0.65 0.81. 14 15.8 8.6 0.931
19
U 20 10.9 1141.8 230.7 0.936 0.15 0.899 114.1 114.8 0.989
21 11.6 136.8 253.5 0.9140 0.70 0.8614 10.8 11.3 - 0.9614
21 9.7 133.7 255.6 0.9147 0.70 0.859 13.1 13.0 22.3/20.9 0.963 Loss analysis run -
part of run 21
22 9.2 121.1 290 0.914 2.90 0.783 18.3 18.3 26.1/30.8 1.001
23 9.7 138.0 295.9 0.350 1. IêO 0.799 16.0 15.9 - 0.985
2h 12.3 125.9 250.8 0.927 0.140 0.813 12.14 12.1 23.1 1.561
25 3.8.7 135.1 271.5 0.907 0.35 0.739 ‘8.6 17.9 - 1.1489 Split caustic flow
26 13.0 123.14 293.6 0.930 0.20 0.770 22.2 21.5 28.9 0.938 Split caustic flow
27 9.6 115.5 289.8 0.928 0.29 0.759 25.2 214.2 18.5 0.955 High liquid circu-
lation rate
28 13.5 1142.7 280.0 0.93 0.147 0.795 15.1 3.14.8 214.1 0.983
29 20.7 1149.1 277.0 0.909 0.13 0 792 14.7 12.1 114.0 0.908
30 9.7 170.9 253.7 0.953 0.25 0.793 10.7 8.9 8.0 0.962
31 12.0 170.1 2145.8 0.930 0.31 0.793 9.4 8.7 15.7 0.939
32 7.8 1614.1 2149.3 0.9143 0.09 0.776 10.2 8.14 - 0.936
33
3/. 11.2 132.0 216.8 0.926 0.32 0.785 13.1 12.14 23.0* 0.9 1e5
*Ouestionable acid
1ow measurement
-------
TABLE F—2
DISTRIBUTION OF OXIDATION AND SO 2 REMOVAL BENEEN TOP AND BOTTOM ABSORPTION STAGES
Average for Rim
NO 2 SOI 1 . S02 Overall Oxidation Stage 1 (Bottom )
Inlet Gas in Removed 5014 SO 2 SOli.
SO 2 Rate NaOH g niol Formed Removed Formed
Run p Lb/Hr g inol/L Mm Oxidation g moi/inin g mol/min g mol/min Oxidation
10-1 2,265 8,631, 1.205 11.143 6.5 .288 .898 .26 .071 27.0
10-5 11,20 11.0 . 1 462 .925 .50 .103 20.7
13-5 1,550 8,716 1.317 2.92 9.5 .277 .882 .02 .128 100+
13-9 2.76 13.6 .376 .918 .08 .1914 100+
20 2,1113 6,1129 1.3145 3.76 16.8 .632 .966 .87 .2115 28.1
20-11 3.68 16.1 .592 .927 .117 .306 65.6
21—11 2,3143 5,336 1.5146 3.014. 11.0 .335 .9115 .36 .161 143.9
21—16 2.62 111.6 .383 .911.]. .30 .230 75.8
26-3 2.52 211.0 .605 .9211. 1.32 .11.27 32.2
26-5 1,7211. 7,666 1.628 2.90 19.3 .559 .930 1.55 .1402 26.0
26-8 2.87 22.0 .631 .920 1.73 .1130 214.8
28-3 3.98 20.3 .809 .931 .50 .396 79.6
28-5 2,669 6,1179 1.6711 14.17 114.2 .591 .936 .115 .2611 58.9
29-10 2,328 6,3311 1.5110 3.21 9.2 .297 .907 .23 .065 28.3
31-11 3.80 8.6 .327 .9140 .27 .161 60.0
31111 3,050 5,689 1.1475 11.18 10.2 .11 .26 .91411 .30 .1147 149.0
-------
TABLE 7—2 (Continued)
DISTRIBUTION OF OXIDATION AND SO 2 RE) )VAL BEWEEN TOP AND BOTTOM ABSORPTION STAGES
Stage 2 (Tc m
State 1
Stage 2
202
R oved
g mol/min
2014
Formed
g mol/ndn
%
Oxidation
10-i
.8140
10—5
.828
13—5
.833
13-9
.78].
20- ].
.820
20—4
.811
21-U
.799
21—16
.825
26-3
.766
26—5
.772
26-8
.738
28—3
.726
28—5
.773
29-10
.762
11.17
3.70
2.90
2.68
2.89
3.2].
2.68
2.32
1.20
1.35
1.111
3.148
3.72
2.98
3.53
3.88
• 217
.359
• 119
.182
• 387
.286
.1714
.153
.178
.157
.201
.327
.232
.166
.279
pof
R iova1
Stage ( 2 )
941.1
88.1
99.5
96.9
76.8
87.3
88.0
88.14
147.14
116.7
39.6
87.1
89.2
92.7
92.9
92.8
5.2
9.7
14.1
6.8
13.14
8.9
6.5
6.6
114.8
11.6
17.7
11.8
8.8
7.8
11.7
7.2
% of
Oxidation
Stage C 2 )
75.3
77.7
113.0
148.11
61.1
118.3.
52.0
110.0
29.11
28.1
31.7
50.8
55.0
511.6
51.0
65.5
Na 2 SO 3
NoR 803
Na 2 SO4
Na 2 SO 3
Ne l l 803
Na 2 SO14
25.2
9.2
161.9
1113.1
206
186
39.7
27.7
139.5
127.0
198
183
20.5
109.8
211
32.8
108.3
205
111.6
123.9
213.
1111.1
93.7
190
3.2
11.14
]is3.2
139.5
220.6
2112
20.3
38.0
117.1
].].1,8
222.11.
212
10.8
130.7
2411
1114.9
110.3
2142
10.2
127.5
288
35.8
109.7
267
9.0
12.6
117.8
127.5
288
282
1 18.6
11.7.7
90.9
98.3
268
259
114.2
123.3
300
53.2
84.1
291
11.2
137.3
316
66.1
89.7
2914
3.2.2
142.8
273.
50.3
105.8
260
20.8
1119.85
257
59.8
108.14
253
31—11 .7143
31—14 .775
8.8 173.3
7.11 172.3
2112
67.0
60.9
lol&.8 233
113.1 238
2144
-------
TABLE F-3A
OVERALL GAS MASS TRANSFER COEFFICIENTS FOR SO 2 ABSORPTION
IN SODIUM SULFITE—SODIUM BISULFITE SOLUTIONS
502
Inlet Percent SO 2 App
pp Removed Oxidation Removed g mol
Run Dry Overall Total g mol/min Mm
10-1 2,1100 87.0 6.5 11,113 1137
10—5 2,250 86.5 11.0 11.20 3.96
13-5 1,560 86.5 9.5 2.92 2.92
13-9 1,508 93.11. 13.6 2.76 2.92
20-1 2,11.90 95.2 16.8 3.76 3.148
20-14 2,310 89.6 16.1 3.68 3.113
21-1]. 2,520 91.1 11.0 3.011. 2.92
21-16 2,150 8140 111.6 2.62 2.1 1 .0
26—3 1,560 83.7 21+.O 2.52 2.55
26—5 1,680 89.3 19.3 2.90 2.97
26-8 1,650 92.7 22.0 2.87 2.96
28-3 2,5114 911.11 20.3 3.98 3.17
28—5 2,835 92.6 114.2 14.17 3.62
29-10 2,310 87.9 9.2 3.183 3.303
31-U 2,880 95.8 8.6 3.80 11.19
31 111 3,053 89.11 10.2 14.18 11.144
161
-------
TABLE F-3B
OVERALL GAS MPSS TFLkNSFER COEFFICIENTS FOR ABSOPi.L ON
IN SODIUM SULFITE—SODIUM BISULFITE SOLUTIONS
StaRe One (Bottom)
S02
X 1O S02 x 1O Rem x 1O
s/c pp so 2 Ye S/C Top Ye g mol Fraction Y Y
Run Btm nm Hg Btm Hg Top M n Removed Btm
10-1 .898 0.67 8.8 .893 0.67 8.8 0.26 .059 21.1 20.0 .0911 106.7
10-5 .925 1.05 13.8 .8314 0.225 3.0 0.50 .119 19.8 17.8 .205 148.7
13—5 .882 0.50 6.6 .838 O.21i0 3.2 0.02 .007 13.7 13.6 .012 8611.5
13-9 .918 0.92 12.1 .908 0.76 10.0 0.08 .027 133 12.9 .208 118.2
20—1 .966 2.00 26.14 .956 1.70 22.14 0.87 .231 21.9 17.1 - -
20—4 .927 1.06 114.0 .919 0.92 12.1 0.147 .128 20.3 18.0 .377 26.5
21—11 .9145 1.113 18.8 .937 1.214 ].6.3 0.36 .3.18 22.2 19.8 .696 111.11
21—16 .9141 1.314 17.7 .936 1.22 16.1 0.30 .115 18.9 17.1 1.61i 6.1
26—3 .9211 1.04 13.7 .889 0.56 7.11 1.32 .5214 13.7 7.7 3.15 3.17
26-5 .930 1.10 114.5 .897 0.66 8.7 1.55 .5314 114.8 7.7 - -
26-8 .920 0.95 12.5 .885 0.53 7.0 1.73 .603 i 1 4. 6.li - -
28-3 .931 1.12 114.8 .915 0.87 11.14 0.50 .126 22.1 19.5 .338 29.6
28-5 .936 1.214 16.3 .9211 1.00 13.2 0.145 .108 25.0 22.6 .278 36.0
29-10 .907 0.76 10.0 .899 0.69 9.1 0.23 .072 20.3 19.0 .123 81.14
31-11 .9140 1.30 17.1 .931 1.20 15.8 0.27 .071 25 .Ii 23.6 .2214 114.7
31—34 .91411 1•1 18.6 .931 1.20 15.8 0.30 .072 26.9 25.1 .205 118.8
-------
TABLE F-3C
OVERALL GAS MASS TRANSFER COEFFICIENTS FOR SO 2 ABSORPTION
IN SODIUM SULFITE—SODJIJM BISULFITE SOLUTIONS
Stage Two (Top )
so 2
x10 1 4 x1O 1 Rem
S/C pp SO 2 Ye S/C pp SO 2 Ye g mci Fraction
Btm Btm Btm Top Top Mm Removed Y Btm Y Top N0a HO 0
10—i .8140 0.25 3.3 .765 0.071 0.9 14.17 .9141 20.0 2.7 2.59 3.87
10-5 .828 021 2.8 .782 0.093 1.2 3.70 .881 17.8 2.7 2.58 3.88
13—5 .833 0.225 3.0 .768 0.075 1.0 2.90 .993 13.6 1.9 2.97 3.36
13-9 .781 0.105 1.14 .721 0.0314 0.5 2.68 .973 12.9 0.9 3.63 2.75
20-1 .889 0.57 7.5 .830 0.21 2.8 2.89 .769 17.1 1.1
20—4 .820 0.18 2.14 .770 0.077 1.0 3.21 .872 18.0 2.1 2.91 3.li14
21—11 .799 0.126 1.7 .765 0.071 0.9 2.68 .882 19.8 2.0 2.93 3.141
21-16 .825 0.195 2.6 .792 0.11 1.14 2.32 .885 17.1 3.0 2.141 14.15
26—3 .766 0.072 0.9 .7148 0.053 0.7 1.20 -.1476 7.7 2.2 1.57 6.38
26-5 .772 0.080 1.1 .758 0.063 0.8 1.35 .1466 7.7 1.6 2.22 14. i
26—8 .738 0.0146 0.6 .728 0.038 0.5 1.114 .397 6.14 1.1 2.31 14.32
28—3 .726 0.037 0.5 .678 0.017 02 3 ,18 .8314 195 1.2 2.99 3.314
28—5 .773 0,080 1.1 .726 0.037 0.5 3.72 .892 22.5 1.9 2.81 3.56
29—10 .762 0.067 0.9 .723 0.035 0.5 2.98 .928 19.0 2.5
31—11 .7143 0.0119 0.6 .700 0.0214 0.3 3.53 ..929 23.6 1.1 3.141 2,914
31—14 .775 0.0814 1.1 .716 0.031 O.li 3.88 .928 25.1 2,9 2.314 14.28
-------
TABLE F-.
SUMMARY OF MAJOR STREAM METALS ANALYSIS
Sample No. 51& 6 1 4 62 241 2142 2143 2 144 2145 2146 247 —2148 349 350 381 382
location Net Net Acid Strip A B Mid Comb Cell Net Strip Quench Cell Cefl Strip Quench
Draw Draw Bts Anolyte Anolyte Anolyte Catbolyte Teed Draw Rims H ,O Feed Feed Btms H 2 0
At Stks Return T-l At Stks Return
Date 7/7 7/7 7/3.6 7/7 8/22 8/22 8/22 8/22 8/22 8/22 8/22 8/22 io/s 10/5 10/27 10/27
Time 0100 1530 0700 1535 1535 1535 1535 1535 1600 1600 3.600 1600 1600 1320 1320
Cations:
M m 0.2 0.5 0.1 0.2 0.30 0.06 0.36 0.23 1:35 0.35 0.60 0.15 0.25 1.50 2.0 0.1
Mg <1. 3. 1. <1. 2.0 <1. <1. <1. 2. 10. 5. 38. 0.60 0.65 0.62 34.
K 70 75 60 55 lie S 0.6 130 1400 170 250 175 4.o
Al <0.5 5.0 C.5 <0.5 <0.5 <0.5 <0.5 <0.5 <0.5 0.6 0.5 <0.5
Pb 2.5 3.0 8.0 2.0 5.6 3.8 1.5 1.7 1.4 2.0 3.5 <0.02
Zn 0.5 1.0 1.0 0.5 0.63 0.28 0.56 0.514 0.58 0.68 0.62 <0.02
Cr 7.0 8.0 6.0 5.0 0.21 <0.02 0.10 0.15 0.12 0.63 0.113 1.0 0.14 0.3.4
Ca 2.5 13.0 7.0 7.0 5.2 0.5 5.6 1.3 5.6 5.3 6.3 52.0 3.3 3.0 7.8 77.5
Na 100,000 100,000 50,000 75,000 145,000 15. 43,600 90,000 51,000 65,000 50,000 30 514,000 55,000 73,1400 214.
1 Fe Llmtetics 11.0 14.5 2.0 2.0 1.8 2.0 1.14 1.8 i.6 3.6 58.0 3.6 3.2 3.1 33 1.7
Ni 1.9 o.k 1.6 2.2 1.8 2.14 2.1 0.3
a Cu
Co
Mo <0.05 <0.05
Nitrites o.oth 0.00k 0.004 <0.002 <0.002 <0.002 0.006 0.012
Nitrates 130 110 18 30 29 <0.01 13 0.8
C l 152 38 698 161.5 7141 29,000 950 152 10.5 9.3 1.2 <1.
Si 36 6 29 31 29 23 33 6 20 8
Chemical Analysis:
MaR50 3 g/1 Nil 52.1 614.5
Na 2 50 3 ill 26.3 19.6 0.15 115.5
Na 2 50p 4 g/1 171 140.5 0 151 125 176 1143 1714
NeCK g/1 Nil 108
B 5014 g/l 39.2 100 9.8 1.2 0.30 5.4 0.98
9.0 6.1 8.9 1.0 1.0 1.2 13. 8.85 5.75 2.35 2.25 2.3 3.1
p 60 °o 1.20 1.20 1.16 1.125 1.0148 1.13.6 1.184 1.1335 1.1655 0.968 1 4 1.166 0.991
Lj sties Fe Analysis proves incorrect
-------
Mg
K
Al
Pb
Zn
Cr
Ca
Na Limnptics
e
Ni
Cu
Co
Mo
Nitrites
Nitrates
Cl
Si
383 38 ’ . 385 386 387 388 389 902 903 9014 905 906 907 908 909
Net Cell Mid Comb Cell B A Net Quench Strip Comb B A Mid Cell Cell
Draw Feed Anolyte Catholyte Feed Anolyte Anolyte Draw H 2 0 B ns Catbolyte Anollrte Anolyte Anolyte Feed Feed
T-l At Stks Return At Stks At Stks
10/27 10/27 10/27 10/27 10/27 10/27 10/27 1/8 1/8 1/8 1/8 1/8 1/8 1/8 1/8 2/28
1320 1320 1320 1320 1320 1320 1320 0730 0730 0730 0730 0730 0730 0730 0730 1600
Chemical. Analysis:
NaBSO 3 g/l
Na2803 g/l
1218.8 1.014
7.0 1.89
1 1 .8. 1 .
22.95
218 146. 1. 173 229 7 236.14 186
78.20
97.5 145. 0. 10.09 -
8.8 18.9 2.3 1.9 12.8
1.167 1.081 1.157 1.160 1.0 1.183 1.210
52.7 191.8 186.6 230
92.00 1414.28 25.83 -
0.05 1.1 1.14 7.5
1.079 1.162 1.151 1.180
Sample No.
Location
Date
Time
Cations:
TABLE F— / . (Continued)
SUMEARY OF MAJOR STREAM METALS ANALYSIS
U I
0.25 1.1 0.9 1.0 0.18
0.25 0.32 0.30 0.38 0.1.0
0.8
1.7
1.14
0.5
6.2
0.8
2.0
1.6
0.13
1.3
1.0
0.1.1.
0.58
0.187
0.13
0.59
0.140
0.141
0.36
51.
0.1.8
0.15
0.2
0.2
0.’.
14.0
0.5
2.0 6.0 18.0 2.0 2.0
0.51
<0.1
7.1
7.2
6.0
5.2
7.3
2.1
5.9
10.5
105.
10.5
8.8
3.2 8.3 10.0 10.5 3.14
85,000
14.8
70,000
3.9
37,500
3.3
96,800
18.3
75,000
3.8
1,250
1.5
1422,000
3.2
io1e,000
.7/.5
<0.5
<0.5
<0.3
110
1.5/1.6
<0.5
<0.5
<0.3
79,000
1.1/1.5
<0.5
<0.5
<0.3
100,000
0.3/0.1
<0.5
<0.5
<0.3
13,000 60,000 60,000 72,000
0.5/<.01 0.7/<.01 0.7/<.01 0.3/<.01r.C.5/<.Ol
<0.5 <0.5 <0.5 <0.5 18.0
<0.5 <0.5 <0.5 <0.5 0.6
<0.3 <0.3 <0.3 <.0.3 <0.1
352
<1.
5.8
<1.
<1.
<1.
<1.
.
2214
1.22
-
11.6 55.2 105 1314
ii.
16
16
28
16
0.06
0.16
6
114.8
6.3
95.9
11.2 6.9 13.3 12.1
Na 2 SO / . gil
NaOH g/l
H 2 S0 4 5/i
pH
p 60 °C
0.10
180
80.8
20.6
5.18 8.1 1.9 12.9
1.208 1.168 1.201
aLilimetics Fe Analysis proven incorrect
-------
TABLE F-5
SUMMARY — MATERIALS OF CONSTEUCTION TESTING
1
Flue Gas Quench Section (cooling
H 2 0 With pH 2.5)
Aeration: Moderate Spool Located
Near Bottom of Section, Covered
With Liquid 100% of Time
Periodic Operation - System
Drained and Flushed When
Shutdown Occurred
Temp: 120 F Avg (110-130 i)
Exposure Time:- 2,300 Hr
Incolcy 825
Incoloy 825
Type 316 S.S.
Type 316 S.S.
Carpenter 20Cb3
Carpenter 20Cb-3
Hastelloy G
Hastelloy G
Titanium - 0.2%
Palladium
Titanium - 0.2%
Palladium
2
Absorber (Bisulfite Solution
pH 5.5 NaHSO (P 100 g/L;
Na80 3 (P 10 g7L; Na 2 SO14 (P
200 g/L
Spool Located on Top of
Packing, Therefore Exposed
to Spra - of Process Solution
Periodic Operation - System
Drained and Flushed When
Shutdown Occurred
Temp: 110 Avg (100-120 F)
Exposure Time: 1,900 Hr
Extensive Aeration
0 .l Allegheny Ludlum 6x
0.l Allegheny Ludlum 6X
0.l Incoloy 825
0.1* Incoloy 825
0.1 Durimet 20
0.1 Durinet 20
0.1 Type 316 S.S.
0.1 Type 316 3.8.
0.1 Carpenter 2OCb-3
Carpenter 20Cb3
0.1 Chlorimet 2
Cblorimet 2
3
Stripper Overhead (Wet 802
With Some Na ,S0 14 From
Entrainment tVapor Phase)
Spool Located in Vapor Phase
Just Below Condenser
Periodic Operation - System
Drained and Flushed When
Shutdown Occurred
Temp: 230 F Avg (220-2140 )
Exposure Time: 2,300 Hr
Carpenter 2OCb-3
Carpenter 20Cb3
0.1 Hastefloy G
0.0 Hastelloy G
0.0* Type 316 S.S.
0.1* Type 316 S.S.
0.l ” Incoloy 825
0.14*4
0.1*
0.1*
22.l
23.0*4
‘4
Stripper Reboiler (23% Wt
Na 2 SO 14 Solution With 0.014
to 0.1 N H 2 S0 14 ; pH = 2.0
to 2.5)
Spool Located in Stripper
Reboiler Drum (On Bottom)
Periodic Operation - System
Drained and Flushed When
Shutdown Occurred
Temp: 230 F Avg (225-2k5 F)
Exposure Time: 2,300 Hr
0.1 Incoloy 825
0.1 Incoloy 825
0.2’ Carpenter 2OCb-3
0.3 Carpenter 2OCb-3
0.1 Titanium
Titanium
lncipient pitting
44 Few random incipient pita
(1 nil diem and depth)
4 lncipient pitting
** tergreflu].ar etch
4 Numerous small shallow pits
4 lncipient crevice corrosion
4* Random pits
Crevice corrosion up to
8 mils deep
**Intergranular etch
Spool No.
Environment
Test Results
Coupon
Material
Corrosion Coupon Corrosion Coupon
Corrosion Coupon Corrosion
Rate (amy) Material Rate (mpy) Material
Rate
(mpy) Materlai. Rate (mpy)
0.0
0.0
0.2*4 Durimet 20
0.1 Duriinet 20
0.1
0.1
lIt.9
0.1
12. 2-’ *
18.
0.0
0.0
-------
TABLE F_5 (Continued)
SUMMABI — M&TERIALS OF CONSTRIJCTION TESTING
5
Feed Liquor Tank (Na 2 SO
Solution, 200 to 280 g/L
In D I H 2 0, pH = 8.0 - 8.5)
Spool Located Just Off Floor
of Vessel. Subjected to
Both Acid and Base Condi-
tions (pH 1.0 and 13+) and
Temp 150 F on Na 2 SO Makeup
Incoloy 825
Incoloy 825
Carbon Steel 1010
Carbon Steel 1010
6
Caustic Storage Tank (2.0 N
NaOh Solution Plus 17% Wt
Na 2 SOj )
Spool Located on Floor of
Tank; Always in Liquid
Phase
Texnp: 120 F Avg (].OO-lleO F)
Exposure Tine: 218 Days
0.0 Inconel 600
0.0 Inconel 600
Ie.6* Carbon Steel 1010
I l7 Carbon Steel 1010
Monel O0
Monel eO0
Nickel 200
Nickel 200
7
Acid Storage Tank (1.0 N H SO
Solution Plus 17% Wt Na 2 S )
Spool Located on Floor oI’ Vessel;
Always in Licuid Phase
Temp: 120 F Avg (l00-lIeO F)
Exposure Time: 218 Days
0.0 Type 317 S.S.
0.0 Type 317 S.S.
0.1 Incoloy 825
0.1 Incoloy 825
0.3 Durimet 20
0.3 Durimet 20
0.3 Carpenter 2OCb-3
0.2 Carpenter 2OCb-3
0.0*
0.0
0.0
0.0
0.0
0.0
0.0
0.0
Hastelloy 0
Haste].loy G
0.0
0.0
Nonnuiform general corrosion;
broad pits
Allegheny Ludluin 6x
Allegheny Ludluin 6x
*A few random pits 1 mil deep
**Intergranular etch; one piece
partially corroded away
Spool No.
Environment
Test Results
• Temp: 100 F Avg (90-110 F)
I Exposure TIme: 81 Days
-1
Coupon
Corrosion
Coupon
Corrosion
Coupon
Corrosion
Material
Rate
(m v)
Material
Rate
(mi v
Material
Rate
(mov
Type 316 S.S.
Type 316 S.S.
0.0
0.0
0.1
0.1
iO 4 .8 *
86. 5**
Chlorlnet 2
Chlorinet 2
-------
APPENDIX C
IGT TRACE IRON ANALYSIS
169
-------
A
INSTITUTE OP GAS TECHNOLOGY • 3424 SOUTH STATE STREET • lIT CENTER • CHICAGO, ILLINOIS 60616
PHONE 225-9 . 0O t R A COOr. 312
February 4, 1974
Mr. Alex Korosi
Stone and Wcbster Engineering Corp.
225 Franklin Street
Boston 1 Mass. 02107
Dcar Alex:
This is the final report on the analytical work which was undertaken
by our laboratories in order to resolve the problems associated with the trace
iron determination of your “Wisconsin SO 2 Pilot Plant” scrubbing solutions.
The presence of high concentration of Na 2 SO 4 (approx. 3 N) in
your 8 sample solutions seriously interferred with the iron analysis by either
atomic absorption or cOlorimetric methods. We therefore made up a series
of synthetic samples containing known amounts of iron in Na 2 SO 4 solutions to
study the est way of arriving at a suitable method of analysis. The following
eommon t s and conclusions help to explain the significance of the results of
Table I below.
Table I.
CONCENTKATION OF IRON IN SAMPLES OF SODIUM
SULFATE SOLUTIONS TAKEN FROM
WiSCONSIN SO 2 PILOT PLANT
Iron, Ii g/mi
Sample Identification y AA By Ferrozinc
584 0.04 0.54
585 0.01 0.58
586 0.10 0.64
587 <0.01 0.74
588 35 46
589 1.4 1.9
No. 1 3,3 3.9
No. 3A 0.48 1.0
APFIUATEO WITH LLINOIS INSTITUTE 0F ‘I ECIINOLOGY
171
-------
Mr. Alex Korosi
Page Two
I. Atomic Absorption Methods
A. Direct aspiration of the original 3N Na 2 SO 4 solutions resulted in
highly erratic iron values, but by diluting the samples with 4 to 5
parts of water (approx. 0. 60 to 0. 75 N in NazSO 4 ) and using a 3-slot
burner head we were able to obtain results which were within ±5 for
up to lO g/ml of iron present.
B. A solvent extraction scheme (Appendix 1) for iron with subsequent
AA analysis of the solvent afforded the best set of values (Table 1) for
iron, as judged by the complete recovery of known amounts of iron from
a series of standard solutions.
II. Colorimefric Methods
A. The use of I, lO -phenanthroline as colorimetric reagent for iron
in 3 N Na 2 SO 4 solutions again proved unworkable, since the high sulfate
ion concentration caused considerable deviation from Beer’s law. Dilution
of the salt solutions reduced the deviation but the method did not have
enough sensitivity for low iron concentrations.
B. Ferrozine, which is a relatively new colorime t ric reagent for iron,
forms an iron complex possessing 2. 5 times the molar absorptivity of the
similar iron I, lO-phenanthroline complex. Because of this enhanced
sensitivity and better tolerance for sulfate ion concentration. ferrozine
gave much better results for iron in diluted solutions of NazSO 4 (not over
0.3 N) than the comparable 3, lO-phenanthrolino procedure. Neither
reagent, however, is as specific for iron as is the AA method, due to the
formation of other metal iron complexes such as copper and cobalt.
Our conclusion, based on the above vork, is that while afornic absorption
method for dilufe soluUons of NazSO 4 (not to exceed 0. 5 to 0. 7 N) is workable,
it is not as dependable a the solvent extraction - AA procedure for trace iron
ana’ysis in your system.
As a second choice ferrozinc method of colorrnetry may fill a limited
function as a quality control measure provided that ‘he Na SO 4 concentration
INSTITUTE OF GAS TEC hNOLOGY
172
-------
Mr. Alex Korosi
Page 1 ’hree
in the analysis solution does not exceed 0. 3 N and the operator keeps a watchful
eye on the other metal ions and corrosion products which may give a positive or
negative error.
Best regards.
Sincerely yours,
A. Attarf
Manager. Analy+ical
Services
aa/ah
end.
INSTITUTE OF GAS TFCK1 OLOGY
-
-------
Appendix 1.
AA DETERMINATION OF TRACE IRON BY
SOLVENT EXTRACTION OF IRON FROM 3N Na 2 SO 4 SOLUTION
Reagents :
1. APCD soin.: Ig ammoniurn pyrrolidinecarbodithioatc in 25 ml.
water. Must be prepared fresh daily.
2. Ammonium acetate buffer, 5M: dissolve 385 g NH 4 OAc in water,
make to one liter.
3. Standard Fe soin.: dissolve one g. iron in 6N Hcl, make to one
liter (1000Mg/mi).
Procedure :
a. Place a suitable aliquot of sample or standard solution, containing
less than 10 1.’ g iron in less than 25 ml soin.
b. Add 2 ml 5M NH 4 OAc soin., and sufficient NH 4 OH or Hcl to bring
the pH to 3.0 ±0.1 with a pH meter.
c c. Transfer the soin. to a 50 ml volumetric flask, add I ml of
APCD soin. to sample solution.
d. Add 5 ml methylisobutyl ketone (MIBK) and shake vigorously for
one minute. Add water to bring the organic phase into the neck
of the flask and aspirate directly from the flask for AA
det erminations.
INSTITUTE OF GAS TECiHIIOLOGY
174
-------
PHASE lB
DESIGN, FABRIC1 TION, AND PERFORMANCE
OF THE IONICS PROTOTYPE ELECTROLYTIC
CELLS FOR THE STONE & WEBSTER/IONICS S02 R 4OVAL
AND RECOVERY PROCESS
175
-------
I
ABSTRACT - PHASE lB
Phase lB of the EPA—WEPCO sponsored program to evaluate
the Stone & Webster/lonics closed cycle S02 removal system has
been successfully cc upleted. In this phase, prototype
electrolytic cells, Types A and B, having production
capacities per cell more than five times those of the pilot plant
size cells were developed. The performance of both types of
cells was demonstrated; they meet the contract performance
objectives for energy consumption per unit of caustic produced as
well as, in the case of B” cells, the current efficiency for
purge sulfuric acid. A cell packageR concept in which cell
membranes, diaphragms, screen separators, and molded rubber
gaskets for two cells are preassembled was utilized to simplify
field assembly of the cells.
177
-------
II
CONCLIUS IONS
1. The targeted performance objectives for both prototype “A”
and “B” cells were satisfied in the current density range of
120 to 160 ampere per square foot (ASF) studied.
The d—c energy in kilowatt—hours per pound of caustic
produced at 160 ASP was under 2.47 for the prototype “A”
cells and under 2.68 for the prototype “B” cells.
2. The production capacity per cell was increased by more than a
factor of five over the 8 ,c 17 in. cells used in the WEPCO
pilot plant, at the sante electrical energy consumption.
3. In the case of prototype “B” cells, some internal and
external leakage experienced initially was corrected by
manual modification of the molded components. These
modifications can be incorporated into the canponent molds.
4. The “cell package” concept was demonstrated in the prototype
cells. A preassembled “cell package” contains the membranes,
diaphragms, screen separators, and molded rubber gaskets for
two cells. This innovation simplified the field assembly of
the cells.
5. The prototype “A” and “B” cells are ready to be demonstrated
in a larger scale operation.
179
-------
III
RECOMMENDATIONS
1. The prototype “A and B cells should be tested next on the
75 megawatt demonstration plant proposed under Phase II of
this Contract.
2. Superficial modifications to the ccmponent molds, deemed
desirable as a result of the prototype cell tests, should be
made when manufacture of cells for a d nstration plant is
undertaken.
18].
-------
Iv
INTRODUCTION
The objectives of Phase lB of the program were to
provide detailed operating test data on commercial prototypes of
both 3—compartment (Type A) and Ll ccinpartment (Type B) cells.
Phase lB was divided into three subphases: 1) Design
and fabrication of the prototype cell components; 2) Design and
tabrication of a facility to test the prototype cells; and 3)
Operating tests and data reduction. These are discussed in the
following sections of the report.
183
-------
V
DESIGN AND FABRICATION OF ThE PROTOTYPE CELL COMPONENTS
The design objectives were:
1. To increase the production capacity per cell by a factor of
five at the same energy factor as the 8 x 17 in. cells used
in the WEPCO pilot plant.
2. To maintain an internal cell height close to the 17 in. cell
compartment height used in the “tall cell pilot plant.”
3. To obtain good flow distri.bution within cells and between
cells.
4. To simplify the field assembly of the cells.
5. To avoid exposure of the electrodes to the process solutions
in the cell manifolds.
The production capacity was increased simply by
increasing the effective area of the cells from 0.945 sq ft in
the “tall cell pilot plant” to 5 sq ft in the prototype cells.
The effective area ratio increase of 5/0.945 = 5.29 results in a
d—c current equally higher at a given current density. The
production capacity is directly proportional to the current.
The increase in area was achieved by widening the cells
from 8 in. to a nominal 36 in. in the prototype cell. The
internal compartment height is 20 in.
The current efficiency for electrolytic regeneration
depends on good flow distribution. This was accomplished by
using dual ports to feed and exhaust the major cell compartments,
and by equipping selected ports with flow distribution channels.
The components are shown in detail in the drawings listed in
Table 5.1.
To obtain the same electrical energy consumption as in
the tall cell pilot plant the same intercoinponent distances have
been maintained in the prototype cell. The flow rate of anolyte
recirculation was increased by a factor of 5.29 per cell to
achieve the same gas disengagement from the electrode compartment
as in the “tall cell . A similar increase in catholyte
recirculation flow rate was provided.
The field assembly of the cell was simplified by
preassemblying the components into “cell packages.” A “cell
package” contains the membranes, diaphragms, screen separators,
and molded rubber gaskets for two cells. Each unit in the
package is held together by plastic rivets. The cell package is
165
-------
TABLE 5.1
LIST OF DRAWINGS OF PROTOTYPE CELL COMPONENTS
Drawing No.* Title
20043—SH1—8 Cathode Frame
2004 3-SH2—7 Cathode Frame Sections
200411-5 Cathode Insert
20045-SH1—8 Anode Frame
200145—SH2—8 Anode Frame Sections
20046-6 Anode Insert
20047—SH1-5 Mid-Anolyte Compartment - Anode Side
20047-SH2—5 Mid—Anolyte Compartment — Cathode Side
20048-SH1—5 Center Canpaz-tment Frame - Anode Side
20048-SH2—7 Center Compartment Frame - Cathode Side
20049-4 Anode-Cathode Current Collector
20050—5 Anode
20051-5 Cathode
20052-5 Anode Screen-Cathode Side
20053—4 Mid—Anolyte Screen Cathode Side
20054-4 Mid-Catholyte Screen Compartment — Cathode
Side
20066-2 Cell Assembly
20067—4 Flow Diagram—Sulfomat (5 ft2) Test Facility
20068—2 B Cell Extrusion
20069-2 Package Retainer Rivet
20071-4 Hydraulic End Block
20077-2 Anode Diaphragm
20078—2 Cathode Diaphragm
20079-2 An .on Membrane
20080—2 Cation Membrane
20086-1 Cathode Subassembly Layout
20087—3 Anolyte Outlet Flow Distributor
20088-3 Anolyte Inlet Flow Distributor
200 89—1 b• Cell CC Frame Tab Spacer
20090-1 Anode Support Fingers
20092—1 Hydraulic Module Separator
*Drawings available upon request
186
-------
hung over the rigid anode frame component with one cell on one
side of the anode, and the other on the opposite side • For a
module of eight cells, there are four anode frames, four cell
packages, and five cathode frames. The rigid cathode frames are
placed between the cell units that hang fran the opposite anodes.
Therefore, cell assembly is fast and the assembler is not
concerned with individual membranes, diaphragms, and gaskets.
The electrode frames are molded from filled
polypropylene and are similar to filter press components (Figure
5.1). The electrodes fit in from the side and are sealed in
place by a filled polypropylene insert welded over the opening
for the electrode. The electrode does not traverse any place in
the frame where there are manifold holes that distribute the
flows to and fran the individual cells. Thus, exposure of the
electrodes to the manifold electrolytes is not possible. The
rididity of the electrode frames imparts equal rigidity and
strength to the assembled module.
187
-------
CATHOLYTE OUTLET MANIFOLDS
C-C OUTLET MANIFOLDS
ANOLYTE OUTLET MANIFOLDS
MAN IFOLDS,,//
MANIFOLD
LEAD ALLOY ANODE
PLASTIC ANODE FRAME
ANOLYTE INLET MANIFOLDS
M-C FEED
MANIFOLD
C-C FEED MANI
CATHOLYTE INLET MANI
FIGURE 5. I CELL ANODE ASSEMBLY
-------
VI
PROTOTYPE CELL TEST FACILITY
The prototype cells were tested in a pilot facility
constructed in Watertown. The test facility was designed to test
sixteen “A’ cells or sixteen “B” cells at a time at current
densities up to 160 amperes per square foot of effective
electrode area (ASF). As stated previously, each prototype cell
has an effective electrode area of 5 sq ft.
The sixteen cells (either “A” or B cells) were
arranged in two groups of eight cells each. The cells within a
group (“electric module”) were connected in parallel
electrically. The t electric modules were connected in series.
Hydraulically, all sixteen cells were connected in parallel
(Figure 6.1).
The design caustic and acid production rates for the
16 cells at 160 ASP are 0.893 and 0.446 lb—mole/hr. respectively.
Other selected design values for the test facility are summarized
in Table 6.1. Design point material balances for the 16 “A”
cells and 16 “B” cells are given in Figures 6.2 and6.3,
respectively. The change in cell voltage with current density,
a)nsistent with Contract performance objectives, is given in
Figure 6.1$.
lb conserve chemicals, the caustic and acid products
from the cells were recombined on a continuous basis to form new
sodium sulfate feed for the cells.
189
-------
FIGURE 6. I
ASSEMBLED PROTOTYPE CELL STACK
190
-------
TABLE 6.1
SELECTED DESIGN VALUES OF PROTOTYPE CELL TEST FACILITY
Cell Type A B
Active Area per Cell, sq ft 5 5
No. of Stacks 1 1
No. of Cells per Stack 16 16
Total Active Cells 16 16
Cells per Electric Module 8 8
Cells per Hydraulic Module 16 16
Design Current Density, ASF 160 160
Stack Current, amps 6400 6400
Maximum Cell Voltage, (a,b) volts 6.9 7.5
Maximum stack voltage, volts 13.8 15.0
Maximum total power, kw d—c 88.3 96.0
Maximum Design d—c Energy per lb
NaOH(a,b), kwh/lb NaOH 2.47 2.68
Total Production, lb-mole/hr
NaOH in nu.xed caustic 0.893 0.893
Low Sodium H2S04 0 0.2 10
H2S04 in mixed acid 0. 1446 0.236
Total Makeup Water (c)
lb—mole/hr 0.689 12.52
GPM 0.025 0.451
Cooling RequremelLts (d)
Anolyte, sTU/hr 109,000 117,000
Catholyte, BTU/hr 72.600 78,000
Cooling water(e) GPM 12.1 13.0
(a) Basis: 3N Na2SOL& feed solution, 85 percent NaOIi C.h.,
‘40 percent H2S014 C.E., 140 F (60 C)
(b) Minimum Contract performance objective.
(C) Makeup water consists of water vaporized (140 F basis),
elect.rolyzed, and, in the case of “B cells, fed to the
anolyte.
(d) Based on cell voltages (above) less 2 volt reversible
voltage.
Sodium sulfate solution assumed fed at 120 F and makeup
water account but heat losses to the environment neglected.
Total cooling split 60—40, anolyte-catholyte.
(e) Based on 30 F temperature rise of cooling water.
191
-------
Figure 6,2 Design Point Material Balance for 16 Prototype “A” Cells
Na 2 SO 4
H 2 0 Liquid
H 2 0 Vapor
R 2
02
Totals
1
2
3
4
5
6
7
8
9
10
MOL
MOL
‘HR
2.° ?.12 ,o&qz 117.3
5•3 . 30.1 3S7
2 4 3t2 S 4
MOL
21.
LB
9.12
MOL
HR
.c 4Z
ac.7
117.3
MOL
‘93
MOL
9.12
3O. 1474 q4’ 24’s ‘9J2 Z.i4
392
oc2
602
s2’
£ 2l
LB
4
lz
tic
MOL
1403
.I/I1’
4’z .2 3
iic6
3 ZO
73a
19222
MOL
,
i 47
6
MOL
“a
MOL
— — S — —
GPM
Sp. Gr.
2. O
1.22.7
,. 227
2
i.z27
T F-
/.2.27
.t’(i )
/. 42(’j
q. g
1.342.
i.Zff(”)
at 9
/.2Z5
Temp, °L’
100
/t O -l40
/00
/4O
/4 0
l4O
/40
Press, psia
—
—
—
17.0
i 1.o
cFN, OPT
—
—
—
—
a.øç )
Basis:
/ 6A Cells, 5.0 sq. ft. ea.,
Feed Normality
Current Density
NaOH Current Efficiency
Purged H 2 S0 4 Current Efficiency
Operating Temperature
Catholyto CaustiC Nortnat.ity
I no1yte Acid Norma)ity
ULd-Molyte Acid 4o maAi*
S’O sq. ft.
= 4 N Na 2 SO 4
= / 0 ASF
= ___
= 140 °F
. /_14
WJAN
to
-------
Na OH
Na 2 S 04
H 2 0 Liquid
1120 Vapor
112
02
Totals
Figure 6 • Design Point Material Balaitce for 16 Prototype “B” Cells
1.
2
3
4
5
6
7
8
9
10
MOL
MOL
V
‘HR
MOL
LB
MOL
MOL
3S.7 •g
Zc?I 2.O 9’2 . 42 1)7.3 ff. 12. /17.3
%3 53. 3o.l /. 7 .4 3fl 2/5 30.1 j. 74 “#6 a4.f
/2 (4 31.2 02
394
/293
‘1?32
21
MOL
9.iZ
274
2.O.
202
MC,L
• Z 10
lI.2
•q7 .ô5’v
v.42 .2 3
232.
/qy o
MOL
4;
lao
4R
MOL
‘hR
2/3 / .t3
2/3
hR
zz
23 ’.I
‘/37
MOL
23
7
GP: I
2.O4
.o64
V a ?.
.064
. %,)
.F7
•42(1. )
29.9
.42
1.791
Sp. Gr.
1.277
i.TL7
i.’i?..7
1227
/. 1I2(”)
1.342.
,.o (”)
/,Oi . I
/.O0
/.2.7O
Temp, °F
bc
/ ‘o-i4O
/OO
14 0
/4 0
14O
i’ O
-140
Zo
/40
Press, psia
17.0
—
/7.0
—
—
C II I, OPT
4.O a )
2.0( S)
1
AM Dcl
T
Basis:
Cells, 5o
Peed Norn a1ity
Current Density
NaOH Current Efficiency
Purged H 2 S0 4 Current Efficiency
Operating Temperature
Catho1yt. C us tic t4o’rmality
Anolyte Acid Norwality
Mid-Anolvte Acid Moimality
sq. ft. ea., O sq. ft.
4 N Na 2 SO 4
= ASP
= g -
= 4’O %
= , ‘qo 0 p
= al l
= 2N
o.5’? N
‘0
-------
9
—
“B” ELLS
>
—
;;
.
,._—
- ;
TYPE “A ’ CELLS
—
FEED SOLUTION 3N Na2 SQ4
OPERATING TEMPERATURE 140° F
20 40 60 80 100 120 140 160 180 200
CURRENT DENSITY, AMPERES PER SQUARE FOOT (ASF)
FIGURE 6.4 DESIGN CELL VOLTAGE-CURRENT DENSITY
RELATIONSHIPS FOR TYPES”A” AND “B” SULFOMATTM CELLS
8
—I
0
>
l J
(D
b
0
>
-J
0
0
0
-J
-J
I J
C.)
2
0
0
:i 9J
-------
VII
PROTOTYPJ. CELL PERFORI’ ANCE
The perrormance ot the prototype cells was measured in
the test facility described in Sec. 3.0. Sixteen “A” cells were
tested first, then sixteen “B” cells.
7.1 “ A” CELL PEr FOR1 ’IANCE
7.1.1 Current 1 .fficiency
Current efficiency measurements compare the actual
amount of NaOH and H2S04 produced by the cells to the theoretical
anount that could be produced according to Faraday’s law ror the
actual current applied to the cells. The current efficiency
results for the sixteen prototype “A” cells in the test facility
are summarized in Table 7.1. The table shows that the average
current efficiencies for the catXiolyte arid anolyte were
b4 percent and b3 percent, respectively.* Furthermore, the
current efficiencies at a current density of 160 L SF
(C.E.. Nos.13—16) averaged essentially the same as those at
120 ASP. Also, the current efficiencies at a sodium sulfate teed
concentration of 3N were the same, as expected, as those at 2 W.
7.1.2 Energy Consumption
The total d—c energy applied to the bus bars ot the test
facility divided by the production rate of caustic by the sixteen
prototype “A” cells gives the overall specific energy consumption
for caustic production, kwh/lb NaOH. Specific energy consumption
results are included in Table 7.1 and are arranged according to
Na2SO4 feed concentration and current density in Table 7.2.
The “Performance Objective” values in Tables 7.1 and 7.2
for the “NaOH DC Energy Factors” are basea on the Phase I
Performance Objectives (3W Na2SO4 Feed Solution, 140 F operating
temperature, 85 percent NaOli current etficiency) but adjusted
downward to retlect the actual current densities. The adjustnent
was made in accordance with the cell voltage—current density
relationship given in Figure 6,14.
*NOTh: The independently measured catholyte and anolyte current
efficiencies must be equal for a given cell provided
tnere are no undesirable side reactions at the electrodes
and no external leakage of either product. Our prior
experience with these cells indicates that the c atholyte
current efficiency may run slightly higher than the
anolyte current efficiency due to anode side reactions.
195
-------
Table 7.1
SUM.MI4W OF 1ER?O ffiECE TESTS OF WOTUlYPE “8” aia
Catholyto Anolyte consistency Av. 145014 DC
c.a. Feed NaOH 2 °g Ratiofl) Op. Bus Coil voltago Era?.i Factors
Ho. Cone . Coic. C.E. Ccnc . C.E. ( tI80HlH2S04 ) !2.. author C.D. Voltage Actual ’ Tarqcte Measured 4 ___________
— — saps ASP volts volts volts 10 1 1 4/ lb nOR
4 2.2 1.62 83% 1.04 79% 1.05 135 4770 119.3 10.4 5.2 6.01 1.92 2.15
5 2.2 1.62 91% 0.89 88% 1.03 136 4770 119.3 10.3 5.2 6.01 1.73 2.15
6 2.0 1.82 76% 0.95 78% 0.98 134 4770 119.3 10.6 5.3 6.01 2.12 2.15
7 2.0 1.86 76% 0.96 84% 0.90 134 4800 1.20.0 11.0 5.5 6.02 2.22 2.16
6 1.8 2.02 87% 1.09 77% 1.13 135 4800 120.0 11.0 5.5 6.02 1.91 2.16
9 1.9 1.85 94% 0.98 90% 1.04 135 4770 119.3 12.0 6.0 6.01 1.95 2.15
1C 1.9 1.86 76% 0.99 82% 0.92 137 4875 121.9 12.6 6.3 6.07 ,.S1 2.17
1 1 2.0 1.89 89% 0.98 92% 0.98 135 4800 120.0 10.8 5.4 6.02 1.84 2.16
12 2.0 1.87 86% 0.98 81% 1.06 136 4800 120.0 11.2 5.6 6.02 1.97 2.16
13 1.9 1.92 85% 0.96 79% 1.08 137 6390 159.8 13.4 6.7 6.89 2.39 2.47
14 2.0 1.97 89% 1.02 83% 1.O 137 6390 159.8 13.0 6.5 6.89 2.23 2.47
a’ 3.0 2.40 81% 1.33 79% 1.03 140 6375 159.4 11.6 5.8 6.89 2.18 2.46
it 2.8 2q37 86% 1.29 76% 1.05 140 6405 160.1 11.3 5.6 6.90 2.14 2.47
17 2.9 2.27 84% 0.89 86% 0 .97 137 4785 119.6 10.0 5.0 6.02 1.81 2.15
18 3.0 2.22 95% 0.96 87% 0.98 135 4800 120.0 10.2 5.1 6.02 1.82 2 .16
Av. 84% 8w. 03%
a. Patio of total equivalents of lIaOH end 11 5504 involved in the C.B. Measursaent.
b. Averatje cell voltage, i.e. bus voltago-r2.
c. Calculated Eras Va m et • 5.469 x 10 ‘a,bus + 3.4 which is plotted in Fig. 3.
4. Calculated Eras usasured caustic producttcn rate, BC busbar current end
bus voltage -
e. Phase I Performance objective (3M 14a 2 50 4 feed solution, 1400 p operating
terperatute. 85% geoii current efficiency) bat adjusted dowrn,ard in accordance
with the cell voltage-current dansity relaticiiship given in Pig. 3 to
account for the above actual current densities.
-------
TABLE. 7.2
SUMMARY TO PROTOTYPE •A” CELL ENERGY
CONSUMPTION DATA
*I .nase I Performance Objectives (3 Na2SOLI reed solution,
140 F operating temperature, 85 percent NaOH current eff i-
ciency) but adjusted downward in accordance with trie c l1
voltage — currint density relationship given in Figure 6.4
to account tor the actual current densities.
Na2SOU
Feed
Conc .
Current
Density
ASF
120
C.E.
No .
NaO1 DC Energy Factor
Measured
Performar c
Ob1ective c
1.8
—
2.2N
14
5
6
7
8
9
10
11
12
1.92
1.73
2.12
2.22
1.91
1.95
2.51
1.84
1.97
2.15
2.15
2.15
2.16
2.16
2.15
2.17
2.16
2.16
A.v.
2.02
2.16
160
13
14
2.39
2.23
2.47
2.47
Av.
2.31
2.47
2.8
—
3.0
120
17
18
1.81
1.82
2.15
2.16
1 W.
160
1.82 2.16
15
16
Av.
2.16
2.14
2.46
2.47
2.16 2.47
197
-------
Table 7.2 shows that, at both Na2SO4 teed concentrations
and both current densities, the measured NaOH energy factors were
within the periormance objectives. At 2N Na2SO4 feed
concentration, the measured energy factors were better than the
performance objectives by 6.5 percent at both current densities.
At 2.9H Na2SO4 feed concentration, the measured energy tactois
were better than the performance objectives by 12 percent and
16 percent at current densities of 160 ASF and 120 A.SF,
respectively.
7.1.3 Cell Voltage
Cell voltage as a function of current density was
determined for the sixteen prototype “A” cells in the test
facility with a Na2SO4 feed concentration of 2.9N and an
operating temperature of 135—140 F. The results are plotted in
Figure 7.1 together with the Contract performance objective curve
for “A 9 cells, taken from Figure 6.4. The plotted cell voltage
is the average voltage of the two electric nodules into which the
sixteen cells are divided in the test tacility.
Figure 7.1 shows that the measured “A” cell voltages are
significantly lower than the target performance objective
voltages at all current densities coverea by the rieasurements.
Also, the figure shows that the voltage—current density curve is
essentially linear in the current density range of practical
interest ( 80 to 160 ASP) with a zero—current intercept of
3.4 volts.
7.2 “ B CELL PERFORMANCE
7.2.1 Cell Voltage
Cell voltage as a function of current density was
determined for the sixteen prototype “B” cells in the test
facility with a Na2SO4 feed concentration of 2N, and an operating
temperature of 135—140 F. The results are plotted in Figure 7.2
together with the Contract performance objective curve for
“B” cells, taken from Figure 64. The plotted cell voltage is
the average voltage of the tw electric modules into which the
sixteen cells are divided in the test facility.
Figure 7.2 shows that the measured “B” cell voltages are
significantly lower than the target performance objective
voltages at all current densities covered by the measurements.
Also, the Figure shows that the voltage—current density curve i
essentially linear in the current density range or practical
interest, i.e., above 80 ASP.
198
-------
10
U)
I-
-J
0
>
U i
-J
0
>
-J
-J
Ui
C-)
CURRENT DENSITY,AMPERES PER SQ. FT
FIGURE 71 CURRENT-VOLTAGE CURVE FOR PROTOTYPE “A”CELLS
9
8
7
6
5
4
3
2
0
TARGET PERFORMANCE OBJECTIVE
FOR “A’CELLS(FROM FIGURE 6.3)
—
-
-
;
L
w
4
— - ______
o 2.9N Na2SO4FEED 1 II-27 —
o 300N Na 2 SO 4 FEED 1 I2-3
A 287N NO 2 SO4FEED,12-4 —
OPERATING TEMPERATURE I35-I4O°F
I I I I I
0 20 40 60 80 100 120 140 160 180
200
1.’- -?
-------
40 60 80 100 120 140
CURRENT DENSITY 1 AMPERES PER SO. F.
FIGURE 7.2
CURRENT-VOLTAGE CURVE FOR PROTOTYPE “ CELLS
8
7
6
C l ,
IiJ
1 .
-J
0
>
w
U
5
4
TARGET PERFORMANCE OBJECTIVE
FOR”B”CELLS(FROMFIGURE 6.3)
‘
2 &P
No 2 $04 FEED CONCENTRATION :2N
OPERATING_TEMPERATURE__135-140°_F
0
0
20
160 80
• ifl
-------
7.2.2 Current Erficiency and Enerqy Consumption
As originally assembled. the current efficiencies of the
prototype “B cells were relatively low due to the presence of
crossleaks.
Corrective measures were taken by hand revising some
components of two cells. Tests with the revised cells showed
improved performance as sixwn in Table 7.3.
The table shows that the measured NaOH energy factors
were within the performance objectives. At current densities of
90 and 125 ASF, the measured energy factors were bettEr than the
performance objectives by 1 and 2 percent, respectively. At
150 ASP, the measured e nergy factors were better than the
performance objectives by an average of 10 percent.
ifl aciclition, the “B anolyte current efficiencies ot
Table 7.3 averaged 42 percent. This also met the performance
objective tor the current efficiency of purge sulfuric acid.
201
-------
Table 7,3
swoaRr OP IERPORMANCt TESTS OP t’aniv “9” LLS
Na 2 SO 4 Catholyte Mid-Anolyte Anolyte Total Cons istency Av. Macc DC Energy Factors
C.E. Feed NaOM N 2 S0 4 I I 50 scid Op. Current Ce ll h/ Ferfornang ,
No. Co c . Conc . C.E. Conc. £ .5. C.E • C . 5. laaolLnI 2 So 4 ) Density Voltage Measured ecttve —
i ir — — _ _ _ _ _ _ _ _ _ _ °j ASP volts 180*/lb 5a80
at 3.1 2.17 84% 0.51 36% 2.19 45% 8 1% 1.04 128 150.0 6.3 2.33 2.59
82 3.3 2.49 80% 0.83 37% 2.34 41% 79% 1.02 134 151.0 6.2 2.44 2.60
83 3.6 2.52 79% 0.60 36% 2.21 42% 77% 1.03 118 125.3 5.9 2.32 2.37
84 3.2 2.45 75% 0.77 40% 2.07 38% 78% 0.96 100 90.2 5.1 2.02 2.04
25 3.5 2.12 80% 0.73 39% 2.13 42% 82% 0.98 123 149.7 6.0 2.19 2.59
86 3.5 2.55 80% 0.74 36% 2.26 43% 79% 1.01 124 150.0 6.1 2.38 2.59
a/ Ratio of total equivalents of Mao)! and 02504 involved in the C.E. Measurement.
Calculated frcst measured caustic production rate, CC busbar current and bus boltage.
/ Iltase I Ferformance jective (3! Na 2 SO 4 feet ) solution, 1400 F operating
tenperature, 85% NaOlI current efficiency) but adjusted d ,nvard in accordance
with the cell voltage—current density relationship for “B . cells given in
Figure 3 to account for the above actual current densities.
-------
PHASE IC
PROTOTYPE PLANT COST ESTIMATE, PR0JEC’r OPERATING
COSTS, AND DESIGN AND CONSTRUCTION SCHEDULE
FOR PRO’flYWPE PLANT
203
-------
I
GENERAL DESCRIPTION OF PROCESS WORK
The SWEC/lonics S02 Removal Unit is designed to continuously
desulfurize boiler stack gas from WEPCO Valley Plant boilers
3 or 4. The process for remval of S02 from stack gases is
a closed system using electrolytic regeneration of basic
process liquids, caustic soda and sulfuric acid. Absorption
of flue gas S02 i s accomplished in a process tower using
recycle caustic soda as the absorption medium and forming
sodium compounds. TOwer bottoms consisting of these sodium
compounds are reacted with dilute recycle sulfuric acid
which releases SO2 and fonus sodium sulfate. S02 is then
stripped from the sodium sulfate in a stripper tower and the
sodium sulfate processed in an electrolytic cell to produce
caustic soda and sulfuric acid. These basic process liquids
are recycled to the ocess streams.
The SWEC/lonics process unit for SO2 removal therefore
consists of an absorption/stripping section and an
electrolytic cell section plus necessary ancillary process
equipments as shown in the Figure I in the Executive
Summary -
Design of a S02 Prototype Removal Unit was based om the
results of the S02 Removal Pilot Plant.
205
-------
II
COST hSTIMATE. FOF .. SWEC/IONICS S02 REMUVAL UNIT FOR WEPCO
75 MW POWER FbP NT
Development of the cost estimatE for the SWEC/lonics
502 Removal Unit for W PCO’s 75 MW Valley Power Plant given
in the Cost Summary Sheets is summarized as follows:
2.1 Design Basis Cost Summary Sheet Dated 8/22/74
This cost represents an optimized chemical process with
modification to front—end process equipments for lowered
costs. The cost estimate, excluding escalation, contingency
and fee is S 14,475,000. This cost provides for an
S02 Removal Unit for a 75 MW Plant which incluaes:
A. A 4250 lb/hr 502 removal capacity from boiler
stack gas
B. A. t -story cell system building
C. Automated cell room syste -a with automatic
control valves
D. S02 drying and storage system
2.2 Design Basis 3100 lb/Hr S02 Removal Unit Cost Summary
Sheet Dated 11/18/74 — Aiternate to 8/22/74 Summary
This cost is for a 3100 lb/hr 502 Removal Unit alternate
S02 Removal Unit with reduced size equipment. The cost
estimate, excluding escalation, contingency and fee is
12,229,563. This cost provides an alternate, smaller S02
Removal Unit. The cost reflects a tightened design basis
for the S02 Removal Unit with cell room design or:
A. 3100 lb/r r S02 regeneration capacity
B. ieduction of capacity of equipment items for
3100 lb/hr design basis
C. Subsitution of semiautomatic system in cell
room requiring greater operator attention
during start—up and shutdown
It should be noted that the cost of the 4250 lb/hr Unit and
the 3100 lb/hr Unit include the costs of:
A. The Q—213 Dilute Acid Storage Tank ($261,720)
( EPC0 holding tank for accumulation for
shipping, or for WEPCO dilute acid use) ana
B. The S02 arying and storage system ( 1,127,0O0).
207
-------
Table 2.1
PROCESS PLANT COST SUMMARY
&5 1S0 is •,P,ONTI
WISCONSIN LI.ECTRTC VALI.EY PLANT, NILW., WISCONSIN
12302.03
DATE
Nrv 8—22.74
--
SO 2 RF2 .IOVAL I ’LANr: SUMIIARY SI0.ET 4250 #lhr Removcl
-m
NO FACTORS
DESCRIPTION ————. MATERIAL LASOR TOTAL
::_ ‘• -
MAN HOURS
PROCESS EDUIPHEST I
I — 544.250: 3,100 547,350 365
Dou. ,Ds SiIFAT(DL I I
[ 1 PROCESS Pu.c.cc; I
GFG ALLCOU , P’ aC ’ .T -________ — 35 ,3flQ 52..flQQ 410.300 5.775
— .2..100..QDO I72 . )0j ,a7j 00 JLU.1_...
. ._! ““ — — 4. .8flp 143.2Q0 55Q _
I a — ._.....25.1450_ 4.QO0 . .2 1_ 50 470
!. — Z. Q ._1L Q00 144 2 .200_.
• COUPSISIOSI ,,C..,,. • L_i 2O , fl .O [ O6po 2.3 O
• STACAS j I
T D C EEC SOCRU 2 18.7 .flO 2.900 211.600 30
— 27 . 60O__4.175. QP_._ . . . . .j9.985
•SOCCSSMATCSPALS I
— — p . ,00 0_ 850,000j ,680 ,000 93,407
— 16Q ,000_ 50.000 2jQ ,000
I . 80 .00O_. . 2 LI q 5 900 56,846
I.!. OIL , 17 ,71’.
S RUOIOTS 375 ,O . Q_L5jQQ__ 1 opp I ,251
I — . —. - - - A..
-- . - -
50 OO0 - 50.000 - —
I IL 2 6 3R,O0fl 555,OO0 229,936
I TOTAL DIRECT COSTIOD 05.5 7.815.900 2.914,600 10,730,500 2(.O,92 1
IT.15 T 1 ACCOUNTS — I
L .t 5’
VA i OQ 186.220
TLAPORADV co ..NTc. cno r CsuricS 2.685
14,900 361,200 3i6 iOO
.1 . . COVSTRUCflOH rOOSS AND rou•p . ,CT.T 421 ,200 94 , 700 5 1 , 900 II • 850
o cross DIs ,s ,su,Arn. . TEd 6,500 551 • 100 557 • 600 15 ,022._
.o , .. S32 ,900 1 • 279 • 600 1 ,812 ,300 ?Q • 535
12,542,800
U INDIRECT ACCOUOYS
, r NC I SCC E I T.r____________________ 109,700 619,700 529 400 3C )O0
I 5Q , —
84,4QQ9 ..k.QQQ173 QO LQ. Q
I ...___..._ Ji.6.70O 6 tj0Q..._________
T DT€SHEA _ E . ._$81.J.O0
TOTAL SN0 ,SE,DEADOUARTESS OPPICE 883.200 1.049.000 1.932.200 96.965
TOTAL PEESEST OAT PRICES . . . .. .9.J32.000 S243 ,000 14.471 .000
I PICALATION I I
TOTAL COSt IICLUOSNO FC C AND CONTISOCNCT I , I
ESTIMATE OF SUSCOMTKUCT COSTS INCLI,OISC AIIOVC I . I
Pea ITEMS INCLUOFOOR CrCLUDCD P 5CM TOE ESIIMAEI SEE OTHER SIDE
208
-------
Table 2,2
i, .o...C? ACCDUNY
rhn.ntEN ,NdI
cE,’oN
ur*Iou . ,, Io ,PIC,
_ L_
I CM.AYPON
Y 01
508.5201 34,500
178.4001 10,Q4
4t,900U
— I
S..
fl!
I.CLUCINC YC WD COP .TI .flVGY
PROCCU PtANT COST SUMMARY
£I S IS (PT)
, ‘ 23O2.O3
V.—
flu ,.,
D TF
VISCON5IN ELEcTRIC POWEl COMPAZI?, MEWAUKEE, wxscoNsrM
. .. . ..
INTCGRATED SO RE1!0V31. 9.ANT: SU) RY SHEET 3100 0/hr Remor,t
NOV 18,1974
8. UAUE*
cu o,I P j ,,4
DESCRIPTION
NO
FACTORS
MATEflIAI.
LAROR
TOTAL
.IAN.HOURS
PUOCESS tUUUPflM?
A TOWp S
544,250
3,100
567 5OI
lu PIOPLCi b NC*TC I
162.800
—
38,600
— —
— —
201,200 4,263
P PO)Cr$SPU NACCI
Q
I. flrACTO U
iIu. .
—
—
—
• I
—
7O ICS CELL_IUSTALLA oN
TO?Au.P OC(5SOUUPN NV 1
P OCIU4 M*Tl M.S
C PuPINC
96,700
1O5 ,550_
81,C00
142.800
2,000
10.500
18,200
—
LO JJ.00 394
107.550 230
91 .500 1.fl3_
161.000 2.092
—
l& SOO 224
3 ,402 ,O’3 -
S/C
182.500
3,402.063
2,000
-
4.717.663 77.600 ‘, ,195,2631 8,800
I
46 0OO 523 9O 96 3 O0OJ 57,470
146 LOO 46,500 193,3001 .d2L
j 7,7OO — 1.457 —
24O 3OO 77 OO__4 , Q
129,500 1 96,9001 226,41)O 1 L6Ol
I .TRUCTU . .$_______________
• ,.cc,uc.,.
‘ 7E
. .u ,o. ,. ,..
CuvH.
K •PI TflUPNV
.
53.1OO’ )0
S.. 1 .Z. .!..
N IPPI L IIOI4 I I 304 • IO U I JU’I. IOU I
•PITA PWOCCIS M TCNl l.1
VI
IOT S. OIVCT COlT io.i’,OI .I
87tW36T i1O37 .U
DI TIIu4UT*SLE •CCOUNfl
—
.
‘ ‘ ‘
.r .-fl I —
I
J, 3JZ,IUU
P POsPLI. £ ‘IO STATITAUNI
I
V
1 I
CONITIUCTIO ’I TOOLS ANO IOUIPMSNT
103. L 1
t.
37, 100
I?NUCTPONP.CILITIIS
71,Z0
18 , 200
- W3 .iUU
yeT. .I.T.I.uy..tc - -
0
OTlIPI Ou$IUPPU?*Ul.I jIM.
5 ,300
368
14,900
LUU •
I ’
334, 200
61,600
i’i9, WV I
3,4)3,( .b3
1,9b3,600
343,300
300,3001 1,100
9Th ,UUU
16.370
- 26,355
OULflHL*S ALLOWANCI
109,700
398,820
.1
86.400
TOTM INOmIC ? (MSOOUANVCIP OP ?PCCI
94.000
59.600 —
000
9. 2M . 663
9/6,300
2,939,900
1,817.,
- -
°°! 90,165
12,229
P.. ITIuuN NCLUDCDON i Ci.UOtO PNON IMP CSTIU TC. III O•Ntfl suol
209
-------
III
PROJECTk D MONTHLY OPERATING COSTS
A cost estimate was generated for monthly operating
costs for the first year of Phase III operation of the
75 MW Demonstration (Prototype) Plant. Both SWEC and
WEPCO labor will be used to conduct tests througbout the
year • Graduate Engineers including a supervisor will
constitute the SWEC labor requirement. Test equipment
and material costs have been included in tne estimate
along with the utilities and needed chemicals.
The SWEC/lonics process generates ailute sulfuric acid
and hydrogen as well as the S02 removed from the stack
gasses. These streams, and return of condensate used in
the stripper reboiler to the power plant, have been
considered to have an economic value favorable to the
SWEC/lonics process. A credit for these streams has
therefore been factored into the operating and
maintenance cost estimate. The total monthly operating
and maintenance costs for each month of the first yearss
operation is tabulated in Table 3.1 which shows that
when the year test period is concluded and SWEC
test/operating labor are no longer required, the
S02 Removal Unit can provide an income as a result of
the SWEL/lonics process.
211
-------
TABLE 3.1
ESTIMATE OF OPERP.T1NG AND MAINTENANCE COSTS - 75 MW DEMONSTRATION PLANT
FOR 3,100 LB/HR S02 REMOVAL (NOTE 1 )
DRIYPIPPYPR STAMP OPER 55 ION PHASE ITT
Typical for
Typical for
Typical for
hach Month
Each Month
Each Month
January February (Mar.Apr,May)
(June ,July ,Aug)
(Sept,Oct ,Nov) Decembur
$ 53,375.00 S 53,375.00
8,000.00 8,000.00
$ 53,375.00
4,000.00
5 53,375.00
£4,000.00
19,150.00 19,150.00 19,150.00
24,167.00 24,167.00 24,167.00
5 45,750.00
11,000.00
Months:
OPERATING LABOR
Supervisory Personnel (S6W)
(Note 2)
WEPCO Operators at $148,000 / rr,
3 Lhifta (1 Man/Shift)
MAINTENANCE
For SEW Section at 4% of Fixed
Capital Investment (Note 3)
For tonics Section (Note 4)
UTILITIES
Water
CooLing 1124 Ga]. at
50.116/1, 000
Deionized MM Gal at
$1. 00/1 .000
Steam Tens at 60.67/1,000 lb
(Note 5)
Electricity
Cells 4,500 xv at 80.00781/
Xwhr
Other Equipment 1,433 Ew at
80.0078 1/Xwhr
Teats and Inspection at
$0.50/Ion SO2
PLANT ADMINISTRATION (Note 6)
CHEMICALS
50% (Vt) Caustic Soda at $85/Ion
93% (Vt) Sulfuric Acid at
$33/ron
35 vt% Hydrogen Peroxide at
$300/Ion (Note 7)
Filter Aid Material at $135/N?
ESTIMATEN CRNDII’S
Condensate from Stripper Reboiler
(Note 8)
Dilute Sulfuric Acid (Note 9)
Hydrogen (Heating Value) (Note 10)
Sale of Liquid 502 (Note 11)
TOTAL MONTBLY OPERATING AND
MAINTENANCE COSTS
19,150.00 19,150.00
24,167.00 244,167.00
$ 53,375.00
8,000.00
19,150.00
24,167.00
6 • 8124/
$789.00
508 • 00
1,640.00
10,250.00
6,528.00
225.00
20,000.00
11 lbns/
$935.00
7.2 Tons,
$238.00
551.00
123.00
140.00)
3,625.00)
990.00)
24,989.00)
29,774.00)
144.9/
$1,728.00
952.00
3,279.00
20,501.00
6,528.00
450.00
20,000.00
8/11680.00
a. 8/6 158. 00
762 • 00
170.00
111 • 9/
$1,728.00
952.00
3,279.00
20,501.00
6,528.00
450 • 00
20,000.00
20.1/
62,332.00
952.00
3,279.00
20,501.00
6,528.00
450.00
20,000.00
8/6680.00
4 . 8/5 158 .00
762.00
170.00
10.2/
$1,183.00
762.00
2,460 .00
15,354.00
6,528.00
337.00
20,000.00
7/8595.00
4.2/5139.00
592.00
132.00
200.00)
5,438.00)
1,485.00)
37,483.00)
£414,606.00)
13.6/
$1,578.00
952.00
3,279.00
20,501.00
6,528.00
450.00
20,000.00
8/$680.00 8/1680.00
4.8/5158.00 4.8/5158.00
762.00
170 • 00
280.00) ( 280.00) ( 280.00)
7,250.00) ( 7,250.00) ( 7,250.00)
1,979.00) ( 1,979.00) ( 1,979.00)
49,952.00) 1 149,952.00) ( 449,952.00)
59,461.00) ( 59,461.00) ( 59,461.00)
TYPICAL NORMAl. OPER?.PION (.641 COSTA5 NTH
(With Credits and Less SEW Test Labor)
897,043.00 — $53,375.00 543.668.00
See tootnotes next page.
762.00
170.00
280.00)
7,250.00)
1,979.00)
49,952.00)
59,461.00)
(Less Credits)
(With Credits)
146,479.00
116,735.00
152,774.00
108,168.00
159,900.00
100,439.00
156,5014.00
97,043.00
155,900.00 148,125.00
96,439.00 88,664.00
212
-------
ot.s to Table 3.1
1) Costs predicated on 3,100 lb/hr SO, removal on continusits basis.
Capability for 4,250 lb/hr SO rdoval peaking capacity for 33 hr
is incorporated into p1 t dJign. Second quart.r 1974 piLe...
2) C.stsinclud.salaryandperdi.afora3Oandl/2daymonth,
7 day p.r v.ek operation. One ohi.f op.rator, four operating
advisor. (four man cov.ring 24 hours), one chamist, and on.
proo.ss engf.n.er. Two w 1 14 ti”n.i man—months should be inolud.d
for report writing for month. 13 and 14.
3) Coats predicated on basis of 3,100 lb/hr SO removal ca ta1 costs
estimat. and axoludes escalation, contingeJy, and fee.
4) Monthly costs based’ on cell maintenance materials yearly cost of
$2lO plus cell system maintenance of $80,000 p.r year.
5) Cost i. for steam rate of 8,370 lb/hr for an average 583 hr month
(2,447 tons per month), ixoept for start-up in months 1 and 2.
6) Plant taxes end pl.nt insurance ar. not included du. to diversity
in p’ant location, and aoao uitiug methods.
7) Based on yearly cost of $9,150 for 30 and 1/2 tons at $300/ton.
8) Credit baud on return of condensat, t. power pl-”t at rate of
approximately 8 gpo and cost of $1.00 p.r 1,000 gal. (Total
$3, 320/year)
9) Credit for use of dilute ‘ 1 °L by power plant for treating d 4eralisers
on 100% basis. Rate of 0.356”tons/hr (100% basis) at $35/ton.
total $87,220/year).
10. Credit for use of H 2 as fuel (heating vain.). 3.9 M Btu/hr,
$O.87/ O( Btu (LET) (Total $24,COO/y.ar)
1].. Credit for the ..l. of SO 2 . Credit based on 2,635 lb/hr x 7,000 =
9,223 tons/year. Ourrent market pric, for SD, i. $110/ton.
Arbitrary conservative price of $65/ton used for this estimate.
213
-------
Iv
DESIGN, ENGINEERING, AND CONSTRUCTION SCHEDULE FOR WEPCO
75 MW UNIT
The proposed schedule for installing the 75 MW
demonstration at the WEPCO Valley Station site appears
as Figure 4.1. The schedule extends over the estimated
28 months required to reach mechanical completion from
the date of Contract Award. Upon reaching mechanical
completion, the Operating and Test Program described
under Phase ID would begin.
215
-------
Figure 4.1
Q CNT WI SCO1.J5It.J t LECT IC POwER
‘YPE O UNIT 75 MW DrM P.1 T ATION LAt4T
LocA-ruoN MILW \IJ E WI.JSi .4
ØONTl S /IFT ‘
O..TQAeT &WAOO
ZF I4ELC .iC
Pt t
I
L.. - - 4-’ -
: ;; I I JI
• •• •I
p. -
14. .
rLL .-L
-I--4
DL Q 2TICN OF wo 1 TT’T. t.
PROPOSAL 5( III )ULE
•10 . • Wt C COOO• T .Of
r —
J_I L
LE ,ENO: X-D4Qu’Qf Jb ’. NO. t2302.OS
Q - QUOTE S
p -
- DEUV I 1 ii SiTE
DArE JUi .E IO)I 7L
TrLJImfl I1
I—
H
a’
[ tf
1 ±iTflTTt t .
1-’ T
PLOT rt N L 1 1 fy
J I EF G_ [ .j I J j
ii i_ iiti I
Pf..1LiOWE J
— I—, I
KS 99 r4 ±i
r’c -’IaL EQUIP - :—TT’ —
E’ .t L.
-. 1. jdT4
iT LC P( cr - r — -. - —
Li
F( .J ..C.VDN
I II E.(ECT
-
E M_L TZ’ EQUI?(IN-EL’Xi
-
LLECT : C uiP (oTHE
‘ J ELECT ,I45T u..ETC1 - I -
F [ U 1TITI.T T 11
It h*j2 L 1 L14 ’ 1
tLLLiH i - -; i: i11
‘ TfEITfIJTf1I 1 : 1 JT 1
111 1 1 I 1 1 1 j T rr r’
- - I 1-f-I A Iti -
I -h - J j tkh ±
I 4}
: 1 ]J: t :
Li : Ijfri 4 i # + 1 4 4f4fl
:t
- - - -I- I
1t
-
. C •,I-Lr
:t
[ 11 I1]
I
-
-------
PHASE ID
TEST PROGRAM AND OPERATING SCHEDULE
FOR 75 MW PROTOTYPE PLANT
217
-------
I
INTRODUCTION
This test program and operating schedule is intended as a
supplement to the Test Program and Operating Schedule Manual
prepared for the Pilot Plant Test Program under EPA contract
No. 68—02—0297. This schedule specifies the differences between
the 75 Mw Demonstration Test Program and the Pilot Plant Test
Program.
During the Pilot Plant Program, it was possible to vary some
operating conditions in the absorption section. In the
demonstration plant, it will not be possible to vary absorption
conditions in a predetermined manner. The demonstration plant
operating conditions will largely be determined by the prevailing
operation of WEPCO No. 4 boiler.
The Test Program for the demonstration plant will necessarily be
limited in scope. The Test Program is primarily directed toward
the evaluation of process technical performance on a larger
scale, determination of process economics, process availability,
and maintainability.
219
-------
II
TEST PROGRAM OBJECTIVES
A summary of the major test program objectives, excluding the
Electrolytic Section, is presented in Table 2.1. The test
program for the Electrolytic Section will be similar to the
program as detailed in the Pilot Plant Test Program and Operating
Schedule Manual.
2 • 1 PROCESS PERFORMANCE
It will be important to establish the process material balance at
an early date. This will help to identify and quantify system
chemical losses.
This time will also allow for:
1. Checking accuracy of Instrumentation
2. Familiarization with analytical procedures
3. Technical personnel training period
2.2 AbSORPTION ShCTION
Pilot plant operation indicated a problem with closure of the
system water balance. One explanation for the lack of closure
may have been that the inlet flue gas was not completely
saturated in the quench section. Should a similar conditior.
occur in the demonstration plant, it is planned to take humidity
measurements of the gas before and after the quench section.
Entrainment, if any, of process liquor in the absorber overhead
will be more closely followed. It is planned to follow the same
measurement technique as used by EPA in the pilot plant
determination.
Oxidation of S02 in the absorber was found to be relatively
insensitive to those variables studied in the pilot plant, except
for stage recirculation rates. The demonstration plant will have
less holdup than the pilot plant; i.e., the packing volume per
cubic foot of gas will be 50 percent of that used in the pilot
plant.
In the demonstration plant, it is proposed to explore more
closely:
1. The effect of NOx on oxidation
2. The effect of heavy metal cations on oxidation
3. The amount of S03 contained in the gas that
contributes to the overall oxidation
14• The effect of flue gas 02 concentration
5. The NOx removal capabilities of the system
221
-------
TABLE 2.1
TEST PROGRAII OBJECTIVES
TASK
I. Establish process material balance
and identify process chemical losses
k REQUIRED
1. Flue gas flow
2 • 502 concentration to and from absorber
3. Recycle acid—caustic flows and compositions
4. Process liquor losses and concentrations
5. Quantity and composition of recovered S02
6. Tank inventories
7. Amount of oxidation
8. Weak acid production rate and concentration
XX. Absorption Section
1. Establish water balance on scrubber
2 • Measure entrainment from scrubber
3. Measure S02 removal efficiences
4 • Measure flue gas system pressure drops
5 • Demonstrate feed-forward control system on
caustic flow for maintaining desired 502
exit concentration.
6. Oxidation
a. Determine amount of S02 removal and
oxidation in each packing stage
b. Monitor concentration of trace heavy
metals in feed/effluent of absorber
a. Monitor ?10x to and from scrubber
d. S0302 concentrations in feed gas
7 • Determine effectiveness of hydroclones for
removing particulate matter from quench water
XIX. Stripping-Recovery Sections
1 • Determine minimum stripping steam rate
vs S02 concentration and pif in bottoms
1. Gas moisture content before and after
quench section, gas temperatures
2 • As defined by EPA procedure
6.
a. S02 concentrations in gas phase
to and from each stage
Liquor analyses to and from each stage
b. Appropriate liquor samples for
Cu+2, Fe+2, Al+3
7. Solids concentration of inlet/outlet streams
1. Steam, reflux and feed rates
concentration of S02 and H2S04
feed liquor composition
pH of bottoms stream temperatures
XV. Feed Liquor Treathent
1 • Establish optimum peroxide consumption
2 • Establish cycle tii m, pre—coat type and
quantity, body aid feed quantity for
removal of Fe/Al hydroxides.
V. Process Characterization Factors
1. Operator requirements and costs
2 • Maintenance requirements and costs
3 • Establish utilities and chemical
make—up recuirements.
1. S02 and Fe concentrations
222
-------
2 • 3 STRIPPING SECTION
The most important variable in this section is the minimum amount
of steam necessary for a given concentration of S02 in the
stripped cell feed liquor.
As indicated during the pilot plant operation, the remaining S02
in the stripped solution must be oxidized with hydrogen peroxide.
If this is not done, divalent iron will not be oxidized to
trivalent iron for subsequent removal as the hydrous oxide.
Complete iron remo ral is necessary to maint in satisfactory cell
operation.
Minimum steam consumption versus S02 concentration in the cell
teed liquor will be measured. The economic optimum S02
concentration in the stripped solution will then be determined.
2.14 FEED LIQUOR TREA TMENT
Iron and aluminum will be removed as the hydrous metal oxides on
a pressure leaf, precoat type filter. Cycle times, solids
loading, body and feed concentration, and the specific flow rate
will be optimized based on the field operation.
Selection of the proper filter aid will be determined by field
testing.
2.5 DATA COLLECTION AND EVALUATION
2 • 5 • 1 Technical Program
Data collection in the demonstration plant operation will be
similar to the pilot plant operation.
Additional effort will be placed on obtaining more data in the
following areas:
1. Sodium/sulfate ratios in absorber
feed and draw liquors
2. 502 gas phase inlet/outlet and interstage
concentrations
3. S03 concentration in flue gas
14• I bsorber entrainment
Determination of the height of a transfer unit in the absorber
bottom stage will be done by integrating the mass trQnsfer
equation. The equilibrium curve—operating lines are not straight
over the concentration range in the bottom section. In the
second stage. the mass transfer coefficient may be found more
simply by assuming linear relations for the equilibrium—operating
lines.
223
-------
As was done in the evaluation of pilot plant data, we shall use
the Stone g Webster Regression Analysis Program for data fitting.
A multiple regression analysis routine is also available.
Normal data evaluation will specify a 95 percent confidence belt,
unless it is decided to broaden or narrow the limits after
analyzing the data.
2.5.2 Process Maintenance
The following information will be recorded in connection with
maintenance requirmuents:
1 • Number and types of WEPCO personnel
2. 3ubcontractor labor
3. Materials cost
4. Man—hours by craft category and supervisory
requirement
2.5.3 Process Operating Costs
Detailed records shall be maintained to properly record all
operating costs • Areas that are to be monitored are:
1. Raw materials — caustic soda, sulfuric acid,
filter aid
2. Productive labor — operating personnel
3. Z ianufacturing expenses — water, steam,
electricity
4. Other expenses — waste disposal, tests and
inspections, plant transportation, and taxes
22.4
-------
TECHNICAL REPORT DATA
fMca:e rr.d INwuctioni on the reutu before completb ,zJ
1. REPORT NO. 2.
EPA-650/2-75-045
I. RECIPIENTS ACCUIIOINO.
4. TITLE AND SUBTITLE
tone &Webster/Ionics 502 Removal and Recovery
Process--Phase I, Final Report
I. REPORT DATE
ay 1975
i PERFORMING ORGANIZATION CODE
7. AuTHOR(S) Tonics, Inc. (Watertown, Mass.) and
Stone &Webster Engineering Corp. (Boston, Mass.)
I. PERFORMING ORGANIZATION
I. PERFORMING OR6ANIZATION NAME AND ADDRESS
Wisconsin Electric Power Company
231 West Michigan
Milwaukee, Wis cons in 53201
10. PROGRAM ELEMENT NO.
lABOl3; ROAP 2IACX-082
11. cONTRACT/GRANT NO.
68-02-0297
12. SPONSORING AGENCY NAME AND ADDRESS 13. TYPE OF REPORT AND PERIOD COVERED
EPA, Office of Research and Development - 12/74
NERC-RTP, Control S 7stems Laboratory
Research Triangle Park, NC 27711
IS. SUPPLEMENTARY NOTES
‘ “ ‘ ne report covers Phase I of a potential three-phase program to evaluate
the Stone & Webster/lonics process at 1 MW pilot plant scale with the option to scale
up and demonstrate process viability at the 75 MW prototype level. The report cites
the objectives, approach, results, and conclusions, and discusses a program that
included: the design, construction, and operation of, and completion of a test program
for, the pilot plant; the design, construction, and testing of prototype-size electro-
lytic regeneration cells; the design, engineering, and estimation of construction and
operating costs of the 75 MW prototype; and preparation of a test program and oper-
ating schedule for the prototype. An executive summary Includes the background and
objectives of the overall program and pilot-scale effort, and highlights significant
results and conclusions. Although technical feasibility was demonstrated at the pilot
scale, the economics of a 75 MW prototype plant at the site of the pilot plant do not
appear favorable. There are no current plans to continue into Phase II (detailed
design, procurement, and installation of the 75 MW prototype) or Phase III (12-
month start-up and operational test period for the 75 MW prototype).
7. KEY WORDS AND DOCUMENT ANALYSIS
I. DESCRIPTORS
b.IDENTIFIEP’IS/OPEN ENDED TERMS
C. COSATI FteldlGrovp
Air Pollution Scrubbers
Combustion Desulfurization
Flue Gases Sulfuric Acid
Sulfur Oxides
Electric Power
Plants
Sodium Hydroxide
Air Pollution Control
Stationary Sources
Stone & Webster/Tonics
Process
Electrolytic Regener-
ation
13B 07A
21B 07D
07B
lOB
18. DISTRIBUTION STATEMENT
Unlimited
19. SECURITY CLASS fThliReponf
Unclassified
21. NO. OF PAGES
257
20. SECURITY CLASS (ThIs pop)
Unclassified
22. PRICE
EPA Form 2220.1 (5.73) - —
225
------- |