U.S. Environmental Protection Agency Industrial Environmental Research EPA~600/7~77~050C
Office of Research and Development Laboratory riTT
Research Triangle Park. North Carolina 27711 May 1977
FINAL REPORT: DUAL ALKALI TEST
AND EVALUATION PROGRAM
Volume III. Prototype Test Program-
Plant Scholz
Interagency
Energy-Environment
Research and Development
Program Report
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S.
Environmental Protection Agency, have been grouped into seven series.
These seven broad categories were established to facilitate further
development and application of environmental technology. Elimination
of traditional grouping was consciously planned to foster technology
transfer and a maximum interface in related fields. The seven series
are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
. 5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
This report has been assigned to the INTERAGENCY ENERGY-ENVIRONMENT
RESEARCH AND DEVELOPMENT series. Reports in this series result from
the effort funded under the 17-agency Federal Energy/Environment
Research and Development Program. These studies relate to EPA's
mission to protect the public health and welfare from adverse effects
of pollutants associated with energy systems. The goal of the Program
is to assure the rapid development of domestic energy supplies in an
environmentally—compatible manner by providing the necessary
environmental data and control technology. Investigations include
analyses of the transport of energy-related pollutants and their health
and ecological effects; assessments of, and development of, control
technologies for energy systems; and integrated assessments of a wide
range of energy-related environmental issues.
REVIEW NOTICE
This report has been reviewed by the participating Federal
Agencies , and approved for publication. Approval does not
signify that the contents necessarily reflect the views and
policies of the Government, nor does mention of trade names
or commercial products constitute endorsement or recommen-
dation for use.
This document is available to the public through the National Technical
Information Service, Springfield, Virginia 22161.
-------
EPA-600/7-77-050C
May 1977
FINAL REPORT: DUAL ALKALI TEST
AND EVALUATION PROGRAM
Volume III. Prototype Test Program-
Plant Scholz
by
C.R. LaMantia, R.R. Lunt, J.E. Oberholtzer,
E.L Field, and J.R. Valentine
Arthur D. Little, Inc.
Acorn Park
Cambridge, Massachusetts 02140
Contract No. 68-02-1071
Program Element No. EHE624
EPA Project Officer: Norman Kaplan
Industrial Environmental Research Laboratory
Office of Energy, Minerals, and Industry
Research Triangle Park, N.C. 27711
Prepared for
U.S. ENVIROMENTAL PROTECTION AGENCY
Office of Research and Development
Washington, D.C. 20460
-------
ii
-------
ABSTRACT
This report presents the results of the Dual Alkali Program
conducted by Arthur D. Little, Inc. (ADL) for the Industrial
Environmental Research Laboratory, Research Triangle Park,
(IERL, RTF) of the U.S. Environmental Protection Agency (EPA).
The purpose of the program was to investigate, characterize
and evaluate the basic process chemistry and the various
modes of operation of sodium-based dual alkali processes.
The work was carried out at three levels of investigation:
• Task I - Laboratory studies at ADL and IERL, RTP.
• Task II - Pilot Plant Operations in a 1,200 scfm
system at ADL.
• Task III - Prototype Test Program on a 20-megawatt
Combustion Equipment Associates (CEA)/
ADL dual alkali system at Plant Scholz,
Southern Company Services, Inc./Gulf
Power Company.
Various modes of operating dual alkali systems on high- and
low-sulfur fuel applications were investigated, including:
• Concentrated and dilute sodium scrubbing systems
• Lime and limestone regeneration
• Slipstream sulfate treatment schemes.
In each mode, the objective was to characterize the dual alkali
process in terms of S02 removal., chemical consumption, oxidation,
sulfate precipitation and control, waste solids characteristics
and soluble solids losses.
This is Volume III of the final report covering the Prototype Test
Program, Task III. Volume I is the Executive Summary; Volume II
covers Tasks I and II, the Laboratory and Pilot Plant Programs.
iii
-------
iv
-------
VOLUME III
TASK III, PROTOTYPE TEST PROGRAM
TABLE OF CONTENTS
Page
Chapter No.
LIST OF FIGURES ix
LIST OF TABLES x
ABSTRACT iii
ACKNOWLEDGEMENTS xiii
APPLICABLE CONVERSION FACTORS
ENGLISH TO METRIC UNITS xv
I SUMMARY 1-1
A. PURPOSE AND SCOPE 1-1
B. PROGRAM DESCRIPTION 1-2
1. System Design 1-2
2. System Operation 1-2
C. SYSTEM PERFORMANCE 1-3
1. S02 Removal 1-3
2. Particulate Removal 1-4
3. Lime Utilization 1-4
4. Oxidation/Sulfate Control 1-4
5. Waste Cake Properties 1-6
6. Sodium Makeup 1-7
7. Power Consumption 1-7
8. Operability/Reliability Potential 1-7
II INTRODUCTION II-l
III SYSTEM DESCRIPTION III-l
A. PROCESS CHEMISTRY III-l
B. SYSTEM CONFIGURATION III-2
1. Flue Gas Scrubbing III-5
2. Absorbent Regeneration III-8
3. Solid/Liquid Separation and Solids
Dewatering 111-10
v
-------
TABLE OF CONTENTS (cont)
Page
Chapter J^
IV OPERATING HISTORY IV~1
A. GENERAL OPERATING CONDITIONS IV~1
B. DESCRIPTION OF OPERATING PERIODS IV~3
1. Operating Period 1 — Startup and
Initial Operations IV~3
2. Operating Period 2 — Low- to
Medium-Sulfur Coal Operation IV-5
3. Operating Period 3 — High-Sulfur
Coal Operation IV-7
V SYSTEM PERFORMANCE V-l
A. AVAILABILITY V-l
B. PROCESS PERFORMANCE ~ STABLE LOAD
OPERATION V-2
1. Overall Operation V-2
2. S02 Removal V-10
3. Lime Utilization V-15
4. Oxidation and Sulfate Control V-17
5. Waste Cake Properties V-25
6. Sodium Makeup V-31
7. Process Operability V-33
C. PROCESS PERFORMANCE — FLUCTUATING LOAD V-35
1. S02 Removal - V-39
2. System Chemistry (Sulfite Oxidation) V-39
3. Lime Utilization V-39
4. Waste Solids Properties V-40
5. System Operability V-40
D. PROCESS PERFORMANCE ~ PARTICULATE TESTING V-40
1. Particulate Removal V-43
2. S02 Removal V-48
vi
-------
TABLE OF CONTENTS (cont)
Chapter
V (cont)
VI
VII
3. Chloride Removal
4. Lime Utilization
5. System Chemistry (Oxidation
and Sulfate Control)
6. Waste Cake Properties
7. Liquor Entrainment
8. Overall Operability
E. MECHANICAL PERFORMANCE
1. Equipment
2. Instrumentation
REFERENCES
GLOSSARY
APPENDICES
Appendix A - Daily Coal Analyses
Appendix B - Process and Boiler Outages
Appendix C - Process Operating Data
Appendix D - Typical Process Flows
and Stream Compositions
Appendix E - Solution Composition/pH
Correlations
Appendix F - Scrubber System Efficiency
Estimates
Appendix G - Equipment Problems
Page
No.
V-48
V-48
V-49
V-49
V-50
V-50
V-50
V-52
V-56
VI-1
VII-1
A-l
B-l
C-l
D-l
E-l
F-l
G-l
vii
-------
vili
-------
VOLUME III
LIST OF FIGURES
Figure Page
No. No.
HI-1 CEA/ADL Dual Alkali System at the Scholz
Steam Plant - Process Flow Diagram III-4
III-2 Gas Scrubbing System III-6
III-3 Absorbent Regeneration and Chemical Makeup System III-9
III-4 Solids Dewatering System III-ll
V-l Availability of the CEA/ADL Dual Alkali System
at Scholz VI-3
V-2 S02 Removal as a Function of pH - Low/Medium-
Sulfur Coal V-ll
V-3 S02 Removal as a. Function of pH - High-Sulfur Coal V-12
V-4 Oxidation in the Scrubber System as a Function of
Flue Gas Oxygen Content (Periods 1 and 2) V-19
V-5 Oxidation in the Scrubber System as a Function of
Flue Gas Oxygen Content (Period 3) V-20
V-6 (CaSOit/CaSOs) Ratio in Waste Solids as a Function
of (Na2SO^/Na2S03> Ratio in Regenerated Liquor V-22
V-7 Filter Station - Showing Filter Cake in Storage/
Transfer Bin Below Filter V-26
V-8 Loading Filter Cake for Transport to Disposal Pond V-27
V-9 Gas Load Schedule for Fluctuating Load Testing V-36
V-10 Schematic of the Electrostatic Precipitator on
Boiler #1 at Scholz Steam Plant V-41
V-ll Schedule of Precipitator Operation and Scrubber
Tests V-44
V-12 Liquor Entrainment vs Gas Flow Measured by York
Research (6/76) V-51
IX
-------
VOLUME III
LIST OF TABLES
Table Pa§e
No. No-
III-l Design Basis III-l
IV-1 Summary of Operating Periods TV-2
IV-2 Summary of System Inlet Conditions -
Operating Period 1 (2/75-7/75) IV-4
IV-3 Summary of System Inlet Conditions -
Operating Period 2 (9/75-1/76) IV-6
IV-4 Summary of System Inlet Conditions -
Operating Period 3 (3/76-7/76) IV-8
V-l Summary of Prototype System Availability V-4
V-2 Summary of System Operating Conditions -
Period 1 (3/75-6/75) V-5
V-3 Summary of System Operating Conditions -
Period 2 (10/75-12/75) V-6
V-4 Summary of System Operating Conditions -
Period 3, Stable Load Testing (4/76-5/76) V-7
V-5 Summary of Overall System Performance
During Stable Load Testing (4/75-6/76) V-9
V-6 S02 Removal Performance During Stable
Load Operation V-14
V-7 Lime Utilization During Stable Load Operation V-16
V-8 System Sulfate Balances V-24
V-9 Summary of Filter Cake Properties V-28
V-10 Summary of System Operating Conditions
Fluctuating Load Testing V-37
V-ll Comparative Performance Characteristics
Stable vs Fluctuating Load V-38
-------
LIST OF TABLES (cont)
Table Page
No. No.
V-12 Summary of System Operating Conditions
Particulate Testing (6/15-7/1/76) V-42
V-13 Summary of Scrubber Test Conditions V-45
V-14 Summary of System Performance V-46
V-15 Summary of Particulate Test Results V-47
xi
-------
xli
-------
ACKNOWLEDGEMENTS
The work under this program was performed over a four-year period from
May 1973 through May 1977, with contributions from many individuals
representing several organizations. Persons involved at Arthur D. Little,
Inc. were:
Principal Investigators
Charles R. LaMantia - Project Manager
Richard R. Lunt - Pilot Plant and Prototype Program Manager
James E. Oberholtzer - Laboratory Program Manager
Edwin L. Field - Data Analysis Manager
James R. Valentine - Chemical Analysis Manager
Contributing Staff
Itamar Bodek
Lawrance I. Damokosh
Bruce E. Goodwin
George E. Hutchinson
Michael lovine
Bernard Jackson
Indrakumar Jashnani
C. Lembit Kusik
Stephen P. Spellenberg
Robert A. Swanbon
Frank J. Tremblay
Lawrence R. Woodland
The EPA Project Officer for the entire four-year progam, Norman Kaplan,
made continuing and important technical and management contributions to
the program. Michael MaKwell and Frank Princiotta at EPA, through their
involvement in the review and planning, helped to guide the program over
the four-year period. The earlier part of the EPA laboratory program
was conducted under the direction of Dean Draemel, now at Exxon. EPA
laboratory work was carried on and completed by James MacQueen and Robert
Opferkuch of Monsanto Research Corporation under contract to EPA.
The cooperation and important contributions and support of Gulf Power
Company and Southern Company Services, Inc. (SCS) to the prototype test
program were invaluable. Randall Rush, responsible for coordination of
the program at SCS, made important technical contributions to the test
program and to the preparation of this report, in addition to this con-
tinuing support throughout the program; the value of Mr. Rush's dedica-
tion and commitment cannot be overstated. In addition, we would like to
thank Reed Edwards of SCS and James Kelly of Gulf Power for their on-site
xiii
-------
assistance in the operation of the system. We wish to acknowledge the
cooperation of George Layman of Gulf Power and William Harrison of SCS,
individually and as representatives of their organizations, in making
the prototype system available and for the operation and maintenance of
the system during the program.
The cooperation, support and contributions of Combustion Equipment
Associates, Inc. (CEA) and its personnel were important to both the
pilot plant and prototype test programs. With the cooperation of CEA,
both systems were made available to the program. Tom Frank, the CEA
Project Manager for prototype system, and Richard White, on-site for
maintenance and operations, were importantly involved in the prototype
test program. The cooperation of Richard Sommer is gratefully acknowl-
edged for CEA's participation and support in this program.
xiv
-------
APPLICABLE CONVERSION FACTORS
ENGLISH TO METRIC UNITS
British
5/9 (°F-32)
1 ft
1 ft2
1 ft3
1 grain
1 in.
1 in*
1 in3
1 lb (avoir.)
1 ton (long)
1 ton (short)
1 gal
1 Btu
Metric
°C
0.3048 meter
0.0929 meters2
0.0283 meters3
0.0648 gram
2.54 centimeters
6.452 centimeters2
16.39 centimeters3
0.4536 kilogram
1.0160 metric tons
0.9072 metric tons
3.7853 liters
252 calories
xv
-------
I. SUMMARY
A. PURPOSE AND SCOPE
This report presents the final results of the Dual Alkali Test Program
conducted on the 20-megawatt (nominal capacity) dual alkali prototype
system at Gulf Power Company's Scholz Steam Plant in Sneads, Florida.
The prototype system was designed and built by Combustion Equipment
Associates, Inc. (CEA) and Arthur D. Little, Inc. (ADL) for Southern
Company Services, Inc. (SCS) and the Gulf Power Company. The test program
represents Task III of the Dual Alkali Program conducted by ADL for the
Industrial Environmental Research Laboratory (IERL) of the U.S. Environ-
mental Protection Agency (EPA).
The purpose of the test program was to characterize and evaluate the
performance of the dual alkali process operating in a concentrated
active sodium mode with lime regeneration. The formal test program
lasted a total of about 14 months and covered a variety of conditions,
including operating with low-, medium-, and high-sulfur coal. The effects
of both fluctuating gas loads and simultaneous particulate removal were
also tested in conjunction with high-sulfur coal operation.
The operation of the system was evaluated with regard to the following
performance characteristics used to evaluate all dual alkali modes:
• S02 removal efficiency;
• lime utilization;
• oxidation/sulfate control;
• waste solids properties;
• sodium makeup requirements and degree of closed-loop operations; and
• overall system operability and reliability potential.
While the overall operability and reliability were a principal concern,
the system was not intended to be a-demonstration unit to test the ulti-
mate availability of such systems in full-scale applications. The test
program was focused on evaluating the viability of the process technology
and defining process capabilities and limitations. The process reliability
and operability were, therefore, of importance primarily as they reflected
process chemistry and operational problems related to process chemistry.
1-1
-------
B. PROGRAM DESCRIPTION
1. System Design
The 20-megawatt prototype system was installed on Unit No. 1, a 40-mega-
watt (nominal capacity) Babcock and Wilcox pulverized-coal-fired power
boiler. The boiler is equipped with a high-efficiency, sectionalized
electrostatic precipitator. The system consisted of basically three
process sections: scrubbing; absorbent regeneration; and waste solids
dewatering. The scrubbing system contained a venturi followed by an
absorption tower with two trays and a demister. The scrubber system
was designed with the flexibility of operating either in a direct lime
or limestone scrubbing mode as well as dual alkali. The venturi was
included for testing simultaneous particulate and S02 removal. Modi-
fications to the scrubber system following startup of the system pro-
vided for operation of the venturi alone by bypassing regenerated liquor
around the absorber.
The regeneration system consisted of the CEA/ADL two-stage reactor system.
Provisions were made for feeding dry or slurried hydrated lime to either
or both reactors.
The waste solids dewatering system consisted of a thickener and a single
rotary drum vacuum filter equipped with wash sprays. The thickener was
sized to handle 40 megawatts of capacity in contrast to the scrubbers,
reactor, and filter, which were designed for 20 megawatts.
The system was designed to operate in the concentrated active sodium mode
on medium- and high-sulfur coal. In this mode, sulfate removal cannot be
accomplished by precipitation of gypsum (CaSOit • 2H20); rather, calcium
sulfate is precipitated along with calcium sulfite, resulting in a mixed
crystal of the two salts.
2. System Operation
The prototype system was started up on February 3, 1975 and was operated
over a period of 17 months, through July 2, 1976. The EPA Test Program
formally began in May 1975 and was completed in July 1976, after which
the system was shut down. This report covers the entire 17 months of
operation, including system startup and shakedown.
The operation of the system can be logically broken down into three dis-
crete periods as defined by coal composition, flue gas conditions, and
the characteristics of the system operation. The first period, from
February through July 1975, covered system startup and shakedown. During
these first six months of operation the boiler burned low-sulfur coal
(average sulfur content corresponding to approximately 2.6 Ibs S02/MM
Btu). Sulfur dioxide concentrations in the flue gas averaged 1,050 ppm
(range = 600-1,550 ppm) and oxygen levels averaged 7.5% (range = 5.0-11.0%) -
conditions well outside the range for which the process was originally de-
signed. This represented a difficult test for a system operating in the con-
centrated active sodium mode because of the levels of oxidation experienced.
1-2
-------
In the second operating period, lasting from September 1975 through early
January 1976, the system was tested under relatively stable load conditions
with the boiler firing a combination of medium- and low-sulfur coals. (The
average sulfur content of the coal fired corresponded to approximately
3.1 Ibs S02/MM Btu.) During this period the electrostatic precipitator
was maintained in full service. Repairs to the boiler combustion air pre-
heater and better control of combustion resulted in improved flue gas con-
ditions in comparison to operations in Period 1. Sulfur dioxide levels
in the flue gas during Period 2 averaged 1,250 ppm (range = 800-1,700 ppm),
and oxygen concentrations averaged 6.0% (range = 4.5-9.5%).
In Period 3, which lasted from March through early July 1976, the system
was tested on high-sulfur coal (average sulfur content in the coal cor-
responding to about 5.7 Ibs S02/MM Btu). Flue gas S02 levels averaged
about 2,250 ppm (range = 1,500-2,800 ppm), and oxygen concentrations
averaged about 6.5% (range = 4.5-9.0%).
In addition to 10 weeks of operation at relatively stable load, testing
during Period 3 consisted of three weeks of operation at fluctuating gas
loads and two weeks of particulate testing. The fluctuating load testing
involved adjusting the gas flow to the system to four different levels
according to a prearranged schedule roughly representative of the normal
load swings of the Scholz boilers. The average gas rate handled during
fluctuating load testing was 65%, as compared with 85-90% during the
stable load periods. Particulate testing was performed during the last
two weeks of the program to evaluate the effects of fly ash on the system
performance (S02 removal, scale formation, oxidation, and waste cake prop-
erties) and to assess particulate removal efficiency and mist eliminator
performance. During these two weeks the operation of the precipitator
ranged from fully activated to completely deactivated.
C. SYSTEM PERFORMANCE
Overall, the performance of the system was excellent. The system demon-
strated high S02 removal efficiency, high lime utilization, excellent
waste cake properties, and very good overall availability. The various
aspects of system performance are discussed below.
1. SO? Removal
S02 removal efficiencies at Scholz confirm the high S02 removal capability
of sodium solution scrubbing in the concentrated active sodium mode. With
sodium solution scrubbing, achieving a given outlet S02 level (within the
limit of the number of contact stages used) is essentially a matter of
adjusting the operating pH of the scrubber system (by adjusting the feed
forward rate or regenerated liquor pH). Over the 15 months between
April 1975 and July 1976, the average S02 removal using both the venturi
and absorber (with two trays) was 95.5% (for low-, medium-, and high-sulfur
coal); with the venturi alone (low-sulfur coal only) S02 removal efficiency
averaged 90.7%.
1-3
-------
For the most part, when both the venturi and absorber were operated
together, the venturi was used principally for quenching the gas, and
the venturi pressure drop was maintained in the range of 4.5 to 7.0 inches
of water. Under these conditions the pH of the venturi bleed liquor was
maintained between 4.8 and 5.9 to ensure better than 90% S02 removal.
With the low inlet S02 levels of Periods 1 and 2 (600-1,700 ppm) this
resulted in outlet S02 levels generally ranging from 15 to 100 ppm. At
the higher inlet S02 levels of Period 3 (1,500-2,800 ppm), the outlet
S02 typically ranged from 25 to 150 ppm.
When the venturi alone was used (10-16 inches of water pressure drop),
the bleed liquor was generally maintained at a pH above 5.7 to keep outlet
S02 levels below 100 ppm.
2. Particulate Removal
The particulate removal capability of the scrubber system was tested with
the venturi operated at both 12 and 17 inches of water pressure drop fol-
lowed by the absorber containing two trays. Three ranges of inlet partic-
ulate loadings were tested (by partially or wholly de-energizing the
precipitator): 0.015-0.025 grains/standard cubic foot dry (grs/scfd);
0.03-0.085 grs/scfd; and 2.3-3.6 grs/scfd. In general, outlet particu-
late loadings increased slightly with increasing inlet loadings, as would
be expected. Outlet loadings ranged from 0.010-0.015 grs/scfd at the
lowest inlet loadings to 0.024-0.037 grs/scfd at the highest inlet load-
ings. However, there was no statistical difference in outlet loadings
between operations at 12 and 17 inches of pressure drop across the venturi
throat (undoubtedly due, at least in part, to the two trays). At these
high venturi pressure drops, the S02 removal increased to an average of
about 98% for the particulate test period (using high-sulfur coal).
3. Lime Utilization
Lime utilization throughout all three operating periods was quite good.
Under normal conditions, lime utilization ranged from 90% to 100% of the
available Ca(OH)2 in the raw, hydrated lime and typically ran 93% to 97%.
(The Ca(OH)2 fraction of the delivered lime ranged from 87% to 93% wet
basis.) There was no discernible effect of fluctuating load, simultaneous
particulate removal, or manner of lime feed (dry vs. slurry) on lime uti-
lization. Utilization during operations with high-sulfur coal, though,
tended to average closer to 93%; while in operations with low-sulfur coal,
it ran 96% to 97%. This slight difference has been attributed to the
shorter holdup time in the reactor system at the higher feed forward
rates with high-sulfur coal.
4. Oxidation/Sulfate Control
a. Oxidation
Oxidation rates experienced in the prototype system were slightly lower
than those observed in the pilot plant under similar conditions. As in
1-4
-------
the pilot operation, though, oxidation in the scrubber circuit accounted
for 80-95% of oxidation throughout the system. The principal variable
affecting oxidation was the oxygen content of the flue gas, although the
flue gas rate and the type of coal fired also had some effect.
With the low- and medium-sulfur coals fired in Periods 1 and 2, oxidation
rates in the scrubber system ranged from about 180 ppm equivalent S02
(^ 230 ppm in the entire system) at 5% oxygen in the flue gas to about
370 ppm equivalent S02 (^ 420 ppm in the entire system) at 9% oxygen in
the flue gas. The total system oxidation rates correspond to a range of
20% to 45% of the S02 removed for the average inlet levels for these
periods. However, there was considerably more gas/liquid contacting
provided in the scrubber system (venturi + two trays) than would normally
be incorporated in an absorption system for a low-sulfur coal application.
This not only resulted in very low outlet S02 levels (typically less than
50 ppm) but also unnecessarily high rates of oxidation (as a percentage
of S02 removed).
With the high-sulfur coal in Period 3, the absolute rate of oxidation in
the scrubber system increased slightly. At 5% oxygen in the flue gas,
oxidation in the scrubber system averaged about 200 ppm equivalent S02
('v 250 ppm throughout the system) ; and at 9% oxygen , oxidation in the
scrubber ran slightly over 500 ppm ('v 550 ppm throughout the system) .
These higher oxidation rates, though, represent lower percentages of
oxidation in terms of S02 removed. For the average S02 removal in
Period 3 these oxidation rates correspond to about 10% and 25% of the
S02 removal, respectively.
As would be expected, the absolute rate of oxidation (mols/min) decreased
with reductions in gas flow, although the percentage of S02 oxidized in-
creased slightly. No effect of fly ash on oxidation was apparent during
the particulate testing period.
b. Sulfate Precipitation
Precipitation of calcium sulfate measured in the reactor system showed
that calcium sulfate could be coprecipitated with calcium sulfite at
levels as high as 25% of the total calcium sulfur salts, indicating
that the system was capable of keeping up with such levels of oxidation.
The correlation of sulfate/sulfite content of the precipitated calcium
salts to sulfate/sulfite concentrations in the reactor liquor was found
to be:
/mols CaSOi
reactor Bolldfl
This degree of sulfate coprecipitation corresponds to about 85% of that
observed in the pilot plant.
1-5
-------
There was also a slight decrease in the calcium sulfate/sulfite ratio in
the filter cake in comparison with that in the reactor product solids.
The data indicate about a 15% decrease in the sulfate content between the
reactor and filter. This is probably due to some dissolution of calcium
sulfate during the long holdup in the thickener.
Overall, the sulfate formation (oxidation)/sulfate precipitation data
show that the system is capable of keeping up with oxidation rates of
up to 25% of the S02 removed — oxidation rates much higher than those
anticipated for most medium- and high-sulfur coal applications. The
operation at the widely fluctuating conditions demonstrated the stability
of the system chemistry and its ability to "self-adjust" to handle any
oxidation rate up to 25% without operator intervention. As oxidation
changed, the ratio of sulfate to active sodium in the liquor changed
accordingly to increase or decrease the amount of calcium sulfate
precipitated.
5. Waste Cake Properties
a. Solids Content
The solids content of the waste filter cake varied from 41% to 77% of
the total cake weight. In general, the solids content of the cake varied
with calcium sulfate content (decreasing with increasing calcium sulfate
levels) and with variations and upsets in the filter operation. During
stable load conditions the average solids content of the filter cakes
produced in each successive operating period increased from 48% in
Period 1 (low-sulfur coal) to 54% in Period 3 (high-sulfur coal). The
inclusion of fly ash during simultaneous particulate removal in Period 3
increased the average solids content to about 57%. These averages in-
clude periods of minor filter upsets and partial loss of vacuum.
Under most all conditions the cake had the appearance and handling prop-
erties of a moist soil. It was easily transferred from the storage pile
to dump trucks using a front-end loader for transfer to the disposal pit.
b. Solubles Content
Wash efficiency tests performed on the prototype filter verified pilot
plant results regarding the washability of the cake. The results show
that the soluble solids levels in the cake can be readily reduced to
2-3% (dry cake basis) under controlled filter conditions using a wash
ratio of about 2.5 (gals wash water/gal water occluded in the cake).
The solubles levels actually achieved on a continuous basis, though,
were higher due to the limited capacity of the spray nozzles, system
upsets, and inadequate operator attention to cake washing. Soluble
solids levels in the cake throughout the program ranged from as low as
1.2% to as high as 12% of the dry cake weight, depending upon the degree
of washing and the solids content of the filter cake. The average losses
1-6
-------
estimated for each operating period based upon cake analyses and overall
material balances ranged from 4% (Period 2) to as high as 8% (estimated,
Period 1). The average solubles losses, though, were biased upward by
the fluctuating wash conditions. Long periods of adequate cake washing
were more than offset by short periods of poor cake washing (due to high
rates of cake withdrawal required to catch up with the single filter, and
inattention to wash water rates).
6. Sodium Makeup
The rate of sodium makeup to the system in comparison to the estimated
sodium value losses in the filter cake provide a measure of the degree
of closed-loop operation (as well as accountability in the overall mate-
rial balances). Soda ash feed rates were closely monitored only during
Periods 2 and 3. During Period 2 soda ash makeup rates represented about
8% of the total S02 removal (mols Na2C03/mol AS02) compared with about
4.5% soda ash requirements based upon cake losses. The difference is
attributed to pump seal leaks, a small thickener leak that developed
during Period 2, and errors in the overall material balance. Entrain-
ment losses of sodium (in entrained liquor) were negligible. As measured
in both December 1975 and June 1976, entrainment losses were equivalent
to less than 0.1% of the AS02 (as soda ash required). During Period 3
the material balance on sodium was almost completely closed. The soda
ash required to make up for cake losses was about 7% of the S02 removal
(using an average wash ratio of 1.8) versus a soda ash feed rate of 8%
of the S02 removal.
While soda ash requirements were slightly higher than desired due primarily
to inadequate control of the filter operation, the relatively small soda
ash requirements and the degree of closure in the soda ash material balance
reflect a tight, closed-loop operation.
7. Power Consumption
Since the process included a venturi scrubber as well as additional pump
capacity for operation in a direct lime or limestone mode, the power
consumed in the dual alkali mode was greater than that which would be
consumed in a system designed specifically as a dual alkali system. When
operating at or near design gas flow, the system power consumption aver-
aged 2.5-3.0% of the boiler output. Correcting for the additional pressure
drop included with the venturi and the unnecessary pump capacity, the power
consumed by the equipment actually required for this application was about
1.0% of the capacity of the boiler.
8. Operability/Reliability Potential
a. Availability
While the system was not operated for the purpose of achieving a high
availability figure, the availability record of the system is impressive.
Over the 17 months of operation the system logged more than 7,100 hours
1-7
-------
of operation, which corresponded to an overall availability of slightly
higher than 70%. Most of the downtime occurred between the operating
periods and resulted from equipment problems of a mechanical nature or
problems caused by operation of the system well outside the design con-
ditions. The availability during the operating periods averaged about
90%. This availability is impressive, particularly in light of the fact
that the only spare equipment was pumps (and replacement parts for un-
spared equipment were minimal), and that the system was called upon to
operate about 70% of the time at conditions outside those for which it
was designed.
The longest single outage (1,460 hours) occurred between Periods 1 and 2.
During the end of Period 1, oxygen levels in the flue gas were running in
the range of 8-10%, with inlet S02 levels depressed to 850-950 ppm. Because
of the resulting high oxidation levels (as a percentage of S02 removal)
the system was allowed to drift into a dilute active sodium mode, a mode
for which it was not designed. The result was precipitation of gypsum
and formation of some gypsum scale in the reactor tanks and piping. At
the same time, mechanical problems in the scrubber required a shutdown
of the system, and it was decided to await repair of the preheater and
higher-sulfur coal prior to restart of the system. There was also some
delay in replacement parts, so the system remained down from mid-July
through mid-September 1975. Such delays would not normally be encoun-
tered in full-scale applications with adequate sparing of equipment and
maintenance of a reasonable inventory of spare parts.
Between Periods 2 and 3 the boiler was shut down for scheduled maintenance.
The system remained out of service an additional month following boiler
startup, again due to delays in shipments of replacement parts and equip-
ment being overhauled.
b. Equipment Performance
Most of the problems encountered with equipment and instrumentation during
the course of the test program were mechanical in nature and reflected
design or fabrication oversights commonly associated with a prototype
system. All but a few were resolved during the course of the test program
by simple operational adjustments and/or equipment modifications.
Equipment
The most significant equipment problems encountered in the system involved
the filter, vessel linings, scrubber control and block valves, and solids
buildup in the first-stage reactor. Collectively, these accounted for
the bulk of mechanical-related downtime and maintenance.
• The filter was the largest source of problems in the prototype
system, but the problems resulted in few system outages due to
the solids holdup capacity in the thickener, which allowed suffi-
cient time for most filter-related maintenance work. Normally,
filters do not require an inordinate amount of maintenance.
1-8
-------
However, a large part of the filtration equipment in the prototype
system was fabricated out of fiberglass and plastic because of both
anticipated corrosion problems from the high chloride levels achieved
in the tight, closed-loop operation (3,500-11,000 ppm Cl~) and to mini-
mize the cost for the short-term prototype test program. Fiberglass
is not as sturdy as stainless steel, and there were failures at stress
points in the construction as well as erosion of some of the plastic
and fiberglass parts. Most of the problems occurred during Period 1
and the early part of Period 2. Modifications of the filter drum and
tub by plant personnel, and overhaul of the filter drum by the manufac-
turer between Periods 2 and 3 either eliminated the problems or re-
duced them to routine, low-level maintenance.
Lining erosion or lining cracks and pinholes occurred in various
vessels in the system. Cracks and pinholes occurred in the ab-
sorber recycle tank and the thickener floor and walls. These were
patched during interim periods and did not reoccur during the re-
mainder of the test program. Erosion of lining occurred beneath
the agitator in the second-stage reactor vessel and on the liquor
redistribution shelf in the venturi. These linings were also patched
in the interim between Periods 2 and 3, and the venturi tangential
nozzles modified. No further erosion at either location was observed.
There was also deterioration of the lining in the area of the quench
zone at the gas inlet to the venturi. The cause of the failure may
have been a combination of factors including poor application, inade-
quate surface preparation, and severity of temperature and chemical
attack. This failure suggests that corrosion-resistant metal alloys
may be most suitable in such areas.
Erosion and "debonding" of rubber linings occurred in control and
block valves in the scrubber system. These failures were traced to
the high degree of throttling to the control flow. (The valves were
sized to accommodate the higher flows associated with direct lime
slurry scrubbing.) These valves were replaced with 316 stainless
steel valves prior to Period 3 and no further erosion or debonding
occurred in the valves. There was also no corrosion or erosion of
the 316 valves after the three and one-half months of service in
Period 3.
Buildup of product solids occurred in the first reactor throughout
the test program. Through adjustments made to the reactor system
and simulation of the operation in the ADL pilot plant, the cause
of the problem was traced to poor agitation and operation during
severe upset conditions (e.g., gross overfeeding of lime). While
the buildup was never serious enough to cause a shutdown, it did
require occasional cleaning. Improved agitation and better process
control should reduce such buildup to, at worst, a semi-annual main-
tenance item. Such maintenance would not require system shutdown in
large-scale systems where parallel reactor trains can be used, or the
first reactor temporarily bypassed.
1-9
-------
Instrumentation
Instrumentation problems primarily involved pH units, level transmitters,
and the soda ash feed solution control system.
• The flow-through pH probes originally installed in the system were
prone to plugging and/or erosion and failure of probe tips. The
flow-through unit in the reactor system was replaced with an im-
mersion unit, which proved to be much more reliable. Modification
of take-off lines for flow-through units in the scrubber system and
increasing the flow rate minimized problems with these units.
• The level transmitters originally installed were unreliable and re-
quired an inordinate amount of instrument maintenance. These were
eventually replaced with Foxboro units, which proved to be much more
reliable and less prone to failure of critical parts.
• A number of difficulties were encountered with the soda ash feed
control system, some of which were related to the wide turndown
range for which it was designed. None of the problems affected
the operability of the system, since continuous, accurate control
of makeup soda ash is not required to replace the small sodium
losses in the cake. The principal impact of the difficulties in
the feed control system were in the accuracy of the material bal-
ances on sodium.
c. Ease of Operation
Ease of system operation was assessed throughout the program either during
planned tests of system capabilities under differing conditions or indirectly
through inadvertent process upsets and equipment malfunctions. The planned
testing included:
• stable operation with low-, medium-, and high-sulfur coal;
• fluctuating load testing (30-100% of design gas rate) with
high-sulfur coal; and
• simultaneous particulate removal with high-sulfur coal.
The results of this testing have been discussed.
Indirect measures of the system operability were also obtained during
upset conditions. Upset conditions encountered included:
• wide, short-term fluctuations in inlet SC^;
• wide swings in inlet oxygen concentration;
• inadvertent substitution of limestone for lime in the
chemical storage silo;
1-10
-------
• gross underfeeding and overfeeding of lime; and
• short-term outages of the filter, first-stage reactor and
various instruments with continued operation.
Operation during both the planned variations in system conditions and
upsets served to demonstrate the basic stability of the system and the
inherent ability of the concentrated lime mode dual alkali technology
to withstand sudden (and extended) changes in operating conditions with-
out loss of performance. Of particular note is the fact that close control
of pH throughout the system is not required to ensure high S02 removal
efficiency and prevent scaling. In fact, during some extended periods
lasting up to a few days in length the system flows and makeup chemical
feed rates were set by inlet and outlet S02 and trimmed according to pH's
of samples taken from the reactor and scrubber twice per shift.
As oxidation rates changed (due to changes in inlet S02 or oxygen concen-
tration of the flue gas), the system chemistry adjusted accordingly. The
ratio of sulfate to active sodium in the system liquor simply shifted to
effect the appropriate rate of calcium sulfate precipitation required
(up to about 25% oxidation).
Similarly, operator errors in setting system flows or makeup chemical feed
rates rarely had any immediate effect on system performance, and the effects
were usually completely reversed simply by re-establishing proper system
conditions.
d. Scale Potential
Due to the low calcium concentrations maintained throughout the system,
there was little potential for scale formation. Other than the deposition
of solids in the first-stage reactor (previously discussed) the only occur-
rence of scale formation in the system was the precipitation of calcium
carbonate in the absorber during two extended periods when the scrubber
system was inadvertently operated well outside the specified pH range.
The calcium carbonate was completely dissolved within a few hours by
returning the system to normal operating conditions and had no effect
on system performance in any way.
The low scale potential, particularly in the scrubber system, is evidenced
by the operation of the mist eliminator in the absorber. The mist eliminator
was operated without any wash sprays (fresh water or liquor) for the last
two operating periods (4,600 hours). No deposit of solids or scale of any
form could be found on the mist eliminator following completion of the test
program. Similarly, there was no deposit of solids on the reheat gas dis-
tributor downstream of the mist eliminator.
1-11
-------
II. INTRODUCTION
This is Volume III of a three-volume report for the EPA/ADL Dual Alkali
Program. Volume I is an Executive Summary. Volume II covers Task I and
Task II, the laboratory and pilot plant tasks of the program of EPA con-
tract No. 68-02-1701 (with IERL, RTF). Volume III covers Task III, the
Prototype Test Program performed on the 20-megawatt (nominal) CEA/ADL
dual alkali system at Plant Scholz, Gulf Power Company/Southern Company
Services, Inc.
The 20-megawatt, prototype dual alkali process was designed and installed
by Combustion Equipment Associates, Inc. (CEA) and Arthur D. Little, Inc.,
(ADL) at the Scholz Steam Plant of Gulf Power Company near Sneads, Florida.
This system is one of three advanced technology, prototype flue gas desul-
furization systems installed at the plant as part of a technology evaluation
program being conducted by Southern Company Services for The Southern Company
(an electric utility holding company including Alabama Power Company, Georgia
Power Company, Gulf Power Company, Mississippi Power Company, Southern
Electric Generating Company, and Southern Company Services, Inc.).
The process was developed and designed jointly by CEA/ADL. Early labora-
tory research on the process, performed by ADL and sponsored by the Illinois
Institute for Environmental Quality, dealt exclusively with characterizing
the nature of the regeneration reaction. Based upon the laboratory re-
sults, a 2,000 cfm dual alkali pilot plant was constructed at ADL's facil-
ities by CEA/ADL and an eight-month test program sponsored by CEA prior to
initiation of the EPA contract was conducted to generate the design data
for the prototype system. The pilot system contained the complete dual
alkali process loop involving: gas scrubbing; absorbent regeneration;
and solids separation. Results of the laboratory program and pilot oper-
ations for generation of the prototype system design have been reported
previously in the literature.1
The laboratory and pilot plant investigation of dual alkali technology
has continued at ADL in this program for the U.S. Environmental Protection
Agency's (EPA) Industrial Environmental Research Laboratory at Research
Triangle Park, N. C. The program involved characterization of the basic
process chemistry and the various modes of operation of sodium-based dual
alkali processes. The work covered a wide range of flue gas conditions,
liquid reactant concentrations, and process configurations including the
use of both lime and limestone for regeneration of the sodium scrubbing
liquor. This work is reported in Volume II.
Task III, the Prototype Test Program consisted of a one-year test of the
20-megawatt CEA/ADL dual alkali system at Gulf Power. The objective of
the program was to characterize the important aspects of the prototype
process performance:
• S02 removal efficiency;
• oxidation and sulfate formation and control;
II-l
-------
• lime utilization;
• waste solids properties;
• sodium makeup requirements and degree of closed-loop
operation; and
• process operability and reliability potential.
Construction of the prototype system was completed and the system put
in operation in early February 1975. In mid-May 1975, the test program
was initiated by ADL, Southern Company Services, CEA, and Gulf Power
Company as part of the EPA/ADL Dual Alkali Program.
This report describes the performance of the system over the 17-month
period from startup in February 1975 through operations ending in
early July 1976.
II-2
-------
III. SYSTEM DESCRIPTION
A. PROCESS CHEMISTRY
The CEA/ADL dual alkali S02 control process at Scholz Station was an aqueous
sodium solution scrubbing system in which the absorbent solution was regen-
erated using hydrated lime.
The absorption of S02 was accomplished using a solution of sodium sulfite,
sodium hydroxide and possibly some sodium carbonate (makeup) producing a
spent sodium sulfite/bisulfite liquor:
2NaOH + S02 •*• Na2S03 + H20 (2 )
Na2C03 + S02 •* Na2S03 + C02 t (3 )
Na2S03 + S02 + H20 £ 2NaHS03 (4 )
During absorption, and to a lesser extent through the remainder of the
system, some sulfite was oxidized to sulfate:
2Na2S03 + 02 -»• 2Na2SOi( ( 5 )
converting an "active" form of sodium to an "inactive" form. Oxidation
in the scrubber was generally a function of the scrubber design, oxygen
content of the flue gas and the scrubber operating temperature. At excess
air levels normally encountered in utility power plant operations (25-40%)
burning medium- or high-sulfur coal (> 2% sulfur), the level of oxidation
is expected to be on the order of 5-10% of the sulfur dioxide removed.
The scrubber solution was regenerated by reaction with lime which precipi-
tates a mixture of calcium sulfite and calcium sulfate solids for disposal,
as shown by the following overall reactions (shown for lime):
2NaHS03 + Ca(OH)2 -»• Na2S03 + CaS03 • 1/2H20 + + 3/2H20 (6)
Na2S03 + Ca(OH)2 + 1/2H20 + 2NaOH + CaS03 • 1/2H20 4- (7)
Na2S04 + Ca(OH)2 * 2NaOH + CaSO^ + (8)
After regeneration, the solids were separated from the regenerated liquor
and washed. The clear liquor, containing very low amounts of suspended
and dissolved calcium, was returned to the scrubbers. Soluble calcium
levels are generally less than 100 ppm in concentrated dual alkali
processes. Any sodium value lost with the washed waste solids was
III-l
-------
replaced by the addition of sodium carbonate to the regenerated liquor,
although sodium hydroxide could also have been used since carbonate
softening is not required. Since sodium sulfate is also reacted with
lime in this system to regenerate sodium hydroxide, it should be possible
to use sodium sulfate as the sodium makeup source when low to moderate
oxidation levels are encountered.
The system was designed to operate in the concentrated active sodium
mode (active Na+ concentration greater than 0.15M). In this mode,
sulfate removal cannot be accomplished by the precipitation of gypsum
(CaSOit • 2H20), since the high sulfite levels prevent the soluble calcium
concentration from reaching that required to exceed the gypsum solubility
product. However, calcium sulfate (in a form other than gypsum) is pre-
cipitated along with calcium sulfite (CaS03 • 1/2H20) in the regenera-
tion reactor, resulting in a solid solution of the two salts. The amount
of sulfate precipitated in this form is a function of the concentrations
of species in solution and the reactor pH. Under normal operation, with
sulfate levels up to 1.5M SOij, a concentrated mode dual alkali system
is capable of keeping up with sulfite oxidation rates equivalent to
25-30% of the S02 absorbed without becoming saturated in calcium sulfate.
Additional details of dual alkali chemistry and terminology are given
in recent publications2'3 and in Volume II of this report.
B. SYSTEM CONFIGURATION
The prototype system at the Scholz Steam Plant was installed on Boiler
No. 1, a 40-megawatt (nominal) Babcock and \vfilcox pulverized coal-fired
boiler. The boiler is capable of operating at peak loads of up to 47
megawatts. The prototype was sized to handle 50% of the flue gas from
the boiler. The boiler was retrofitted with a sectionalized, electro-
static precipitator capable of 99.5% particulate removal. Sections of
the electrostatic precipitator could be de-energized to study the impact
of particulate input on process operation.
The design basis for the prototype system is given in Table III-l; a
schematic flow diagram is given in Figure III-l. The system, designed
as a prototype, incorporated a high degree of flexibility aimed at gen-
erating design and operating information for a wide variety of applica-
tions. Although the basic mode of operation of the system is a dual
alkali process with lime regeneration, the system was designed to accom-
modate regeneration with limestone alone and regeneration with a combina-
tion of limestone and lime. The system was also designed to enable
operation as a direct limestone or lime scrubbing system including
simultaneous particulate and S02 removal. As a consequence, the system
contained equipment and piping in excess of that required to operate a
dual alkali system on a boiler already equipped with adequate particu-
late control.
The venturi scrubber was included in the system to investigate the oper-
ation of that type of scrubber in a dual alkali mode (or for direct lime
III-2
-------
TABLE III-l
DESIGN BASIS
Flue Gas Inlet
Flow Rate (acfm)
(nominal Mw equivalent)
Temperature (°F)
02 Concentration (% volume, dry)
Particulate Loading (gr/scf, dry)
S02 Concentration (ppm, dry)
75,000
20
275
6.5 (maximum)
0.02 (precipltator
energized)
1,800-3,800
Design Performance
S02 Removal (% reduction)
Maximum S02 Removal Rate (Ib/hr)
Particulate (gr/scf, dry)
Power Consumption (% power output)
with venturi, full spray absorber pump
requirements
without venturi, with tray absorber pump
requirements
90 (minimum)
1,530
0.02 (no increase
in loading)
2.5-3.0
1.0-1.5
III-3
-------
M
M
SCRUBBED GAS
MAKE-UP
S02 Io2 OUTLET PROBE H2°
SOLIDS
A) CAKE
DISCHARGE
V AIR
ALTERNATE FLOWS
NUMBER IN SYMBOL INDICATES
NUMBER OF PUMPS OF) BLOWERS
IN PLACE
FIGURE III-l CEA/ADL DUAL ALKALI SYSTEM AT THE SCHOLZ STEAM PLANT - PROCESS FLOW DIAGRAM
-------
or limestone scrubbing) when simultaneous particulate and S02 removal
is required. The second scrubber, an absorption tower, was designed
for operation as a tray scrubber, as in a dual alkali system; or as a
spray tower, for direct lime or limestone scrubbing. Normally, when
simultaneous particulate and S02 removal is not required, the system
would include only an absorber. In addition to this redundancy in
scrubbers, sufficient pump capacity was provided to operate the venturi
at an L/G of 25 gallons/Macf of saturated gas and for an absorber L/G
of 60 when operating in a spray tower configuration.
An additional storage silo (for limestone), a mix tank and other assorted
tanks, pumps, controllers and piping were included in the system to ac-
commodate the high degree of flexibility desired in the prototype system.
Except on occasions when limestone was erroneously delivered to the silos
instead of lime, the system was operated only in the lime dual alkali
mode as shown in Figure III-l.
The system design was based upon removal of at least 90% of the S02 in
the flue gas for medium- to high-sulfur fuels (up to 5% sulfur). With
the precipitator energized, the system was specified not to increase
particulate loadings in the scrubber outlet above those in the inlet
flue gas.
The power consumption of the system (not including oil for reheat) was
equivalent to 2.5-3.0% of the power generated by the unit in producing
the flue gas load to the system, with the system operating at full fan
capacity (full gas flow at a system pressure drop of 20 inches water)
and at full venturi and absorber liquid recirculation capacity. This
consumption figure is calculated based upon a nominal 20 megawatts at
full load. Correcting for the excess fan and pump capacity, the power
consumed by the equipment actually required in this application (tray
tower at an L/G of 5-10) is roughly 1.0-1.5% of the power generated at
the design conditions. In a full-scale dual alkali system designed for
S02 removal only (without the venturi), the power consumption should be
in this range.
The dual alkali system can be conveniently broken down into three process
subsystems: gas scrubbing; absorbent regeneration; and solids dewatering.
The design and operation of each of these subsystems is discussed in the
following sections.
1. Flue Gas Scrubbing
The gas scrubbing system consisted of a variable-throat, plumb-bob type
venturi scrubber followed in series by an absorption tower. A photograph
of the scrubbing system is shown in Figure III-2.
Treated flue gas flows through both scrubbers. No provision was included
to bypass the gas around either of the scrubber units. Each of these
scrubbers was equipped with a removable liquid entrainment separator
III-5
-------
Absorption Tower And Stack
Venturi Scrubber
Forced Draft Fan
FIGURE II1-2 GAS SCRUBBING SYSTEM
-------
(although only the entrainment separator in the absorber was Installed),
an enclosed recycle tank to contain the scrubbing liquor, and two recycle
pumps (one operating and one spare) . The venturi could be used for ab-
sorption and/or particulate control and could be operated on a separate
liquor loop from the absorption tower or in series with the absorber
liquor loop. The absorption tower can be operated as a tray tower
(with up to four trays), as a spray tower, or as a de-entrainment
separator.
Gas from the exit of the electrostatic precipitator was forced through
the scrubbing circuit using the booster fan provided with the dual alkali
system. The fan and motor were designed for a total system pressure drop
of 20 inches 1^0 at maximum flow. Under normal conditions with the pre-
cipitator in service, the venturi was used only for gas saturation (with
some attendant S02 removal), and a pressure drop of roughly 5 to 10 inches
of water was maintained across the venturi.
Gas from the fan entered the venturi scrubber, flowed downward over the
wetted approach section and into the high velocity venturi throat. Re-
cycled scrubbing liquor also entered the top of the scrubber through
tangential pipe inlets and through a number of vertical bull nozzles
equally spaced around the center of the venturi.
After passing through the throat, the flue gas and scrubbing liquor con-
tinued downward through the internal downcomer and the liquor was collected
at the bottom of the scrubber in the internal recycle tank. At the bottom
of the downcomer the flue gas made a 180-degree upward turn and contacted
the chevron-type entrainment separator (when used). The entrainment
separator was equipped with wash sprays above and below which could be
operated continually or on a sequential timing cycle. Either scrubbing
liquor or fresh water could be used for washing. During dual alkali
operations, this entrainment separator system was not installed.
Following the venturi scrubber, the saturated flue gas entered the bottom
of the absorption tower. This tower was designed to operate either as a
spray tower, with one or two sets of sprays, or as a tray tower, with
from one to four trays. The bottom tray was equipped with a spray under-
neath to wet the bottom side of the tray. Gas passed upward through the
trays and then through a final de-entrainment separator, with a wash system
similar to the venturi demister wash. The wash system was used initially,
but found to be unnecessary and its use was discontinued.
The clean flue gas leaving the tower was finally reheated by the injection
of hot gas from a burner fired with No. 2 fuel oil before being discharged
through the stack on top of the absorber. Oil-fired reheat was specified
for all the prototype systems in order to reserve steam for power generation.
Regenerated absorbent solution, containing sodium hydroxide, sodium sulfite,
sodium sulfate and some sodium carbonate, was mixed with recycle liquor and
fed to the top tray of the absorber. The solution flow was countercurrent
to the gas through the tray system (two trays) and was collected at the
III-7
-------
bottom of the absorber in the internal recycle tank. This collected
liquor supplied solution both for spraying the bottom tray and for re-
circulation to the top tray, as needed for pH control in the scrubber
or to maintain liquor flow across the trays.
A bleed from the collected tray tower bottoms was sent forward to the
venturi recirculation loop for additional S02 removal. In the venturi,
the gases with the highest S0£ concentration were contacted with the
most acidic liquor. A.continuous bleed stream was drawn from the venturi
recirculation loop and sent to the absorbent regeneration system.
2. Absorbent Regeneration
Spent scrubber solution from the venturi recirculation line was bled
(on level control) to a two-stage reactor system where it was reacted
with hydrated lime. A photograph of the reactor system, chemical storage
silos and feed tanks is shown in Figure III-3.
The first-stage reactor, to which both the scrubber bleed and lime were
added, had a holdup time of two to ten minutes depending upon the feed
forward rate. Slurry from the first stage was fed by gravity overflow
to the second-stage reactor, which had a holdup time ranging from 20
minutes to one hour. Both reactor stages were baffled, cylindrical
vessels equipped with center-mounted agitators. Initially, the first-
stage reactor was a trapezoidal-shaped vessel mounted inside the second-
stage reactor. In February 1976 this was replaced with a separate vessel
of conventional, cylindrical design.
Hydrated lime was used in this small prototype system for simplicity.
It was fed to the first reactor either as a dry solid directly or as
a slurry containing from 12% to 32% suspended solids. In a full-scale
system, quicklime would normally be slaked and fed as a slurry of 20%
to 30% Ca(OH)2. The rate of lime feed was usually adjusted to maintain
the pH of the reactor effluent within a preset range (usually 11.5-12.5).
The lime neutralized the bisulfite acidity in the scrubber bleed and
further reacted with sodium sulfite and sodium sulfate to produce sodium
hydroxide. These reactions precipitated mixed calcium sulfite and sulfate
solids resulting in a slurry containing up to 5 wt % insoluble solids.
This two-stage reactor system design has been shown to generate clusters
of sulfite/sulfate crystals which are spherically-shaped agglomerates
(rosettes) rather than the needle-like or platelet crystals generally
associated with calcium sulfite precipitation. These crystal clusters
have good settling, filtration, and washing properties and can be gen-
erated in the reactor system over a wide range of flue gas and process
conditions. The multistage reactor system is part of a process patented
by CEA/ADL.
III-8
-------
H
M
M
VO
Regeneration Reactor And
Dry Lime Storage Silo
Auxiliary Lime Slurry Tank
And Storage Silo
Soda Ash Solution Tank
And Storage Silo
FIGURE 111-3 ABSORBENT REGENERATION AND CHEMICAL MAKEUP SYSTEM
-------
3. Solid/Liquid Separation and Solids Dewateriug
Slurry from the regeneration reactor system was fed to the center well
of the slurry thickener. The 40-foot diameter thickener was sized to
handle the solids produced from the treatment of flue gases from the
full 40 megawatts of Boiler No. 1 (in contrast to the scrubber and
reactor systems which were sized to handle 20 megawatts). Figure III-4
shows a photograph of the thickener at Scholz.
The thickened slurry from the bottom of the settler was sent to a rotary
drum vacuum filter, also shown in Figure III-4. The solids content of
the underflow was maintained below 30% for ease of underflow pumping.
The slurry was recirculated past the filter in a recycle loop that re-
turned the slurry to the solids zone in the settler. The feed to the
filter was drawn as a bleed from this recirculation loop. The filter
surface area was 75 ft2.
The cake was washed on the filter using from two to four water spray
bars arranged in series. This wash removed a large fraction (up to 95%)
of the occluded soluble salts from the cake and returned these salts to
the system, thereby reducing sodium losses and minimizing sodium carbon-
ate makeup.
Solids from the filter were retained in a solids storage enclosure directly
under the filter from which they were loaded into dump trucks for transfer
to the disposal area. The mixed filtrate and wash liquor from the filter
were returned to the thickener.
The disposal area for the dual alkali waste cake was a one-acre pit
(450 ft x 100 ft) approximately 12 feet deep. The bottom and sides
of the pit were lined with a layer of clay covered by a double thick-
ness of polyethylene liner. The polyethylene was reinforced between
the sheets with a mesh of nylon fiber. The floor of the pit was sloped
to a single collection drain constructed of PVC from which leachate was
discharged to the ash pond. On top of the polyethylene liner was a two-
inch layer of sand and gravel.
Makeup sodium carbonate was fed to the thickener in order to allow for
easy removal of any CaC03 precipitated. This carbonate was not intended
for use as a softener, since soluble calcium concentrations in the re-
generated liquor generally run less than 100 ppm. However, some CaCOs
precipitation occurs due to its very low solubility limits. The amount
of CaC03 formed will generally be very small, a few hundred pounds per
day at design conditions, or less than 0.5% of the total cake produced.
Provisions were made to prepare soda ash solution using either river
water or solution liquor.
Clear liquor overflow from the thickener is collected in the thickener
hold tank which acts as surge capacity for the absorbent liquor feed to
the scrubber system. Water was normally added to this hold tank to make
up for the difference between total system water losses (evaporation and
cake moisture) and total water inputs from other sources (sodium makeup
solution, pump seals, lime feed, cake wash, and demister wash).
111-10
-------
M
I
Thickener And Thickener Hold Tank
Filter Station
FIGURE 111-4 SOLIDS DEWATERING SYSTEM
-------
IV. OPERATING HISTORY
A. GENERAL OPERATING CONDITIONS
The prototype system commenced process startup on February 3, 1975 and
was operated over a period of 17 months through July 2, 1976, when the
system was shut down at the completion of the test program. During
these 17 months the system logged 7,128 hours of operating time with
shutdown periods of varying length for system maintenance and modifi-
cations and when the boiler was taken off line. The 17 months of oper-
ation can be logically broken down into three discrete operating periods
as defined by the flue gas composition and coal characteristics, and by
the mode of system operation. These three periods of operation are
summarized in Table IV-1.
During the first two operating periods, covering the first 11 months of
operation, the system treated flue gas with lower sulfur dioxide concen-
trations and higher oxygen levels than the range for which the system
was originally designed. This represented a difficult test for the
prototype system. In sodium-base dual alkali systems for a given
scrubber design and soluble solids levels, the rate of sulfite oxida-
tion (mols per unit time) is a strong function of the oxygen concen-
tration in the flue gas and is relatively independent of the rate of
sulfur dioxide removal (mols per unit time). As oxygen concentrations
increase and S02 concentrations decrease, a higher percentage of the
S02 removed from the flue gas is converted to sodium sulfate rather
than sodium sulfite/bisulfite in the scrubber liquor. This higher
percentage oxidation requires an increase in precipitation of sulfate
relative to sulfite and a higher calcium sulfate content in the pre-
cipitated calcium sulfite/sulfate to enable closed-loop operation with
no intentional purging of sodium sulfate.
Sulfur dioxide concentrations in the flue gas averaged between 1,100 and
1,200 ppm, compared with a minimum design concentration of 1,800 ppm during
the first two operating periods. The actual concentrations varied over a
range of from 600 to 1,600 ppm, often fluctuating daily and even hourly
over the entire range of concentrations. Oxygen concentrations in the
flue gas entering the scrubber ranged from 5% to 11% by volume (equiva-
lent to 30-100% excess air) with oxygen levels increasing with decreasing
boiler load (higher excess air). During the first operating period, air
leaks in the boiler combustion air preheater and coal feed tubes contrib-
uted from 0.5 to 2.0 volume % to the oxygen concentrations. Oxygen levels
in the flue gas were reduced from the 7-10% range down to the 5-7% range
in September 1975 after air preheater repairs and when burner box pressure
was brought under better control.
During the third period of operation the system was tested on high sulfur
coal. In addition to stable load testing, as conducted during Periods 1
and 2, testing also included fluctuating load testing and particulate
testing during which time the electrostatic precipitators were partially
or completely de-energized.
IV-1
-------
TABLE IV-1
SUMMARY OF OPERATING PERIODS
Operating Periods
123
Inclusive Dates 2/3/75-7/18/75 9/16/75-1/2/76 3/16/76-7/3/76
Operating Hours 2,537 2,180 2,411
Coal Low Sulfur Low/Medium Sulfur High Sulfur
Testing Startup & Shakedown Stable Load Stable Load
Fluctuating Load
Particulate
IV-2
-------
The general process operating conditions during each of the three oper-
ating periods are discussed in the next section.
B. DESCRIPTION OF OPERATING PERIODS
1. Operating Period 1 — Startup and Initial Operations
This first operating period extended from initial startup on February 3,
1975 through July 18, 1975, when the system was shut down for a two-month
period for modifications and repairs, and to await receipt of replacement
parts. During this operating period, the system was operated approximately
2,540 hours, 79.0% of the time that the boiler was in operation. More than
half of the entire system downtime of 670 hours relative to the boiler is
accounted for by a three-week shutdown in mid-April, after the start-up
period, when necessary maintenance was performed and equipment modifica-
tions and adjustments were made prior to the start of the formal EPA test
program in mid-May. These modifications were to enhance the system oper-
ability and performance under the low-sulfur coal conditions — conditions
for which the system was not originally designed. Equipment changes are
described in Chapter V.
Table IV-2 contains a summary of the fuel and flue gas characteristics
encountered during this operating period. The boiler was fired with a
combination of several medium- and low-sulfur coals ranging in sulfur
content from 0.9 to 2.2 wt %. A plot of coal analyses for daily coal
samples is presented in Appendix A. The weighted average sulfur content
of the coal as burned was 1.6 wt %, producing an average S02 level in
the flue gas of about 1,050 ppm. Since it was not possible to segregate
and fire the different coals selectively, the S02 levels in the flue gas
fluctuated daily and often hourly from a low of 600 ppm to a high of
1,550 ppm.
The flue gas load to the prototype system was kept within a range equiva-
lent to 15-23 megawatts (39,000-60,000 inlet scfm, wet basis)* and typi-
cally ran about 17 megawatts (44,000 inlet scfm). However, during Period 1
the boiler load varied from 15 megawatts to 47 megawatts with attendant
variation in excess air rates.
In addition to the unfavorable process conditions resulting from low-
sulfur coal and the high oxygen concentrations, there were frequent
process upsets. These upsets included occasional carryover of fly ash
with the flue gas and the inadvertent contamination of the lime supply
with limestone (caused by a mix-up in the lime and limestone deliveries
to the prototype systems at the plant). While the presence of fly ash
in the system caused little or no effect on operation or performance,
the limestone did produce temporary changes in the process chemistry —
particularly when a significant fraction of the calcium feed to the
reactor was limestone.
The limestone usually appeared mixed with lime at levels of up to 50% of
the total feed; however, on a number of occasions (for periods lasting
IV-3
-------
TABLE IV-2
SUMMARY OF SYSTEM INLET CONDITIONS
Operating Period 1 (2/75 - 7/75)
Range Average
a
Coal Fired!
Sulfur (wt %) 0.9-2.2 1.6
HHV (Btu/lb coal) 11,500-13,000 12,000
Chloride ( wt %) 0.05-0.10 0.08
Ash (wt %) 4.2-19.2 12.9
Inlet Gas:
Gas Load Treated (Mw equiv.) 15-22 17
S02 Level (ppm-dry basis) 600-1,550 1,050
02 Level (% dry vol.) 5.0-11.0 7.5
Particulate Loading (grs/scfd) <0.02
£L
Based on analyses of daily coal samples (as fired).
Based upon a nominal 20 Mw design load.
IV-4
-------
up to one day) the feed to the reactor system was pure limestone. During
periods when only limestone was fed to the reactor, the pH in the reactor
system dropped, S02 removal decreased somewhat and waste cake properties
deteriorated slightly. Normal process conditions were re-established
after the limestone passed through the system.
There was a general rise in the flue gas oxygen concentration over the
operating period due to worsening leakage in the boiler air preheater
section. By mid-July oxygen concentrations were in the 8-10% range,
with dilution of the already low S02 levels down to about 850-950 ppm.
At the resultant high oxidation rates (30-50% of S02 removal), active
sodium levels were inadvertently allowed to drop below 0.15M, into the
range of dilute mode dual alkali operation. Under these conditions,
soluble calcium levels in the regeneration system rose and the regen-
erated liquor became, saturated in calcium sulfate, resulting in some
gypsum scaling of the piping at the outlet of the regeneration reactor.
At the same time, mechanical problems in the scrubber system (failure
of two control valves and a pinhole leak in the absorber recycle tank)
required that the system be shut down for repairs.
Based upon this experience, it was decided that in future operations
the active sodium concentration would be maintained well above the
0.15M level by operating with a slightly higher total sodium concen-
tration in the system. This would allow the sodium sulfate concentration
to increase in the system during periods of increasing rates of oxidation.
Sulfate levels would continue to rise relative to the sulfite until the
rate of sulfate precipitation as a calcium salt equaled the rate of oxida-
tion. This can be accomplished without reverting to dilute mode operation
at oxidation rates up to about 25-30% of the S02 removal rate.
The system was shut down from mid-July until mid-September. Repairs and
revisions made during this period were of a mechanical nature rather than
involving process changes and are discussed later. After shutdown it was
decided to await the repair of the preheater leak rather than attempt to
continue operating at the low S02 concentrations and increasing oxygen
concentrations. About two-thirds of the shutdown period was to await
replacement parts for the valves that had failed.
2. Operating Period 2 — Low- to Medium-Sulfur Coal Operation
The system was put back in operation on September 16, 1975. From mid-
September to mid-October repairs were made to the air preheater during
boiler outages and adjustments were made in the boiler operation, re-
ducing flue gas oxygen levels down to the 5-6% range. For the remainder
of the test period through January 2, 1976, oxygen concentrations were
generally kept in the 5-6% range. As shown in Table IV-3, S02 levels
were slightly higher than those encountered during the first operating
period, with an average level of about 1,200 ppm (a plot of coal analyses
for daily coal samples is presented in Appendix A). The gas load to the
system was in the same range as that for Period 1.
IV-5
-------
TABLE rV-3
SUMMARY OF SYSTEM INLET CONDITIONS
Operating Period 2 (9/75-1/76)
Range Average
Coal Fired:a
Sulfur (wt %) 1.5-3.1 2.0
HHV (Btu/lb coal) 11,900-14,100 13,000
Chloride (wt %) 0.02-0.14 0.08
Ash (wt %) 8.8-21.0 13.9
Inlet Gas:
Gas Load Treated (Mw equiv.)b 16-19 18
S02 Level (ppm-dry basis) 800-1,700 1,250
02 Level (% dry vol.) 4.5-9.5 6.0
Particulate Loading (grs/scfd) <0.02 r«-
a
Based on analyses of daily coal samples (^s fired).
Based upon a nominal 20 Mw design load.
IV-6
-------
Active sodium concentrations were maintained in a range for concentrated
mode operation by the slightly higher S02 levels and lower oxygen concen-
trations and by operating with a slightly higher total sodium concentration.
Continuing improvements were also made in the mechanical performance of
system components, particularly the filter. One unprogrammed period of
regeneration with limestone occurred in this interval.
During this period, until shutdown on January 2, 1976, the prototype
system logged approximately 2,180 hours of operation, which corresponds
to 97.1% of the time the boiler was in operation.
3. Operating Period 3 — High-Sulfur Coal Operation
During the boiler maintenance outage in January and February 1976 a number
of system modifications were made. These included equipment adjustments
to improve operability and provide for additional flexibility in testing
during Period 3, as well as inspection and maintenance of all equipment.
Among the improvements were the installation of a new first-stage reactor
vessel to replace the original one (to improve mixing) and modification
of the lime slurry feed system for use during the third test period. The
original first-stage reactor had been mounted inside the second-stage
reactor and had a trapezoidal cross-section; the new reactor was of
standard cylindrical design and was mounted external to the second
reactor.
Operating conditions during the third operating period are summarized
in Table IV-4. The coal fired was a high-sulfur, Alabama coal ranging
in sulfur content from 2.4% to 5.1% and averaging 3.5% (based upon
analyses of daily grab samples). Inlet S02 levels experienced ranged
from as low as 1,500 ppm (on rare occasions) to almost 3,000 ppm.
Typical inlet S02 levels were 2,250 + 300 ppm.
Oxygen levels in the flue gas during Period 3 were similar to those ex-
perienced in Period 2 after repairs on the air preheater were completed.
Oxygen concentrations varied from 4.5% to as high as 9.0%, but typically
ranged from 5.5% to 7.5%.
Testing during Period 3 included 10 weeks of operation under relatively
stable conditions, three weeks of operation at fluctuating gas load
(30-100% of design load) and two weeks of particulate testing (with
the precipitator either partially or completely de-energized). ^Over
this period the system operated a total of 2,411 hours, or 93.2% of
the time that the boiler was in service. The total downtime during
this period is due to a number of short-term outages (e.g., to repair
a broken coupling on the agitator in the second-stage reactor).
IV-7
-------
TABLE IV-4
SUMMARY OF SYSTEM INLET CONDITIONS
Operating Period 3 (3/76-7/76)
Range Average
Coal Fired:a
Sulfur (wt %) 2.4-5.1 3.5
HHV (Btu/lb coal) 10,000-13,200 12,200
Chloride (wt %) 0.03-0.20 0.11
Ash (wt %) - 9.3-22.3 12.3
Inlet Gas:
Gas Load Treated (Mw equiv-) 6-20
SOz Level (ppm-dry basis) 1,500-2,800 2,250
02 Level (% dry vol.) 4.5-9.0 ^7
c
Particulate Loading (grs/scfd) 0.01-3.2
a
Based on analyses of daily coal samples (as fired).
b
Gas load averaged about 18 Mw equivalent during stable load testing and
13 Mw during fluctuating load testing (based upon a nominal 20 Mw design load),
c
Particulate loadings were <0.02 grs/scfd except during particulate testing.
IV-8
-------
V. SYSTEM PERFORMANCE
The purpose of installing the prototype system was to test the process
chemistry and design on this relatively small scale in order to evaluate
the viability of the process technology for large-scale utility applica-
tions. The prototype system represented a scale-up factor of about 40
over the CEA/ADL dual alkali pilot plant. The modular units for a large-
scale utility system would correspond to a further scale-up factor of 5
to 10 over the prototype system design capacity. While the overall re-
liability and operability of the process were a principal concern, the
system was not intended to be a demonstration unit nor a test of the
ultimate availability of such systems when applied full scale. The
test program at Scholz, therefore, was focused on characterizing the
process chemistry under different operating conditions, verifying
process and equipment design parameters, and defining the process capa-
bilities. Overall the system performance was excellent. During the
more than 7,100 hours of operation, the prototype system demonstrated
the capability of the concentrated mode dual alkali system to handle
flue gas from a wide range of coals (from as low as about 1.5 wt %
sulfur) with no loss in S02 removal efficiency or significant deteri-
oration in operability or waste cake properties. There were, of course,
equipment problems as would be expected with a test system. Some of
these caused short-term outages and loss of availability; however,
none were symptomatic of any deficiencies in the process technology.
The following discussion of the system performance covers the entire
17-month period of operation from startup in February f975 through
completion of the test program in July 1976. The discussion is broken
down into three parts: (1) availability; (2) process performance; and
(3) mechanical (equipment) performance. Emphasis has been placed on
process performance, since it ultimately defines the capabilities and
viability of the concentrated mode dual alkali technology. System
availability and mechanical (equipment) performance are of principal
interest with regard to how these reflect and affect the overall
projected reliability, operability and cost of the technology when
applied full-scale.
A. AVAILABILITY
In reviewing the system availability, it is worth noting that the only
spares in the prototype system were pumps. There was no redundancy or
spare capacity in any of the major process units. There was one scrubber
train, one reactor system, one lime feed system (slurry or dry), one
thickener, and one filter. Multiple trains and/or spare capacity for
the units would normally be incorporated in full-scale applications.
Such redundancy would allow downtime at convenient, scheduled times
for even the most trivial maintenance requirements without causing
loss of availability or even necessarily curtailment of power gen-
eration.
V-l
-------
Despite the lack of redundancy in major pieces of equipment and the
difficult flue gas conditions encountered during the first operating
period, the availability record of the prototype system over the entire
17 months of operation is impressive.
Boiler and prototype system operating hours are graphically displayed
by month in Figure V-l and summarized by operating period in Table V-l.
Appendix B contains a list of boiler and system outages, the duration
of each outage, and the reason for the shutdown.
The availability during the first operating period, 79.0%, includes the
initial startup of the system and the three-week shutdown period for
system adjustments and modifications prior to the start of the EPA test
program. These adjustments included normal process refinements expected
during system shakedown as well as modifications required to handle the
low-sulfur coal conditions — conditions for which the system was not
designed. During the second operating period, system availability was
97.1%; and during the third operating period, system availability was
93.2%. These availability figures correspond closely to the PEDCo
definition of system operability since they reflect only hours the
system was actually in operation.
During the two months between the first and second operating periods,
the system was down for maintenance. This outage was prolonged by
delays in shipment of spare parts and completion of repairs to the
boiler air preheater to correct excessive air leakage. In January
and February 1976 the boiler was down for six weeks for scheduled
maintenance. The system was down for an additional four weeks to
complete modifications and repairs begun in January.
The last two weeks of system downtime resulted from a delay in the return
of the filter drum which was being modified. Including these extended
shutdown periods in the availability calculation, overall system avail-
ability for the entire program was 70.2%. Exclusion of these two interim
periods from the calculation yields availability during operating periods
of 88.6%.
B. PROCESS PERFORMANCE — STABLE LOAD OPERATION
1. Overall Operation
The general process conditions for the three operating periods are sum-
marized in Tables V-2, V-3, and V-4. Control plots (plots of operating
conditions versus time) for the entire 17 months of operation are given
in Appendix C. Appendix D shows process flows and stream compositions
for typical days during each operating period. Appendix E gives corre-
lations of solution composition with pH based upon analyses of the
scrubber bleed and reactor effluent liquors.
The system operating philosophy during Period 1 was to maintain the active
sodium concentration in the system at or above 0.2M Na+ and to maintain a
V-2
-------
800
600
<
O
I
-------
TABLE V-l
SUMMARY OF PROTOTYPE SYSTEM AVAILABILITY
Availability
Operating Hours Prototype Hours
as percent of
Dates Prototype Boiler8 Boiler Hours
Operating Period 1 2/3/75-7/18/75 2,537 3,211 79.0
Interim Period . 7/18/75-9/15/75 0 1,460 0
Operating Period 2 9/16/75-1/2/76 2,180 2,244 97.1
Interim Period 1/2/76-3/15/76 o 658 0
Operating Period 3/16/76-7/3/76 2,411 2,588 93.2
Total Operating Periods 7,128 8,043 88:6
Total Program 7jl28 10,161 70.2
Hours boiler available to dual alkali system
V-4
-------
TABLE V-2
SUMMARY OF SYSTEM OPERATING CONDITIONS - PERIOD 1
(3/75 - 6/75)
Range
Inlet Gas:
Q
Gas Load Treated (Mw equiv.)
S02 Level (ppm - dry basis)
02 Level (% dry vol.)
Regenerated Liquor Composition:
PH
Na+act (M)
SO^ (M)
Cl~ (ppm)
I |
Ca (ppm)
^ased upon 20 Mw (nominal) design basis.
Hot including periods when limestone fed to reactor.
^a"1"1" levels above 250 ppm occurred during the period just prior to and
during the dilute mode operation.
15-22
600-1,550
5.0-11.0
b
Range
10-12.6
0.10-0.35
0.60-1.05
3,000-7,000
20-800+C
Average
17
1,050
7.5
b
Typical
11-12.5
0.2-0.25
0.7-0.9
4,000-5,500
50-200
V-5
-------
TABLE V-3
SUMMARY OF SYSTEM OPERATING CONDITIONS - PERIOD 2
(10/75 - 12/75)
Range
Inlet Gas:
Gas Load Treated (Mw equiv.)'
862 Level (ppm - dry basis)
02 Level (% dry vol.)
Regenerated Liquor Composition:
pH
Na+ , (M)
act
SO" (M)
Cl~ (ppm)
i j
Ca (ppm)
16-19
800-1,700
4.5-9.5
b
Range
Average
18
1,200
6.0
b
Typical
10-12.8
0.25-0.6
0.6-1.05
1,000-2,100
30-160
11-12
0.3-0.35
0.8-1.0
1,300
70
fl
Based upon 20Mw (nominal) design load.
Not including periods when limestone fed to reactor.
V-6
-------
TABLE V-4
SUMMARY OF SYSTEM OPERATING CONDITIONS - PERIOD 3. STABLE LOAD TESTING
(4/7.6 - 5/7.6)
Range
Inlet Gas:
Gas Load Treated (Mw equiv.)
SC>2 Level (ppm-dry basis)
Q£ Level (% dry vol.)
Regenerated Liquor Composition:
PH
Na+act (M)
S04 (M)
Cl~ (ppm)
Ca""" (ppm)
16-19
1,500-2,800
4.5-9.0
Range
10-12.8
0.35-0.45
0.65-0.75
7,500-11,500
(30-100)b
Average
18
2,300
6.5
Typical
11.5-12.3
0.35-0.4
0.7-0.75
9,000-11,000
(60)b
Sased upon 20 Mw (nominal) design load.
^Estimated — only a few data points taken.
V-7
-------
total sodium concentration sufficient to allow soluble sulfate concen-
trations high enough to precipitate calcium sulfate roughly at the rate
of sulfate formation. This resulted in a fairly stable operation at
about 2.0M total sodium (^ 0.8M SO") until the end of June, when the
inlet S02 level fell to 850-950 ppm and oxygen levels rose to 8.5-10.0%.
Under these conditions, without an increase in the total sodium concen-
tration by increasing the sodium carbonate feed rate, the system chemistry
drifted into the range of a dilute dual alkali mode with active sodium
concentrations dropping to below 0.15M Na+. Except for the two-week
period of dilute mode operation in July, the thickener liquor typically
contained 0.20-0.25M active sodium, 0.7-0.9M sodium sulfate, and 0.10-0.15M
sodium chloride (4,000-5,500 ppm Cl~).
At the start of Period 2, it was decided to maintain the active sodium
concentration in the-system above 0.3M Na+ and to allow the sodium sulfate
concentration to fluctuate to any level necessary to maintain an equilib-
rium between sulfite oxidation in the system and sulfate precipitation.
This would prevent any possibility of deterioration of the system chemistry
to that of a dilute mode.
With the slightly improved flue gas conditions (higher inlet S02 and
lower oxygen levels) stable conditions were achieved at sodium sulfate
concentrations in the range of 0.8-l.OM. Chloride concentrations lev-
elled out at 0.03-0.06M (1,300 ppm Cl~ average). Routine analyses of
coal samples performed for the plant from both Periods 1 and 2 show about
the same chloride content, approximately 0.08 wt % on the average. How-
ever, the accuracy of the analytical technique was only about j^ 0.05 wt %.
During the third operating period with high-sulfur coal the system was
operated under relatively stable flue gas conditions (gas load and com-
position) from mid-March to the last week of May 1976, when fluctuating
load testing was begun. At the high inlet S02 levels experienced during
Period 3, active sodium concentrations stayed above 0.35M Na+ with about
the same total sodium concentrations as in the prior two operating periods.
This reflected the lower oxidation rate (as a percentage of 862 absorbed)
and the lower sulfate concentration required to keep up with oxidation
(as will be discussed later). Chloride concentrations, though, rose to
over 10,000 ppm due to the increased chloride content of the high-sulfur
coal.
Table V-5 shows the overall system performance in terms of SC>2 removal,
makeup chemical requirements, and waste solids properties for material
balance intervals in each operating period. The performance data for
Periods 2 and 3 are based upon overall material balances derived from
flue gas analyses, waste cake properties and production rates, and
makeup chemical inventories. Complete material balances were not
attempted for Period 1 because installation of the truck scale for
weighing cake was not completed until the end of June 1975. The ac-
curacy of material balances was limited by the fluctuating gas load
and S02 levels, contamination of the lime supply with limestone (Peri-
ods 1 and 2), and occasional instrumentation problems. The accuracy
V-8
-------
TABLE V-5
SUMMARY OF OVERALL SYSTEM PERFORMANCE
f
VO
Balance Period
No . Dates
1 4-6/75
2 11-12/75
3 4-5/76
DURING
(April
SO 2
Coal Sulfur Removal Efficiency
Content (% of inlet)
Low 93a
Low/Medium 96b
High 97b
STABLE LOAD TESTING
1975 - June 1976)
Lime Feed Soda Ash Feed Waste Cake
Stoichiometry Stoichiometry Solids Content
(mols Ca(OH)9/mol ASO?) (Net mols Na9COP/mol ASO,) (wt ,%)
* 0.95 C 48
0.95 0.085 51
0.985 0.080 54
3.
Data represents weighted average of periods with venturi only
and venturi and absorber
Venturi and absorber
Q
Insufficient data
-------
of short-term material balances was also affected by the inability to
measure silo inventories accurately, particularly for sodium carbonate,
for which there was a large inventory and a low feed rate.
The various aspects of the process performance during each period are
discussed in the following sections, according to the following six
performance characteristics:
• 862 removal;
• lime utilization;
• oxidation and sulfate control;
• waste cake properties;
• sodium makeup; and
• process operability and reliability potential.
2. SOa Removal
The scrubber system was operated using two different configurations for
SC>2 removal: the venturi and absorber together in series (with two trays)
and the venturi alone. In the latter configuration the trays were not
removed from the absorber; rather, the regenerated liquor feed to the
top tray was diverted either to the absorber recycle tank (from which
it was transferred to the venturi) or bypassed directly to the venturi
through a line installed in May 1975. (The bypass line was not in the
original design, since the operation on low-sulfur coal was not antici-
pated.) When the absorber was not used, the recycle flow to the top
tray and the flow to the spray on the underside of the bottom tray were
both discontinued; however, the absorber pumps were maintained in opera-
tion to transfer liquor collected in the absorber tank to the venturi.
In order not to exert excessive back-pressure on the absorber pumps, a
recycle was maintained through an open spray header during intervals
when there was no feed to the trays. This open recycle did offer a
minimal amount of gas/liquid contacting in addition to the venturi,
since the recycle resulted in a falling film of liquor.
During Period 1, both operational configurations were used at different
times. However, during Periods 2 and 3, only the combined venturi and
absorber configuration was used.
The S02 removal efficiency achieved with each of these configurations
is shown in Figure V-2 for intervals when the inlet S02 concentration
ranged from 1,050 to 1,250 ppm (Periods 1 and 2). Figure V-3 shows
the S0£ removal achieved as a function of pH at the higher inlet SC>2
levels (1,900-2,200 ppm) in Period 3. Points shown on these plots
represent data taken during the normal course of operations when inlet
and outlet S02 levels and system pH's were simultaneously available and
V-10
-------
Active
Na+, (M)
Venturi
Ap,(in. H2O)
5-7
8-11
8-11
Operational Configuration
Venturi + 2 Trays
Venturi + 2 Trays
Venturi (No Feed to Trays)
0.25-0.4
0.15-0.3
0.15-0.3
Inlet S02 = 1050 - 1250 ppm
5.5 6.0
Scrubber Bleed Liquor pH
Figure V—2 SO2 Removal as a Function of
pH - Low/Medium-Sulfur Coal
V-ll
-------
250
200
I
100
50
2400±200
2000±200
VenturiAPC'HjO)
Inlet S02
(ppm) 4.5-7 11-12.5
•
o
A
A
4.0
4.S
5.0 5.5
Scrubber Bleed Liquor pH
6.0
6.5
FIGURE V-3 SO2 REMOVAL AS A FUNCTION OF pH-
HIGH-SULFURCOAL
V-12
-------
confirmed. In most cases, each data point represents at least a few
hours of operation at the condition shown. Where such a continuum of
data exists, outlet S02 levels and pH's have been averaged and rounded
to the nearest 5 ppm and 0.05 pH units, respectively. Table V-6 sum-
marizes the overall S02 scrubbing performance at Scholz from April 1975
through June 1976, when the outlet S02 monitor was operational.
The data in Figures V-2 and V-3 and Table V-6, which reflect the general
operating experience at Scholz, confirm the high S02 removal capability
of sodium solution scrubbing systems operating in a similar concentration
range of active sodium (0.2-0.4M Na+). Achieving a given outlet S02 level
(within the limit of the number of contact stages in use) was essentially
a matter of adjusting the operating pH of the scrubber system by changing
the feed forward rate and/or pH of the regenerated liquor. Over the 15
months of operation Between April 1975 and July 1976, the average S02
removal using both the venturi and absorber was 95.5% (with low- and
high-sulfur coal); and with the venturi alone, was 90.7% (low-sulfur
coal only).
As indicated in Figures V-2 and V-3, outlet S02 levels and S02 removal
efficiency were primarily a function of the inlet S02 level, scrubber
configuration (venturi versus venturi plus trays), pH and, to a lesser
extent, venturi pressure drop. At inlet S02 levels of 1,050-1,250 ppm,
and using both the venturi (AP = 5-11 inches of water) and absorber
(two trays), outlet S02 levels below 50 ppm could be easily achieved
at venturi bleed liquor pH's above 5.2. At these low inlet S02 levels,
a 50 ppm outlet level corresponds to better than 95% removal efficiency.
Operation at higher pH's resulted in higher S02 removal efficiency; how-
ever, there was little to be gained by operating at pH's much in excess
of 6.0. Above a pH of 6.0, outlet S02 levels dropped to 20 ppm or
less (> 98% removal), taxing the accuracy and operating range of the
outlet S02 monitor.
At higher inlet S02 levels (1,900-2,800 ppm) the experience was much
the same, although achieving the same outlet S02 level (which would
correspond to a higher removal efficiency) required a slightly higher
pH. To achieve 50 ppm outlet S02 (^ 98% removal), a pH of about 5.5
was required at about 2,400 ppm inlet S02 compared with a pH of 5.2
at 1,200 ppm.
For the most part, when both the venturi and absorber were operated to-
gether, the pH of the venturi bleed liquor was maintained between 4.8
and 5.9 to ensure better than 90% S02 removal. With the low inlet S02
levels of Periods 1 and 2 (600-1,700 ppm) this resulted in outlet S02
generally ranging from 15 to 100 ppm. At the higher inlet S02 levels
of Period 3 (April and May 1976), the outlet S02 typically ranged from
25 to 150 ppm.
The effect of venturi pressure drop on S02 removal with both trays in
operation is also shown in Figures V-2 and V-3. At low inlet S02 levels
there was no discernible effect of venturi pressure on S02 removal over
V-13
-------
TABLE V-6
SO? REMOVAL PERFORMANCE DURING STABLE LOAD OPERATION
i
*-
Operating Period
No. Dates
1 5/12-6/11/75
4/2-4/13/75 |
6/12-6/30/75/
2 9/15-12/31/75
3 3/1-5/31/76
Scrubber Conditions
Approximate Venturi
Operating Hrs. AP ("H^O)
520 5-11
610 8-12
2,100 5-11
1,650 4-7
No. Trays
2
0
2
2
Inlet SO 9
Range
850-1,400
690-1,310
800-1,750
1650-2,800
a a
(ppm) Outlet SO, (ppm)
Overall
Avg . " Range
1,120 5-370
1,000 5-450
1,220 5-320
2,300 10-470
Typical
Range
15-100
20-180
15-100
25-150
Avg. S02 Removal,
Avg. Efficiency (%)
50 95.5
90 90.7
60 95.2
95 95.9
basis
Integral average weighted by gas flow
-------
the range of pressure drop from 5 to 11 inches of water. At high inlet
relvaT6 ff Jni?eri0£ V1161" r a SmaU but "^"icant increase in S02
removal efficiency by increasing the venturi pressure drop from 4 to 7
inches of water to about 12 inches of water. The increase in pressure
drop reduced outlet loadings by 10 ppm to 30 ppm over a PH range of
4.3 to 5.5.
Use of the venturi alone for S02 removal (no feed of regenerated liquor
to the absorber) was tested only at the low inlet S02 levels of Period I
With just the venturi, 95% S02 removal efficiency (less than 50 ppm out-'
let S02) required a bleed liquor pH of about 6.5. However, better than
90% removal could still be quite easily achieved at pH's on the order
of 6.0. When the venturi alone was used, the bleed liquor pH was gen-
erally maintained above 5.7 to keep outlet S02 levels at or below 100 ppm.
Appendix F contains estimates of the stage efficiencies in the scrubber
system. The analysis is based upon operation with both the absorber and
venturi during stable periods at high and low inlet S02 levels.
3. Lime Utilization
Lime utilization throughout all three operating periods was quite good.
Under normal conditions, lime utilization ranged from 90% to 100% of the
available Ca(OH)2 in the raw, hydrated lime, and typically ran about 94%.
The Ca(OH)2 fraction of the delivered lime ranged from 87% to 93% (wt %
basis).
Table V-7 summarizes the estimates of lime utilization and overall lime/
AS02 feed stoichiometry for balance intervals during each of the three
operating periods. Feed stoichiometries are calculated from integral
average S02 removal rates, and from lime deliveries and lime silo in-
ventories. Lime utilization estimates have been calculated by two
methods — first, from the analyses of the filter cake solids; and
second, from the overall feed stoichiometry taking into account sodium
carbonate feed (and sodium-sulfur losses). The small differences between
the two different estimates of utilization indicate the closure of the
material balance. And the difference between the lime feed stoichiometry
and unity reflects both utilization and soda ash makeup. Lower lime
utilization increases the stoichiometry, tending toward values greater
than unity; higher soda ash feed rates decrease the stoichiometry, tend-
ing to reduce stoichiometry below unity.
Accurate estimates of lime/AS02 stoichiometry are not possible for extended,
uninterrupted intervals during operating Period 1. The outlet S02 monitor-
ing was not operational until April 1975, and accurate accounting of lime
deliveries and silo inventories was not initiated until June 1975, when
the filter cake weigh scales were installed. In addition, there were
frequent periods when unknown quantities of limestone were inadvertently
charged to the lime silo. Therefore, estimates of lime utilization from
filter cake analyses during periods when little or no limestone contami-
nated the lime supply have been used to back-calculate lime feed stoichiometry.
V-15
-------
TABLE V-7
LIME UTILIZATION DURING STABLE
Balance
Period Interval
1 4
2 11/3
3 4/11
- 6/75
- 12/23/75
- 5/23/76
Coal
Fired
Low-Sulfur
Low /Medium-Sulfur
High-Sulfur
Approximate
Duration of
Balance
Period
(hours)
(2,000)
1,050
840
LOAD OPERATION
Lime Feed
Stoichiometry
(Ca(OH)2/AS02)
(^0.95)
0.95
0.985
Lime Utilization
(% of Ca(OH)9)
From Lime
Feed
Stoichiometry
96.0
93.0
Filter Cake
Analysis
(92-98)
96.5
93.0
H-1
-------
In Period 2, detailed material balances were computed for the last two
months -- November and December 1975. The overall lime/AS02 feed stoi-
chiometry of 0.95 and lime utilization of 96-97% (based upon available
Ca(OH)2) are representative of the entire three and one-half months of
Period 2.
The balance interval for Period 3 covers the five-week period beginning
the second week of April when stable system chemistry was achieved and
ending in May just prior to starting the fluctuating load testing. The
closeness of the lime utilization estimates calculated both from the
overall lime/AS02 stoichiometry and from the analyses of filter cake
samples reflects the accuracy of the material balance during this period.
The slightly lower utilizations estimated for Period 3 with the high-sulfur
coal (93%) versus that for Period 2 with the mixed low- and medium-sulfur
coal (-v96%) does not represent any basic change in the system chemistry
or capability of the technology. In fact, utilization would be expected
to improve with higher sulfur content in the coal due to lower soluble
sulfate concentrations (lower system oxidation rates). The slightly
lower utilization figures for Period 3 can be attributed to two factors:
first, the average extent of regeneration (pH) during Period 3 (pH =
11.5-12.3) was slightly higher than in Period 2 (pH = 11-12); second,
the reactor holdup time during Period 3 ranged from 20-50 minutes and
averaged about 30 minutes, while during Period 2, the average holdup
time was longer, averaging closer to 40-45 minutes. The shorter holdup
time in Period 3 (due to higher liquor rates with the higher-sulfur coal)
in combination with the slightly higher operating pH would have a tendency
to depress lime utilization slightly. The higher chloride content (0.25-
0.35M Cl~ in Period 3 versus 0.03-0.06M in Period 2) may also have had a
minor effect, although prior work at ADL has not shown any significant
difference due to chloride at these concentrations.
It is worth noting that during Periods 1 and 2, lime was fed to the reactor
system only as a dry solid, while in Period 3, lime was fed at different
times as a dry solid or as a slurry containing from 12% to 32% solids.
There was no discernible effect of the difference in the form of lime
feed during Period 3.
4. Oxidation and Sulfate Control
a. Oxidation
As would be expected, the major portion of the sulfite oxidation in the
system occurred in the scrubber circuit, and the single most important
variable affecting the oxidation rate was the oxygen content of the flue
gas. Estimated oxidation rates in the scrubber circuit ranged from a low
of about 150 ppm equivalent S02 (at 4.5-5.0% oxygen in the flue gas) to
as high as 450 ppm (at oxygen levels up to 9-10%). By contrast, oxidation
throughout the remainder of the process is estimated to have been 25-40 ppm.
Thus, in general, oxidation in the scrubber circuit accounted for more than
85% of the total system oxidation.
V-17
-------
Figures V-4 and V-5 are plots of estimated oxidation rates in the scrubber
system in equivalent ppm of S02 (design gas flow basis) as a function of
oxygen content of the flue gas. Design gas flow is used as a normalizing
factor since oxidation was not significantly affected by small changes in
gas flow.
Data in Figure V-4 cover operation with mixed low- and medium-sulfur coals
and both scrubber configurations used during Periods 1 and 2. Figure V-5
shows data with high-sulfur coal operation during Period 3 using the
venturi/two-tray combination. The data shown were screened to ensure
relatively stable operation (gas flow and oxygen content, and liquor
flow and composition) to minimize random data scatter.
It would be expected that oxidation rates would be somewhat lower using
the venturi alone than with the combined venturi and absorber, due to
the decreased gas/liquid contacting. However, such a decrease in oxida-
tion is not apparent in Figure V-4. This can be attributed to three
factors: first, the carryover of entrained liquor from the venturi to
the absorber; second, contacting of gas with liquor recirculated through
the open spray header; and finally, the flow of a small amount of regen-
erated liquor onto the top tray through a leaky shutoff valve.
Considering all data regardless of scrubber configuration, the data scatter
exhibited in Figure V-4 amounts to a range of 80 ppm of oxidation at a
given flue gas oxygen level. This data scatter can be accounted for by
differences in operating temperature (120-135°F), liquor flow, slight
differences in pressure drop, and error in the oxygen analyzer, as well
as sampling and analytical errors.
Assuming that the median value of oxidation represents the average oxida-
tion experienced, then oxidation in Periods 1 and 2 ranged from 175 +^ 40 ppm
at 5% oxygen to 380 + 40 ppm at 9% oxygen. For a typical S02 level of
1,200 ppm with 95% S02 removal and 90% gas load, these oxidation rates
correspond to 17% and 37% of the S02 removed, respectively. Adding in
the oxidation through the remainder of the system, the total oxidation
amounted to 20% of the SC>2 removed at a 5% oxygen level and 40% at 9%
oxygen. It should be noted that these oxidation rates represent opera-
tion with three active contact stages (apparently even when the venturi
was operated alone). Normally in a low- or medium-sulfur coal application
only one, or at most, two stages should be required. This reduction in
the number of stages would significantly reduce oxidation rates.
The oxidation data plotted in Figure V-5 for high-sulfur coal operation
show an increase of about one-third in oxidation over that observed with
the lower-sulfur coals. However, as a percentage of the SC>2 absorbed,
there is a significant decrease owing to the higher inlet S02. At a 6%
oxygen level the estimated median oxidation rate with high-sulfur coal
in Period 3 was about 270 ppm of 802(^13% of the S02 removed), compared
with 210 ppm operating with the mixed low- and medium-sulfur coals in
Periods 1 and 2 (20-25% of the S02 removed). At 8% oxygen the estimated
average oxidation rate in Period 3 was about 430 ppm of S02 (v/21% of the
V-18
-------
_ 500
lrt
1C
00
« 400
(D
a
c
D)
'in
CD
a
1 300
CM
O
C/5
"o
£
n
3 200
c
g
CO
'x
£ 100
•£
W
o
I T 1 1
-
/
D^ D >*
/ ^^
/ D ^^ _
• ^x^
^ ^^r
''' ' ^SQuu **''
^^^ sj^s^' u
^-•^* BX^ * ^ -
**•* ***"^* • ^^^ ^
^ T Q ^^^ B Q ^^
• • ^^^ ^^ _^
^" •""*•* • 1"^
. • ^-^
• • •"""*
"*" Venturi Operating
Operational Configuration AP (in. h^O) Period ~
• Venturi + 2 Trays 5-7 2
• Venturi + 2 Trays 8-11 1
D Venturi (No Feed to Trays) 11-13 1
i iii
0.07
0.06
0.05
-£
.—
^>
"• —
_^2
0.04 I
J3
C
g
0.03 |
._
\f
O
0}
^:
0.02 15
0.01
0
4 5 6 7 8 9 10
FIGURE V-4
Oxygen Content of Flue Gas (% dry vol.)
OXIDATION IN THE SCRUBBER SYSTEM AS A FUNCTION
OF FLUE GAS OXYGEN CONTENT (PERIODS 1 AND 2)
V-19
-------
500
O
c
O)
TJ
I
-------
S02 removed), compared with 310 ppm in Periods 1 and 2 (29-37% of the AS02
removed). This increase in the absolute oxidation rate in Period 3 is
not readily explainable. It may have been due, in part, to the higher
active sodium levels (0.25-0.35M Na+active in Period 2 versus 0.35 to
0.45M Na actlve in Period 3). This would be a factor if oxidation were
partly reaction rate limited. Subtle differences in trace contaminants
in the flue gas due to differences in coals may also have played a part.
The oxidation of sulfite is known to be sensitive to the catalytic effects
of trace levels of heavy metals and inhibitory effects of certain organics.
It is important to note, though, that the oxidation levels experienced
in Period 3 are well within the range that can be handled by the system
capability for sulfate precipitation.
b. Sulfate Precipitation
At steady state, the total sulfate being formed must be removed from the
system at an equivalent rate by the precipitation of calcium sulfate and
attendant sodium sulfate losses in the filter cake. Calcium sulfate levels
measured in the product solids ranged from as low as a few percent of the
total insoluble calcium-sulfur salts (during early startup) to as high as
30% during periods of high oxygen levels and low inlet S02 (when the ratio
of sodium sulfate to active sodium in the system liquor was fairly high).
The typical range of calcium sulfate in the waste cake during Period 1
was 15% to 25% (mol basis) of the total calcium-sulfur salts; in Period 2,
calcium sulfate in the cake ranged from 12% to 20%; and in Period 3, the
calcium sulfate fraction in the cake ranged from 8% to 17%.
These levels of calcium sulfate precipitation indicate that the system
is capable of keeping up with oxidation rates of up to 25-30% of the S02
removal in the concentrated mode of operation without excessive sodium
losses. In operations where the cake is thoroughly washed, essentially
all of the sulfate losses would be as calcium sulfate. Sodium sulfate
losses, which usually amount to 50-80% of the total sodium value occluded
in .the cake, would represent less than 2% of the S02 absorbed. Depending
upon the levels of oxidation experienced and the degree of cake washing,
the system liquor composition would simply readjust so that the ratio of
sodium sulfate to active sodium would allow precipitation of the amount
of calcium sulfate required to keep up with oxidation.
Figure V-6 shows the relationship between the ratio of calcium sulfate
to calcium sulfite in the waste solids versus the ratio of soluble sulfate
to sulfite in the regenerated liquor. The data plotted are taken from
stable operating periods lasting at least a few days. The dashed line
represents the relationship derived from the pilot plant operations for
calcium sulfate in the reactor effluent solids. The data for the calcium
sulfate content of the reactor effluent solids in the prototype system
are slightly displaced from the pilot plant correlation. Based upon the
data shown in Figure V-6, the correlation of the sulfate/sulfite content
of the precipitated calcium salts to the sulfate/sulfite concentrations
V-21
-------
0.30
0.25
to
.c
a.
&
c
o
3
CO
ro
O
0.20
0.15
0.10
0.05
Pilot Plant
Correlation
Scholz Filter Cake
Scholz Reactor
Effluent Solids
Reactor Effluent Solids
Filter Cake Solids
I
I
4 6 8 10
Mols Na2S04/mols Na2S03 in Regenerated Liquor
12
FIGURE V-6 (CaS04/CaS03) RATIO IN WASTE SOLIDS
AS A FUNCTION OF (Na2SO4/Na2SO3)
RATIO IN REGENERATED LIQUOR
V-22
-------
in the reactor liquor can be approximated by the following.relationship:
mols CaSOij \
- 0.031 T^rr (l)
solids \ LbU3J /
reactor liquor
This level of sulfate precipitation is about 85% of that observed in the
pilot plant operations.
There was a small, but significant decrease in the calcium sulfate in the
filter cake compared with that in the reactor solids. This corroborates
previous laboratory results indicating that a small amount of calcium
sulfate apparently does redissolve on prolonged contact with the regen-
erated liquor. The redissolution of calcium sulfate amounted to an
average of about 13% of the total calcium sulfate originally precipi-
tated. This was determined during a particularly stable two-week interval
in Period 3 when the CaSC\/CaSOx mol ratio in the reactor solids averaged
11.7% (CaSQ^/CaSQz = 0.132) and the filter cake contained a CaSO^/CaSOx mol
ratio of 10.1% (CaSO^/CaSOs = 0.112). This differential corresponds to a
decrease of 13.5% in calcium sulfate. A small portion of the decrease is
due to continued reaction and precipitation of calcium sulfite at the
higher reactor operating pH's. The data shown in Figure V-6 suggests
about a 15% drop in calcium sulfate in the cake from that in the reactor
effluent solids.
Sulfate balances were performed over three separate intervals in Periods 2
and 3, each lasting three to five weeks. Table V-8 summarizes these bal-
ances. The estimated total system oxidation shown is the sum of the oxida-
tion in the scrubber system (based upon the average flue gas oxygen content
over the interval) and in the combined reactor/dewatering systems. The
sulfate account is based upon the average filter cake analyses and total
cake discharged, sodium inventory in the liquor and the total soda ash fed.
In all cases the relative error in the sulfate balance (difference between
sulfate formation and sulfate account) is less than 10%. While this alone
does not guarantee a high degree of accuracy in the various estimates of
oxidation, soda ash feed, and calcium sulfate precipitation, it does in-
dicate that overall the material balances are substantially correct.
During November, 20% of the S02 absorbed was oxidized to sulfate. Of
this, 13% left the system as insoluble calcium sulfate. The remaining
7% (as calculated from the total soda ash feed and the composition of
the system liquor) left as sodium sulfate. Based upon analyses of filter
cake samples, about half of the sodium sulfate was lost from the system
in the liquor occluded in the filter cake (^5% sodium salts on a dry cake
basis). The other half of the sodium sulfate losses are unaccounted for.
Leaks from pump seals and tanks, and entrainment losses (which are dis-
cussed later) were all quite small and realistically could only amount
to, at most, half of the unaccounted for losses. The remaining 2-3% of
V-23
-------
TABLE V-8
SYSTEM SULFATE BALANCES
Average Inlet S02 (ppm)
Average S02 Removal (%)
Average Inlet 02 (vol. %)
PERIOD 2 PERIOD 3
11/3-11/23/75 12/2-12/23775 4/11-5/23/75
1,265
95.5
5.5
1,135
96.5
6.0
2,335
96.4
6.5
In Cake -
Otherc
ate Formation (%AS02)
count (%AS02)':
- CaSOi,
NazSOt,
a
in System Liquor Inventory
T/i*-o1
20.0*i
13.0
4.5
-0.5
4.0
on n *
23.0*
15.0
3.0
2.5
3.0
97 R .*
l/.^<
10.5
5.0
0.0
1.5
1 7_n<
A negative value represents a decrease in sulfate concentration in the
system liquor over the balance interval.
Other represents the sodium sulfate losses from the system that are not
accounted for by cake losses and changes in sodium sulfate inventory
(accumulation or depletion) in the system liquor. Other includes sulfate
losses from pump seal leaks and inadvertent spills and is calculated inde-
pendently of the total sulfate balance from the soda ash makeup rate and
the average solution chemistry, as follows:
Other
Soda Ash Makeup x
( Average solution sulfate concentration
\ Average total Na ~H" concentration
- sodium sulfate losses in the cake - accumulation of sodium
sulfate in the system liquor
Since the estimate is based upon the soda ash makeup, it inherently includes
any errors in the estimate of the soda ash feed.
V-24
-------
the sodium losses, therefore, may be due to unreported spills or, more
likely, to inaccuracies in the estimates of soda ash makeup and poor
washing of the filter cake during some periods when it was not sampled.
In December, due to the slightly higher average oxygen content of the
flue gas and the slightly lower inlet S02 levels, 23% of the S02 absorbed
was oxidized to sulfate. About 15% left the system as insoluble calcium
sulfate in the filter cake; the remainder was accounted for by sodium
sulfate losses based upon the total soda ash feed. The losses included
3% as sodium sulfate in the filter cake (^3.5% sodium salts - dry cake
weight); 2.5% sodium sulfate accumulation in the liquor inventory; with,
again, 3% unaccounted for. Because of the relatively large unaccounted
for sodium losses during Period 2, attempts were made during the third
balance interval (in Period 3) to ensure good, representative cake sampling
and minimize (or report accurately) spills and leaks.
With the high-sulfur coal in Period 3 and oxygen concentrations ranging
from 5.5% to 9% in the flue gas (averaging about 6.5%), total system
oxidation averaged about 17.5% over the five-week material balance in-
terval. The calcium sulfate fraction of the filter cake averaged 10.5%
of the AS02 and sodium sulfate in the cake averaged about 5.0% of the AS02
Other losses of sodium sulfate as calculated from the total soda ash makeup
amounted to 1.5% of the AS02. Of this 1.5%, approximately 0.25% is esti-
mated to have resulted from known spills (0.2%) and entrainment losses
(0.05%). The remaining 1.25% is unaccounted for and represents pump
seal and piping leaks; unreported spills; and perhaps most importantly
errors in the estimate of the total soda ash makeup, and errors in cake
loss estimates. This 1.25% discrepancy is much lower than the discrep-
ancy in Period 2 and reflects the efforts made to control and monitor
the sodium balance.
If all system leaks had been stopped or returned to the system and the
cake had been washed to the same level of soluble solids, then the
sulfate-to-active-sodium ratio in the liquor would have increased
slightly (as shown in Figure V-6) to allow more calcium sulfate to
precipitate, thereby re-establishing the equilibrium between sulfate
formation and losses.
5. Waste Cake Properties
a. Solids Content
The solids content of the filter cake produced throughout the program
typically ranged from 45% to 60% solids. However, over all operating
conditions encountered during the 17 months, filter cakes containing
as high as 71% and as low as 41% solids were produced. For the most
part, during normal filter operation, the filter cake had the consis-
tency and handling properties of a moist powder, as shown in the photo-
graphs in Figures V-7 and V-8. Table V-9 summarizes the filter cake
properties during each stable load period.
V-25
-------
f
FIGURE V-7 FILTER STATION - SHOWING FILTER CAKE
IN STORAGE/TRANSFER BIN BELOW FILTER
-------
ro
FIGURE V-8 LOADING FILTER CAKE FOR
TRANSPORT TO DISPOSAL POND
-------
TABLE V-9
SUMMARY
Range
(2-7/75)
Coal Sulfur Content (wt %) 0.9-2.2
Filter Cake Properties:
Solids Content (%) 41-77
CaSO^/CaSO (mol ratio) ^0.02-0.31
X
Solubles (wt % dry cake)
"f No. Displacement Washes
ho
oo
No. of Spray Banks
OF FILTER CAKE PROPERTIES
i
Average Range
(5-6/75) (9-12/75)
1.6 1.5-3.1
48a 45-71
(0.17)c 0.08-0.17
(8)C
(0.5-l)d
2
0 T
Average Range Average
(11-12/75) (3-5/76) (4-5/76)
2.1 2.4-5.1 3.7
.
51a 48-62 54b
0.15 0.05-0.14 0.11
4 — 5
"2 — 1.8
3 3&4
a Includes intervals of limestone contamination of lime supply.
b Includes intervals of filter testing (drum speed, filter cloth, wash sprays).
c Estimated based upon analyses of 8 to 12 cake samples.
d Estimated from available wash spray data.
-------
In general, solids content (and, therefore, cake quality) was a function
of three factors:
• normal variations in process chemistry — principally the
level of calcium sulfate in the waste solids;
• filter operating variables — principally drum speed and
type of filter cloth, both of which were varied during the
program; and
• system upsets — including process upsets such as contamination
of the lime feed supply with limestone and gross overfeeding of
lime, as well as mechanical problems with the filter which affected
the filter performance or with the reactor system which affected
solids properties, (see Section V.E., MECHANICAL PERFORMANCE).
Data shown in Table V-9 include all of these effects for the operating
intervals reported.
Most system upsets, both process and mechanical, occurred during the
first operating period. While these upsets sometimes reduced solids
content of the filter cake by as much as 5-10%, for the most part they
were discrete incidents with identifiable causes (e.g., limestone con-
tamination of lime, cracking of plastic support grids in filter, etc.)
peculiar to the Scholz operation and with little significance to full-
scale applications. The two most important upsets in terms of effects
on filter cake properties were the presence of large quantities of lime-
stone in the lime feed supply and the mechanical problems with the filter
that reduced filter vacuum. During Periods 2 and 3 there was only one
incident of limestone contamination and there were few mechanical failures
in the filtration system.
While the data in Table V-9 include upset conditions, the general effect
of the level of sulfate content of the cake (or system oxidation) is
apparent. As sulfate precipitation (oxidation) increased, the solids
content of the filter cake decreased. Although the effect is small, it
is real and confirms a similar trend observed in the pilot plant operations
(see Volume II). A quantitative correlation of filter cake solids content
with either calcium sulfate fraction or soluble sulfate-to-TOS ratio similar
to that developed from pilot plant operations is not possible, though, due
to the fluctuations in operating conditions and variations in filter vacuum
resulting from filter mechanical problems and filter testing.
b. Solubles
Wash efficiency tests conducted at Scholz during Period 1 demonstrated
that the soluble solids content of the cake could be readily reduced to
2-3% (dry cake basis) under controlled conditions using a wash ratio of
about 2.5 (gals wash water/gal water occluded in the cake). These tests
were performed using two spray banks in series. Two banks provided
sufficient water to achieve a wash ratio of up to 4 at the low cake
V-29
-------
discharge rates normally experienced with the low-sulfur coal in Period 1.
However, mechanical problems and improper filter operation frequently pre-
vented a reasonably continuous operation of the filter with a thin enough
cake to allow consistently high wash ratios (> 1.5) required for low solu-
bles losses. There were also periods when the cake was not washed. In
Period 1, the levels of solubles in the filter cake ranged from 2% to 12%
(dry cake basis) and typically ranged from 5% to 8%. The 12% solubles
occurred during periods when the cake was not washed.
In order to increase the available wash water for periods o'f high solids
withdrawal, a third bank of spray nozzles was added to the filter prior
to the start of the second operating period. Then, following the first
month of operation in Period 2, the capacity of the sprays was increased.
These adjustments roughly doubled the total wash capacity from 3-3.5 gpm
to 7 gpm. As a result, in November 1975, the level of soluble solids in
the cake averaged 4.8% (dry cake basis) with a wash ratio of 1.5-2.0. In
December, the wash rate was increased to a wash ratio of 2.0-2.5 and the
average soluble solids level in December decreased to 3.3% (a range of
1.5% to 6.0% dry cake basis). The solids content of the filter cake
during both months averaged 51%.
In anticipation of higher solids production rates during higher-sulfur
coal testing in Period 3, a fourth bank of high capacity sprays was added
to the filter during the interim between Periods 2 and 3. This resulted
in a further increase in the wash capacity from 7 gpm to about 10 gpm.
Some difficulty was experienced with the higher capacity nozzles. The
high water velocities from these full-cone nozzles tended to wash cake
off the cloth, which contributed to cloth blinding. These nozzles were
later replaced and total wash capacity reduced to 8-9 gpm. Improved
mechanical performance of the filter following overhaul, though, did
allow a fairly continuous operation of the filter and wash ratios were
typically maintained between 1.5 and 2.0. The average wash ratio during
April and May, as calculated from total cake discharged and total wash
water fed, was about 1.8. The average solubles loss in the cake with
this wash rate was 5% (dry cake basis). The range of soluble levels
experienced ran from 1.5% to a high of 11.5% (very little washing).
As in other periods of operation, the solubles content of the cake varied
with cake thickness. Even during stable, continuous operation, differences
in cake thickness across the filter resulted in a range of sodium in the
cake discharged from different points of the filter. Typically, the cake
at the ends of the drum (6-8" on each end) ran about twice the thickness
of the cake in the middle of the drum and usually contained 1.5 to 2 times
the solubles levels. This difference in cake thickness was accounted for
by taking a mixed sample of different thicknesses or taking what appeared
to be the average cake thickness. Occasionally, a number of samples from
across the cloth were taken as a check.
V-30
-------
c. Filter Testing
Filter testing was conducted at Scholz both on the Dorr-Oliver vacuum
drum filter and on a pilot-scale rotary drum vacuum filter supplied by
the Bird Machine Company. The filter testing performed with the Dorr-
Oliver unit, over and above that related to solving mechanical problems,
involved the evaluation of different filter cloths, and general optimiza-
tion of filter control (drum speed, wash efficiency, feed slurry concen-
trations, etc.).
The filter cloth testing began midway through the first operating period
and continued throughout the remainder of the program. The purpose of
the testing was to minimize or eliminate blinding and increase cloth
life without sacrificing high cake solids content or filtrate clarity.
A number of different nylon, polyester, and polypropylene cloths in
both monofilament and multifilament weaves were tested. Porosities of
the cloths tested ranged from 10 cfm to 100 cfm (air flow per square
foot of cloth measured at a pressure drop of 1/2 inch of water). Over-
all, the polypropylene monofilament with a porosity of 20-50 cfm was
found to be most suitable. The cake release was very good and the
filtrate clarity was excellent (at most, a few hundred ppm of suspended
solids). Depending upon the care taken during installation and opera-
tion, this cloth had a life up to a month or more. Based upon this
experience, a projected life of at least a few months on a full-scale
system would be reasonable. The multifilament, simply because of the
greater strength imparted by the weave, was more durable but tended to
blind more readily (requiring cleaning of the cloth two to three times
per week). Nylon was not acceptable at Scholz due to the occasional
excursions to low pH.
Testing with the Bird filter was performed during May and June 1976.
While the Bird filter design was significantly different from that of
the Dorr-Oliver filter, the performance was roughly equivalent in terms
of filtration rate, dewatering capability, and wash efficiency. There
was a slightly higher carry-through of solids into the filtrate which
has been attributed to the thinner cake with the Bird filter. The
thinner cake allows more fines to be pulled through the cloth. Suffi-
cient time was not available, though, to optimize the Bird filter fully,
particularly with regard to the type of cloth.
6. Sodium Makeup
The sodium makeup requirement in the operation of a dual alkali system
is determined simply by the rate of sodium loss, both controlled (solu-
bles in the filter cake) and uncontrolled (pump seal leaks, tank spills,
etc.). Under normal operating conditions in a tight, closed-loop opera-
tion the losses in the filter cake are the single most important sodium
loss, and sodium makeup should equal or only slightly exceed the quantity
contained in the cake (on a mol equivalent basis).
V-31
-------
As previously indicated, wash efficiency tests performed at Scholz during
Period 1 showed that 2-3% solubles losses in the cake (weight %, dry cake
basis) could be easily achieved with a wash ratio of 2.5, and that solids
contents as low as 1.5% could be attained at slightly higher wash ratios.
Solubles losses of about 1.5% probably represent a reasonable lower limit
for continuous operation because of limitations in water balance and the
need to purge chlorides at the rate at which they enter the system. These
solubles losses translate almost directly into an equivalent soda ash
makeup requirement because of similarities in molecular weights of the
sodium and calcium-sulfur salts. A 1.5% solubles loss usually corre-
sponds to a soda ash makeup requirement of between 1.5% and 2.0% of the
S02 absorbed (mols Na2C03/mol AS02). Soda ash makeup to the system was
consistently higher than this lower level throughout the operation.
During Period 1, mechanical problems with the filter and insufficient
washing limited the level to which soluble sodium salts could be washed
from the cake over an extended period to a minimum of about 5%. In
addition to the cake losses, there were also inadvertent spills as new
operators gained familiarity with the system, as well as the normal
leaks from pump packing, piping, etc. Thus, soda ash makeup require-
ments frequently exceeded 10% of the S02 absorbed on a mol basis during
Period 1.
The improvements in the filter operation and increased wash efficiency
in Period 2 substantially reduced sodium losses in the cake to 4.8%
solubles in November 1975 and 3.3% solubles in December. These sodium
losses correspond to soda ash makeup requirements equivalent to 5.8%
and 3.6% of the S02 absorbed, respectively. The net soda ash makeup
(not including that used in increasing the inventory of sodium in the
system), however, was 10.5% in November and 7% in December. The higher
makeup requirement in November was partially due to contamination of
the lime with 25 tons of limestone, decreasing the washability of the
waste and increasing the amount of waste solids and associated sodium
losses.
The apparent difference of 3-4% between the estimated soda ash feed and
sodium losses in the cake during November and December can be attributed
to leakage from the system, errors in estimates of the soda ash feed,
and sampling error in the cake. Very few samples were taken during
evening shifts when there was a greater tendency for improper filter
operation, including poorer washing.
Entrainment losses of sodium were very small. In particulate sampling
conducted during December (by Guardian Testing Services), the total
weight of sodium salts in the scrubbed gas averaged 0.002 grains/scf.
This represents a liquor loss through entrainment of less than one
gallon per hour. The soda ash makeup required to replace this entrain-
ment loss is less than 0.1 mol % of the S02 absorbed. Additional gas
sampling performed during Period 3 (June-July 1976) showed even lower
entrainment losses of sodium than measured during the Guardian testing
(see Section V.D. PROCESS PERFORMANCE - PARTICULATE TESTING)
V-32
-------
Because of the inability to close the sodium balance during Period 2,
efforts were made to minimize the potential for accidental spills and
leaks during Period 3, and to monitor soda ash makeup, cake losses and
cake washing more accurately. As an aid in monitoring cake wash water,
a wash water totalizer was installed prior to Period 3.
Analyses of cake samples throughout April and May (in Period 3) indicated
a solubles loss in the cake averaging about 5% of the total dry cake
weight. Based upon total cake withdrawn from the system, this corre-
sponded to a soda ash makeup requirement of about 7.0% of the S02 ab-
sorbed (mols NazCO3/100 mols ASOz) — a requirement which was about
0.5-1.0% higher than normally expected due to the high quantity of
waste solids withdrawn compared to S02 removed- The higher estimated
cake rate in part can be attributed to a higher lime feed stoichiometry
in Period 3 (slightly decreased utilization as previously discussed) and
the low available Ca(OH)2 in the lime; but it also reflects errors in
estimates of S02 absorbed, in recorded cake weights, and small errors
introduced in data averaging.
Despite these errors, the 7.0% is still quite close to the net soda ash
makeup rate of 8.0% of the S02 absorbed estimated for April and May. As
in Period 2 the difference of 1.0% cannot be explained by pump seal leaks
and known (small) spills; rather this difference reflects errors in cake
sampling and inadequate washing of the cake during periods when the cake
was not sampled (evenings and filter startup and shutdown). The tendency
in the filter operation was to underwash the cake and for wide fluctua-
tions in wash ratios. While the average wash ratio may have been 1.8,
extended operation at a wash ratio of 1.0 would require an equivalent
period of operation with a wash ratio of about 5 to result in the same
total sodium losses as with a steady wash ratio of 1.8. This effect is
magnified by the fact that most periods of higher wash ratio were when
less cake was being discharged.
Problems of inadequate washing can be avoided to a great degree in larger
systems simply through the use of multiple filters which would normally
be required. A multiple filter system would not allow one filter to run
alternately starved and overloaded. And interlock controls on wash water
and filter operation would reduce the tendency to underwash the cake.
7. Process Operability
Process operability refers to the overall ease of process operation, in-
cluding the resistance of the process chemistry to operational upsets,
the sensitivity of the process performance to small changes in the process
chemistry, and the potential for scaling in the process equipment. In
this regard, it is to be differentiated from the mechanical/equipment-
related problems previously discussed.
In all respects the process operability was excellent. The system was
successfully operated over a range of widely fluctuating inlet S02 levels
and flue gas oxygen concentrations during each period of testing with
V-33
-------
little or no change in the S02 removal efficiency, cake properties, or
lime utilization. The system demonstrated an ability for continuous
operation with large and frequent variations in pH in both the scrubber
circuit (pH = 4-7) and regeneration/dewatering section (pH = 6.5-13).
And the low soluble calcium levels throughout the system even during
most upset conditions resulted in a low potential for scale formation,
particularly in the scrubber circuit. During Period 1 with active
sodium concentrations between 0.15Mand 0.30M, soluble calcium levels
generally fluctuated between 20 ppm and 250 ppm and were usually less
than 150 ppm. In Periods 2 and 3, when active sodium levels were main-
tained above 0.3M, soluble calcium concentrations typically ran about
70 ppm and were consistently below 100 ppm. The low potential for scale
formation or solids deposition in the scrubber was further demonstrated
in Periods 2 and 3, when operation of the wash sprays on the demister
was discontinued. During the last three months of Period 2 and through-
out Period 3 (including particulate testing) the wash sprays were not
used. Inspection of the demister after Period 2 and following comple-
tion of the test program showed the demister to be free of scale or
deposit of any kind — after about 4,500 hours of operation.
There were only two problems with scale formation/deposition of solids
in the system. First, there was a continuous, slow buildup of solids
in the small first reactor during all periods of operation. This
problem was closely linked to major upsets in the lime feed and reactor
agitation and is discussed later as a part of equipment performance.
At worst, such buildup if it were to occur would be expected to require
quarterly or semiannual cleaning as a part of a regular maintenance
program. In a full-scale system the cleaning would not require a shut-
down of the system, since flow could be switched to another reactor
train or the first reactor bypassed.
Second, there was precipitation and deposition of gypsum scale in the
reactor vessels and reactor outlet piping in July 1975 (Period 1). This
resulted from a simultaneous drop in inlet S02 levels to 850-950 ppm and
an increase in flue gas oxygen levels to 8.5-10% — a condition which
persisted for several days. The system was allowed to drift into a
dilute mode of operation (active sodium levels decreased to below 0.15M)
and the low soluble sulfite concentrations resulted in calcium super-
saturation with respect to gypsum. While this did not cause a shutdown
of the system, the scale in the piping did reduce the feed forward capacity
by about 20%. Most of the gypsum scale was removed during the interim
between Periods 1 and 2. A second parallel (polybutylene) pipeline was
also installed between Periods 2 and 3 to ensure sufficient feed forward
capacity for high-sulfur coal testing.
In addition to the fluctuating S02 levels and oxygen concentrations, the
system also experienced a number of process upsets during stable load
testing. These upsets included the carryover of small amounts of fly
ash (Period 1), the contamination of the lime feed supply with limestone
(Periods 1 and 2), and occasional overfeeding of lime (all periods). Only
the limestone had any significant .effect on the process performance and
V-34
-------
this effect was temporary. During periods when pure limestone was fed
to the reactor system, the pH of the regenerated liquor fell to below 7
causing a loss of S02 removal efficiency, from above 90% down to 80-85%.
The solids content of the filter cake also decreased slightly and sodium
losses in the cake correspondingly increased. Within a day or two after
the limestone had passed through the system, all of the effects were
reversed and the system operation returned to normal.
C. PROCESS PERFORMANCE—FLUCTUATING LOAD
Following completion of the stable load testing on high-sulfur coal in
May 1976, the system was operated with a fluctuating gas load for a
period of two weeks (6/1-6/14/76). The purpose of this testing was to
evaluate the effects of regular wide swings in gas flow (boiler load)
on system performance. While the gas flow to the system had been cur-
tailed to as low as 50% of design capacity for brief periods during
stable load operation, prior to fluctuating load testing there was no
attempt to follow or simulate swings in boiler load.
There were three primary objectives of fluctuating load testing: first,
to determine the effect of the changes in gas flow on S02 removal effi-
ciency; second, to evaluate the operability of the system both during
load changes and overall (as a result of the fluctuating conditions);
and finally, to determine the effect of fluctuating load conditions on
the system chemistry and process performance.
During fluctuating load testing, gas flow to the system was regularly
adjusted to four different levels between 30% and 100% of design load
according to a prearranged schedule roughly representative of the normal
operation of the Scholz boilers. The gas load schedule used is shown
in Figure V-9. The average gas rate to the system corresponded to about
65% of design load.
During fluctuating load testing, inlet S02 levels ranged from 1,550 ppm
to 2,700 ppm and averaged about 2,100 ppm. Oxygen levels in the flue
gas varied from 5.5% to 9.0% and typically ran about 7.5%. The pressure
drop across the venturi throat was maintained between 4 and 7 inches of
water (average of about 5 inches of water) to ensure saturation of the
gas. The pressure drop across the two trays and demister varied with
gas load from a low of 2.2 inches of water to a high of about 4.7 inches
of water. Liquor recirculation rates about the venturi and tray absorber
were maintained essentially constant at the different gas rates for con-
venience. Total liquor feed to the tray tower, therefore, ranged from
an L/G of about 5 (at 100% load) to 9 (at 30% load). Similarly, the
venturi L/G ranged from about 20 to about 60. Table V-10 summarizes
the operating conditions.
Overall, the system performance during fluctuating load testing was ex-
cellent. There were no problems of scale formation, decreased S02 removal
efficiency or deterioration of reactor performance. Table V-ll summarizes
V-35
-------
inn
90
80
v.o 70
O
II
| 60
1
< -=50
u> .5>
1
o 40
1
-------
TABLE V-10
SUMMARY OF SYSTEM OPERATING CONDITIONS
FLUCTUATING LOAD TESTING
(6/1-6/14/76)
Range Average
Inlet Flue Gas:
Flue Gas Rate, (equiv. Mw.)
(% of design load)
SO 2 Level, (ppm)
< 02 Concentration, (vol. %)
i
OJ
^ Temperature, (°F)
Fly Ash, (grains /scf dry)
Venturi Scrubber:
Pressure Drop, (inches of water)
Liquor Recirculation Rate, (L/G - gals./Macf)
Tray Tower (Absorber) :
Pressure Drop, (inches of water)
Liquor Recirculation Rate, (L/G - gals./Macf)a
6-20
30-100
1,650-2,600
5.5-9.0
280-365
<0.02
4-7
20-60
2.2-4.7
5-9
13
65
2,150
7.5
^315
—
5
40
3.5
7
Liquor recirculation rates were not adjusted during fluctuating load testing.
-------
TABLE V-ll
00
COMPARATIVE PERFORMANCE CHARACTERISTICS
STABLE VS
Inlet S02, (ppm-avg.)
S02 Removal, (% of inlet-avg.)
Inlet 02, (vol. %-avg.)
Thickener Liquor Composition:
Na+act, (M)
SO^, (M)
C1-, (M)
System Oxidation, (% of AS02~avg.)
Lime Utilization, (% of avail, lime-avg. )
Waste Solids, (wt % solids-avg.)
FLUCTUATING LOAD
Stable Load
(4-5/76)
2,250
96.5
6.5
0.35-0.4
0.7-0.75
0.25-0.33
17.5
93
54
Fluctuating Load
(6/1-6/14/76)
2,150
97. 5a
7.5.
0.33-0.38
0.65-0.74
0.32>0.22
^20b
91C
52d
Net Change
During Fluctuating
Load Testing
Decrease
Increase
Increase
No Change
No Change
Decrease
Increase
Decrease
Decrease
Curing periods when SO2 monitors operating and no mechanical problems.
Estimated oxidation based apon a few data points.
GSlight decrease in the estimated lime utilization is not considered significant - see text.
Decrease due in part to filter adjustments and non-optimal filter operation during filter tests.
-------
the principal performance characteristics compared with the stable load
operation during Period 3.
1. SOg Removal
In general, there was little significant effect of gas load on S02 re-
moval efficiency during the fluctuating load testing. S02 levels rarely
exceeded 100 ppm and typically were below 50 ppm. The trend, of course,
was for lower outlet loadings (higher removal efficiencies) at the low
flue gas rates due to the tendency to overfeed absorbent liquor slightly.
Frequently, outlet S02 levels ran below 20 ppm at 30% gas load. Of greater
significance was the ability to control S02 removal during transition peri-
ods when gas load was being adjusted. Transition periods typically lasted
about five minutes and during these periods there was no difficulty in
maintaining or adjusting outlet S02 levels as required. Throughout the
period, S02 removal averaged better than 97%, which is slightly higher
than that achieved during the stable load testing.
2. System Chemistry (Sulfite Oxidation)
The system chemistry was quite stable and liquor compositions were almost
identical with those of the stable load period. There was no apparent
trend in the concentration of any species other than chloride, which de-
creased from about 0.33M (^11,500 ppm) to 0.23M (^8,000 ppm) in the re-
generated liquor. There was a slight drop in chloride levels measured
in the coal; however, the decrease in chloride corresponding to the drop
in liquor concentration was within the error band for the analytical
method used for determination of coal chloride content.
The absence of any significant change in the liquor composition suggests
that the rate of sulfite oxidation experienced during fluctuating load
testing was roughly the same (as a percentage of S02 absorbed) as during
the stable load testing. Attempts were made to measure oxidation rates
in the scrubber system based upon species balances across the venturi
and absorber. Taking into account slight variations in oxygen levels
during the balance intervals, estimated oxidation rates are only a few
percentage points higher (as a percent of S02 absorbed) than oxidation
rates at 90% load.
3. Lime Utilization
The average of the analyses for six cake samples taken during the fluc-
tuating load tests shows a slight decrease in utilization (91%)^compared
with the average utilization during the stable load testing (93%). While
there was a tendency to overfeed lime slightly during low load periods,
it would not be expected that short periods of such overfeeding would
significantly change utilization. Since the 91% is within the variance
in the utilization for the stable load testing and because the average
lime analyses for the entire third period of operation were used in cal-
culating utilization, the 2.0% difference in utilization between stable
and fluctuating load is not considered significant.
V-39
-------
4. Waste Solids Properties
Only intermittent operation of the filter was required to keep up with
the rate of solids production. Usually, one or two intervals lasting
from four to eight hours each were sufficient. To some extent, the
rate and timing of cake withdrawal were dictated by filter test con-
ditions.
The filter cake produced during the fluctuating load testing was much
the same as that produced during the prior two months of stable load
testing. There was a slight decrease in the average solids content of
the cake; however, this is attributed to adjustments made in the filter
and non-optimal operation during filter testing (which was conducted
throughout the fluctuating load interval). Sodium losses in the cake
were in the same range as those during stable load testing.
5. System Operability
Overall system operability was excellent. Few problems were encountered
in the system operation either during changes in load (or inlet S02 con-
centration) or with operation at the extreme load levels (30% and 100%).
One minor problem that did occur in the operation of the thickener over-
flow hold tank was due to inadequate surge capacity. The capacity of the
hold tank at Scholz was sized for only a 10-minute holdup at design flow
(5 minutes of surge with a 50% normal level). On two occasions, when all
system tank levels were high due to excessive rainfall, there was insuffi-
cient surge capacity in the hold tank to accommodate the rapid change
from 100% to 30% gas load. This resulted in overflow of liquor from the
hold tank for a few minutes until flow rates and tank level set points
could be adjusted to compensate for the overflow.
Based upon this experience, a minimum 30-minute holdup should be provided
in the thickener hold tank.
D. PROCESS PERFORMANCE--PARTICULATE TESTING
The last three weeks of the test program were devoted to testing the
effects of fly ash on the process chemistry and overall system perfor-
mance. Particulate loadings in the inlet flue gas were increased by de-
activating all or part of the electrostatic precipitator. As shown in
Figure V-10, the precipitator consisted of six sections, each of which
could be independently deactivated. Extensive flue gas sampling was
performed by York Research Corporation during these tests. Sampling
included total particulate loadings, chloride concentration and total
sodium levels. Chloride sampling was conducted to determine chloride
absorption efficiency, and sodium analyses were performed to confirm the
entrainment data obtained by Guardian Testing Services during Period 2.
General operating conditions over the three-week period are listed in
Table V-12. Principal control variables adjusted during the period
V-40
-------
Flue Gas In
V
A
B
C
E
D
F
i K
1 >
V
Flue Gas Out
j
Electrostatic Precipitator Sections
FIGURE V-10 SCHEMATIC OF THE ELECTROSTATIC PRECIPITATOR
ON BOILER #1 AT SCHOLZ STEAM PLANT
V-41
-------
TABLE V-12
SUMMARY OF SYSTEM OPERATING CONDITIONS
PARTICULATE TESTING
(6/15-7/1/76)
Range Average
Inlet Flue Gas:
Flue Gas Rate, (equiv. Mw) 11-20 19
(% of design load) 55-100 95
S02 Level, (ppm) 1,500-2,700 2,100
02 Level, (vol. %) 5.5-8.5 7
Temperature, (°F) 290-370 ^335
Fly Ash, (grains/scf dry) 0.02-3.6
Venturi Scrubber:
Pressure Drop, (inches of water) 12-17 —
^
*- Liquor Recirculation Rate, (L/G-gals./Macf) — ^20
Tray Tower:
Total Pressure Drop, (inches of water) 3.5-5 4.5
Liquor Recirculation Rate, (L/G-gals./Macf) — 7
Thickener Liquor Composition:
pH 11.8-12.4 12.2
Active Sodium (M Na+) 0.35-0.42 0.38
SO", (M) 0.65-0.90 0.80
Cl~, (ppm) 8,000-9,500 8,700
-------
were the number of precipitator sections in service, gas flow rate and
venturi pressure drop. Figure V-ll shows the precipitator operation
and Table V-13 lists the conditions for the specific scrubber tests
performed. All scrubber tests were conducted with the flue gas re-
heater off.
Particulate loadings measured in the inlet flue gas were found to be
fairly insensitive to the number of precipitator sections deactivated,
but rather were a function of whether or not the precipitator was in
service. With sections A and B in operation, inlet loadings ranging
from 0,017-0.025 grains per dry standard cubic foot (grs/scfd) were
measured. With section A alone, loadings were only slightly higher,
0.034-0.084 grs/scfd. But with all sections deactivated, measured
loadings ranged from 2.29-3.60 grs/scfd at the scrubber inlet.
Inlet S02 levels during the last three weeks decreased slightly from
levels experienced in the previous fluctuating load testing. Levels
ranged from as low as 1,500 ppm to as high as 2,700 ppm, and averaged
about 2,100 ppm. Oxygen concentrations in the flue gas varied from
5.5% to 8.5% and typically ran about 7.0%.
Except for one brief interval during gas sampling for entrainment losses
at low load, the flue gas rate was maintained at or above 85% of design
load. The average flue gas rate ran about 95% of design load.
As in other operating periods, thickener liquor composition reflected
the variations in inlet flue gas conditions. Soluble sulfate levels
ranged from 0.65M to 0.90M, increasing at low inlet S02 levels and high
oxygen concentrations (as system oxidation increased). Typically, sulfate
concentrations ranged from 0.7M to 0.85M. Active sodium levels and chloride
concentrations remained fairly constant throughout — 0.35-0.42M Na+ active
and 0.22-0.26M Cl~. This range of chloride levels was established with
inlet chloride concentrations of approximately 30 ppm (^1.5% of the
average inlet
Table V-14 summarizes the general system performance. Important perfor-
mance characteristics are discussed in the following section.
1. Particulate Removal
The particulate removal capability of the scrubber system was tested
with the venturi operating at a pressure drop of both 12 inches and
17 inches of water and at three inlet particulate loadings. The results
of the particulate measurements for each test condition are given in
Table V-15. Standard EPA sampling and analytical procedures were used
in all tests.
As would be expected, outlet particulate loadings increased slightly
with increasing inlet loadings, but there was no statistical differ-
ence between outlet particulate loadings (or particulate removal effi-
ciencies) measured at venturi pressure drops of 12 inches and 17 inches
of water.
V-43
-------
Test Nos.
1-23-456-7 8 9-11 12-15 16-21
j>
JS
15 16 17 18 19 20 21
22 23 24 25 26 27 28 29 30
Day of the Month (June/July, 1976)
FIGURE V-11 SCHEDULE OF PRECIPITATOR OPERATION AND SCRUBBER TESTS
-------
TABLE V-13
Test
No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
SUMMARY
Gas Flow
(% of design)
100
100
100
100
100
100
100
100
100
55
55
100
100
85
85
100
100
100
100
100
100
OF SCRUBBER TEST
Precipitator
Sections in
Service
A&B
A&B
A&B
A&B
A
A
A
A-F
A-F
A-F
A-F
A-F
A-F
A-F
A-F
None
None
None
None
None
None
CONDITIONS
Venturi AP
("H,0)
12
12
12
17
17
12
12
12
12
12
12
5
5
12
12
17
17
17
12
12
12
Flue Gas
Sampling
Particulate
Particulate
Particulate
Particulate
Particulate
Particulate
Particulate
Sodium
Sodium
Sodium
Sodium
Chloride
Chloride
Sodium
Sodium
Particulate
Particulate
Particulate
Particulate
Particulate
Particulate
V-45
-------
TABLE V--14
SUMMARY OF SYSTEM PERFORMANCE
Inlet Flue Gas:
S02, (ppm) 1,500-2,700
02, (ppm) 5.5-8.5
C1-, (ppm) ^30
Particulate, (grs/scfd) 0.017-3.60
Performance:
S02 Removal, (avg. % of inlet) ^98
Outlet Particulate, (grs/scfd) 0.010-0.037
Cl~ Removal, (avg. % of inlet) 99+
Lime Utilization, (% of available Ca(OH)2) 97
Filter Cake, (wt % solids-avg.) 56
(wt % acid insolubles) 0.2-41
(mols CaSOif/mol CaSO ) 0.09-0.13
X
Liquor Entrainment—102% load, (gpm)a 0.005Ja
92% load, (gpm)b 0.013b
83% load, (gpm)a 0.0037a
57% load, (gpm)a 0.0030a
a
Gas sampling by York Research Corporation (June 1976) , average of two data
points each
Gas sampling by Guardian Systems (December 1975) , average of five data
points
V-46
-------
TABLE V-15
SUMMARY OF PARTICULATE TEST RESULTS
General Conditions:
100% gas load
Reheater off
Venturi L/G = 20 gals./Macf sat'd
Precipitator
Sections in
Service
A&B
A&B
A
A
None
None
Inlet Loading Test
(grs/scfd) Nos.
0.017-0.025 1-3
0.018 4
0.034, 0.047 6,7
0.084 5
2.95-3.60 19-21
2.29-3.34 16-18
Average Outlet Loading
AP = 12" H»0 AP = 17" H00
0.010-0.015
0.011
0.021, 0.027
0.026
0.024-0.037
0.033-0.037
V-47
-------
Normally, it would be expected that the increased pressure drop would
increase removal efficiency, particularly at the higher inlet loadings.
That the outlet loadings were the same at both pressure drops can only
be attributed to the presence of the absorber (with two trays) follow-
ing the venturi. The average removal efficiency of the combined ven-
turi and absorber for periods when the precipitator was out of service
was 98.9%.
The quantity of fly ash added to the system ranged from 10-50 pounds per
hour (0.5-2.5% of the dry cake weight) with the precipitator in service,
to 1,000-1,600 pounds per hour (25-45% of the dry cake weight) with the
precipitator out of service.
2. SO2 Removal
Outlet S02 levels throughout the particulate testing period rarely ex-
ceeded 100 ppm and were consistently below 50 ppm. As a result, the
average S02 removal efficiency during these last three weeks was better
than 98%. On only half-a-dozen occasions, for periods up to a few hours,
did the outlet S02 rise above 200 ppm. These occurred either because of
temporary mechanical problems or because of calibration/certification
testing of the S02 monitors when it was necessary to maintain outlet S02
levels on the order of 200 ppm.
3. Chloride Removal
In the two tests run during June (Tests 12 and 13, Table V-13), chloride
levels of approximately 25 ppm were measured in the inlet flue gas. Cor-
responding outlet levels ran below 0.25 ppm — a chloride removal efficiency
of 99+%. This level of chloride absorption is consistent with the equilib-
rium level of chloride established in the system liquor, as dictated by
chloride absorption and losses of chloride in the waste cake. The mea-
sured inlet chloride concentration, however, is about half that expected
based upon chloride and sulfur analyses of the coal. The measured chloride
levels in the gas correspond to about 1% of the inlet S02 level (25 ppm Cl~
and 2,100 ppm S02). Coal analyses show about 0.1 wt % Cl~ and 3.5 wt %
sulfur on the average, or a chloride content of about 2.5% of the sulfur
value on a mol basis. Since very little chloride was found in the ash,
this discrepancy must be attributed to analytical error in the coal
analyses due to the low accuracy in the analytical procedures.
4. Lime Utilization
Lime utilization during particulate testing averaged approximately 97%
of the available Ca(OH)2 in the raw lime feed based upon cake analyses
and average analyses of the raw lime throughout Period 3. Based upon
lime sample analyses for the last three weeks only, the utilization was
95%. As in the case of fluctuating load testing, the particulate test-
ing period was too short to calculate accurate overall system material
balances that would confirm these utilization figures. However, based
upon the fact that all but one of the cake analyses showed a utilization
higher than the average utilization for the stable load period (April-May,
V-48
-------
1976), it Is probable that there was a slight improvement in utilization
during particulate testing. This can be attributed to the slightly lower
operating pH s (extent of regeneration) and slightly lower inlet SO, levels
which allowed slightly more holdup time in the reactor.
5. System Chemistry (Oxidation and Sulfate Control)
There was very little change in the system chemistry throughout the
particulate load testing, and very little real difference between solu-
tion compositions during particulate testing and stable load testing.
The system operated slightly more concentrated in total sodium during
particulate testing. The sulfate level in the system increased on the
average by about 10-15%; however, the active sodium concentration in-
creased also. The active sodium-to-sulfate ratio was, at most, 10%
higher during particulate testing than during stable load testing.
Since oxygen levels in the flue gas were relatively the same during both
stable load and particulate testing, the absence of any significant change
in the system chemistry suggests no major effect of fly ash on oxidation.
This was confirmed by the few data points that could be obtained on scrub-
ber system oxidation (by scrubber system material balances). The absolute
oxidation (Ib mols/min) agreed quite closely with those obtained during
stable load testing.
There was a slight increase, though, in oxidation relative to S02 removal
due to the lower inlet S02 levels during particulate testing. On the
average inlet S0£ levels (and amount of S02 removed) decreased about 10%
from stable load to particulate testing. Therefore, about a 10% increase
in oxidation would be expected relative to S02 removal. This is confirmed
from CaSOtt/CaSOx ratios in the cake. The average CaSOn/CaSOx ratio (mol
basis) in the cake for particulate testing was 0.11 (0.09-0.13) compared
with about 0.10 for stable load testing.
6. Waste Cake Properties
The solids content of the waste cake increased slightly during particu-
late testing. The average cake solids content for stable load testing
was 54%, while for particulate testing it was 56%. Some of the cake
withdrawn during particulate testing had very little fly ash and some
was filtered at unusually low filter vacuum — 8 inches to 12 inches
during the particulate testing period versus 11 inches to 16 inches
during stable load testing. Discounting the periods of low vacuum,
the average solids content of cake containing greater than 20% ash
was about 4-7% higher than comparable cake without ash.
The washability of the cake with fly ash was comparable to or possibly
slightly better than that without ash; however, there is insufficient
data to provide a clear-cut correlation. The soluble sodium losses in
the cake at roughly equivalent wash ratios (and filter vacuum) were
slightly less when the cake contained significant fly ash.
V-49
-------
Even though the cake had a higher solids content and appeared to wash
more easily, it did feel somewhat wetter. This apparent wetness may
have been due to the added lubricity from the spherical fly ash particles.
It is interesting to note that when dumped and left at the disposal pond,
the cake containing fly ash tended to agglomerate into chunks and harden
somewhat, while cake without ash did not. This hardening was also noticed
in the solids that remained in the thickener about a month after the system
was shut down following completion of the test program.
7. Liquor Entrainment
Measurements of entrained liquor in the scrubbed gas leaving the absorber
were made in December 1975 by Guardian Testing Services and in June 1976
by the York Research Corporation using sodium as a tracer. The York data
are summarized in Figure V-12. The highest levels of entrainment measured
(@ 92% gas load by Guardian) averaged approximately 0.013 gpm. This
corresponds to about 0.022 grains of liquor per standard cubic foot of
dry gas (<0.0025 grains of solids/scfd), which is equivalent to a sodium
makeup requirement (as Na-COn) of less than 0.1% of the S02 removal (mol
basis). In subsequent sampling performed by York, entrainment was measured
to be only about one-third this level (M).008 grains of liquor/scfd). The
sodium losses by entrainment, therefore, are so small that they can be
considered negligible.
In the sampling performed in June it was also found that the grain loading
of liquor (or sodium salts) was almost constant over the range of gas rates
tested. Therefore, the volume of liquor lost decreased almost linearly
with gas flow. This suggests that the mist eliminator was operating
optimally at design gas flow — removing all but a fine mist that also
was carried through at lower gas flows.
8. Overall Operability
There were no deleterious effects of the fly ash either on the process or
mechanical operability or reliability. In fact, the fly ash had some
beneficial effects (e.g., improved cake properties). However, particu-
late testing was of too short a duration to test long-term effects of
the fly ash on materials of construction.
E. MECHANICAL PERFORMANCE
The equipment performance in terms of overall reliability is reflected in
the system availability, the level of maintenance required, and the over-
all operability (ease of operation) of the system. In general, the equip-
ment performance was quite good, although there were equipment and instru-
mentation problems. Most of these problems were mechanical in nature and
resulted from design or fabrication oversights commonly associated with
a first-of-a-kind prototype system. There were only a few problems en-
countered that reflected process chemistry, and these required simple
operational and/or equipment adjustments.
V-50
-------
12
10
-------
Appendix G contains a detailed list of equipment and instrumentation
problems encountered during the program. Included in the list are a
number of problems of the type that would normally be expected to occur
during startup or items normally associated with routine maintenance.
The fact that some of the problems shown were not immediately corrected
is an indication that they were either of minor significance or did not
cause important operational difficulties. The more important problems
are discussed in the following sections.
1. Equipment
The mechanical equipment problems and maintenance items of primary concern
are reviewed by process subsection. The order in which each subsection
is discussed generally reflects the frequency or number of problems en-
countered or degree of maintenance required.
a. Filter
The filter was the largest source of problems in the prototype system.
Because of anticipated corrosion problems associated with high chloride
levels, a large part of the filtration equipment was fabricated out of
fiberglass, which is not as sturdy as stainless steel and more prone to
failures at stress points in the construction. Filter problems during
Period 1 included:
• erosion of the fiberglass scraper blade, resulting in
jagged edges which tore the cloth
• erosion of the bridge valve due to solids carried through
cloth holes
• loss of vacuum due to cracks in the internal drum trunnion
tubes
• cracking of the plastic caulking strips, allowing retention
ropes to loosen and releasing the cloth panels
• failure of the fiberglass rocker arm used to agitate the
slurry in the filter tub
Modifications were made to the filter during the latter part of Period 1
and the early part of Period 2. These modifications included replace-
ment of the fiberglass scraper blade with one fabricated out of stainless
steel, reinforcement of the rocker arm with stainless steel plates, and
the design of a new method for retaining the filter cloth panels. These
changes along with regular inspection and maintenance and increased famil-
iarity of Gulf Power personnel with the equipment significantly improved
filter performance.
During the scheduled boiler shutdown for annual maintenance in January
and February 1976, the filter drum was overhauled by the manufacturer
V-52
-------
and permanent cloth retention strips of the new design were installed.
Following this overhaul (Period 3) there were fewer problems with the
filter operation.
bi Reactor System
Two equipment problems of note were encountered in the reactor system:
plugging of the dry lime feed chute when dry lime was fed directly to'
the first reactor, and buildup of solids in the first reactor. The
chute plugging problem was caused by hot vapors rising from the first
reactor into the chute entrance port, condensing and wetting the lime.
The problem was resolved by installation of a vibrator on the feed chute
with provision for injection of air near the flange connection to prevent
vapors from rising into the chute. Of course, no chute plugging occurred
when lime was fed as a slurry as it would be fed in a full-scale application.
Solids buildup in the first reactor was a function of the manner of lime
feed (dry versus slurry lime), dry lime feed chute location, the degree
of agitation, and operation of the reactor under upset conditions (e.g.,
gross overfeeding of lime). Deposition of solids occurred both above
and below the liquid surface, particularly in the area near the feed
chute. While the problem was never serious enough to cause shutdown
of the system, it did require occasional cleaning. The rate of buildup
averaged about 0.75-1.0 inch per month.
A simulation of the conditions in this reactor at the CEA/ADL pilot plant
in Cambridge was successful in confirming the source of the solids buildup
and a new first reactor was installed prior to the start of the .third
operating period. The lime slurry feed system was also activated as an
alternative to the feeding of dry lime. (Feeding a slurry of slaked lime
would be the normal practice in a full-scale application because of the
high cost of direct purchase of dry hydrated lime.) The new reactor and
use of slurried lime minimized, but did not eliminate, the buildup. Agi-
tation was still poor and process upsets continued. However, with improved
agitation and process control, it is expected that such buildup would, at
worst, be a semi-annual maintenance item. Such maintenance would not re-
quire system shutdown since flow can be switched to a parallel reactor
train, or the first reactor can simply be bypassed.
An additional problem was the abrasive wear of the flake lining on the
floor of the second reactor in the area directly beneath the agitator.
This lining was replaced with Heil Rigifloor 413G prior to Period 3. The
Rigifloor showed no evidence of wear after three and one-half months of
service in Period 3.
c. Scrubber
The principal equipment problems in the scrubber system involved control
and block valves, and vessel linings.
V-53
-------
Valves
Erosion and debonding of rubber linings in the bleed control valves and
recycle block valves on both the absorber and venturi occurred during
Periods 1 and 2. These failures were due to the high degree of throttling
required to control flow. The valves were sized to accommodate the much
higher flows associated with direct slurry scrubbing. The rubber-lined
control valves were replaced with 316 stainless steel valves prior to
Period 3, and the block valves replaced either with spool pieces or new
rubber-lined valves. Inspection of the control valves following comple-
tion of the test program in July 1976 (three and one-half months of
service) showed them to be free of corrosion, erosion, or scale.
Vessel Linings
Lining problems occurred in both the venturi and absorber. For the most
part, these were corrected by patching or replacement of the lining with
a different material.
In the absorber a few pinholes and hairline cracks occurred in the Heil
4850 Flakeline on the vessel walls, particularly near pressure and tem-
perature tap holes. These appeared during the first period of operation
(in the first five months). The holes were patched prior to resuming
operation in Period 2 and no similar failures occurred. However, after
system shutdown at the completion of the test program, discrete points
of lining wear were found on the sharp edges of the duct expansion joint
connection between the venturi and absorber. As expected, the carbon
steel at these points was corroded.
In the venturi there was erosion of the original Heil 4850 Flakeline on
the liquor distribution shelf under the tangential nozzle discharge. The
shelf was relined with Rigifloor 413G (the same material used on the
floor of Reactor 2) prior to Period 3. After three and one-half months
of service (March-July 1976) the Rigifloor was in excellent condition.
The lining on the venturi gas inlet in the quench zone deteriorated. The
area was relined with Plasite 4030 prior to starting Period 3. After
three and one-half months of service the Plasite lining on the outside
of the downcomer showed no signs of failure; however, on the inside where
it was in direct contact with hot gas, the lining was deteriorating. The
reason for the failure is unclear (whether it was poor application, high
temperature, or both), but the failure suggests that a corrosion resistant
metal alloy may be most suitable for such areas.
Corrosion
In addition, there was evidence of stress corrosion cracking and pitting
of stainless steel in the absorber internals. Most of this occurred on
304 stainless steel which was inadvertently substituted for 316/316L
(316/316L stainless steel was specified throughout). However, some
V-54
-------
rt f°Und t0 be low in molybdenum
stitutid f£8316L S°me " applications, 316 was inadvertently sub-
A slight corrosion and ash buildup on the fan during extended shutdowns
was diagnosed to be due to flue gas leaks through the isolation damper.
Since the fan was located upstream of the system and was constructed
out of carbon steel, significant corrosion was not expected. The cor-
rosion and ash buildup were minor and the fan rotor was easily cleaned
and rebalanced when it occurred. Correcting the leak through the isola-
tion damper by installing a new damper or use of an air seal system was
considered to be unwarranted.
Pump Seals
Leakage of solution through the venturi and absorber pump seals was of
concern primarily because it represented a loss of sodium from the system.
Tests were conducted by Southern Company Services and Gulf Power personnel
to determine the size of the leaks under various pump packing conditions
and to determine how best to eliminate or minimize the leaks. A Teflon
impregnated packing was installed just prior to the end of operating
Period 2 and minimized packing failures and improved maintenance.
d. Thickener
The major problems with the operation of the thickener were plugging of
the underflow lines (Period 1) and leaking of liquor through the bottom
due to lining failure (Period 2). Plugging of the thickener underflow
lines was partly due to the design of the underflow piping, and partly
due to the frequent downtime on the filter which allowed thickener under-
flow slurry concentrations to exceed 30% solids. Redesign of portions of
the underflow lines using flexible piping and adjustments to the opera-
tional procedures to maintain the underflow slurry concentration in the
15% to 25% range (by backflushing with clear hold tank liquor to dilute
the underflow when necessary) effectively eliminated underflow plugging
problems.
The leak that developed in the bottom of the thickener during the early
part of the second period of operation was small but grew worse through-
out Period 2. After shutdown in January all the solids were removed
from the thickener and the liquor drained. Inspection of the inside of
the thickener showed sections of flake lining to be failing due to poor
curing or poor application by the lining supplier. The leaks were lo-
cated and patched, and the sections of failed lining on the thickener
walls replaced with Heil 4850. Holes were also drilled in the bottom
of the thickener, and steel bolts were welded in to support the bottom
and prevent flexing. These areas were patched with Rigifloor 413G.
These measures essentially eliminated any further leaking during Peri°d J,
and inspection of the lining following completion of testing showed the
lining to be in good condition.
V-55
-------
When the thickener was drained, pieces of rope and a flattened paint can
were found in the area of the cone. These items undoubtedly contributed
to difficulties with the thickener underflow system. A piece of rope
had previously been extracted from a seized underflow pump during opera-
tion in Period 2.
During Period 3 no significant problems were encountered with the thick-
ener or thickener filter interface.
2. Instrumentation
Instrumentation problems primarily involved the pH units, level trans-
mitters, and soda ash feed solution control system. Other instrumentation
problems occurred, but for the most part, these were minor.
a. pH Probes
The piping for the flow-through pH units originally installed throughout
the system had a tendency to plug and the electrodes tended to coat with
a fine film, causing a drift in pH readings. In the case of the pH unit
in the reactor system, the lines and probe chamber plugged completely
with solids at low slurry flow rates and the probes eroded and broke
at high flow rates. The flow-through unit on the reactor system was
replaced with an immersion unit fitted with a sonic cleaner midway
through Period 1. This proved to be much more reliable; however, there
were problems keeping the sonic cleaner mounted on the probe casting.
The unit required routine checking and recalibration about every week
to two weeks. A Uniloc immersion pH unit (Model No. 321) was installed
without a sonic cleaner for testing during Period 3. This unit proved
to be more reliable and accurate than the Leeds and Northrup unit with
the sonic cleaner.
The take-off lines for the flow-through units on the scrubber system were
relocated, and higher flow rates were maintained through the probes to
prevent plugging of the piping. Since close pH control is not required,
the problems with the pH units were not critical. The scrubber system
can be operated on either the venturi bleed pH or outlet S02 (or, for
that matter, inlet S02 and feed forward flow) with only occasional checks
of the bleed pH to verify the S02 monitor. In fact, during a few weeks
in the early part of Period 1, when all pH units and S02 monitors were
out of service (due to delays in obtaining replacement parts), the system
was successfully operated by taking hourly pH readings with a portable
pH unit.
b. Level Transmitters
The original level transmitters installed in the system were generally
unreliable. The particular type of unit used could not be completely
serviced in the field. Thus, there were periods when inaccurate level
indication and/or control caused operational problems. Starting in
Period 2, these units were replaced as they failed with Foxboro trans-
mitters. The Foxboro units performed acceptably.
V-56
-------
c. Soda Ash Feed Control
The soda ash solution feed control system was also unreliable. This, in
combination with occasional plugging of the feed gate on the dry soda ash
feeder, made it difficult to close the overall material balance on sodium
for short periods. During the last half of the second operating period,
the sodium makeup rate was determined using frequent checks on the specific
gravity of the soda ash solution and recalibration of the flow indicator
along with the inventory of the soda ash silo and the quantity of soda
ash delivery. Inventory and delivery information alone were not suffi-
cient over the short term due to the large storage capacity in the silo
and the low soda ash feed rates. Inventory and deliveries were used,
though, for longer-term material balances during Period 1.
V-57
-------
VI. REFERENCES
1. LaMantia, C.R.; R.R. Lunt; J.E. Oberholtzer; E.L. Field; and
N. Kaplan , EPA Dual Alkali Program—Interim Results, Proceedings
of the Fifth Flue Gas Desulfurization Symposium, Environmental
Protection Agency, Atlanta, Georgia, November 1974, pp. 549-665.
2. Kaplan, N. , Introduction to Double Alkali Flue Gas Desulfurization
Technology, Proceedings of the Sixth Flue Gas Desulfurization
Symposium, Environmental Protection Agency, New Orleans, Louisiana,
March 8-11, 1976, pp. 387-422.
3. LaMantia, C.R.; R.R. Lunt; R.E. Rush; T.M. Frank; and N. Kaplan,
Operating Experience—CEA/ADL Dual Alkali Prototype System at Gulf
Power/Southern Services, Inc., Proceedings of the Sixth Flue Gas
Desulfurization Symposium, Environmental Protection Agency, New
Orleans, Louisiana, March 8-11, 1976, pp. 423-69-
4. Johnstone, H.F.; H.J. Read; and H.C. Blankmeyer, Recovery of
Sulfur Dioxide from Waste Gases: Equilibrium Vapor Pressures Over
Sulfite-Bisulfite Solutions, Industrial and Eng. Chem., 30 (1):
101-9, 1938.
VI-1
-------
VII. GLOSSARY
Active Sodium - Sodium associated with anions involved in S02 absorption
reactions and includes sulfite, bisulfite, hydroxide and carbonate/
bicarbonate. Total active sodium concentration is calculated as
rollows :
[Na+1active = 2 x (CNa2S03] + [Na2C03]) + [NaHS03] + [NaOH] + [NaHC03]
Active Sodium Capacity - The equivalent amount of S02 which can be theoreti-
cally absorbed by the active sodium, with conversion to NaHS03.
Active sodium capacity is defined by:
[Na+] active caPacity = [Na2S03] + 2 x [Na2C03] + [NaOH] + [NaHC03]
Calcium Utilization - The percentage of the calcium in the lime or lime-
stone which is present in the solid product as a calcium-sulfur salt.
Calcium utilization is defined as:
mols (CaS03 + CaSO^) generated
Calcium Utilization = - x 100%
mol Ca fed
Concentrated Dual Alkali Modes - Modes of operation of the dual alkali
process in which regeneration reactions produce solids containing
CaS03>J5H20 or a mixed crystal containing calcium sulfite and calcium
sulfate hemihydrates, but not containing gypsum. Active sodium con-
centrations are usually higher than 0.15M Na+ in concentrated mode
solutions.
CSTR—Continuous Stirred Tank Reactor - A well-agitated, baffled reactor
vessel having a uniform concentration of species throughout. At
any time the concentrations in the effluent from a CSTR are equiva-
lent to those within the vessel.
Dilute Dual Alkali Modes - Modes of operation of the dual alkali process
in which regeneration reactions produce solids containing gypsum
(CaSOit-2H20). Active sodium concentrations are usually lower than
0.15M Na+ in dilute mode solutions.
VII-1
-------
Sulfate Formation - The oxidation of sulfite. The level of sulfate forma-
tion relative to 862 absorption is given by:
mols S0| oxidized
Sulfate Formation » x 100%
mol S0£ removed
Sulfate Precipitation - The formation of CaSO^'XI^O in soluble solids.
The level of sulfate precipitation in the overall scheme is given
by the ratio of calcium sulfate to the total calcium-sulfur salts
produced:
mols CaSOi*
Sulfate Precipitation =
mol CaSOjj
TDS—Total Dissolved Solids - Equivalent to the sum of all soluble species.
TOS—Total Oxidizable Sulfur - Equivalent to the sum of all sulfite and
bisulfite species.
VII-2
-------
APPENDIX A
DAILY COAL ANALYSES
A-l
-------
DAILY COMPOSITE OF COAL FIRED IN UNITS 1 & 2
, 10 n 12 13 14 IS 16 17 II It 20 21 22 23 24 25 26 27 28 29 30 31
6.0-
5-0-
cc a 4-0-
"- CE
^ a 3 fl-
irt jj-
2.0-
10-
S 16,000-
J ? 14,000-
o:
2 Q
5 5 12.000-
S 10,000-
026-
0 24 -
0 22 -
020
£ 0 18
~ S 016
%i 014
X ". 0 12
0 10
008-
006
004
0 02
,,!,,., . ... , ... 1 1 1 1 1 _l 1 1 1 1 1 1 — 1 1 1
s~—^^ ^_^ — --— ^^~^ — — — ^
^/ X/ ^^^X"^^1
•5.0
•4.0 « ,
o!
2 ,
•3,0 ii
>
•2.0 *
•1.0
H6.000 ;
| H
^\, ^^ ~* ~ • - "^- — ^ _ ^ "•*"^* nS
• +- ^. — *• «. •- " — *- •* ( ' l^
U,ooo2"
A
A
A A
A
•026
•0.2(
-0.22
•020
•018 \
• 0 16 ?
0.14 S
-0.12 s
-0.10 S3
-0.08
-0.06
-0.04
0.02
-0
1 234 56 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
KEY- % TOTAL MOISTURE
FEBRUARY 1975 —PLANT ANALYSIS AVE6M
A OUTSIDE LABORATORY RANGE 4 .70-9.01
DAILY COMPOSITE OF COAL FIRED IN UNITS 1 .& 2
5.0-
VI
<
tt E 40-
£ >
-, 0
v, ^ 3.0-
i
2.0-
1.0-
f, 16,000-
3 3
< " 14.000-
t3 g
? 3 12.000-
a
— 10.000J
0.26-
0.24-
022-
0 20 •
| 018
0 » 0.16 •
§ g 0.14 -
5*- °«-
£ 010-
o.oe-
ow-
004-
002-
o-i
1 2 3 4 5 6 78 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
/\
/ \ x*\ ^r^*\ A
^ V/N. /r^~^ \A/^~^~~~~£----*^***^S^~~~^* *~~-4- — *-~-^_- -*^A _^r~^ *^-^.
V" ^ ^-^_^-A
» ^ A
^\^ — *^^~~~* — ~»-__^xx^ *^~-^-^ _, — '*' i^^ i A • A -^^^j
—^^~~^ ~~ -^
A
A A A A
A
5.0
i
-40 *
0
31
-3.0 •
z
~™
-1.0
-16,000 g
I
-14,000 t
V
-Itf-j
.10,™ "
o.zc
0.21
•0.2!
0.20
-0,8
• 0.16
0.14
• 0.12
• 0.10 |
-0.01 '
•0.06
•O.M
• 0.0!
1 2 3 4 5 6 7 1 B 10 11 12 13 14 IB 1« 17 18 19 20 21 22 23 24 26 26 ' 27 28 ' 29 30 31
UAQCU 1OTE KEV * TOTAL MOISTURE
MARCH 1975 —•— PLANT ANALYSIS AVE7.«4
Prepared by Southern Company Services Inc ^ OUTSIDE LABORATORY RANGE 3.821047
A-2
-------
DAILY COMPOSITE OF COAL FIRED M UNITS 1 t 2
6.0-
6.0-
M
3 4.0-
e a
3 ° 3.0-
t
- 1.0-
1.0-
s »-«°l
33
<; 14.000-
P j 12,000-
23
I h
8 lo.oooJ
026
0.24-
0.22-
_ 0.20-
S 0.18 •
SS 016
3 S °'14
5* «•«•
g 0.10-
o.oei
0.06-
0.04'
0.02
0-
* .A
* * *^T«r-^ «-— ~* * ^ ' •* «-"~"^. *-— ^__J_--A ' .^-—— -£ • •
^-, ± A- — ^ A __
— -»-^ ~* A , -^.
A A A A A
A A
•6.0
•5.0
5
•4.0 * w
D E
55
• 3.0 m X
S
•2.0 "
•1.0
pie.ooo _
3 i
-14,000 g 5
.12.000 < g
5 r-
•10.000 —
•026
•024
•0.22
020
•018 I
-0.16 Jig
0.14 g 5
-
-o.io S"
-0.08 *
-0.06
0.04
0.02
0
234 56 7 I ( 10 11 12 13 14 16 16 17 II 18 20 21
APRIL 1975
22 23 24 25 26 27 28 29 30 31
KEV: % TOTAL MOISTURE
-•— PLANT ANALVSIS AVE 7 99
A OUTSIDE LABORATORY RANGE 5.23-15 8
DAILY COMPOSITE OF COAL FIRED IN UNITS 1 & 2
10 11 12 13 14 16 16 17 16 19 20 21
23 24 26 26 27 28 29 30 31
5.0
si 4.0
P 3,
3 £ 3.0
|
2.0-
1.0-
^3 '•."»•
5> 14.000-
ll I
SB «••
- 10.066-
0.26
OJ4
0.22
, OJO
i an
?J OH
§i «
i«t 0.11
1 "'
ui'
64I-
6JI-
6Jt-
6-
^^^ A A^ ^* * *~^^^
. __<^ »^ • ^x' ' » — *—_^ — ~*\ ^—-^~^ A 4i _-• — •- — ^^^ ^~*1 — •
^
^ ^ ^X— ~-^- _^_
A A A
A A
A
-5.0
-4.0 ? <*
\\
•3.0 n>
2
-2.0
•1.0
16,000 5
C S
-» ^
-14,000 £ S
s^
12.000 < 5
Ii
10,000 -
0.26
0.24
0.22
0.20
CHLORIDE
ft! %. DRV BASli
« « J w o
b b b b b
0.06 ~
0.06
004
002
0
'' ' 1 1 ' ' •* ' r— — I • •— •'„•-'„'.'..',,',.' 22 23 24 26 26 27 21 26 JO 31
)« • « 7 • t It 11 W 13
t by Southern Company Service*, Inc.
17 II »
MAY 1975
,EY % TOTAL MOISTUAE
-•— PLANT ANALYSIS AVE 7 M
A OUTIIDE LAKMUTORV RANGE 4.JH4 S
A-3
-------
DAILY COMPOSITE OF COAL FIRED IN UNITS 1 * 2
1
D c
i
_
TING VALUE
.B. DRV BASIS
< ^
I ?
ffi
%
sl
S a
3 o
SH
-
6.0-
50-
3.0-
2.0H
1.0-
i 1 1 1 1 1 1 1 1 1 1 1 1 1
__A— -— ^—~ _^— A^_____A-_ *--*-- A. /*^^— A A
* — SS\ ^S* * ""*" ~~~~~* **^ j^^ ^^*v ^*~~~-~-~ £\ *^^**^* • T
^"^^
16.0OO-)
14.000-
12,000-
10,000-
026-
024-
0.22-
0.20-
016
0.16
0.14
0.12-
0.10-
008-
O.M
0.04
0.02
0
•—-A—. v. ^»-~-*~~^*~ — A«.A.i_A,t'A »— — -^- A r — " — A—--.—— »—_ A^— •
•6.0
§ r
•< c
-3.0 5 »
S
"
•2.0 "
•1.0
-If , .
HEATING VAI
TU/LB, DRV S
i 1
u c
MO.ODO "
A
A A A
A
-0.26
-0,24
-0.22
- 0.20
-0.1! |
-0.16 «S
0.14 So
-0.12 ll
-0.10 i
-0.08 -
•006
OH
0.02
-0
34 5 6 7 I » 10 11 12 13 14 16 16 17 II 19 20 21 22 23 24 25 26 27 21 29 30 31
KEY: * TOTAL MOISTURE
—•— PLANT ANALYSIS AVE 5.26
A OUTSIDE LABORATORY RANGE 3.43-7.66
JUNE 1075
DAILY COMPOSITE OF COAL FIRED IN UNITS 1 & 2
3 4 6 6 7 8 > 10 11 12 13 14 IE 16 17 II II 20 21 22 23 24 25 M 27 26 29 30 3t
e.u-
50-
M
CO 4.0
3 >
3 x 3.0-
~ 2.0-
1.0-
s 11.000-
33
J > 14.00B-
5 3 12.000-
- 10,000'
0.91
024
0.22
020
5 "•"
S i aw
li °14
5 2 ° u '
b ow
OJI-
OJi
Ui'
MB-
0-
4 A^__^^— -4k— «— -~4u-^^-^^^^— ^-Ar— *-^^-^— A— ^^J.
^^x\^-^ -*- • *
^^^__^__^____^____^_^^
*r
A A A A
A A A A A
A
6-°
-60
I
-40 -"g
-3.0 »"
-20
-t.D
16.000 5
14.000 SH
SI
5
12000
-------
fi 4.0
li
3 j 3.0
i
1.0
; 11,000
ii
i 12.MO
f
'• »MtJ
Ml
MI
Ml
i ».*
K "•
E "•
MiH
Ml-
I
DAILY COMPOSITE OF COAL FIRED IN UNITS 1 t 2
1 2 3 4 B « 7 « »
6.0-1 1 1 1 1 1 1 1 1_
3.0
h2.0
s
3§
> >
,. e
iS
ui D
S
S
S2
°Q
5'
j
~
16,000-
14.000-
12.000
10.000-
0.26
0.24
0.22
0.20-
0.18.
0.16
0.14
0.12
0.10-
o.oa-
O.M
0.04
0.02
0
*^~ w*^^*^~~~*~~~^±—+~^ A A —>• A ~^ — *^~- ^A—^_ ^-^
^^X
A
^
"
,-16,000 -
CD
•H I
-14.000 g5
-12.000
-------
DAILY COMPOSITE OF COAL FIRED IN UNITS 1 * 2
:UR
Y BASIS)
3 °.
|
»
U
Sg
H m'
< ^
I K
S
J
0 us
K >
5*
t
6.0-
5.0-
4.0-
3.0-
2.0-
1.0-
16.000-
14,000-
12.000-
10.000-
026
0.24 •
0.22-
020-
0.18 •
0.16 •
0.14
012-
0.10-
O.OD-
0.0«-
0.04-
002-
0"
1 ' ' ' ' ' ' ' ' ' '
^ A^^^^ "*^" ^ *^ '** * ' * — »-— _ ^ .S ^ — * fc~^~~*
A ^ A^>^^*^
^-*--^_ _. ^ . . A ^^^^>^->»__^>*~--*----^_^__^
. — A-^*'^ A "^ — *^ -* ^^r--"*
A
A .
A . A
•5.0
1
•4.0 .*«
,., J!
J;
•2.0 "
•1.0
-16XMO 5^
-14,000 £ 5
1^
-12.000 Jj
|S
-10,000 -
•0.24
-0.24
-0.22
-020
-018 |
- 0.16 .* J
0.14 S 0
.lo.***"1
0.21
•0.22
0.20
•0.11 .'l
Sr
j
•0.10 I"
•001
•O.M
•IJ4
.00)
Prepared by Southern Company Services, Inc.
NOVEMKR 1978
A-6
a 30
KEY. K TOTAL MOHTURE
-•— PLAMT AMALYOM AVE Ml
A OUTIIDE LAOOBATORY RANGE XtM.f1
-------
DAILY COMPOSITE OF COAL FIRED M UNITS 1 A 2
«
.i
ij
i!
S
EC a
3).
3°.
"s
w
D§
«*
Si
X (-
S
i
Q ID
J O
5*'
1
60-
5.0-
40
3.0-
2.0-
1.0-
16,000-
14.000-
12.000-
10.000-
0.26 •
0.24
0.22
020-
018-
0.16-
0.14-
0.12-
0.10-
0.06-
0.06-
0.04-
0.02-
0-
A^_ ^ -____^__^
» * *— -"* " * --"^ *"~"*^»A__^— " * ^~~^
1— *— -H— - A A A
J .-— -Tfc-- ^_fc- - A -, .
A
.
A
A A . A A
A
•6.0
•60
-40
-30
-2.0
-1.0
•16,000
-14,000
• 12.000
-10000
-026
-024
-0.22
-0.20
-0.18
-0.16
0.14
-0.12
-010
-0.06
-0.06
004
-002
-0
f
£c
5^
sS
-
—
H I
>
D ^
Z
P
2"
1
it "
So
P
—
» 10 11 12 13 14 16 16 17 16 19 20 21 22 23 24 26 26 27 26 29 30 31
KEY: * TOTAL MOISTURE
—•— PLANT ANALYSIS AVE 5.80
A OUTSIDE LABORATORY RANGE 3.93 8.85
DECEMBER 1075
DAILY COMPOSITE OF COAL FMED M UNITS 1 A 2
10 11 12 13 14 It 16 17 II 16 JO 21 22 23 24 26 26 27 26 29 30 31
6.0-
6.0-
4.0-
3.0-
2.0-
1.0-
H.OOO
14.0W-
12.000-
10JHJ-
BJt
OJ4-
0-22
0.2*
O.W
r
Ui-
m-
OJt-
e-
1 1 1 1 , ! 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
^ A /-^ + _ _ ^
^ ~^^- _^-*-^
^ ^_^_^____^_ ^^-^_^_^ .— * -_A-_
A T A . * ^^^~
A
^ A A
A
A
~~l ' • ' ' ' • ' ' ' •.•'„'..'.,'..'.« JB H a 2124262627262630 31
-60
•5.0
3
-4.0 .' |»
5?
-3.0 ox
•2.0 ~
•1.0
•16.000 |^
•14.000 J *
•12.000 <{
V r
2 m
10,000 ~
0.26
•0.24
•0.22
OJO _
• 0.16 J
•016 .*S
014 SS
•012 «J
-0.10 1
• am
•0.06
•0.04
• o.«
•0
1 »»4 •• 7 • t W It
Prepared by Southern Company Servicesp Inc.
12 It 14 II M 1' '* «
JANUARY 1076
MY:
-•— PLANT ANAL YM
A OOTIIDE LAIOHATOHY
* TOTAL MOUTUM
AVE S 67
HAMOE X67-»H
A-7
-------
DAILY COMPOSITE OF COAL FIRED IN UNITS 1 * 2
60-
_ 5.0-
1
X X 4.0-
» * 3.0-
1
2.0-
1.0-
^ 16,000
= i
i «
g >. 14.000-
P 3 12,000-
i 2
— 10.000-
026
0.24
0.22
0.20
VI
Z 0.18
S m 0.16
is -
5 * 0-12
j 0.10
008-
OM
0.04
002
0
_ 50-
\
=j§
" / 3.0-
I
2.0-
1.0-
s 11,000 •
2l
;J 14.000-
aS
l°- 12QQi.
s|
- 10.010-
0.26
OM
022
020
3 0.11
£ 3 0.11
z >
J « 0.14 •
5* «"•
1 .,..
Ml-
OJI-
0«-
OJI-
0-
A _
^./ \^ *. A
"^^^^-^^^^^^^^ /^x _--^* ^V^-— ^*— «^ N ^^ *^/^«
-^^/ ^~^^
t
" ' "• ""• — ^^* — *^ ""
1 234 56 7 1 • 10 11 12 13 14 16 It 17 11 19 20 21 22 23 24 25 26 27 21 29 30 31
KEY % TOTAL MOISTURE
FEBRUARY 1976 -»- PLANT ANALYSIS AVE 5.56
A OUTSIDE LABORATORY RANGE 348-8 1
DAILY COMPOSITE OF COAL FIRED W UNITS 1 ft 2
A
^ A J^^^~^—~^ ~-^A^— ^A_^/^_ ^^
/*—-._ v'
• ••— -^-"* — " *\ £/ ~^~~*~^,/
X-^^
A_ A ^ A A • A «-^"
A- ., — A -. -^ . ,--*--, A .
A
A
A
A A
A A
•5.0
1
•4.0 )K „
•3,
-------
DAILY COMPOSITE OF COAL FIRED M UNITS 1 t 2
6.0
5.0-
5
% 4.0-
So 3.0-
I 2.0
1.0J
- 16,000
P j 12,000-
iu3
* 10.000
026
0.24
0.22
0.20
I «•«
0« 0.16
51 0.14
0.12
J 0.10
008
o.n
0.04
0.02
0
10 11 12 1]
11 17 II 11 20 21 22 23 24 26
21 27 28 M 30 31
*•
6.0
5.0
£
-I
4.0 .X o
•1.0
•16.000 —
10.000 -
0.26
0.24
0.22
0.20
0.18 I
0.16 Xj
0.14 i 0
012 ?5
010 £
0.08 5
0.06
0.04
0.02
10 11 12 13
14 16 IS 17 II 19
APRIL 1976
21 22 23 24 26 28 27 2» 28 30 31
KEY: * TOTAL MOISTURE
—•— PLANT ANALYSIS AVE 5.46
A OUTSIDE LABORATORY HANGE 2.79-1444
DAILY COMPOSITE OF COAL FIRED IN UNITS 1 * 2
6.0'
5.0-
4.0
1.0
5 11,000
3> 14.000-
UNO,
OJ*
OJM
::
466 7 ( « 10 11 12 13 14 IS 1* 17 II 1* 20 21 22 23 24 2E 2* 27 2t 29 30 31
i i ... i • 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 • 1 1 1 1 1 (-60
A
A
A
A
rS.O
I
r4.o .*
1-1.0
16,000 5
.14,000 E ?
.12.000 J >
Is
• 10,000 S
OM
•0.24
-0.22
020
Oil J
0.11 ' S
0.14 S|
0.10 §
0.01 *
O.M
004
002
0
Prepared by Southern Company Sarvicea, Inc.
12 11 14 IS II 17 II 11 » « » ^ M *
MAY 1976
21 27 21 21 » 31
% TOTAL MOISTURE
AVE 6.0S
OUTIIDC lAKHIATOHY HAHGE 2 W7 31
A-9
-------
DAILY COMPOSITE OF COAL HMD M UNITS 1 * 2
£
i!
s°-
*
_
=i
fi
^ °
< ^
s
a
CHLORIDE
IWT V DRV BASI
50-
40-
3.0-
2.0-
1 0-
16,000
1 4,000"
12.000-
10.000-
0 26-
0.24-
0.22-
0.20 -
0 IB -
0.16
014
012-
0.10-
008-
oot-
004
0.02-
0 -
*~"\ ^^~^v— ^— ^"^^t-^^^A^^^^X V^~*
\f A ^~~^^^ V^^^A— - A~^^_ — . — ^ — ^^
J^ ^^ ^-_ ^ ^*— *-—^_ . A
t.
A A A A A
KEY % TOTAL MOISTURE
•6.0 —
•4.0 0|
'« I'
S
•2.0
•1.0
rie.ooo 5
M
-14.000 E-i
'nz
Sn
•12,000
5*
t
50-
4.0
3.0-
2.0-
1.0-
11.000-
14.0001
12.000
10,000-
O.M
OM
0.22
029
0.1*
0.1«
014
012
0.10-
OJ«-
Mi-
OJ».
O.B-
-A.
. / Nw_ _^^^*^
• -S ^* , . , , ^ -+ «^
~- •—^^^^-—~ ' '
. »_ , . T
• 6-0
r» |
r40 S*
EC-
•3.0 £
»
-2.0
•1.0
•16.000 i
-14,000 ( H
'si
.12.000
-------
APPENDIX B
PROCESS AND BOILER OUTAGES
B-l
-------
SYSTEM AND BOILER OPERATING HISTORY'
CEA/ADL DUAL ALKALI PROCESS
BOILER OUTAGES
CAUSE/EXPLANATION
TUBE LEAKS & LOAD ORDERS AFTER REPAIRS
LOAD ORDERS
2SEE EXPLANATION FOR PROCESS OUTAGE NO. • (BOILER ON LINE )
LOAD ORDERS
LOAD ORDERS
LOAD ORDERS REPAIRED AIR PREHEATER SEAL LEAKS
LOAD ORDERS
PERIOD
2/18/75 2007 2/24/75 0449
2/28/75 1642 3/3/75 0410
3/24/75 1600 4/3/75 1300
5/3/75 0821 5/4/75 0544
5/7/75 1627 5/12/75 0510
7/4/75 2034 7/7/75 0435
7/19/75 1945 7/21/7S 051G
DURATION
(MRS)
128.70
59.47
238.00
21 .38
108.72
56.02
33.52
NO. MO.
•Ti
• H"1
* El —
iiM
I A
' H^
|;
• l~
• r"! M
HI A
^1 Y
•
• J
• :
6DT
^H u "
' Hi'
f
^H u
1
B-2
Prepared by Southern Company Services, Inc.
-------
SYSTEM AND BOILER OPERATING HISTORY1
CEA / ADL DUAL ALKALI PROCESS
— •
PROCESS OUTAGES
MO]
MMB
E |B
BE
^H
»C
M H
A H
R H
— 1
^H
A ^H
p H
R ^
— 1 —
^M
M ^"
A LH
Y ^1
L_
^Z
1
4
N H
-I
J h— 1
L *
L
~~ —
A
U
G
.
NO.
t
^4
5
1
— f
7
8
9
10
11
12
13
v
y
\
\
15
(HRS) 1 PERIOD
J2/3/75 1730
2.5O
1.26
2.67
13433
92.75
(S.2S
238.00
375.00
66.5
138.25
77.50
53.00
64.25
47.75
g / 13.00
$ ^ 33.00
/ 32.50
3 1
2\ 39.00
^ 1442.50
2/6/75 1430 2/6/75 1700
2/12/75 1415 2/12/75 1530
2/13/75 0930 12/13/75 1210
2/18/75 1930 2/24/75 0950
2/28/75 1500 3/3/75 1145
3/24/75 1045 3/24/75 1600
3/24/75 1600 4/3/75 1300
4/16/75 0000 5/1/75 1500
5/1/75 1630 5/4/75 1100
5/7/75 1700 5/13/75 1116
5/21/75 0830 5/24/75 1400
5/25/75 1200 6/27/75 1700
7/4/75 1900 7/7/75 1115
7/7/75 1300 7/9/75 1245
7/975 15OO 7/10/75 0400
7/10/75 0400 7/11/75 1300
7/15/75 0930 7/16/75 1800
7/16/75 1800 7/18/75 0900
7/18/75 0900 9/16/75 1130
CAUSE/EXPLANATION
— .
PRnPF« «wi iTn
UOWN WHILE PLANT REPAIRED A LOOSE PRECIPITATOR CABLE (A PRECAUTION ONLY
CLEANED CiC03 FROM THE FEED FORWARD CONTROL VALVE FROM THE THICKENER TO THE ABSORB
ER. (N,2C03 WAS FED TO THE THICKENER HOLD TANK DURING STARTUP BUT SWITCHED TO THE
THICKENER FOR LATER OPERATION
ESn'iFP.JKfJPyX- FIRED CHEATER (DAMAGED REFRACTORYI
BOILER DOWN TO REPAIR TUBE LEAKS & LOAD ORDERS AFTER REPAIRS
BOILER DOWN FOR LOAD ORDERS
ACID CLEANING TO COMPLETE CaCOjOISSOLUTION CAUSED BY NijCO^EED TO
THICKENER HOLD TANK. (SEE OUTAGE 2) (U. RECYCLE WITHOUT LIME FEED) COUNTED
AS AN OUTAGE SINCE SO, REMOVAL EFFICIENCY IS LOW. GENERALLY LESS THAN (OK.
LOAD CONTROL REQUESTED A BOILER OUTAGE AT THE PLANT. UNTIL THIS TIME CEA AND CIC HAD
BEEN RUNNING ON UNIT 18. 2 RESPECTIVELY. GULF PREFERRED TO SHUTDOWN UNIT 2. CIC HAD
HAD LITTLE OPERATING TIME TO DATE AND WAS READY TO RUN; HOWEVER, THE GAS SPLIT CONTROL
SYSTEM THAT ALLOWED TWO PROCESSES TO OPERATE OFF ONE UNIT WAS NOT OPERATIONAL.
CEA WAS SHUTDOWN TO ALLOW CIC TO RUN. THE BOILER WAS NOT AVAILABLE TO CEA AS INDICATED.
SHUTDOWN TO MAKE SYSTEM MODIFICATIONS FOR THE EPA TEST PROGRAM. THE MAJORITY OF THE
OFF LINE MODIFICATIONS WERE REQUIRED BECAUSE THE COAL SULFUR CONTENT AND BOILER EXCESS
AIR WERE FAR OUTSIDE DESIGN LIMITS. GAS SPLIT CONTROL SYSTEM MADE OPERATIONAL OURING
THIS OUTAGE. (SEE OUTAGES)
FLUE GAS BLOWER WAS UNBALANCED DUE TO RUST CAUSED BY FLUE GAS LEAKAGE ACROSS THE ISO-
LATION DAMPER DURING OUTAGE 7.
BOILER DOWN FOR LOAD ORDERS
SHUTDOWN TO BALANCE SYSTEM FAN. OUTAGE LENGTH EXCESSIVE DUE TO INEXPERIENCE AND LACK
OF MAINTENANCE PERSONNEL. (SEE OUTAGE t)
SHUTDOWN SYSTEM DUE TO A LARGE HOLE IN THE FILTER CLOTH INEXPERIENCE AND A LACK OF
SPARE PARTS LED TO THE OUTAGE.
BOILER DOWN TO REPAIR AN AIR PHEHEATER SEAL LEAK
FLUE GAS BLOWER WAS OUT OF BALANCE AFTER START-UP. ISEE OUTAGE 8.1 PROBLEM CAN BE ELIMI-
NATED ON FUTURE SYSTEM WITH AN AIR SEAL SYSTEM.
ACID CLEANING
HOLE IN VENTURI RECYCLE VALVE. RUBBER LINER FAILED DUE TO THROTTLING F LOW WITH A VALVE
DESIGNED FOR BLOCK SERVICE. REPLACED WITH A SPOOL PIECE AND THROTTLED WITH DOWN-
STREAM VALVES
NOPTLEACuE£FoABT uNA^Tv^TROL STvSoiftE HAVE BY PASS P.PING «.OUND THEM.
FUTURE SYSTEMS SHOULD INCLUDE
ACID CLEANING. ATTEMPTING TO CLEAN ANY SCALE BUILDUP CAUSED BY OPERATING
OUTSIDE DESIGN CONDITIONS.
NITIAL REASON FOR THE OUTAGE WAS A HOLE THAT DEVELOPED IN THE ABSORBER FROM
ACID ATTACK AT A PINHOLE LINER FAILURE. PRINCIPLE REASON FOR THE DURATION OF
OUTAGE WAS LATE ARRIVAL OF REPLACEMENT VALVES. SEE THE TEXT.
B-3
Prepared by Southern Company Services, Inc.
-------
SYSTEM AND BOILER OPERATING HISTORY1
CEA/ADL DUAL ALKALI PROCESS
Continued
BOILER OUTAGES
CAUSE/EXPLANATION
REMOVED BLANK IN CEA GAS DUCT
BALANCE CEA I.D. FAN
LOAD ORDERS
UNIT NO. 1 I.O. FAN FAILURE
3 TRANSFORMER PROBLEMS (SEE EXPLANATION FOR PROCESS OUTAGE NO. 17)
LOAD ORDERS
5-YEAR TURBINE INSPECTION
BALANCE TURBINE
BALANCE TURBINE
BALANCE TURBINE
BALANCE TURBINE
BALANCE TURBINE AND GOVERNOR WORK
BALANCE TURBINE
WORK ON ADMISSION VALVES AND CAM
WORK ON ADMISSION VALVES AND CAM
PERIOD
9/7/7S 1028 9/7/75 1346
9/13/75 1353 9/15/75 0416
9/30/75
10/6/75
2/9/76
2/3/76
2/10/76
2/10/76
2/10/76
2/11/76
2/15/76
2/22/76
1707 9/29/75 OB26
1002
033B
10/6/75 0338
10/7/75 1600
11/26/75 1938 12/1/75 0514
2337 2/8/76 0946
1047
2007
1002
1556
2058
1942
0752
OB03
2/9/76
2/10/76
2/10/76
2/10/76
2/11/76
2/11/76
2/16/76
2/22/76
1653
0041
1339
1912
1432
2232
1917
1732
DURATION
(MRS)
105.60
874.15
Prepared by Southern Company Services, Inc.
B-4
-------
1
SYSTEM AND BOILER OPERATING HISTORY
CEA/ADL DUAL ALKALI PROCESS
Continued
PROCESS OUTAGES
MO.
DURATIOr
(HRS)
PERIOD
CAUSE/EXPLANATION
B9.SO
175.00
9/25/75 2130 9/29/75 1000
9/30/75 1000 10/7/75 1700
BOILER DOWN FOR LOAD ORDERS
BOILER DOWN, ID FAN FAILURE. BOILER WAS NOT AVAILABLE TO THE PROCESS UNTIL 1600 ON 10/7/75
DUE TO TRANSFORMER PROBLEMS IN THE PLANT.
11/24/75 0800 12/1/75 09OO
HOLE IN FIRST STAGE REACTOR BOTTOM CAUSED BV FEED LIQUOR ENTRANCE BEING TOO CLOSE TO
TANK BOTTOM. THE ENTIRE REACTOR DESIGN CHANGED DURING JANUARY OUTAGE SEE THE TEXT.
1/2/76 2300 3/12/76 OB30
BOILER WAS SHUT DOWN FOR FIVE YEAR TURBINE INSPECTION. MAJOR REPAHIS AND MODIFICATIONS
1AIERF MADE TO THE PROCtS.S SEE THE TE XT
Prepared by Southern Company Services, Inc.
B-5
-------
SYSTEM AND BOILER OPERATING HISTORY1
CEA/ADL DUAL ALKALI PROCESS
Continued
BOILER OUTAGES
CAUSE/EXPLANATION
PERIOD
DURATION
(HRS) I N0-
MO.
REMOVED CEA BLANK IN CE A CAS DUCT
SEE EXPLANATION FOR PROCESS OUTAGE NO. 25 (BOILER ON LINEi
CLEANED UNIT NO. 1 AIR PREHEATER
ECONOMIZER TUBE LEAK
INSTALL SAMPLING NIPPLE IN INLET DUCT (BOILER ON LINE)
3/10/76 1055 3/10/76 1242
4/25/76 0730 4/28/76 1120
5/22/76 2050 5/23/76 0333
,'10/76 0044 6/11/76 0420
6/19/76 0900 6/19/76 1530
M
A
R
A
P
R
M
A
Y
J
U
N
J
U
L
Prepared by Southern Company Services, Inc.
B-6
-------
SYSTEM AND BOILER OPERATING HISTORY1
CEA / ADL DUAL ALKALI PROCESS
Continued
-• -1
^. •
MO.
«»•>
M
A |H
R |B
B
1
p ^s
R ^B
-E
•1
^^
Y B^
B
•
^^•i
^B
|B
^=
i ^^^
V^BHi
U ••
N JS
•
^H
B*
— Bi
^
J
U
L
PROCESS OUTAGES
NO.
21
23
24
25
26
27
28
Y*
L
k\
r\ 32
\\
K \ 33
\ M
IV *
\^~ 36
(HRS) 1 PER'OD
0.58
26.28
0.13
26.50
1.50
75.67
81.00
8.50
3.25
1.33
3.75
0.75
1.50
32.25
20.50
g K(J
14.92
3/19/76 3/19/76 1110
3/25/76 1320 3/26/76 1537
3/29/76 0858 3/29/76 0906
4/9/76 1530 4/10/76 1800
4/18/76 1948 4/18/76 2115
4/25/76 0730 4/28/76 1120
5/14/76 0930 5/17/76 1830
5/22/76 2030 5/23/76 1500
5/28/76 1215 5/28/76 1530
6/4/76 1600 6/4/76 1720
6/5/76 0900 6/5/76 1245
6/5/76 1500 6/5/76 1545
6/9/76 0900 6/9/76 1030
6/10/76 0030 6/11/76 0845
6/12/76 1800 6/13/76 1430
6/19/76 0900 6/19/76 1530
6/29/76 2116 6/30/76 1210
7/3/76 0930
CAUSE/EXPLANATION
BROKEN LINE TO REACTOR FROM VENTURI (CORRODED BOLTS]
BROKEN COUPLING ON REACTION TANK AGITATOR (WELDED]
INSTALLED SAMPLE PORTS ON FLUE GAS INLET DUCT
AGITATOR SHAFT BROKE AT NEW WELD (SEE OUTAGE 21)
UNIT PRECIPITATOR DOWN MATERIAL BALANCE PERIOD IN PROGRESS, THEREFORE SHUT DOWN TO
KEEPPARTICULATESOUTOFTHE SYSTEM.
PLANT TRANSFORMER OUT. POWER AVAILABLE AGAIN AT 1 100.
REACTOR AGITATOR COUPLING BROKE AT PREVIOUS WELD (SEE OUTAGE 21). MACHINIST NOT
AVAILABLE UNTIL E/17/76 AT 0800. OUTAGE AFTER 1830 ON 5/15/76 DUE TO LACK OF MAINTE-
NANCE.
UNIT NO. 1 DOWN TO CLEAN AIR PREHEATER
REPLACE REACTOR BLEED CONTROL VALVE. RUBBER LINED PLUG FAILED. (SEE NOTE FOR
OUTAGE 15)
SAUERISEN 33 CEMENT LINER THAT WAS INSTALLED IN AN UNSPECIFIED LOCATION IN THE
VENTURI CAME LOOSE AND PASSED THROUGH THE VENTURI PUMP. DESTROYED THE PUMP LINER
AND IMPELLER'
LEVEL CONTROL PROBLEMS. FOUND ABSORBER PUMP SEAL WATER WIDE OPEN. CORRECTED PUT
SYSTEM BACK ON LINE
SAME PROBLEM AS OUTAGE 29. ALL LOOSE MATERIAL REMOVED FROM VENTURI
CHANGED INLINE SO2 ANALYZER FILTERS. PROBE IS LOCATED AT A FAN DISCHARGE (i.e. HIGH GAS
PRESSURE). OUTAGE WOULD NOT HAVE BEEN NECESSARY IF THE PROBE LOCATION WERE DIFFERENT.
BOILER DOWN FOR ECONOMIZER TUBE LEAK
VENTURI LEVEL CONTROL VALVE STICKING IMETAL-TO-METAL. FILED & PUT BACK IN SERVICE)
(SEE NOTE FOR OUTAGE 151
INSTALL SAMPLING NIPPLE IN INLET DUCT
HOLE IN SCRUBBER RECYCLE SPOOL PIECE - REPLACED
FINAL SHUT DOWN
NOTES
1. DARK AREAS INDICATE PERIODS IN SERVICE
LIGHT AREAS INDICATE PERIODS OUT OF SERVICE
2. ALTHOUGH THE BOILER WAS ON LINE IT WAS NOT AVAILABLE TO THE PROCESS AS INDICATED
3 SINCE THE POWER SUPPLY SYSTEM WAS ADAPTED FROM THE EXISTING PLANT FACILITIES AND
SUFFICIENT SPARE CAPACITY WAS NOT AVAILABLE FBOM THE STATION SERVICE TRANSFORMERS
THE BOILER IS COUNTED AS NOT AVAILABLE TO THE PROCESS.
Prepared by Southern Company Services, Inc.
B-7
-------
APPENDIX C
PROCESS OPERATING DATA
C-l
-------
CEA OPERATING DATA
MOOO-
60000
t_ 5 in » * 40000'
S^K§ 30000
20000-
10000-
,y ^ | DEt|gH QAC RATE
kr-w^V^S/v^^'A^--^^^ ^-^^^
-60000
-««.»,?.
3-.fSp
•20000
10000
3000
g c 2000'
& Q
z { 1000
0
100
0
a; n
i- •»
1= ,0
S •
Z GO
I«l5
§ss«
< u
> ^ 3 J«
0 u g O 7
5
18
16
°£ U
aS- i2
"> Q *" —
Hi"
q ct
S °-
i
i
.
iii
~
1.4
i 1J
w w 1-0
i s °-«
i a "
£ * • 0.4
g OJ
§ "
! is "
8 * | 1 «3
K o g "
g S > " »J
*" 0.1
04
S "
Ms •
i§ 30
"
UA A r_Ax^ " "V ^^-^.'
-•
WjHCMDVAL EFPtCltNCIII
LACK Of INtTnUMMlT MAIIiTIHANCl
r "p i
' • ' Li
||] ''';;.' || U' ,, . |
1 II II III '
crr^- ^c^
ACGtMATI fH DATA MOT HKXMMD
, . . . .
. i
^ -^—_
"*
^ • *""*^
-3000
20* IP
83
•-» P
-0
100
» s
•0 — .*
70 ~|
•0 Jjj
SO >
lS3!
So OH
.,, • i
9 Sib
' ls|s
z
5
IB
16 >
14 j|
12 IflK*
10 i 52 i 5
8 s™ss
•4 •* 5
z
2
a
14
12
10 —
• 8 S "*
6 S J
4 -
2
0
1.4
12 «
1.0 r .,
0* H 1
0.6 j |
0.4 ~ 1
LOS g
•o* 5
04 hi
•" s|i 1
sis i
" 5|S 5
.,., i' 8
•0.0
. --^^^^ r» S
r" *'l
MOTES HY,.
' REFER TOO^AOE?UMifR FEiWIAHY 1«75 fflMURj O|W1 ~~THICICiMER OVFRFI ffl. cnNCENTIIATIIIM
LTZZIVENTURI ' CHLO"IDE
Pr«p«r«d by South.rn Caff my S.rvlc... Inc. ES3*"0"'1" 4^—* ACTIVE tooiim
S REACTOR EFFLUENT -LAI 0»f
» ttRUMEREFFLUENT • LAI
C-2
-------
. CEA OPERATING DATA
,-jj
i|
"g
is
8'
60000'
50000'
.-
ill* 40000
Si
8 30000
20000'
10000
BOILER
PROCESS
3000
£i 2o00
tu >
ii "™i>
0
•- HE^MRATE ' ' ' '
^~r^v'Vv>-~<'^^— ^J~ •*r~~J*\ -S\f*-S~~
\
\A fr -"•— .^A IV srt
-^ ^V V J - -- •"*"• v.'W*' ^|lr -~-~~
-10000
-50000
-40000 l^?.
.* > S?
-30000 m " ^-*
"
-20000
L 10000
•cS
BOILER 23
S £
PROCESS §|
•3000
2000 f ?
0
looo 5^
0
ill
A"-
'
?I
o.« - S
MOTES
1 NUXBEM IN CIHCLES
REFER TO OUTAQE NUMBER
• 10 11 1J 13 14 16 M 17 11
MARCH 1975
0.1
I-0.0
•fiEjmRE PROFS
EZ3v£NTURI
Pr«p»r«d by Southm Coquny Services, inc.
24 2G 21 IT U » 30 31
K|V«:
THICKENER OVERU f~ COHCENTRATIONl
•• « CHLORIDE
A A ACTIVE 1ODIUM
9 REACTOR EFFLUENT LAI
a KRLMER EFFLUENT LAI
- OPERATION LOOS
C-3
-------
CEA OPERATING DATA
10 11 12 13 14 16 16 17 II It 20 J1 23 23 24 » 2« 27
S =
90000
60000
f 40000
si 30000
20000
10000
BOILER
PROCESS
3000
s
1000
0
100
80
H
JUS*
0.3
02
0.1
0.0-
70
®
®
DESIGN OAS RATE _
60000
40000 K
30000 S™
3
20000
10000
I BOILER
PROCESS
•3000
1000 |$
0
100
- I
w _»
70 -|
" S
so
I
1 ' i I ' T-—i r——i 1 -r 1 1 1 , 1 1 1
12 13 14 16 1« 17 1| 1J 20 21 22 23 24 » 21 27 M 2« 30
0.4 > S
" s:! I
sis s
" "ii I
.. I'i
APRIL 1975
Frepurad by Southam Conpany S«rvicett Inc.
KtYS:
THICKEMER OVERFLQ" COMCEMTRATIOIIt
EZZZJvtNTu,, -— CHLO"IDE
* 4 ACTIVE •ODIUM
S»
» REACTOR EFFLUENT -LAI OfiMTMHI-O
9J KNUHER EFFLUENT • LAi
C-4
-------
CEA OPERATING DATA
a « » a 27 a » so 31
£ 90
Is «
I- TO
s »
M*
.1
si*
P1
OUTLET ANALYZfHS
tl*K CHUTE PLUGGED
VCNTUBI • ITRAVS
= 2 r ^
go o ^
z<= zo
^ w 5
z
_ll
a<|s
z
m *"
I i
1
i
, i
SSB |
Egil
5
HD1EI
.„.-...
1 " HUMVKFW
•IfEHTOOUTAOl
,0 ,1 12 11 14 « M « W 11 20 21 B
MAY 1875
EZZ] VEHTUIM
Frapucd by South*rn Co«p«ny Scrvieu, Inc.
THICKENER OVERf LOW CONCE«TI>ATION$
* * ACTIVE KOtUH
JH
« REACTOR EFFLUENT V« Ol
II EFFLUENT • LA*
C-5
-------
CEA OPERATING DATA
1 M it 11 17 11 it 10 21 22 a M
IZZZ2vENTI«l
ES3*
THKHIMH OVtHf LOU COMCEMTRATKMB
•• « CHLOHIDi
* * ACTIVE SODIUM
3S.
» REACTOR EFFLUtNT. LAI Of EHATIWI LOOS
••CIIUMER EFFLUENT LAI
C-6
-------
CEA OPERATING DATA
60000
"". 8 30000
20000
10000
"sv
p S BOILER
£ S PROCESS
5
3000
?
sg 200°
zi low
0
Q 10°
$ M
i; »
I- 70
£ ~
SO
ipl
< 11
2
25 5g »
| £ 2 -*'
5iz§ 7
8~ -
s
18
z 16
9g 14
£§>- "
Z LLJ S U I"
'Si- •
Is 6~
S°- 4-
2
0-
14
~ 10
£fc* a
S*" "" B •
4 ,
2
0
1.4
* !l
y ui 1-0'
1 2 OJ
j a "•*
z "
jj oj
§ M
! it "•
« si s "•
I Sss.^.
1 s§
0.1
0.0
» 70-
!ft "•
30'
11
. ' ' ' • L ' • 1 • * i, -i ._ . . . ' " *."_"." * . » » " » « M 31
' '" . DEBJflN.OAIJIAIE_
A - N\
— — ® , « . « ..^ _„. ®
^, ^ - ^.^^ ^^V^.
- ~ ^^^-y%^
- A , W&tAIMTENANCB AVAILAliE
V-^\ . / ,. j
W r - — H
1 i
\v
\ V HVhHAL IWT«UMEMT PNIMLEHC
1 MOtMtNTCNCE AVAILABLE
CITOT THAVEflUB
1 \
1 Nj
s^ r— ^
~~^v/_\| | \
r— IMT'UIMENTATKlMniOM.EMC
/ MO MAIMTErtCE MV4IUW1.E
i n
v — " — ^ r
11 ^— HUCttM IFFtlKNT
|
|
^^^^ X^™"1 — ' -t^-_ -
\
^- KHUMfF. EFFLUVNT
* .
A
4-
1 i * 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 i ' I ' I * l ' ' ' ' ' '
INUttBCM IN CIRCLE* JULY 1975 PRUKIRE PROM TMICKEMER OVERFLOW COHCEN1
REFER TO OUT AOENWRMR trrfl vtmmi * * CHUMIIM
•torn
50000
40000 5ar-
•3000. ^§5
20000
10000
BO.LER | |
PROCESS 1 f
a
3000
•2000 ||
< ™
1000 =^
"""
0
•100
» §
•M _ =
•70 ~|
•« s
L50
^
5? pi
„ M«l
i
H ° S
-0.3 S So J
-0.2 •< 1 j 5
«• i
-0.1 "
-0.0
-70 o
•30
RATION!
Prepared by Southern Owpaa; Senrieee, Inc.
« REACTOR EFFLUENT.LAB 'OPERATIC* LOGS
• KRUWER EFFLUENT-LAB
C-7
-------
CEA OPERATING DATA
< MOOO
3 £
%$£* 40000-
ilSS
i g 30000
20000
10000
DEBIONOURATE
IBOOOO
•60000
•30000 |"g3
20000
10000
"«„ ,s
S 5 PROCESS ® PROCESS § S
fc* §1
3000 ,
ft —
i\ 1000
0
100 1
u
$ H
8 BO
f~ 70
S »
60
'-
HI OUTMI MO. l«
!§3*
Sfii
< n
i •»>
i i I ° *'
s|!jj '
5
18
16
Z
2ff 14
Sir: 12
w a *) — .-
pl «
1* 6
2-
0-
14
12
5 S "
~ 4
2
0
1.4
i '•*
. ; 1.0
| S OJ
* ::
I oj
§ °-*
!., -
i :ii -
$ = • w
£ *
^ 0.1
0.0
8, "
is*
* fc t so
O jo
' 1 1 ' ' > 1 ' " ' ' 1 ' T I 1 1 1 1 1 1 1 1 1 1 1 1 J 1 1
3000
2000 ?g
1000 *8
0
rlOO
at
-BO ^B
•U £
O
L50
•8 s|
g5 OH
11 ' z
~-|
9 Mfs
7 iff-
5 ~"S
IB
16 fr
14 3S
12 ~ M *
10 J m | g
8 S" J J
' "11
Z
2
0
14
12
10 s
I s*i
«
4 -
2
0
14
1.2 „
1.0 P H
« 5 s
a. » S
0.4 - »
DJ 0
0.5 5
0.4 > I
0.1 S S § |
" 5ll 1
i* i
0.1 ~ »
0.0
70 fli
2 B m
'" si|
X
NOTO KEV*
i.u.m*ii>c«ric> AUQLUT 1A7S »cMlriFrr^t THtfttrnrn mrrnn rm rrarrn
MFMTOOUTAOIMUIMII —mriWF. 1 RMtntn OVtHf LgH CQHCEN-
rTTTI VENTUfll * CHLORIDE
[BAJJOJ&
r\Tl AMOKBIR * * ACTIVE tOOIUM
Pnptnd by Southern CoapniT terrluit Inc. ^!!
• HEACTM if FLUENT -LAB OPERATIO
I BCTUBMR EFFLUENT-LAB
C-8
-------
CEA OPERATING DATA
40000
soooo
p
J;3ju* 40000
u? U 30000
20000
10000
h- . ' . ' , ' , B V. V .".".» " , u " '• » » » » 2, » ,, „ , „ „ „ „ „
/\ A A A7^
J\^\~>> ^/v^A^/w ^J - ^/-
00000
50000
40000 S !2
•30000 m > ~*
-
20000
toooo
_2
B
3000
al -"
"1 1000
5t
0
A/v ,v AA« A"V/
^^^-~ ^ ^f-s^
uuivzinour
^^OMtTAHTHJr
0 "»
; go
S M
Is "
f~ 70
c«
S "
-, BO
pi!
< 11
2
sMs 9
6
IB
16
2 g 14
u a ,2
S £ >
z°£5 10
fl~
0
u
12
- 10
s >
Si* '
- 4
2
0-
].«
I "'
* ; 1.0
1 < 01
< ^ 0.1
z Oj*
u U
^^
VENTUHI - J TBArS
c______
c==\_— ~~ — I =»
INCTMJHEMt ATIOM. AND . \
MOTETMATl^MEMOVAL /*— J^ * 1 l/^U AB. .^%B if --1 •
"""""° \t \»MwHftA N/ 1
v 1 \J\j^i '< \J\JV' v '
"N^x ^S*~'^ "^*^~*"\» ^" .
_^___ •
*
». — L ~~
§ «*1
g
I .8 a*'
i s|""
" 0.1-
IK ™
3
30'
^^^-^ .
^^X_ s
•3000
2000 Jjt
* -I
1000 5*>
0
too
•0 _ 3
70 ~§
00 m
0
50 fr
9S 5 H
. z 9
n ' *
9 |S|Q
I ^V
•D °
S ^
ilfi
5Ii
?
s w
;5s
» ^ *
2
<•
M
1.2 u
1.0 r _
«•• 5 J!
o.« 1 S
0.4 - J
0.2 O
" 1
... > |
0.3 SSS 1
i» 1
0.1 ~ ••
0.0
•70 n
•> ^^
NOTES KEY1:
miiiarnf iHrinrin RFPTEMRFH 1ff7ff rniWIIIE ff""** THKKENE* OVERFLOII CONCENTRATIOM
HCFHTOOUTMINWIUR
Fr.)>u«l by South.tn
Swvlc... Inc.
CHLMIOt
ACT.Vi.XH.*.
9 REACTOK EFFLUENT -LAI OfERATlON (JOGS
a ICRUHER EFFIUENT LAB
C-9
-------
CEA OPERATING DATA
10 11 is 11 M ii 11 17 it it » ji a a u a. M 27 a a x 31
20000
- 10000
BOILER
PROCESS
3000
foe 2000
11 1000
0
100
! »
!a :
i «0
Ills
00J
TO
1
DESIGN GAS HATE
30000
20000
10000
I BOItER mm
2000
1000
- MVITCNCD TO VCMTUPHOMLVOHHATIOM
AMOMf • Of A WAKIOTa H MUH. INOWATIM CO, MOHPTION ON
TOf TM V FHOH »M *IIM TOO HIGH. IMCHUUf O VINTUHI &r TO MOID
MI 30 91
pREnuRt anon
EZ3 VENTURI
HfTk
TMICKENEB OVCHFLQM COMCEHTRATtOm
» • CHLORIDE
* 4 ACTIVE IODIUM
l!l
« REACTOR EFFLUENT LAI DURATION LOO
« KRUHER EFFLUENT-LAI
c-io
-------
CEA OPERATING DATA
K Ik
% 30000
B a M
"Sg
si
• 10000
BOILER
PROCESS
- -; -= BESlStLOAJ BATE
f\ f\s\ — — ^^^ — 1 —
I.NUMMMINCIHCta
REfEdiaOUTMENUMKR
12 13 H II II W II « »
NOVEMBER 1975
60000
"""SsEp
•30000 i
•20000
PROCESS | g
SB
•REHURi DROH
\///A VENTURI
Pr«p«r.d DT Sooth.ro Co-P«ny 8«rvlc.», Inc.
KEYS;
TUie.CM. OYJRFLOIII CONCEMTBATIOHt
» « CHLORIDE
A * ACTIVE nOIUM
BH
• REACTOR EFFLUENT -LA. OFERATION LOOS
IIKRUMER EFFLUENT • LAS
C-ll
-------
CEA OPERATING DATA
10 11 II 19 14 II U 17 II II 20 21 12 21 M
27 21 21 30 11
5 5 PROCESS
S*
sg
EMOVE
1*1
ION
TION
FIGU
ION
1st
SE
DR
EYI
1
I
I i
8 Hi"
I is:
s Is s .
s
NOTCt
I
I (OILED
I 'BOCEB
Uf *K>TE I AT rMf EMO
VfNTUMfZTMA**
4M MIHUII MKV MOT C
I J
\
0.1 s;
HI
i'i
REFER TO OUTAOE NUMWK
11 12 11 14 W II 17 II
DECEMBER 1975
HTt;
OVIRFLOIICOMCEN
Pr*p«r«d by Southern CoafiBj tanrlcM, Inc.
• • CHUWIDC
A * ACTIVE IOOIUM
• HUCTOR EFFLUENT LAI -
• InUMfR EFFLUENT. LAB
C-12
-------
CEA OPERATING DATA
•M
s 3 25
«*• j£
"£« ' „
pg •
wooo
moo
40000
90000
20000
. . . ' ' ' V.».".1. 1»|M ..1. "_!._.._..-. » . M , , » , „
_DH!as.a«RATj_
^ —
inm-i
9
50000
40000 K ? _
10000 I™!™
20000
loom
<2 „ }9V»IITUn,NEIWECT«M -, *
SS PROCESS m^mmm" J BOILER mm
is "*
5*s
il
~
O
Us
oc ""*
£
z
Ills
|ili
si 3|
O tt lu .
* 5 i _.
u
|
u 2 -
lp?
fit 2 —
§E
'
i J
S S
» ;
S *
z
1
g
1 s
I fls
| S* a
I 8§
g
PE
PROCESS S 2
3000-. "X
2000
1000
0
1-3000
- '\
100
00
BO
70
60
u
— II
1000 |^
"~
0
', », AMAkvlEM
/ «W«tB»»
rlOO
(a
-00 |
-BO -B
-70 *|
•60 m
-60
WEMTUfll • J,
1. """J xRaS
11
9
7
S
18
16
14
12
10
a
6
4
2-
0-
14
12-
10-
B
6-
4
2
0<
14
1J-
1.0
OJ
01
04
Oi
OJ
04
OJ
oa
0.1
u>
1 1
70
to
°S53
?zi
n *
<-s
• 9 O Z S o
r |T j x
•' go|§
i
&
18
16
" Jo
10 i»il
~ i S S
* a^
z
0
14
12
10 -
9 KB H
fi " m
4 ^ *
2
0
M
1-1 g
1.0 r H
«•» 5 H
O.B 5 S
0.4 ~ J
« ?
O.E 5
0.4 > 1
" ffip 1
oj 3 1 - S
*• i
0.1 - *
0.0
70 n
u 5 H H
«jr
»H 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 ' 1 ' ' T r -r i I i • —
•jQTgg KEYS!
1-NUHKRiiNcmci.u JANUARY 1976 DEMURE PROM THICKENER OVERFLOW CONCENTRATIONS
REFER TO OUTAOC NUMMR E3viNTUIII *" ' ChlLO"IDE
Pr.p.r.
-------
CCA OPERATING DATA
00000
60000
*" &t 30000
20000
10000
F"o
£ Q BOILER
<3
£ £ PHOCEIS
S'
3000
ffi —
z | 1000
0
a 10°
S — M
ec ~ 70
1 -
BO
fill
* 11
ijij !
5
18
|!|! :
g LU ~ 6
g" 4
* 2
0
14
12
- 10
s >
£5S •
E i ::
2
0'
1.4
s "•
« * 10
S 5 0.1
C tfc
i 2 "•
5 " 04
o 04
8 '•*
i s 0.4
> « S
THICKENER OVEI
CHLORIDE Wl
ACTIVE SODIUM
SEE KEY
S S S
0.0
!J
3 39
N
OUION OAI RATE
rMDUWTBMLtHOIJIUaOUE fO
y TUIMIM avMHMJL
e
HI OUT Ad I 11
Bill KEY1:
1NUMiER» IN CIRCLES FEIMIAMY 1t7C nil»MH Bflflft THKUMEROVEIIFLI
•0000
40000
30000
20000
10000
BOILE
PROCS
•3000
2000
1000
0
•100
•00
w
70
r*0
-60
11
- 7
3
0
OJ
0.1
0.4
-OJ
as
0.4
-03 3
' S
-
-------
CEA OPERATING DATA
» M II M „. „
!i
is
8*
§
fljl
Ili5
.1
sls
If:
x/v \^J V V ./^
WMl HAND MAT FMOM MUCTBD
tutvuuitR nttmiMB » TMI PLANT
CIC ON (MIT I.RIDUCIOaAIIUTt
LMM _
I'
n
L
30000 m
20000
10000
ON UNKNOWti
BOILEfl
FROCKS
H
if
W HAItED IME ASOVt)
VEHTUHI - 1 THAVI
~^i I, ,,i
NIXiD LMf/LWESTOMl
— .
iii|
'•SiM
s "
i i
&
? i
:: i
sgs
u
-------
CRA OPIMATWO DATA
10 11 » 11 u 11
17 »»»naiz>24»n27»aioai
5000Q-
P
f 40000
U 30000
20000'
10DOO-J
, — LOW niDucie WHIN OK TAKI*J OFF LIM DttlKlN OAI MTE
r\V
-Vyv
•• LOW MDUCtO DUE TO THAMFOftMf R
FHOBLIU IN fUUTT
L/
BOILER I
PROCESS I
a
l
i .§
i.o
0-1
i*
03
t3
MTUHI • ITRAVfl
10 11 II 13 14 II U 17 II II a 11 21 29 M
1-NUMEMINCIKLEI
REFER TO OUTAOINWIHR
Irmftitt by Soutlwrn COBpmT tarricu, Inc.
1t7«
^n
• iar
I!
.«.
0
100
•» 5
BO *
•
If
*
- I
" i
on s
02
-------
CEA OPERATING DATA
' ' » « • '• 7 I I 10 . 11 11 13 14 It 1| „
» a it a a
EOOOO
fiU
a!s«
. PUIOM OH RATE
si »»
H-
I BOILER 5 J
I PROCESS
2000
1000
*^~>—— »— -•
^^ _____
-^
i • 7 • • 10 11 i! i] 14 it 11 17 11 n a 21 a a 14 a 2f z7 a a
NOTU
l-NUHUnWCIHCLES
MCFIHTOOUTMEWJHKII
Pra»«r*d by South«rn
MAY 1t70
StrrlcM, Inc.
KEY!;
THICKENER OVERFLOW COMCEMTBATmH
» -• CHUMIIK
A * ACTIVE OOOIUU
• RIACtOK EFFLUENT -LAO
1 EFFLJENT LAI
OKRATION LOOS
C-17
-------
CRA
DATA
HUTU
1 NUHKUt IN CIKUS
n ran TO OUTAIM HUHKR
by Soutkm
Mtttt 197«
torrtcM, Inc.
« CHLOfllDC
— * ACTIVE IOOIUM
— O»»ATIOI(U
IFFIUEHT • IA!
C-18
-------
CEA OnilATINO DATA
BOOM
UOOO
3U
UJJ 40000
Mil
i § 90000
20000
10000
j! BOILER
[s PROCESS
j
3000
si —
»-
0
o 1CO
« n
«?. "
f "
(0
IjN .„
\\\\ °
8Ni 7
5
11
5» »
'o
2 „ £ 3 10
z-
0-
14
12
I > 10
Si* B
> Uj i
S 4
2
0
1.4
1 £
j °-t
S °'*
04
0.2
OJ
.1 "'
'• <
0,1
0.0
» 1
li>
II
^ ^o^ ;
•"•]»—•„.
V
""^ \f~~ WUTOOHPl F°"
'
ITIIAVl ~
1
1 1
=n
B
BU! IE01
...»»»»u».u JULY 1*7* Miauiiiiiiion THtcmrwi BVMFLOU CONCEHI
••MOO
•40000 K ^ _
•30000 m > H
-
•20000
10000
BOILER J |
rnocsss § 2
S-
9000
1000 J,f
0
•100
•BO ^
•BO -S
•70 -§
•* s
.., ii
9 nl$
7 SsM
i6i
6
11
16 >
14 3 §
10 5^i~
4 *5
2
0
14
12
10 _
• s» |
6 m f>
4 ^ *
2
0
14
1.2 „
1.0 n j
Tl ?
04 s ~
0.1 - IS
z S
04 * J
OJ J
o.s s
" §8 ^
«asM
0.0
70 J
3§S
•30
HATIOMS
REFI* TO OUTWH NUMM*
Pnptr«4 by foutlun Coffmf SoinleM, Inc.
EZZZlv
• DUCTMIFFLUINT -LAI Off RATION UXt
«R Iff LUtHT - LAI
C-19
-------
APPENDIX D
TYPICAL PROCESS FLOWS
AND STREAM COMPOSITIONS
D-l
-------
TABLE D-l
STREAM COMPOSITION, JUNE 4. 1975 - (PERIOD 1)
O
GAS STREAM
Location
Flow Rate - (acfm)
Temperature (°F)
SO2 (ppm)
O., (Vol. %)
Pressure Drop ("WG)
LIQUOR STREAM
Location
Flow Rate (gpm or Ibs/min)
Composition:
% Suspended Solids
Liquid Analysis (M)
PH
SO,
sol;
HCOj
OH~
Solids Analysis (morals/gin)
Ca
"8
Ha
S03
SO,,
C03
Cl
2
Fan Exit
65,000
NAa
1150-1400
6-8
-^^->
9
11
Absorber
Feed
90
0
52.3 diluted
V
0
0
0
0
0
0
0
0
3
Scrubber
Exit
NAa
4
Absorber
Exit
NAa
45
7
Stackb
Gas
NAa
45 avg.
's i~3^4?3
12
Absorber
Recycle
1080
0
—
—
—
"
0
0
0
0
0
0
0
0
13
Tray
Feed
160
0
(V7.5)
—
-
—
0
0
0
0
0
0
0
0
15
Scrubber
Recycle
1250
0
5.2
0.01
0.245
0.965
.0.0
0.0
0.0
(0.01)
0
0
0
0
0
0
0
0
16
Reactor
Bleed
LCC
2-2.5
11.3
0.14
0.0
0.93
0.0
0.0
0.01
(0.01)
—
—
—
5.53
1.06
—
—
—
17 18
Thickener Filter
Underflow Cake
(45)
21-35 45
0
0
0
0
0
0
0
0
—
„
0.97
5.18
1.96
._
—
—
23
Thickener
Overflow
LCC
<0.01
12.1
0.09
0.0
0.84
(0.015)
0.0
0.045
(0.01)
—
~
—
—
—
—
—
—
25
Dry
Lime
(10.5)
—
0
0
0
0
0
0
0
0
(12.6)
—
—
—
—
—
(23.0)
—
28
Dry
Na2C03
2.5
0
0
0
0
0
0
0
0
32 33 35
Soda Ash Demister Cake
Water Sprays Wash
7 4 1.5
— —
_
_ _
_
— __ _
__
— — —
(9.45)
aThenDcouples Hot Operating
1 Reheater Not Operated
cFlow on Level Control or Gravity
( ) - Estimated or Daily Average
-------
Inlet Flue Gas (2)
Stack Gas
Gas (7)
Absorber
Exit (4)
Venturi Recycle
(15)
o
Dry Lime (25)
Reactor
System
Venturi
Scrubber
Reactor Bleed (16)
Scrubber Exit (3)
Air (6)
« No. 2 Oil (5)
Demister Wash (33)
r
Tray
Absorber
Makeup Water
(31)
Tray Feed (13)
Absorber
Recycle (12
Water Dry Na2CO3
(3:
Hold Tank
Absorber Feed (11)
Thickener Overflow (23)
(28)
Soda Ash
Mix Tank
Filter Cake
(18)
Thickener Underflow (17)
FIGURE D-1 SYSTEM CONFIGURATION-JUNE 4,1975 (PERIOD 1)
-------
TABLE D-2
GAS STREAM
Location
Flow Rate - (acfm)
Tenperature (°F)
S02 (ppm)
02 (Vol. %)
Pressure Drop ("WG)
LIQUOR STREAM
Location
Flow Rate (gpm or Ibs/min)
Composition:
Z Insoluble Solids
Liquid Analysis (K)
pH
sol
HS03
so;
col
HCO5
OH~
Cl~
Solids Analysis (mmols/gm)
Ca
Mg
Na
S03
SOi,
C03
OH
Cl
2
Inlet Gas
67,000
320-330
1100-1200
STREAM
3
Scrubber
Exit
135
COMPOSITION, DECEMBER 20, 1975 - (PERIOD 2)
4
Absorber
Exit
129
7
Stack
Gas
12 8a
35 avg.
5-5.5 NA
11
Absorber
Feed
•vlOO
0
11.75
0.13
0
0.90
(0.005)
0
0.05Q
(0.055)
0
0
0
0
0
0
0
0
12
Absorber
Recycle
1210
0
—
—
—
0
0
0
0
0
0
0
0
13
Tray
Feed
225
0
—
—
—
0
0
0
0
0
0
0
0
15
Scrubber
Recycle
1520
0
5.1
0.0
0.30
1.035
0
0
0
0.065
0
0
0
0
0
0
0
0
16 17 18
Reactor Thickener Filter
Bleed Underflow Cake
LCb 140 (55) c
2-2.5 (15-20) 47
11.8 — 0
0.145 — 0
0 — 0
0.985 — 0
0 — 0
0 — 0
0.045 — 0
(0.06) — 0
7.15
(0.4)
0.51
5.54
1.09
—
—
23 25
Thickener Dry
Overflow Lime
LCb (12.5)
<6.01
11.75 0
0.135 0
0 0
0.945 0
O.OOj 0
0 0
0.055 0
0.06 0
(12. 95)
(0.8)
.
—
—
—
(24. 85)
aReheater Not Operated
bLC - Flow on Level Control or Gravity
cEst'd Rate While Filter Operating
( ) - Estimated or Daily Average
28
Dry
Na2C03
1.0
—
0
0
0
0
0
0
0
0
32
Water to
Na2C03
5
0
—
—
—
—
—
—
—
—
35
Cake
Wash
6C
0
—
—
—
—
~
—
—
9.45
-------
Stack Gas (7)
(1)
O
I
Ln
Lime Slurry (25)
Inlet Flue Gas
(2)
Absorber Exit
(4)
Venturi
Recycle (15)
Reactor
System
Venturi
Scrubber
Scrubber Exit (3)
Air (6)
No. 2 Oil (5)
r
Tray
Absorber
Tray Feed (13)
J Absorber Feed (11)
Makeup Water
(31)
Dry
Water |\|a9COo (28)
(32)
Hold Tank
Absorber Recycle
(12)
Thickener Overflow (23)
Soda Ash
Mix Tank
I
If
Thickener
JJ
M
Wash Water (35)
Reactor Bleed (16)
Filtrate (19)
I
Filter Cake
(18)
Thickener Underflow (17)
FIGURE D-2 SYSTEM CONFIGURATION-DECEMBER 20,1975 (PERIOD 2)
-------
TABLE D-3
GAS STREAM
Location
Flow Rate - (acfm)
Temperature (°P)
S02 (ppm)
02 (Vol. %)
Pressure Drop ("WG)
LIQUOR STREAM
Location
Flow Rate (gpm or Ibs/min)
Composition:
% Insoluble Solids
Liquid Analysis (M)
PH
503
HSOj
SOl;
co|
HCOj
OH~
Cl"
Solids Analysis (mols/gm)
Ca
Mg
Na
S03
SO,,
C03
OH
Cl
STREAM COMPOSITION, MAY 12, :
2
Flue Gas
69,000
300-340
2400
5.5-6
11
Absorber
Feed
155-170
0
11.95
0.15
0
(0.71)
0
0
0.055
(0.30)
0
0
0
0
0
0
0
0
3
Scrubber
Exit
—
149
—
^— v~—
4.5
12
Absorber
Recycle
2400
0
—
—
—
—
~~
0
0
0
0
0
0
0
0
4
Absorber
Exit
138
~
13
Tray
Feed
265-300
0
—
—
—
—
—
0
0
0
0
0
0
0
0
7
Reheated
Gas
—
172
115a avg.
15
Scrubber
Recycle
1290
0
5.15
0.36
0.36
0.82
0
0
0
0.34
0
0
0
0
0
0
0
0
1976 -
16
Reactor
Bleed
LCb
(2.5-3)
12.1
.0.155
0
0.76
0
0
0.07Q
(0.32)
7.51
--
—
6.13
0.81
—
—
—
17
Thickener
Underflow
(15-30)
—
—
—
—
:
—
—
—
—
—
—
—
18
Filter
Cake
(i£0)c
55
0
0
0
0
0
0
0
0
7.56
(0.3)
0.64
5.88
0.83
—
—
—
23
Thickener
Overflow
LCb
<0.01
12.0
0.155
0
0.74
0.01Q
0
0.06Q
0.31
0
0
0
0
0
0
0
0
25
Lime
Slurry
(24)
(19)
—
—
—
—
:
13.05
(0.6)
—
—
—
—
25.05
—
28 35
Dry Cake
Na fO 3 Wash
2.0 8C
—
—
—
—
__
_
—
—
—
— —
9.2
—
__ __
aReheat Gas Adds Net 5-10 ppm SO, to Exit Gas
Flow on Level Control or Gravity
cEst'd Rate While Filter Operating
( ) - Estimated or Daily Average
-------
Inlet Flue Gas
(2)
Stack Gas
(7)
Absorber Exit
(4)
Venturi
Recycle (15)
a
Dry Lime (25)
Venturi
Scrubber
Scrubber Exit (3)
Reactor
System
Reactor Bleed (16)
Air (6)
i
No. 2 Oil (5)
Tray Feed (13)
Makeup Water
31)
f
Tray
Absorber
Absorber Feed (11)
c
I Absorber Recycle
(12)
f
Hold Tank
M-
Thickener Overflow (23)
Dry
Na2C03 (28)
Soda Ash
Mix
Tank
Thickener
Filtrate (19)
Thickener Underflow (17)
V
Wash Water (35)
f.
I
T
T
•*-,
I
I
( Filt
I
v^
1
I
Filter Cake (18)
FIGURE D-3 SYSTEM CONFIGURATION-MAY 12,1976 (PERIOD 3)
JUNE 22,1976 (PERIOD 3)
JUNE 29,1976 (PERIOD 3)
-------
TABLE D-4
o
oo
GAS STREAM
Location
Flow Rate (acfm)
Temperature (°F)
SO (ppm)
Cl (ppm)
0 (Vol. %)
Particulate (grs/scf dry)
Pressure Drop ("WG)
LIQDOR STREAM
Location
Flow Rate (gpm or Ibs/min)
Composition:
% Insoluble Solids
Liquid Analysis
pH
SOj
HS03
so"
C03
HCOj
Solids Analysis
Ca
S03
SO,
C03
OH (Alk.)
Cl
Wt. % Acid Insolubles
STREAM COMPOSITION, JUNE 22, 1976 - (PERIOD 3)
2
Inlet Gas
73,500
300-310
1950-2380
(25)
7-7.5
(<0.03)
11
Absorber
Feed
140
0
11. B
0.15
0
0.77
0
0.065
(0.26)
0
0
0
0
0
0
0
0
0
3
Scrubber
Exit
—
134
—
-
—
12.5
12
Absorber
Recycle
(1200)
0
—
—
—
—
:
0
0
0
0
0
0
0
0
0
4
Absorber
Exit
—
132
—
—
—
—^— • ~~
5.5-6
13
Tray
Feed
270
0
—
—
—
;:
"
0
0
0
0
0
0
0
0
0
7
Stack
Gas
—
131a
35 avg.
(
-------
TABLE D-5
a
GAS STREAM
Location
Flow Rate (acfm)
Temperature (°F)
S02 (ppm)
Cl (ppm)
02 (Vol. %)
Particulate (grs/scf dry)
Pressure Drop ("WG)
LIQUOR STREAM
Locat ion
Flow Rate (gpm)
Composition:
% Insoluble Solids
Liquid Analysis
pH
SO^
HSO"
scC
col
HCOl
OH"
ci"
Solids Analysis
Ca
Mg
Na
SO 3
SO,,
CO 3
OH (Alk.)
Cl
Ut. % Acid Insolubles
STREAM COMPOSITION, JUNE 29, 1976 - (PERIOD 3)
2347
Inlet Gas Scrubber Absorber Stack
Exit Exit Gas
75,000
330-355 139 136 136a
2100-2650 — — 110 avg.
(25) — — <1
7.0
3.2 — — (<0.03)
11.5-13 6-6.5
11 12 13 15 16 17
Absorber Absorber Tray Scrubber Reactor Thickener
Feed Recycle Feed Recycle Bleed Underflow
120-180 NAb 250 1220 LCC "40
2.5 5.5
4.8 11.9
_ — 0 O.lBj
0.385 0
0.995 0.945
0 0.01
— — 00
0 0.045
(0.27) (0.25)
„
—
__
18 23 25
Filter Thickener Dry
Cake Overflow Lime
(150) Lff (25)
51
0 12.0 0
0 0.16 0
000
0 0.90 0
0 0.02 0
000
0 0.075 0
0 0.24 0
3.92 — (12. 95)
0.13 — (0.55)
0.48
3.15
0.67
(25.2)
0.06
41.3
28
Dry
Na2C03
2.5
35
Cake
Wash
aReheater Not Operated
bNA - Not Available
CLC - Flow on Level Control or Gravity
( ) - Estimated or Daily Average
-------
APPENDIX E
SOLUTION COMPOSITION/pH CORRELATIONS
E-l
-------
8 _
7 -
6 -
5 -
0.4 0.6 _ 0.8
[HSOg] /[HSO3+ SOg ], (mol ratio)
FIGURE E-1
pH VS. SOLUTION COMPOSITION - SCHOLZ PLANT (SCRUBBER EFFLUENT)
-------
13
0.45 - 1.05 M Na2S04
12
I
D.
11
10
0.1
0.2
[OH'] in Reactor Effluent, (M)
0.3
FIGURE E-2 pH VS. [OH'] IN ABSENSE OF CARBONATE - SCHOLZ PLANT (REACTOR
EFFLUENT)
-------
APPENDIX F
SCRUBBER SYSTEM EFFICIENCY ESTIMATES
F-l
-------
Estimates of scrubber stage efficiencies have been made based upon
operating data obtained on the prototype system during stable load
testing in Periods 2 (medium sulfur coal) and 3 (high sulfur coal).
During these periods the venturi and absorber (with two trays) were
operated in series. The system operating data used in calculating
the stage efficiencies are summarized in Table F-l. Vapor-liquid
equilibria data for SO above sodium solutions were calculated from
the equations of Johnstone1*:
PS02 (mm Hg) = M ~ ; (F-l)
where, log M = 4.519 - 1987/T(°C); (F-2)
S = mols dissolved SO /100 mols of water; and
C = mols active sodium/100 mols of water
The method used in estimating the stage efficiencies is illustrated in
Figures F-l and F-2. The passing stream composition points were first
located on plot of S02 gas composition versus S/C. The passing streams
correspond to (inlet S02 concentration, venturi bleed liquor composition)
and (outlet S02 concentration, estimated tray feed liquor composition).
The tray feed liquor was not sampled so the composition was estimated
based upon known rates of the regenerated liquor flow to the scrubber
system and tray tower recycle flow; the composition of the regenerated
liquor; and the pH of the recycle liquor. The tray tower and venturi
operating lines were then located. As can be seen in Figures F-l and F-2,
the estimated stage efficiencies are relatively insensitive to variations
in the composition of the tray feed liquor because of the steep slope of
the tray tower operating line and the low vapor pressures of S02 for
liquor compositions in this range. Thus relatively large errors in
estimating the tray feed liquor have almost no impact on estimated stage
efficiencies.
The stage efficiencies were then estimated based upon the assumption of
equal efficiencies for all stages by the iterative procedure of assuming
the stage efficiency until the passing stream composition points were
matched. This procedure yielded an average efficiency of 77% for each
stage. Actually, the venturi would be expected to have a stage efficiency
much lower than that of the trays; however, the data are not sufficient
to allow a reasonable differentiation of the different stages. More
operating data over a wide range of pH's and S02 concentrations, or
internal scrubber stream compositions would be required to accurately
define the different stage efficiencies.
F-2
-------
TABLE F-l
STAGE EFFICIENCIES
DATE 11/17/75 5/28/76
S02 CONCENTRATIONS, (ppm)
Inlet 1,160 2,400
Outlet 27 40
GAS FLOW, (scfm, dry) 43,600 44,500
SAT'D GAS TEMPERATURE, (°F) 130 130
TRAY TOWER OPERATION
Regenerated Liquor to Scrubber System, (gpm) 120 160
Total Tray Feed, (gpm) 320 270
Tower Pressure Drop, (inches of water) 6
VENTURI OPERATION
Recycle Flow, (gpm) 1,440 1,290
Venturi Pressure Drop, (Inches of water) 6 5.5
LIQUOR COMPOSITIONS, (M)
Regenerated Liquor
[S03 ] 0.124 0.110
[OH~] 0.046 0.145
[C03=] 0.011 0.016
[Cl~] 0.035
[SO"] O-O82
Venturi Bleed
[S03=] 0.030 0.062
[HS03~] O-250 °-236
[Cl"]
°-94 0.75
Pressure drop includes two trays, demister and intermediate ducting.
F-3
-------
1,500
1,500
1,000
s
c
CM
Q.
Q.
500
C [=] mols Na (active)/100 mols HjO = 0.57
Temperature = 130°F
Efficiencies of All Stages Assumed Equal
Venturi A P = 6 inches W.G.
Date of Operation -11/17/75
SO- Equilibria Data Source — Reference
Inlet SO2
1,160 ppm-*
Venturi
Bleed
Composition
Venturi Operating Line
Tray Tower Operating Line
Thickener
Overflow
Composition
I
I I
Estimated
Tray Feed
Composition
Outlet SOj
27 ppm _
J_
1,000
0.3
0.4 0.5 0.6 0.7 0.8
S/C [=] mols S02 in Liquor/100 mols Na+ (active)
500
Equilibrium
Line
0.9
1.0
FIGURE F-1 STAGE EFFICIENCY ESTIMATES FOR MEDIUM
SULFUR COAL OPERATION
F-4
-------
2,500
2,000
s
CD
O
c
CM
a
a
1,500
1,000
500
C [=] mols Na (active)/100 mols H..O = 0.68 'n'et S02
Temperature = 130° F .2,400 ppm"
Efficiencies of All Stages Assumed Equal
Venturi A P = 5.5 inches W.G.
Date of Operation - 5/28/76
SO Equilibria Data Source-Reference
Venturi
Bleed
Composition
Venturi Operating Line
Tray Tower Operating Line
Thickener
Overflow
Composition
Estimated
Tray Feed
Composition
e = 77%
Outlet SO
40, ppm -*
0.2
2,500
2,000
1,500
1,000
500
1.0
0.3 0.4 0.5 0.6 0.7 0.8 0.9
S/C [=] mols S02 in Liquor/mols Na+ (active) in Liquor
FIGURE F-2 STAGE EFFICIENCY ESTIMATES FOR HIGH SULFUR OPERATION (PERIOD 3)
F-5
-------
APPENDIX G
EQUIPMENT PROBLEMS
G-l
-------
APPENDIX G
EQUIPMENT STARTUP AND MAINTENANCE PROBLEMS
Section
SCRUBBER
Problem
MECHANICAL:
Period of
Occurrence
Action
Resolution
O
to
• Deterioration of refractory in reheat burner 1
chamber
• Slight corrosion and ash buildup on fan during 1,2
extended shutdown periods
• Separation of bond on rubber linings in absorber 1,2
and venturi bleed control valves
• Hairline cracks and pinholes in lining on 1
absorber recycle tank
• Leaking of liquor through pump seals and piping 1,2,3
on pump suction lines
• Deterioration of fan thrust bearing 2
• Deterioration of stack lining due to poor curing 1,2
• Deterioration of flake lining at gas entrance in 1,2,3
venturi
• Erosion of lining on liquor distribution shelf in 1,2
venturi
• Erosion of flake lining on sharp edge of duct at 3
expansion joint connecting venturi and absorber
• Pitting and stress corrosion cracking of some 3
316 stainless in scrubber internals
Replaced (Period 1)
Clean and rebalance fan after extended
shutdowns
Replace with SS valves under warranty
(Period 3)
Patched
Replaced packing and flange gaskets
(Periods 2&3)
Corrected
Corrected
Corrected (Period 1)
Corrected
Replaced (Period 3) Corrected
Patch/replace under warranty Corrected
Relined with different material (Period 3) Unresolved
Relined with different material (Period 3) Corrected
Found during inspection following comple-
tion of test program
Found during inspection following comple-
tion of test program
INSTRUMENTATION:
• Water condensation in pressure taps on level
controllers
• Plugging of flow-through pH units on recycle
tanks
• Poor reliability of Fisher-Porter level
transmitters
1.2 Relocated lines (Period 2) Corrected
1,2 Relocated sampling lines and increased Corrected
flow (Periods 2&3)
1,2,3 Replaced with Foxboro (Periods 2&3) Corrected
-------
APPENDIX
(Cont.)
Section
REACTOR
O
Problem
MECHANICAL:
• Plugging of dry lime feed chute to first reactor
• Buildup of solids in first reactor due to loca-
tion of dry lime feed chute, poor agitation, and
overfeeding of lime
• Broken agitator blades in second reactor
• Broken agitator shaft coupling in second reactor
• Broken shaft on reactor pump due to piece of
rubber lining from agitator blade caught in
impeller
• Failure of isolation valves on reactor pumps
• Erosion of flake lining on floor of second
reactor beneath agitator
Period of
Occurrence
2,3
2
2
2
Action
Installed new first reactor (Period 3)
Rebuilt coupling
Replaced (Period 3)
Resolution
Installed vibrator and air jets (Period 1) Corrected
Intermittent cleaning re-
quired
Replaced agitator shaft and impeller under Corrected
warranty (Period 2)
Corrected
Corrected
Overhauled (Period 3) Corrected
Relined with different material (Period 3) Corrected
INSTRUMENTATION:
• Erosion and plugging of flow-through pH unit
Replaced with immersion unit (Period 1)
Corrected (required weekly
or semi-weekly cleaning/
calibration)
FILTER/
THICKENER
MECHANICAL:
• Erosion of fiberglass scraper
• Erosion of plastic bridge valve due to solids
carried through cloth holes
• Loosening of cloth retaining ropes out of
caulking strips
1 Replaced with SS under warranty (Period 1) Corrected
1,2 Instructed operators to shut down filter Corrected
and repair holes immediately
1,2 Designed and installed new caulking strips Corrected
and reduced blower pressure (Periods 2&3)
-------
APPENDIX G
(Cont.)
Section
FILTER/
THICKENER
(Cont.)
Problem
• Loss of vacuum due to cracks in internal fiber-
glass trunnion tubes in filter
• Erosion/cracking of figerglass rocker arm on tub
agitator
• Insufficient agitation in filter tub to suspend
sand-like particles and grit
• Plugging of thickener underflow lines
• Deterioration of sections of lining in thickener
and thickener hold tank
• Failure of thickener rake motor overload control
Period of
Occurrence
1,2
1,2
Action
Resolution
Patched cracks—new caulking strip reduced Corrected
stress on internals when installing cloth,
drum overhauled prior to Period 3
Reinforced with SS (Period 1) Corrected
Occasionally wash tub—agitator over- Corrected
hauled prior to Period 3
Installed flexible lines and new back- Corrected
flushing provisions (Period 1)
Patch and replace lining under warranty Corrected
(Period 3)
Replaced (Period 3) Corrected
INSTRUMENTATION:
• Poor reliability of Fisher-Porter level control
transmitter on thickener hold tank
1,2,3 Replaced with Foxboro (Period 3)
Corrected
SODA ASH MECHANICAL:
• Clogging of dry feeder gate with lumps of soda ash 1,2,3
Replaced feed control to allow higher
gate position (Period 3)
Improved but still
clogged occasionally
• Failure of circuitry in heat tracing on piping
Replaced (Period 3)
Corrected
INSTRUMENTATION:
• Poor reliability of feed control system for soda
ash liquor
1,2 Redesigned control/operating procedures
(Period 3)
Corrected
GENERAL MECHANICAL:
• Water freeze damage to pump seal water rotameters
Replaced, also adjusted operating pro-
cedures (Period 3)
-------
TECHNICAL REPORT DATA
(rleaseread Instructions on the reverse before completing)
EPA-600/7-77-050C
3. RECIPIENTS ACCESSION-NO.
ITLE AND SUBTITLE FINAL REPORT: DUAL ALKALI TEST
AND EVALUATION PROGRAM; Volume IE. Prototype
Test Program--Plant Scholz
5. REPORT DATE
May 1977
6. PERFORMING ORGANIZATION CODE
C. R. LaMantia, R. R. Lunt, J. E. Oberholtzer,
E. L. Field, and J. R. Valentine
8. PERFORMING ORGANIZATION REPORT NO.
. PERFORMING ORGANIZATION NAME AND ADDRESS
Arthur D. Little, Inc.
Acorn Park
Cambridge, Massachusetts 02140
10. PROGRAM ELEMENT NO.
EHE624
11. CONTRACT/GRANT NO.
68-02-1071
2, SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Industrial Environmental Research Laboratory
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
Final; 5/73-4/77
14. SPONSORING AGENCY CODE
EPA/600/13
15. SUPPLEMENTARY NOTES .
IERL-RTP project officer for this report is Norman Kaplan, Mail
Drop 61, 919/541-2915.
ie. ABSTHACTyolume J-Q of ^Q repor|- covers Task m of a three-task program to investi-
gate, characterize, and evaluate the basic process chemistry and the various oper-
ating modes of sodium-based dual alkali scrubbing processes. The tasks were: I,
laboratory studies at both Arthur D. Little, Inc. (ADL) and IERL-RTP; II, pilot plant
operations in a 1200 scfm system at ADL; and in, a prototype test program on a 20
MW dual alkali system at Plant Scholz. Dual alkali system operating modes on high
and low sulfur fuel applications investigated included: concentrated and dilute dual
alkali systems, lime and limestone regeneration, and slipstream sulfate treatment
schemes. For each mode, the dual alkali process was characterized in terms of SO2
removal, chemical consumption, oxidation, sulfate precipitation and control, waste
solids characteristics, and soluble solids losses.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
COSATI Field/Group
Air Pollution
Alkalies
Sodium
Scrubbers
Desulfurization
Sulfur Dioxide
Calcium Oxides
Limestone
Sulfates
Prototypes
Air Pollution Control
Stationary Sources
Dual Alkali Process
Plant Scholz
13B
07D
07B
07A
08G
13. DISTRIBUTION STATEMENT
Unlimited
19. SECURITY CLASS (ThisReport/
Unclassified
20. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
------- |