DA J.S, Environmental Protection Agency Industrial Environmental Research EPA-600/7-78-055b
•» Office of Research and Development Laboratory- n-yo
Research Triangle Park, North Carolina 27711 MaTCh 1978
WATER RECYCLE/REUSE
ALTERNATIVES IN COAL-FIRED
STEAM-ELECTRIC POWER PLANTS:
Volume II. Appendixes
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 nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the INTERAGENCY ENERGY-ENVIRONMENT
RESEARCH AND DEVELOPMENT series. Reports in this series result from the
effort funded under the 17-agency Federal Energy/Environment Research and
Development Program. These studies relate to EPA's mission to protect the public
health and welfare from adverse effects of pollutants associated with energy sys-
tems. The goal of the Program is to assure the rapid development of domestic
energy supplies in an environmentally-compatible manner by providing the nec-
essary environmental data and control technology. Investigations include analy-
ses of the transport of energy-related pollutants and their health and ecological
effects; assessments of, and development of, control technologies for energy
systems; and integrated assessments of a wide'range of energy-related environ-
mental issues.
EPA REVIEW NOTICE
This report has been reviewed by the participating Federal Agencies, and approved
for publication. Approval does not signify that the contents necessarily reflect
the views and policies of the Government, nor does mention of trade names or
commercial products constitute endorsement or recommendation for use.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/7-78-055b
March 1978
WATER RECYCLE/REUSE
ALTERNATIVES IN COAL-FIRED
STEAM-ELECTRIC POWER PLANTS:
Volume II. Appendixes
by
James G. Noblett and Peter G. Christman
Radian Corporation
P.O. Box 9948
Austin, Texas 78766
Contract No. 68-03-2339
Program Element No. EHE624
EPA Project Officer: Frederick A. Roberts
Industrial Environmental Research Laboratory
Office of Energy, Minerals and Industry
Research Triangle Park, N.C. 27711
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Washington, D.C. 20460
-------
ABSTRACT
This study was conducted under EPA Contract No.
68-03-2339 to investigate water recycle/reuse alternatives in
coal-fired power plants. In the first part of this program, five
typical plants from representative geographical regions of the
United States were studied. The major types of water systems
encountered at these plants were cooling towers, ash sluicing,
and SOz/particulate scrubbing.
Computer models were used to identify the degree of
recirculation achievable in each of these water systems without
forming scale. The effects of makeup water quality and various
operating parameters were determined for each water system using
the models. Several alternatives for minimizing water require-
ments and discharges were studied for each plant and rough cost
estimates were made to compare alternatives.
In the second part of the program generalized implemen-
tation plans for the options identified in the plant studies are
presented. An implementation plan is presented for each major
water system (cooling towers, ash sluicing, scrubbing) where the
plans are divided into phases. The phases include system charac-
terization, alternative evaluation, pilot studies, and full-scale
implementation. The characterization phase for each system in-
cludes discussions of the important process variables to consider
when modifying the system. The alternative evaluation phase des-
criptions include discussions of various recycle/reuse alterna-
tives and present methodologies for evaluating the feasibility of
those options. The pilot studies and full-scale implementation
discussions include descriptions of equipment required and impor-
tant operating Variables to be considered.
The overall report is presented in two volumes. The
first volume discusses the recycle/reuse opportunities for cool-
ing tower, ash sluicing, and SOa/particulate scrubbing systems,
as well as for combined systems. The first volume also contains
the results of the studies to prepare generalized implementation
plans. Volume II presents the detailed studies for each plant
the selection methodology, the results of the laboratory studies
for ash sluicing, the results of kinetic studies for CaC03 and
Mg(OH)2, and a description of the models used in this study.
ii
-------
CONTENTS
PAGE
Abstract ii
Figures xvii
Tables xxii
APPENDIX A - POWER PLANT SELECTION CRITERIA
1.0 INTRODUCTION - A-l
2.0 SELECTION CRITERIA A-2
3.0 SELECTION METHODOLOGY A-4
3.1 Plant Identification A-4
3.1.1 Southwest A-4
3.1.2 Northern Great Plains A-6
3.1.3 Southeast A-8
3.1.4 Northeast A-8
3.2 Plant Selection A-ll
APPENDIX B - CHEMICAL CHARACTERIZATION OF
PLANT WATER SYSTEMS
1.0 INTRODUCTION B-l
2.0 SAMPLING B-2
2.1 Four Corners Generating Station
(Arizona Public Service) B-4
2.2 Comanche Generating Station
(Public Service of Colorado) B-5
2.3 Bowen Generating Station
(Georgia Power Co.) B-6
2.4 Montour SES (Penn. Power & Light) B-7
2.5 Colstrip SES (Montana Power Co.) B-8
3.0 ANALYTICAL TECHNIQUES B-ll
3.1 Calcium, Magnesium, Sodium,
Potassium, and Arsenic B-ll
iii
-------
CONTENTS
(Continued)
PAGE
3.2 Chloride B-ll
3.3 Total Sulfur and Sulfate B-ll
3.4 Carbonate -- B-12
3.5 Nitrate B-12
3.6 Phosphate B-12
3.7 Silicate B-12
3.8 Total Dissolved Solids B-12
3.9 Sulfite B-12
4.0 RESULTS B-14
4.1 Four Corners Generating Station B-14
4.2 Comanche Generating Station B-16
4.3 Bowen Generating Station B-18
4.4 Montour SES (PP&L) B-18
4.5 Colstrip SES (Montana Power Co.) B-21
APPENDIX C - CaC03 AND Mg(OH)2 PRECIPITATION KINETICS
1.0 INTRODUCTION C-l
2.0 LITERATURE SURVEY - C-2
2.1 Basis of the Literature Survey C-2
2.2 Solubility Data C-2
2.2.1 CaCO 3 C-2
2.2.2 Mg(OH)2 c_4
2.3 Solubility Product Data C-5
2.3.1 CaC03 C_6
2.3.2 Mg(OH)2 c_6
2.4 Dissociation and Ion-Pairing Data c-7
2.5 Formation and Kinetic Data C-7
2.5.1 CaCO 3 C-7
2.5.2 Mg(OH)2 C-10
-------
CONTENTS
(Continued)
PAGE
3.0 APPARATUS AND PROCEDURES
3.1 Description of Experimental Apparatus C-13'
3.1.1 Inlet Feed Systems - C-13
3.1.2 Batch-Solid Reactor C-16
3.2 Experimental Procedures C-16
3.2.1 Preparation of Feed Solutions C-17
3.2.2 Preparation of Batch-Solid
Crystallizer C-17
3.2.3 Experimental Run Procedure C-18
3.2.4 Sampling Scheme for Experimental
Run Procedure C-19
3.2.4.1 Inlet Feed Solutions C-19
3.2.4.2 Intermediate Diluted
Filtrate Samples C-20
3.2.4.3 Steady-State Diluted
Filtrate Samples C-20
3.3 Analytical Procedures C-20
3.3.1 Determination of Sodium,
Calcium and Magnesium C-20
3.3.2 Determination of Chloride C-21
3.3.3 Determination for Alkalinity C-21
3.3.4 Determination of Total Carbonate -- C-21
4.0 EXPERIMENTAL RESULTS FOR CaCO 9 C-22
4.1 Kinetics Data Processing C-22
4.2 Results -- C-24
4.3 Discussion of Results C-24
4.4 Conclusions C-27
5.0 EXPERIMENTAL RESULTS FOR Mg(OH)2 C-28
5.1 Kinetics Data Processing C-28
5.2 Results C-30
-------
CONTENTS
(Continued)
PAGE
5.3 Discussion of Results C-3Q
5.4 Conclusions C-33
APPENDIX D - ASH CHARACTERIZATION FOR COLSTRIP
AND MONTOUR FLY ASHES
1.0 INTRODUCTION - D-l
1.1 Background D-l
1.2 Summary D-l
2.0 EXPERIMENTAL D-3
2.1 Technical Approach D-3
2.2 Montour SES D-6
2.3 Colstrip SES D-10
3.0 RESULTS D-14
3.1 Montour D-14
3.1.1 Steady-State Operation D-14
3.1.2 Chemical Analyses D-22
3.1.3 Mass Balances D-23
3.1.4 Conclusions D-27
3.2 Colstrip D-27
3.2.1 Chemical Analyses D-27
3.2.2 Mass Balance D-28
3.2.3 Conclusions D-32
APPENDIX E - COMPUTER MODELS
1.0 INTRODUCTION E-1
2.0 MODEL DESCRIPTIONS E_2
2.1 Ash Sluicing (Bowen, Montour,
Comanche) E_
vi
-------
CONTENTS
(Continued)
PAGE
2.2 Cooling Tower Model E-5
2.2.1 Hot Side Slowdown at Bowen,
Montour and Colstrip E-5
2.2.2 Cold Side Slowdown at Comanche E-8
2.3 S02-Particulate Scrubbing E-10
2.3.1 S02-Particulate Scrubbing
at Four Corners E-10
2.3.2 S02-Particulate Scrubbing
at Colstrip E-15
3.0 SUBROUTINE DESCRIPTIONS E-17
3.1 Input Subroutines E-17
3.2 Equipment Subroutines E-18
3.3 System Balance Subroutines E-31
4.0 CHEMICAL EQUILIBRIUM PROGRAM E-33
4.1 Chemical Species E-33
4.2 System Equilibria E-35
4.3 Material Balance E-35
4.4 Constants E-37
4.4.1 Equilibrium Constants E-37
4.4.2 Activity Coefficients - E-39
4.5 Makeup Water Adiustment E-40
APPENDIX F - RECYCLE/REUSE OPTIONS AT FOUR CORNERS
(ARIZONA PUBLIC SERVICE)
1.0 INTRODUCTION F-l
1.1 Summary F-l
2.0 PLANT CHARACTERISTICS F-4
2.1 Water Balance F-4
2.1.1 Cooling' and Bottom Ash
Sluicing Systems F-10
vii
-------
CONTENTS
(Continued)
PAGE
2.1.2 Particulate Scrubbing System F-10
2.2 Existing Operations F-12
2.2.1 Simulation Basis F-12
2.2.2 Simulation Results F-17
3.0 TECHNICAL ALTERNATIVES F-21
3.1 Alternative One F-22
3.1.1 Simulation Basis F-22
3.1.2 Simulation Results F-24
3.2 Alternative Two F-24
3.2.1 Simulation Basis F-24
3.2.2 Simulation Results F-28
3.3 Alternative Three F-31
3.3.1 Simulation Basis F-31
3.3.2 Simulation Results F-31
3.4 Alternative Four F-35
3.4.1 Simulation Basis F-35
3.4.2 Simulation Results F-38
3.5 Conclusions F-38
4.0 ECONOMICS F-41
APPENDIX G - RECYCLE/REUSE OPTIONS AT BOWEN
(GEORGIA POWER CO.)
1.0 INTRODUCTION G-l
2.0 PLANT CHARACTERISTICS -- G-5
2.1 Water Balance G-5
2.2 Cooling Tower System G-9
2.2.1 Simulation Basis G-10
2.2.2 Simulation Results G-12
viii
-------
CONTENTS
(Continued)
PAGE
2.3 Ash Handling Systems G-16
2.3.1 Simulation Basis G-17
2.3.2 Simulation Results G-20
3.0 TECHNICAL ALTERNATIVES G-22
3.1 Cooling Tower System G-22
3.1.1 Simulation Basis G-22
3.1.2 Effect of Increased Cycles
of Concentration G-23
3.1.3 Effect of Calcium Concentra-
tion in the Makeup Water G-27
3.1.4 Summary of Cooling Tower
Alternatives G-33
3.2 Ash Handling Systems - G-33
3.2.1 Simulation Basis G-34
3.2.2 Once-Throueh Ash Sluicing
System G-34
3.2.3 Recirculatine; Ash Sluicing
System G-37
3.2.4 Effects of Carbon Dioxide
Mass Transfer G-40
3.2.5 Effect of CaSCV2H20
Supersaturation in the Pond
Recycle Water G-43
3.2.6 Summary of Ash Sluicing
Operations G-44
3.3 Conclusions -- G-44
4.0 ECONOMICS G-47
APPENDIX H - RECYCLE/REUSE OPTIONS AT COMANCHE
(PUBLIC SERVICE OF COLORADO)
1.0 INTRODUCTION H-l
ix
-------
CONTENTS
(Continued)
PAGE
2.0 PLANT CHARACTERISTICS --- H-6
2.1 Overall Water Balance H-6
2.2 Cooling System H-10
2.2.1 Simulation Basis H-ll
2.2.2 Simulation Results H-15
2.3 Ash Disposal System H-17
3.0 TECHNICAL ALTERNATIVES H-19
3.1 Cooling Towers H-19
3.1.1 Simulation Basis H-20
3.1.2 Effect of Increased Cycles
of Concentration H-20
3.1.3 Effect of Sulfate Concentration
in the Makeup Water H-25
3.2 Ash Handling Systems H-25
3.2.1 Simulation Basis H-27
3.2.2 Once-Through Ash Sluicing
System H-29
3.2.3 Recirculating Ash Sluicing
System H-33
3.2.4 Effect of Carbon Dioxide
Mass Transfer H-37
3.3 Conclusions H-40
4.0 ECONOMICS H-42
APPENDIX I - RECYCLE/REUSE OPTIONS AT MONTOUR
(PENNSYLVANIA POWER & LIGHT CO.)
1.0 INTRODUCTION I_l
2.0 PLANT CHARACTERISTICS I_5
2.1 Water Balance I_5
2.2 Cooling Tower System I_9
x
-------
CONTENTS
(Continued)
PAGE
2.2.1 Simulation Basis 1-9
2.2.2 Simulation Results 1-13
2.3 Ash Handling Systems 1-15
2.3.1 Simulation Basis 1-15
2.3.2 Simulation Results 1-17
3.0 TECHNICAL ALTERNATIVES 1-21
3.1 Cooling Tower System 1-21
3.1.1 Simulation Basis 1-21
3.1.2 Effect of Increased Cycles
of Concentration 1-22
3.1.3 Effect of Magnesium Concen-
tration in the Makeup Water 1-25
3.1.4 Conclusions 1-29
3.2 Ash Handling System 1-29
3.2.1 Simulation Basis 1-30
3.2.2 Recirculating Ash Systems 1-30
3.2.3 Effect of CaSOtt-2H20
Supersaturation in the Pond
Recycle Water 1-33
3.2.4 Effect of Carbon Dioxide
Mass Transfer 1-34
3.2.5 Arsenic Discharges 1-37
3.2.6 Conclusions 1-37
3.3 Summary 1-38
4.0 ECONOMICS 1-41
APPENDIX J - RECYCLE/REUSE OPTIONS AT COLSTRIP
(MONTANA POWER CO.)
1.0 INTRODUCTION J-l
2.0 PLANT CHARACTERISTICS J-5
xi
-------
CONTENTS
(Continued)
PAGE
2.1 Water Balance J-5
2.2 Cooling Tower System J-10
2.2.1 Simulation Basis J-10
2.2.2 Simulation Results J-13
2.3 Scrubbing System J-16
2.3.1 Simulation Basis J-16
2.3.2 Simulation Results J-20
3.0 TECHNICAL ALTERNATIVES J-25
3.1 Cooling Tower System J-25
3.1.1 Simulation Basis J-25
3.1.2 Cooling Tower Makeup
Treatment Alternatives J-26
3.1.3 Effects of Calcium Concentration
in Makeup Water at 20 Cycles
of Concentration J-28
3.1.4 Effect of Sulfate Concentration
in Makeup Water at 20 Cycles
of Concentration J-31
3.1.5 Summary of Cooling Tower
Alternatives J-31
3.2 Ash Handling (Particulate and S02
Scrubbing) System J-35
~~3.27i Simulation Basis J-35
3.2.2 Effects of Flue Gas
Ash Content J-35
3.2.3 Effects of Slurry Solids
Content j-37
3.2.4 Effects of Makeup Water
Composition J-38
3.2.5 Summary of Scrubbing
Alternatives J-40
3.3 Combined System Alternatives J-40
xii
-------
CONTENTS
(Continued)
PAGE
4.0 ECONOMICS J-42
APPENDIX K - POWER PLANT DATA REDUCTION
1.0 APS FOUR CORNERS STATION K-l
1.1 Four Corners Scrubbing System K-l
1.1.1 Four Corners Scrubbing
System Particulate Removal
Efficiency - -- K-l
1.1.2 Four Corners S02
Oxidation Rate K-2
1.1.3 Four Corners Flue Gas
Composition Calculations K-2
1.1.4 Batch Dissolution Results
with Fly Ash From Four Corners K-5
2.0 GPC PLANT BOWEN K-8
2.1 Bowen Cooling Towers K-8
2.1.1 Climatological Data K-8
2.2 Bowen Ash System (Units 3 or 4) K-10
2.2.1 Fly Ash K-10
2.2.2 Total Ash from All Units K-10
2.2.3 Recirculating Ash System Flows K-ll
2i2.4 Pond Evaporation K-ll
2.2.5 Ash Dissolution K-12
3.0 PSC COMANCHE PLANT K-14
3.1 Comanche Cooling Towers K-14
3.1.1 Inlet Air Composition K-14
3.2 Comanche Ash System K-15
3.2.1 Ash Flows K'1^
3.2.2 Pond Evaporation K~16
3.2.3 Ash Dissolution K~17
xiii
-------
CONTENTS
(Continued)
PAGE
4.0 PP&L MONTOUR - K-19
4.1 Montour Cooling Towers K-19
4.1.1 Montour Climatological Data K-19
4.1.2 Montour Heat Load K-21
4.1.3 Montour ACB Index K-23
4.2 Montour Ash System K-24
4.2.1 Montour Fly Ash K-24
4.2.2 Montour Bottom Ash and
Mill Rejects K-25
4.2.3 Montour Pond Evaporation K-26
4.2.4 Montour Calcium Removal Rate K-27
4.2.5 Montour Ash Dissolution K-29
4.2.5.1 Procedure K-30
4.2.5.2 Results K-30
5.0 MPC COLSTRIP K-34
5.1 Colstrip Cooling Towers K-34
5.1.1 Colstrip Climatological Data K-34
5.2 Colstrip Scrubbing System K-35
5.2.1 Colstrip Flue Gas Composition K-36
5.2.2 Colstrip Effluent Tank
Sample Consistency K-37
5.2.3 Colstrip Ash Dissolution K-39
APPENDIX L - ASH CHARACTERIZATION FOR FOUR CORNERS,
BOWEN, AND COMANCHE FLY ASHES
ABSTRACT L-l
1.0 INTRODUCTION L-2
1.1 Background L-2
1.2 Summary ~L-2
xiv
-------
CONTENTS
(Continued)
PAGE
1.2.1 Carbon Dioxide Sorption L-2
1.2.2 Closed-Loop Sluicing L-3
1.2.3 Leaching and Batch Dissolution L-4
2.0 CARBON DIOXIDE SORPTION TESTS L-5
Z.I Technical Approach L-5
2.2 Experimental L-5
2.2.1 pH = 11, Nonbuffered Test Medium L-5
2.2.2 pH = 11, Buffered Test Medium L-8
2.2.3 pH = 9, Buffered Test Medium L-8
2.3 Data L-8
2.4 Conclusions L-8
3.0 BENCH-SCALE SLUICING TESTS L-12
3.1 Technical Approach L-12
3.2 Experimental L-15
3.2.1 Comanche Steam-Electric Station -- L-15
3.2.2 Plant Bowen L-16
3.2.3 Four Corners Power Plant L-17
3.3 Results L-17
3.4 Conclusions L-17
3.4.1 Comanche Steam-Electric Station -- L-17
3.4.2 Plant Bowen L-25
3.4.3 Four Corners Power Plant L-29
4.0 BATCH DISSOLUTION AND LEACHING TESTS L-31
4.1 Technical Approach L-31
4.1.1 Ash Leaching L-31
4.1.2 Batch Dissolution L-32
4.2 Experimental L-33
4.2.1 Ash Leaching L-33
4.2.2 Batch Dissolution L-33
XV
-------
CONTENTS
(Continued)
PAGE
4.3 Results L-34
4.4 Conclusions L-34
4.4.1 Ash Leaching L-34
4.4.2 Comanche Steam-Electric Station -- L-34
4.4.1.2 Plant Bowen L-38
4.4.1.3 Four Corners Power Plant L-40
4.4.2 Batch Dissolutions L-44
4.4.2.1 Comanche Steam-Electric
Station L-44
4.4.2.2 Plant Bowen L-46
4.4.2.3 Four Corners Power Plant L-46
APPENDIX LA - TEST PARAMETERS AND CHEMICAL ANALYSES
FOR CLOSED-LOOP SLUICING TESTS L-47
APPENDIX LB - CORRELATION PARMETERS FOR CLOSED-LOOP
SLUICING TESTS L-60
APPENDIX LC - pH AND EMF VALUES OF CALCIUM AND
DIVALENT CATION SPECIFIC ELECTRODES
FOR BATCH DISSOLUTION TESTS L-68
XVI
-------
FIGURES
PAGE
APPENDIX C - CaC03 AND Mg(OH)2 PRECIPITATION
KINETICS
Figure 2-1. Visual and pH determination of the
time of nucleation. C-8
Figure 2-2. Time of nucleation versus initial
carbonate concentration. C-8
Figure 2-3. Variation with time, of pH (upper curve)
and of number of particles, during an
experiment. C-ll
Figure 2-4. Dependence of rate of nucleation
(particles/cm3 sec) on solution
concentration expressed as (IP) (aM aQu) • C-ll
Figure 3-1. Experimental system for liquid-phase
reaction study. C-14
Figure 3-2. Batch-solid reactor. C-15
Figure 4-1. CaC03 precipitation kinetics. C-26
Figure 5-1. Mg(OH)2 precipitation kinetics. C-32
APPENDIX D - ASH CHARACTERIZATION FOR COLSTRIP
AND MONTOUR FLY ASHES
Figure 2-1. Bench-scale simulation model of ash pond
facilities. • D-4
Figure 3-1. Sodium concentration in the pond vs.
time. -- D-15
Figure 3-2. Calcium concentration in the pond vs.
time. D-16
Figure 3-3. Sulfur concentration in the pond vs.
time. D-17
Figure 3-4. Magnesium concentration in the pond vs.
time. D-18
Figure 3-5. Chloride concentration in the pond vs.
time. D-19
Figure 3-6. Nitrate concentration in the pond vs.
time. D-20
Figure 3-7. Carbonate concentration in the pond vs.
time. --- D-21
xvii
-------
FIGURES
(Continued)
PAGE
APPENDIX E - COMPUTER MODELS
Figure 2-1. Ash sluicing simulation flow scheme. E-4
Figure 2-2. Hot side blowdown cooling tower
simulation flow scheme. E-6
Figure 2-3. Process simulation scheme for Comanche
cooling tower system. E-9
Figure 2-4. Four Corners scrubbing scheme (existing
operation). E-ll
Figure 2-5. Process model for Four Corners
Alternative Three. E-14
Figure 2-6.' Colstrip scrubbing simulation flow
scheme. E-16
APPENDIX F - RECYCLE/REUSE OPTIONS AT FOUR CORNERS
(ARIZONA PUBLIC SERVICE)
Figure 2-1. Arizona Public Service Four Corners
station water balance. F-5
Figure 2-2. Four Corners scrubbing simulation
scheme (existing operations). F-14
Figure 3-1. Schematic flow diagram for Four Corners
Alternative One. F-23
Figure 3-2. Process model for Four Corners
Alternative One. F-25
Figure 3-3. Process flow diagram for Four Corners
Alternative Two. F-27
Figure 3-4. Process model for Four Corners
Alternative Two. F-29
Figure 3-5. Process flow diagram for Four Corners
Alternative Four. F-32
Figure 3-6. Process model for Four Corners
Alternative Three. F-33
Figure 3-7. Process flow diagram for Four Corners
Alternative Four. F-36
xviii
-------
FIGURES
(Continued)
PAGE
APPENDIX G - RECYCLE/REUSE OPTIONS AT BOWEN
(GEORGIA POWER CO.)
Figure 2-1. Georgia Power Co. Plant Bowen Water
Balance. G-6
Figure 2-2. Process simulation scheme for Bowen
cooling tower system. G-ll
Figure 2-3. Process simulation scheme for Bowen ash
sluicing system. G-18
Figure 3-1. Acid requirements as a function of
makeup water calcium. G-29
Figure 3-2. Acid requirements as a function of
cycles of concentration. G-30
Figure 3-3. Acid requirements as a function of
calcium in circulating water. G-31
APPENDIX H - RECYCLE/REUSE OPTIONS AT COMANCHE
(PUBLIC SERVICE OF COLORADO)
Figure 2-1. Public Service of Colorado Comanche Plant
Water Balance. H-7
Figure 2-2. Process simulation scheme for Comanche
cooling tower system. H-12
Figure 3-1. Gypsum relative saturation as a function
of cycles of concentration at Comanche. H-23
Figure 3-2. Process simulation scheme for Comanche
ash sluicing system. H-28
APPENDIX I - RECYCLE/REUSE OPTIONS AT MONTOUR
(PENNSYLVANIA POWER & LIGHT CO.)
Figure 2-1. Pennsylvania Power and Light Company
Montour Plant Water Balance. 1-6
Figure 2-2. Process simulation scheme for Montour
cooling tower system. 1-10
Figure 2-3. Ash sluicing simulation model. 1-16
xix
-------
FIGURES
(Continued)
PAGE
Figure 3-1. Asbestos cement pipe manufacturers index
as a function of cycles of concentration. 1-26
APPENDIX J - RECYCLE/REUSE OPTIONS AT COLSTRIP
(MONTANA POWER CO.)
Figure 2-1. Montana Power Company Colstrip Plant
Water Balance. J-6
Figure 2-2. Colstrip cooling tower simulation flow
scheme. J-12
Figure 2-3. Colstrip scrubbing simulation flow scheme. J-17
Figure 3-1. Slipstream rate as a function of makeup
calcium concentration at Colstrip. *•---— J-29
Figure 3-2. Slipstream rate as a function of makeup
sulfate concentration at Colstrip. J-32
APPENDIX K - POWER PLANT DATA REDUCTION
Figure 4-1. Fly ash leaching studies at pH 6.0. K-31
Figure 4-2. Fly ash leaching studies at pH 8.1. K-32
APPENDIX L - ASH CHARACTERIZATION FOR FOUR CORNERS,
BOWEN, AND COMANCHE FLY ASHES
Figure 2-1. Test containers. L-6
Figure 3-1. Bench-scale simulation model of ash pond
facilities. L-13
Figure 3-2. Comanche steam-electric station calcium
dissolution rate in mix tank versus ar
in mix tank. ±§__ L-18
Figure 3-3. Comanche steam-electric station calcium
dissolution rate in mix tank versus acaaSOif
in mix tank. L-19
Figure 3-4. Comanche steam-electric station sulfate
dissolution rate in mix tank versus ago^
in mix tank. L-21
xx
-------
FIGURES
(Continued)
PAGE
Figure 3-5. Comanche steam-electric station sulfate
dissolution rate in mix tank versus
an acn in mix tank. L-22
La bu4
Figure 3-6. Comanche steam-electric station CaCOs pre-
cipitation rate in mix tank versus ar_arn
in mix tank. ±*-±\>l L-23
Figure 3-7. Plant Bowen calcium dissolution rate in
mix tank versus ap in mix tank. L-26
Figure 3-8. Plant Bowen sulfate dissolution rate in
mix tank versus aco in mix tank. L-27
bU It
Figure 3-9. Plant Bowen sulfate dissolution rate in
mix tank versus ar aCn in mix tank. L-28
Ca SO it
Figure 3-10. Plant Bowen CaCOs precipitation rate in
mix* tank versus a.n anr. in mix tank. L-30
Ca COs
Figure 4-1. Comanche steam-electric station meq acid/g
fly ash versus time. L-39
Figure 4-2. Plant Bowen meq acid/g fly ash versus time. -- L-41
Figure 4-3. Four Corners power station meq acid/g fly
ash versus time. L-42
Figure 4-4. Four Corners power station meq acid/g fly
ash versus time. L-43
xxi
-------
TABLES
PAGE
APPENDIX A - POWER PLANT SELECTION CRITERIA
Table 3-1. Potential Southwestern Power Plants A-5
Table 3-2. Potential Northern Great Plains
Power Plants A-7
Table 3-3. Potential Southeastern Power Plants A-9
Table 3-4. Potential Northeastern Power Plants A-10
Table 3-5. Selected Plants for Water Recycle/Reuse
Study A-13
APPENDIX B - CHEMICAL CHARACTERIZATION OF
PLANT WATER SYSTEMS
Table 2-1. Selected Plants for Water Recycle/Reuse
Study B-3
Table 4-1. Chemical Analysis of Aqueous Samples of
Four Corners Power Plant, Arizona Public
Service B-15
Table 4-2. Chemical Analysis of Aqueous Samples of
Comanche Generating Station Public Service
of Colorado B-17
Table 4-3. Chemical Analysis of Aqueous Samples of
Bowen Generating Station Georgia Power Co. B-19
Table 4-4. Chemical Analysis of Aqueous Samples of
Montour SES, Pennsylvania Power and Light
Company B-20
Table 4-5. Chemical Analysis of Aqueous Samples of
Colstrip SES, Montana Power Company B-22
APPENDIX C - ANALYTICAL TECHNIQUES
Table 2-1. Solubility of Calcium Carbonate in Water
in Contact with Ordinary Air C-3
Table 2-2. Solubility of Calcium Carbonate in Water
Essentially Free of Carbon Dioxide C-4
Table 2-3. Solubility of Calcium Carbonate at 25°C
as a Function of pH C-4
xxii
-------
TABLES
(Continued)
PAGE
Table 2-4. Solubility of Magnesium Hydroxide in
Water C-5
Table 2-5. Solubility Product Constant for CaC03 in
H20 Saturated with C02 C-6
Table 2-6. Time of Nucleation Resulting from Added
Carbonate --- C-9
Table 4-1. Experimental Data - Precipitation
Kinetics of CaC03 C-25
Table 5-1. Experimental Data - Precipitation
Kinetics of Mg(OH)2 — c-31
APPENDIX D - ASH CHARACTERIZATION FOR COLSTRIP AND
MONTOUR FLY ASHES
Table 2-1. Experimental Conditions (Montour) D-7
Table 2-2. Makeup Water Composition (Montour) D-8
Table 2-3. Initial Pond Water (Montour) D-9
Table 2-4. Experimental Conditions (Colstrip) D-ll
Table 2-5. Makeup Water Composition (Colstrip) D-12
Table 2-6. Initial Pond Water (Colstrip) D-13
Table 3-1. Montour Experiment 1 Regression Results -- D-22
Table 3-2. Final Sample Results (Montour) D-24
Table 3-3. Net Dissolution Rates from Montour Fly Ash D-25
Table 3-4. Montour Fly Ash Reactivity D-26
Table 3-5. Final Sample Results (Colstrip) D-29
Table 3-6. Net Dissolution Rates from Colstrip
Fly Ash D-30
Table 3-7. Colstrip Fly Ash Reactivity D-31
APPENDIX E - COMPUTER MODELS
Table 4-1. Equilibrium Program Key Species E-33
xxiii
-------
TABLES
(Continued)
PAGE
Table 4-2.
JTable 4-3.
Table 4-4.
Table 4-5.
Table 4-6.
Table 4-7.
Table 4-8.
Table 4-9.
Table 4-10
APPENDIX F
Table 1-1.
Table 2-1.
Table 2-2.
Table 2-3.
Table 2-4.
Table 2-5.
Table 3-1.
Table 3-2.
Equilibrium Species
Equilibrium Relations
Hydrate Numbers
Mass Balances
Montour Makeup Water Composition
Adj us tmen t s
Colstrip Makeup Water Composition
Adj us tment s
Four Corners Makeup Water Composition
Ad j us tment s
Comanche Makeup Water Composition
Adj ustment s
. Bowen Makeup Water Composition
Ad j us tmen ts
- RECYCLE /REUSE OPTIONS AT FOUR CORNERS
(ARIZONA PUBLIC SERVICE)
Summary of Recycle/ Reuse Options at Four
Corners
Monthly Variation of Flows from Selected
Streams at Four Corners
Parameters Calculated by the Equilibrium
Program for Four Corners Samples
Input Data for Four Corners Scrubbing
Simulation
Four Corners Scrubbing Simulation Results
for Design Conditions
Four Corners Scrubbing Simulation Results
with 30% Solid Waste Operation
Four Corners Scrubbing Simulation Results
for Alternative One
Four Corners Scrubbing Simulation Results
for Alternative Two
E-34
E-36
E-37
E-38
Ei -i
-41
E-42
E-42
E-43
E-43
F-2
F-8
F-9
F-16
F-18
F-19
F-26
F-30
XXIV
-------
TABLES
(Continued)
PAGE
Table 3-3. Four Corners Scrubbing Simulation Results
for Alternative Three F-34
Table 3-4. Four Corners Scrubbing Simulation Results
for Alternative Four F-37
Table 3-5. Four Corners Water Management Simulations p.39
Summary
Table 4-1. Capital Costs for Water Recycle/Reuse
Alternatives at Four Corners F-42
Table 4-2. Operating Costs for Water Recycle/Reuse
Alternatives at Four Corners F-44
APPENDIX G - RECYCLE/REUSE OPTIONS AT BOWEN
(GEORGIA POWER CO.)
Table 1-1. Summary of Technical Feasible Options
at Bowen G-2
Table 2-1. Parameters Calculated by Equilibrium
Program for Bowen Samples G-8
Table 2-2. Input Data for Bowen Cooling Tower
Simulations G-13
Table 2-3. Bowen Existing Cooling Tower Operations -- G-15
Table 2-4. Bowen Existing Ash Sluicing Input Data G-19
Table 2-5. Bowen Existing Ash Sluicing Operations G-21
Table 3-1. Effect of Increased Cycles of
Concentration in Bowen Cooling Towers G-24
Table 3-2. Relative Saturations of Scale-Forming
Species at 15 Cycles in Bowen Cooling
Towers G-26
Table 3-3. Effects of Makeup Water Calcium
Concentration on Bowen Cooling Tower
Operation G-28
Table 3-4. Bowen Alternative Ash Sluicing Input Data G-35
Table 3-5. Bowen Once-Through Ash Sluicing at 10%
Solids G-36
Table 3-6. Bowen Recirculating Ash Sluicing G-38
xxv
-------
TABLES
(Continued)
Table 3-7.
Table 4-1.
Table 4-2.
Table 4-3.
Effects of CO2 Transfer on Bowen Ash
Sluicing Operations
Capital Costs for Water Recycle/Reuse
Alternatives at Bowen
Operating Costs for Water Recycle/Reuse
Alternatives at Bowen
Capital and Operating Costs for
Eliminating Ash Pond Overflow at Bowen
PAGE
G-41
G-48
G-49
G-51
APPENDIX H
Table 1-1.
Table 2-1.
Table 2-2.
Table 2-3.
Table 3-1.
Table 3-2.
Table 3-3.
Table 3-4.
Table 3-5.
Table 3-6.
Table 3-7.
Table 3-8.
- RECYCLE/REUSE OPTIONS AT COMANCHE
(PUBLIC SERVICE OF COLORADO)
Summary of Water Recycle/Reuse Options
at Comanche
Parameters Calculated by Equilibrium
Program for Comanche Samples
Input Data for Comanche Cooling Tower
Simulations
Comanche Existing Cooling Tower Operations
Effects of Increased Cycles in Comanche
Cooling Towers
Relative Saturations of Phosphate and
Silica Solids in Comanche Cooling Towers
Effects of Makeup Water Sulfate
Concentration on Comanche Cooling Tower
Operation -------------------------------
Comanche Once-Through Ash Sluicing
Input Data ------------------------------
Comanche Recirculating Ash Sluice Input
Comanche Once-Through Ash Sluicing
Simulation Results
Comanche Recirculating Ash Sluicing
Simulation Results
Comanche Fly Ash Sluice Makeup Water
Treatment Effects
H-2
H-9
H-14
H-16
H-22
H-24
H-26
H-30
H-31
H-32
H-34
H-36
xx vi
-------
TABLES
(Continued)
PAGE
Table 3-9. Effects of C02 Mass Transfer in Comanche
Ash Sluicing H-38
Table 4-1. Capital Costs for Water Recycle/Reuse
Alternatives at Comanche H-43
Table 4-2. Capital and Operating Costs for Attaining
Zero Discharge at Comanche H-44
APPENDIX I - RECYCLE/REUSE OPTIONS AT MONTOUR
(PENNSYLVANIA POWER & LIGHT CO.)
Table 1-1. Summary of Technically Feasible Options
at Montour 1-3
Table 2-1. Parameters Calculated by the Equilibrium
Program 1-8
Table 2-2. Input Data for Montour Cooling Tower
Simulations 1-12
Table 2-3. Existing Cooling Tower Simulation Results 1-14
Table 2-4. Montour Existing Ash Sluicing Input Data 1-18
Table 2-5. Montour Existing Ash Sluicing Operations 1-19
Table 3-1. Adjusted Water Makeup Compositions for
Increased Magnesium Levels 1-23
Table 3-2. Simulation Results for Increased Cycles
of Concentration 1-24
Table 3-3. Relative Saturation of Scale-Forming
Species at 20 Cycles of Concentration
in Montour Cooling Towers 1-27
Table 3-4. Simulation Results With Different
Magnesium Concentration (14 Cycles) 1-28
Table 3-5. Recirculating Ash Sluicing Input Data 1-31
Table 3-6. Recirculating Ash Sluicing Results
(C02 Equilibrium in the Pond) 1-32
Table 3-7. The Effect of C02 Transfer in
Recirculating Ash Sluicing Systems
At Montour 1-35
xxvii
-------
Table 4-1.
TABLES
(Continued)
Capital Costs for Water Recycle/Reuse
Alternatives at Montour
Table 4-2. Operating Costs
PAGE
1-42
1-44
APPENDIX J
Table 1-1.
Table 2-1.
Table 2-2.
Table 2-3.
Table 2-4.
Table 2-5.
Table 2-6.
Table 3-1.
Table 3-2.
Table 3-3.
Table 3-4.
Table 3-5.
Table 3-6.
Table 3-7.
- RECYCLE/REUSE OPTIONS AT COLSTRIP
(MONTANA POWER CO.)
Summary of Water Recycle/Reuse Options
at Colstrip
Parameters Calculated by Equilibrium
Program for Colstrip Samples
Inptit Data for Colstrip Cooling Tower
Simulations
Colstrip Existing Cooling Tower
Operations Simulation Results -•
Input Data for Colstrip Scrubbing
S imulat ion
Sample Consistency Errors Around Effluent
Tank at Colstrip -^
Colstrip Scrubbing Simulation Results
for Design Conditions
Treatment Alternatives for Colstrip
Cooling Tower Operation
Simulation Results for Calcium Variations
in the Makeup Water at Colstrip
Simulation Results for Sulfate Variations
in the Makeup Water at Colstrip
Relative Saturations of Scale-Forming
Species for 20 Cycles With Existing
Makeup Water at Colstrip
Effect of Flue Gas Ash Content on
Colstrip Scrubber Operation
Effect of Slurry Solids Content on
Colstrip Scrubber Operation
Effects of Makeup Water Composition on
Colstrip Scrubber Operation
J-3
J-9
J-14
J-15
J-19
J-21
J-23
J-27
J-30
J-33
J-34
J-36
J-37
J-39
xxviii
-------
TABLES
(Continued)
PAGE
Table 4-1. Capital Costs for Water Recycle/Reuse
Alternatives at Colstrip J-43
Table 4-2. Operating Costs for Water Recycle/Reuse
Alternatives at Colstrip J-44
APPENDIX K - POWER PLANT DATA REDUCTION
Table 4-1. Averages for December 1975 and
August 1976 K-19
Table 4-2. Cooling Tower Operation Conditions K-20
Table 4-3. Ash Leaching Results K-30
Table 5-1. Average Climatological Data for
Billings, Montana K-34
Table 5-2. Coal and Flue Gas Compositions K-36
Table 5-3. Ash Leaching Results K-39
Table 5-4. Comparison of Calculated, Sample, and
MPC Fly Ash Reactivity K-40
APPENDIX L - ASH CHARACTERIZATION FOR FOUR CORNERS,
BOWEN, AND COMANCHE FLY ASHES
Table 2-1. C02 Sorption - Nonbuffered pH = 11 L-7
Table 2-2. C02 Sorption - Buffered pH = 11 L-9
Table 2-3. C02 Sorption - Buffered pH = 9 L-10
Table 4-1. Results of Chemical Analysis from Leaching
of Ash Samples at Constant pH for Comanche
Steam-Electric Station L-35
Table 4-2. Results of Chemical Analysis from Leaching
of Ash Samples at Constant pH for Plant
Bowen L-35
Table 4-3. Results of Chemical Analysis from Leaching
of Ash Samples at Constant pH for Four
Corners Power Station L-36
Table 4-4. Leachable Species from Ash Samples at pH 6 L-36
Table 4-5. Ash Dissolution Characterizations - Batch
Dissolution L-37
xxix
-------
APPENDIX A
POWER PLANT SELECTION CRITERIA
1.0 INTRODUCTION
In order to meet the national goal of "zero-discharge"
of pollutants to the environment, steam-electric power plants
must consider all possibilities for recycle and reuse of waste-
water streams. This situation is enhanced by limited availabil-
ity and rising costs of water along with water treatment require-
ments .
This project involved studying the water streams and
possible recycle/reuse alternatives at five typical steam-electric
power plants in the United States. This appendix describes the
selection of the typical plants studied.
First, the criteria used in selecting the plant sites
are discussed. These include geographical location, power plant
cooperation and data availability, site characteristics, and
project timing. Next, the selection of the plants in each area
is discussed. Steam-electric power plants in each geographical
region are identified and discussed with respect to the criteria
outlined above. General data for the final plant sites selec-
ted for study are then assembled and inspected to insure that
a general spectrum of cooling systems, ash handling systems,
and pollution control equipment is included.
A-l
-------
2.0 SELECTION CRITERIA
Representative site selection necessitates establish-
ing criteria pertinent to the overall objectives of the pro-
gram. The four main criteria selected for screening the steam-
electric power plants are:
1) location,
2) availability,
3) site characteristics, and
4) timing.
The first criterion, location, represents the geo-
graphical area in which the plant is located. Four geographical
regions were initially defined: Southwest, Northern Great
Plains, Northeast, and Southeast. These four areas were chosen
to represent regions in the United States where water recycle/
reuse is advantageous due to high water costs, limited water
availability, or wastewater treatment and disposal problems.
By selecting plants located in these geographical areas of the
United States, different types of cooling modes will be exam-
ined. The different climatological conditions associated with
each geographical region determine to some extent the type of
cooling and/or waste disposal methods. For example, cooling
ponds and evaporation ponds are most suitable for hot, dry
climates such as found in the Southwest than the more humid
climates of the Southeast.
The second criterion listed is availability. This
involves both data accessibility and plant cooperation. Suffi-
cient plant water data will be necessary to confirm information
gathered from sampling. This will insure an accurate charac-
terization of each plant's water system and serve as confirma-
tion of the validity of the process simulation model.
Power plant cooperation is an extremely important
factor in selecting typical sites for study. A good working
relationship is essential to the successful completion of the
project. Also, the possible implementation of a demonstration
program following the existing work will depend upon the plant's
cooperation.
A-2
-------
Site characteristics, the third criterion, consist of
the plant cooling, ash handling, and pollution control systems.
Three types of cooling systems may be employed by a power
plant: once-through cooling, cooling towers, or cooling ponds.
Plants with cooling towers or cooling ponds are preferred
since these types of cooling systems are good candidates for
recycle/reuse options whereas once-through systems are not.
Wet ash sluicing is similarly preferred over dry handling for
the purposes of this study.
The pollution control systems of the plants studied
are also taken into account in selecting the plants. The type
of particulate control utilized may play a significant role in
the water management scheme of the plant. For instance,
make-up water for particulate or S02 scrubbers may be provided
by wastewater streams which are normally discharged.
The last criterion, timing, concerns the status of
studies or plant modifications which are planned or underway.
Plants which are conducting studies involving the water system
may be able to supply more accurate data which is pertinent to
this study. Plants which are planning modifications such as
the installation of S02 scrubbers may have additional recycle/
reuse options that can be studied.
The aforementioned criteria were used to examine the
potential plants identified in each geographical area. The
following section contains a discussion of these plants and
presents the data collected in tabular form.
A-3
-------
3.0 SELECTION METHODOLOGY
This section describes the methodology used to select
the plants to be studied in the water management program.
First, potential power plants are identified in each region
based on data concerning the plants' cooling, ash handling,
particulate control and sulfur dioxide control systems._ Then
the potential plants identified are screened on the basis of
the selection criteria discussed in the previous section. The
final selection of the plants to be studied involved contacting
each of the utilities to determine their interest in the pro-
gram. Five plants (Arizona Public Service's Four Corners
Plant, Public Service of Colorado's Comanche Plant, Georgia
Power's Bowen Plant, Pennsylvania Power and Light's Montour
Plant, and Montana Power's Colstrip Plant) were selected for
the study.
3.1 Plant Identification
Identification of potential power plants in each
geographical region was accomplished by assembling published
data concerning coal-fired steam-electric power plants for the
Southwest, Northern Great Plains, Southeast, and the Northeast
regions of the United States. This section presents and
describes the data collected for the potential plants identi-
fied in each geographical area. The data concerning location,
capacity, and the types of cooling, ash handling, particulate
control and sulfur dioxide control systems is presented in
tabular form for each region.
3.1.1 Southwest
The potential power plants for the water management
program in the Southwest are presented in Table 3-1. Two of
the seven plants shown were included in a list of recommended
study sites for the project that was sent to EPA-NERL (Thermal
Pollution Branch) from the EPA Office of Energy Activities in
Denver, Colorado. These two plants are Arizona Public Ser-
vice's Four Corners Plant and Colorado/UTE's Hayden Plant. One
additional plant which was recommended by the EPA Office of
Energy Activities but is not included as a potential plant for
this study is Public Service of Colorado's Cherokee Plant.
This plant was not included in the potential candidate list
since a similar study has previously been completed for Chero-
kee. However, Public Service of Colorado's Comanche Plant is
included in the potential plant site list.
A-4
-------
TABLE 3-1. POTENTIAL SOUTHWESTERN POWER PLANTS
Utility
Salt River Project
Colorado/UTE
Arizona Public
Service
Arizona Public
Service
Nevada Power
Public Service of
Colorado
Southern California
Edison
Plant
Navaj o
Hayden
Four
Corners
Cholla
Gardner
Comanche
Mohave
Location
Page,
Arizona
Hayden ,
Colorado
Fannington,
New Mexico
Joseph
City,
Arizona
Moapa,
Nevada
Pueblo,
Colorado
Laugh lin,
Nevada
Type
Capacity, M«r Cooling
2 , 250 WCT
163 WCT
1,600 CP
114 CP
227 *
350 WCT
1,580 WCT
Ash
Handling2
*
WSB
WSB
WSF
WSB
WSF
*
WSB
WSB
DDF
Part.
Control3
*
ESP,
cyclones
ESP,
venturi
cyclones
*
ESP,
DDF
ESP
S02
Control1*
Planned
None
UC
Limestone
scrubbing
Sodium
carbonate
scrubbing
None
Planned
*Data not found before Arizona Public Service and Public Service of Colorado indicated an interest
in the program.
1WCT = wet cooling tower, CP = cooling pond
2WSB = wet sluicing of bottom ash, WSF = wet sluicing of fly ash, DDF = dry disposal of fly ash
3ESP = electrostatic precipitator
"*UC = under construction
Sources: NA-205, LE-201, PE-161, EL-094, FE-102, DE-165
-------
Of the seven plants listed in Table 3-1, four use wet
cooling towers and two use cooling ponds. No data concerning
the type of cooling system used at Nevada Power's Gardner Plant
was located before interested plants were identified. Data con-
cerning ash handling was found for five of the seven plants (no
data found for Navajo or Gardner). All five of these plants
used wet sluicing for bottom ash disposal but only two used wet
sluicing for fly ash disposal. Fly ash is disposed of in dry
form at Mohave and Comanche. No data concerning fly ash dis-
posal was found for Hayden.
The type of particulate control utilized at the plants
in Table 3-1 is in general cyclones and/or electrostatic precip-
itators with the exception of Four Corners where venturi scrub-
bers are used on three of the five generating units. No data
was located for the Navajo or Gardner Plants concerning particu-
late control.
The only plants with existing sulfur dioxide control
are Cholla and Gardner. Cholla employs limestone wet scrubbing
whereas Gardner uses sodium carbonate scrubbing. Some SO2 re-
moval is observed at Four Corners in the venturi particulate
scrubbers (lime is added to the scrubbing system).
3.1.2 Northern Great Plains
The potential power plants for inclusion in the water
management study in the Northern Great Plains area are listed
in Table 3-2. Three of these plants were included in the recom-
mended plant site list sent to EPA-NERL by the EPA Office of
Energy Activities in Denver, Colorado. These three plants are
Pacific Power and Light's Johnston Plant, Basin Electric's
Leland Olds Plant, and Utah Power and Light's Naughton Plant.
Only one of the plants listed, Minnkota Power Coop's
Young Plant, uses cooling ponds for cooling the condenser recir-
culating water. Two of the plants use once-through cooling
exclusively (Leland Olds and Corette) and three others (Bridger,
Colstrip and Naughton) utilize wet cooling towers exclusively.
Pacific Power and Light's Johnston Plant uses both once-through
cooling and wet cooling towers. Cooling towers are used only
for one of the four units at the plant (Unit #4). The other
three units employ once-through cooling exclusively.
A-6
-------
TABLE 3-2. POTENTIAL NORTHERN GREAT PLAINS POWER PLANTS
Utility
Basin Electric
Pacific Power &
Light
Pacific Power &
Light
Utah Power & Light
Montana Power Co.
Montana Power Co.
Minnkota Power
Coop
Plant Location
Leland Stanton,
Olds N. Dakota
Bridger Rock
Springs ,
Wyoming
Johnston Glenrock,
Wyoming
Naughton Kemmerer,
Wyoming
Corette Billings ,
Montana
Colstrip Colstrip,
Montana
Young Center ,
N. Dakota
Capacity, Type Ash
Mw Cooling1 Handling2
216 OTF WSB
500 WCT WSB
750 OTF, WCT WSB, WSF
707 -WCT WSB
173 OTF WSB
700 WCT WSB (Recir-
culating) ,
WSF
250 CP WSB
Part.
Control3
cyclones
ESP
cyclones ,
venturi
cyclones,
ESP
ESP
venturi
ESP
S02
Control"*
None
None
None
None
None
Lime /alkaline
fly ash
scrubbing
UC
JOTF = once-through fresh water
WCT = wet cooling tower
CP = cooling pond
t\
WSB = wet sluicing of bottom ash, WSF = wet sluicing of fly ash
3ESP = electrostatic precipitator
**UC = under construction
Sources: NA-205, PE-161, FE-102, EL-094
-------
All of the plants listed in Table 3-2 use wet sluicing
to dispose of the bottom ash from the boiler. However, no infor-
mation concerning fly ash disposal was found for the plants with
the exception of Johnston and Colstrip. At Johnston the fly ash
from Unit 4 is disposed of in a slurry form (collected in the
venturi scrubber). At Colstrip, fly ash from both units is^col-
lected by venturi scrubbers and used as a source of alkalinity
for removing sulfur dioxide in spray scrubbers. The final dis-
posal product is a mixture of ash and scrubber solids.
Four of the plants utilize electrostatic precipitators
for particulate control, one of which combines cyclones and
electrostatic precipitators. The Leland Olds Plants uses only
cyclones. The Johnston Plant has cyclones on three of the four
units and a venturi scrubber on the remaining unit for particu-
late control.
One of the potential power plants has sulfur dioxide
control planned. A lime scrubbing system is under construction
at Minnkota Power Coop's Young Plant. Colstrip has existing
S02 scrubbing on both 350 Mw units (three scrubbing trains per
unit).
3.1.3 Southeast
Table 3-3 presents the potential power plants for the
water management study for the Southeast. Three of the four
plants shown are part of the TVA power generation system (Kings-
ton, Colbert, and Paradise). Kingston and Colbert use fresh
water on a once-through basis for cooling. The Paradise Plant
uses both once-through cooling and wet cooling towers with the
provision of operating Units 1 and 2 with either once-through
cooling or with the cooling towers. Unit 3 uses the cooling
towers exclusively. Georgia Power's Bowen Plant employs wet
cooling towers.
All of the plants shown in Table 3-3 use wet sluicing
for disposing of bottom ash and fly ash. All of the plants use
electrostatic precipitators for fly ash collection and none of
the plants employ sulfur dioxide control.
3.1.4 Northeast
The potential power plants located in the Northeast
are shown in Table 3-4. Four of the six plants listed use
A-8
-------
TABLE 3-3. POTENTIAL SOUTHEASTERN POWER PLANTS
Utility Plant
Tennessee Valley Kingston
Authority
Tennessee Valley Colbert
Authority
Tennessee Valley Paradise
Authority
Georgia Power Co. Bowen
>
Location
Kingston,
Tennessee
Pride,
Alabama
Paradise ,
Kentucky
Taylors-
ville,
Georgia
Capacity,
Mw
1,700
1,400
2,558
1,595"
Type
Cooling
OTF
OTF
WCT, OTF
WCT
Ash
Handling2
WSB
WSF
WSB
WSF
WSB
WSF
WSB
WSF
Part.
Control3
ESP
cyclones ,
ESP
ESP
ESP
S02
Control
None
None
None
None
WCT - wet cooling tower
2WSB = wet sluicing of bottom ash, WSF = wet sluicing of fly ash
3ESP - electrostatic precipitator
"Plant Capacity as reported in FPC Form 67 Data for 1972; present capacity is 3200 Mw (4 units)
Sources: PE-161, FE-102, EL-094, NA-205
-------
TABLE 3-4. POTENTIAL NORTHEASTERN POWER PLANTS
Utility
Niagara- Mohawk
Pennsylvania Power
& Light
Pennsylvania Power
& Light
Pennsylvania
Electric
Duquesne
Potomac Electric
Plant Location
Dunkirk Dunkirk,
New York
Sunbury Shamokin
Dam,
Pennsylvania
Montour Washington-
ville,
Pennsylvania
Homer City Homer City ,
Pennsylvania
Phillips South
Heights ,
Pennsylvania
Dickerson Dickerson,
Maryland
Capacity,
Mw
628
410
1,500
1,269
410
587
Type
Cooling1
OTF
OTF
WCT
WCT
OTF
OTF
Ash
Handling2
WSB
WSB
WSF
WSB
WSB
WSB
*
Part.
Control3
cyclones,
ESP
baghouses
ESP
ESP
cyclones,
ESP
ESP
S02
Control
None
None
None
None
Lime
scrubbing
Magnesium
oxide
scrubbing
*Data not found before the plant was eliminated from consideration.
JOTF = once-through fresh water
2WCT = wet cooling tower
2WSB = wet sluicing of bottom ash, WSF = wet sluicing of fly ash
3ESP = electrostatic precipitator
Sources: PE-161, FE-102, EL-094, NA-205
-------
once-through cooling. The other two plants, Pennsylvania Elec-
tric 's Homer City Plant and Pennsylvania Power & Light's Montour
Plant, use wet cooling towers. The Dunkirk, Sunbury, Homer City,
Montour, and Phillips Plants all use wet sluicing for bottom ash
disposal. No data was found concerning bottom ash disposal for
Dickerson, or fly ash disposal for any of the plants except Mon-
tour, which sluices fly ash on a once-through basis.
Particulate collection is achieved by cyclones and
electrostatic precipitators for the Dunkirk and Phillips Plants,
electrostatic precipitators for the Homer City, Montour, and
Dickerson Plants, and baghouses for the Sunbury Plant. Only two
of the six plants have sulfur dioxide control equipment.
Duquesne's Phillips Plant uses a lime scrubbing system which
started up in 1973 (PE-161). Potomac Electric's Dickerson Plant
employs a magnesium oxide scrubbing process which was also
started up in 1973 (PE-161).
3.2 Plant Selection
Ten utilities from the geographical regions discussed
in the previous section were contacted to determine their inter-
est in the water management program. These utilities are:
1) Salt River Project (Navajo Plant)
2) Arizona Public Service (Four Corners
and Cholla Plants)
3) Public Service of Colorado (Comanche
Plant)
4) Pacific Power and Light (Johnston
Plant)
5) Utah Power and Light (Naughton Plant)
6) Montana Power Co. (Colstrip Plant)
7) Tennessee Valley Authority (Paradise
Plant)
8) Georgia Power (Bowen Plant)
9) Duquesne (Phillips Plant)
10) Pennsylvania Power and Light (Montour
Plant)
A-ll
-------
Of these ten utilities, Arizona Public Service, Public
Service of Colorado, Georgia Power, Montana Power, and Pennsyl-
vania Power and Light expressed interest in the program. Arizona
Public Service wanted to include their Four Corners Plant as op-
posed to the Cholla Plant since they have been experiencing
scaling problems in the venturi particulate scrubbing system
at Four Corners.
These five plants (Four Corners, Comanche, Bowen, Mon-
tour, and Colstrip) will provide a general spectrum of cooling,
ash handling, and pollution control systems for the water manage-
ment study. Table 3-5 summarizes the data collected for the
plants. All of the plants use recirculating cooling systems as
opposed to once-through systems. Four of the plants employ wet
cooling towers and one (Four Corners) utilizes a cooling pond.
These types of cooling systems are more conducive to recycle/
reuse alternatives than once-through systems and thus are desir-
able with respect to the overall objectives of the program.
Four of the plants (Four Corners, Bowen, Montour, and
Colstrip) use wet sluicing for fly ash disposal and all of the
plants employ wet sluicing for bottom ash disposal. These wet
disposal operations will provide greater water recycle/reuse
potential for the plants than dry disposal. For example, blow-
down water streams that normally are discharged may be used in
an ash sluicing system and thus reduce or eliminate the need for
fresh water makeup.
All of the plants except Colstrip employ electrostatic
precipitation for particulate control. The Four Corners Plant
has venturi particulate scrubbers on three of the five generating
units and Colstrip has combined particulate and SOa scrubbing on
both units. The modeling studies associated with the scrubbers
will identify water recycle/reuse alternatives at plants using
wet scrubbing for particulate and SO2 control.
Only one of the selected plants has sulfur dioxide
control equipment planned. Construction of a lime scrubbing
unit is underway at Four Corners. Possibilities for recycle/
reuse of normally discarded plant water streams in a sulfur
dioxide wet scrubbing system may be studied at Four Corners as
well as at Colstrip.
A-12
-------
TABLE 3-5. SELECTED PLANTS FOR WATER RECYCLE/REUSE STUDY
Utility
Arizona Public
Service
Public Service of
Colorado
Georgia Power Co.
Pennsylvania Power
and Light
Montana Power Co.
Plant
Four
Corners
Comanche
Bowen
Montour
Colstrip
Location
Farmington,
New Mexico
Pueblo,
Colorado
Taylorsville,
Georgia
Washington-
ville,
Pennsylvania
Colstrip,
Montana
Capacity,
Mw
2,150
700
3,200
1,500
700
Type
Cooling
CP
WCT
WCT
WCT
WCT
Ash
Handling2
WSB
WSF
WSB
WSB
WSF
WSF
WSB
WSB
WSF
Part.
Control3
cyclones ,
ven£uri
ESP
ESP
ESP
venturi
S02
Control1*
UC
None
None
None
Lime/ alkaline
fly ash
scrubbing
WCT = wet cooling tower, CP = cooling pond
2WSB = wet sluicing of bottom ash, WSF = wet sluicing of fly ash
3ESP = electrostatic precipitator
"*UC =» under construction
-------
Appendix B. Chemical Characterization of Plant Water Systems
1.0 INTRODUCTION
In order to perform technical and economic evaluations
of water recycle/reuse and treatment options for coal-fired
power plants, complete characterization of the water of the five
generating stations studied was required. Information from the
following sources was used:
1) operating and analytical data available
from power plants,
2) data from spot sampling and chemical
analyses performed, and
3) data on the chemical reactivity of the
ash from each plant.
This appendix presents the data acquired from chemical
analyses performed on water samples taken at the selected plants.
This information corresponds to data source (2) above. The data
will be used in conjunction with specific power plant operating
and analytical data to:
1) identify quantitatively, existing water
management problems of the specific
plants,
2) identify potential problem areas,
3) establish the reliability of the
simulation model, and
4) be used as inputs for the various
recycle/reuse and treatment options
studies.
The presentation of the collected data is divided into
four sections:
1) sampling,
2) analytical techniques,
3) results, and
4) discussion of results.
B-l
-------
2.0 SAMPLING
Five diverse generating stations were sampled to
acquire additional information necessary for the characteriza-
tion of each of their water systems. These plants are:
1) Four Corners Generating Station of Arizona
Public Service,
2) Comanche Generating Station of Public
Service of Colorado,
3) Bowen Generating Station of Georgia
Power Co.,
4) Montour Generating Station of Pennsylvania
Power and Light, and
5) Colstrip Generating Station of Montana
Power Co.
These five plants provide a general spectrum of
cooling, ash handling and pollution control systems for the
water management study. Table 2-1 summarizes the major systems
at the plants. All of the plants use recirculating cooling
systems as opposed to once-through systems. Four of the plants
employ wet cooling towers and one (Four Corners) utilizes a
cooling pond. Four of the plants (Four Corners, Bowen, Montour,
and Colstrip) use wet sluicing for fly ash disposal and all of
the plants employ wet sluicing for bottom ash disposal. All of
the plants except Colstrip have electrostatic precipitation for
particulate control. The Four Corners Plant has venturi partic-
ulate scrubbers on three of the five generating units. Colstrip
has venturi particulate scrubbers as part of a combined SCW
particulate removal system on both units.
This section will break down the sampling by plant and
will describe:
1) the location of the water streams
sampled,
2) the type of samples collected at each
point, and
3) the sampling methods employed.
B-2
-------
TABLE 2-1. SELECTED PLANTS FOR WATER RECYCLE/REUSE STUDY
Utility
Arizona Public
Service
Public Service of
Colorado
Georgia Power Co.
W
i
*"° Pennsylvania Power
and Light Co.
Montana Power Co.
Capacity , Type
Plant Location MW Cooling1
Four Farmington, 2,150 CP
Corners New Mexico
Comanche Pueblo, 700 WCT
Colorado
Bowen Taylorsville , 3,200 WCT
Georgia
Montour Washington- 1,500 WCT
ville,
Pennsylvania
Colstrip Colstrip, 700 WCT
Montana
Ash Part . S02
Handling2 Control3 Control"*
WSB
WSF
WSB
WSB
WSF
WSF
WSB
WSB
WSF
ESP , UC
venturi
ESP None
ESP None
ESP None
venturi Lime /alkaline
fly ash
scrubbing
1WCT = wet cooling tower, CP = cooling pond
2WSB = wet sluicing of bottom ash, WSF = wet sluicing of fly ash
3ESP = electrostatic precipitator
''UC = under construction
Sources: NA-205, DE-165, PE-161, EL-094, FE-102
-------
2.1 Four Corners Generating Station (Arizona
Public Service)"~
The Arizona Public Service Four Corners Plant is a
2150 Mw coal-fired station located near Farmington, New Mexico.
Four Corners uses a cooling pond and bottom ash wet sluicing for
all units, particulate wet scrubbing for Units 1-3, and electro-
static precipitators for Units 4 and 5 with dry fly ash dis-
posal.
Makeup water for the plant is taken from the San Juan
River and stored in Morgan Lake, which serves as the source for
all water used in the system. A periodic blowdown is taken from
Morgan Lake to control the dissolved solids concentration and
discharged to the Chaco River. Cooling water, bottom ash sluice
water, boiler makeup water, and makeup water for the particulate
scrubbing system are taken from Morgan Lake.
The particulate scrubbing system consists of six
venturi scrubbers (two each for Units 1-3), two thickeners, two
thickener transfer tanks for return of thickener overflow and a
sluice tank for combining thickener underflows and the scrubber
loop bleed stream. The flue gas is contacted with the scrubbing
liquor in the venturi throat and upon leaving passes through a
disengagement zone where the liquid is separated from the gas
and falls into a reservoir at the bottom of the scrubber.
The gas passes through a demister and is then vented to
the stack. The liquor collected in the reservoir is recycled
for further gas-liquid contact after a bleed stream is removed.
The major portion of this bleed stream is routed to a thickener
and the remainder is used to reslurry the thickener underflow in
the sluice tank before pumping it to the ash pond.
A total of eight aqueous samples were taken to charac-
terize the major water systems at Four Corners and are listed
below:
1) plant makeup water from Morgan Lake,
2) effluent liquor from Venturi Scrubbers 1A
and 3A,
3) thickener overflow,
B-4
-------
4) thickener underflow,
5) bottom ash sluice water,
6) sluice tank effluent,
7) ash pond effluent, and
8) ash pond surface water.
Temperature and pH at each sample point were taken on
location. A two-liter grab sample was filtered and acidified
with nitric acid for subsequent chemical analysis to determine
the concentrations of calcium, magnesium, sodium, potassium,
chloride, total sulfur, phosphate, and silicate. Portions of
the filtered liquor were taken prior to acidification for use in
determining carbonate concentration and, in the case of the
scrubber liquor and thickener overflow, the aqueous sulfite con-
centration. Another grab sample was taken at each point to de-
termine the nitrate concentration, weight percent solids, and
total dissolved solids. Representative solid samples of the fly
ash and lime were also taken.
2.2 Comanche Generating Station (Public Service of
Colorado
The Public Service of Colorado Comanche generating
station is a coal-fired system composed of two units, each
having 350 Mw capacity, and is located near Pueblo, Colorado.
Comanche uses wet cooling towers with the blowdown used for
boiler refractory cooling and once-through bottom ash wet sluic-
ing. Hot-side electrostatic precipitators and subsequent dry
disposal are employed for fly ash handling.
The water entering the plant is first taken from the
Arkansas River and stored in a reservoir. From here a small
portion of the raw water is sent to the coal handling facilities
for dust suppression. Another portion is sent to the ash removal
system to sluice bottom ash. The remainder of the raw water
leaving the reservoir is sent to the Comanche lime treatment
facility. The calcium carbonate sludge produced is sent to a
special pond which is kept separate from the bottom ash ponds.
The softened water is used for service water and for makeup water
to the two cooling systems.
The water effluent from the overall operation comes
from the overflow from the final polishing pond which is fed by
B-5
-------
the two boiler blowdown streams, the lime sludge disposal pond
overflow, and the two bottom ash disposal pond overflows. The
final polishing pond effluent is sent to the St. Charles River.
The remaining system water losses are cooling tower evaporation
and drift and other evaporative losses. Characterization of the
Comanche water system was accomplished by sampling the following
streams:
1) cooling tower makeup,
2) cooling tower blowdown,
3) bottom ash sluice,
4) ash pond inlet,
5) ash pond subsurface,
6) ash pond effluent, and
7) polishing pond effluent.
The temperature and pH of each sample were recorded
on site. A two-liter grab sample was filtered and acidified
with nitric acid for subsequent chemical analysis. Portions of
the filtered liquors were taken prior to acidification to deter-
mine carbonate concentrations. Another grab sample was taken to
determine the nitrate, suspended solids, and total dissolved
solids concentrations. A representative solid sample of the
fly ash was also taken.
2.3 Bowen Generating Station (Georgia Power Co.)
The Georgia Power Co. Bowen Station is a 3180 Mw coal-
fired plant located near Taylorsville, Georgia. Bowen employs
wet, natural draft cooling towers and once-through bottom and
fly ash wet sluicing for all four units.
Makeup water for the plant is taken from the Etowah
River and stored in a makeup pond and then used as general
service water, boiler makeup, and cooling tower makeup. The
general service water effluent is split so that about 5% of the
flow returns to the makeup pond and 95% is used as cooling tower
makeup. A portion of the cooling tower blowdown is used to
sluice bottom ash and fly ash to the ash pond. The excess cool-
ing tower blowdown is discharged as is the ash pond overflow.
B-6
-------
The following streams were sampled to characterize the Bowen
water system:
1) cooling tower makeup,
2) cooling tower blowdown, Unit #3,
3) bottom ash sluice,
4) fly ash sluice,
5) ash pond subsurface,
6) ash pond effluent, and
7) plant drainoff.
The temperature and pH of each sample were recorded on
site. Both filtered and acidified and unfiltered, nonacidified
samples were taken for analysis. A representative fly ash sample
was also taken.
2.4 Montour SES (Penn. Power & Light)
The Pennsylvania Power and Light Co. Montour Steam-
Electric Station is a coal-fired plant with two 750 Mw units
located in Washingtonville, Pa. Montour utilizes wet natural
draft cooling towers and once-through sluicing of both bottom
ash and fly ash with cooling tower blowdown.
Water enters the plant through a raw water reservoir
which is fed by the Susquehanna River. Water taken directly
from the reservoir is used for cooling tower makeup, boiler
makeup, and for general service water. Cooling tower blowdown
is used for fly ash, mill rejects, and bottom ash sluicing.
The fly ash, mill rejects, and bottom ash are all sluiced to the
ash pond which is divided into two large sections.
The ash pond overflow is treated with sulfuric acid
for pH control as it enters the detention pond. Other streams
flowing into the detention pond include the coal pile runoff,
and all miscellaneous plant waste streams. The detention pond
overflows into the Chillisquaque Creek. Water losses at the
plant occur through cooling tower evaporation and drift, evap-
orative losses from the ponds, and boiler losses.
B-V
-------
A total of ten aqueous samples were taken at the plant
and are listed below:
1) ash basin at separating dike,
2) cooling tower makeup,
3; cooling tower blowdown, Unit #1,
4) cooling tower blowdown, Unit #2,
5) detention basin overflow,
6) mill reject slurry,
7; fly ash slurry, Unit #2,
8) ash basin overflow,
9) bottom ash slurry, Unit #1, and
10) miscellaneous wastes.
Temperature and pH of each of the ten samples were
taken on location. A two-liter grab sample was taken. One
liter of this sample was filtered and acidified with nitric acid
for subsequent chemical analysis to determine the concentrations
of calcium, magnesium, sodium, potassium, chloride, total sul-
fur, phosphate, and silica. The cooling tower blowdown and fly
ash slurry grab samples were also analyzed for arsenic. A por-
tion of the filtered liquor was taken prior to acidification for
use in determining carbonate concentration in each of the ten
samples. The remaining liter of sample was used to determine the
nitrate concentration, weight percent solids, and total dis-
solved solids. A solid sample of the fly ash was taken by PP&L
personnel. The analytical techniques used for these samples are
explained in Section 3.0, and the results are presented in
Section 4.0.
2.5 Colstrip SES (Montana Power Co.)
The Montana Power Company Colstrip Steam-Electric
Station is a coal-fired plant with two 350 Mw units located in
Colstrip, Montana. The Colstrip Plant employs wet forced-draft
cooling towers and lime/alkaline fly ash scrubbing for S02 and
fly ash removal. A recirculating bottom ash sluicing system is
also used.
B-8
-------
Water from the Yellowstone River is stored in a surge
pond from which all plant water is withdrawn. Water taken from
the surge pond is processed through a lime softening system.
Softened water is used for cooling tower and scrubber makeup
water. Cooling tower blowdown is piped to two 12.6 £/sec
(200 GPM) capacity brine concentrators. A portion of the dis-
tillate provides the demineralizer feed. The remainder of the
distillate is used as scrubber makeup. The concentrated waste
stream produced is disposed of in two one-acre lined ponds.
There are three identical scrubbing trains on each of
the two generating units. The scrubbing system makeup water is
added along with lime to the recycle tank in each train. The
dust-laden, SCh-rich flue gas enters the scrubber venturi sec-
tion at the top of each train and flows down cocurrently with
the scrubber recycle liquor. The gas then is channelled through
a 180° bend and flows upward through the spray section for S02
removal. The scrubbing liquor is sprayed countercurrently to
the gas. The spent scrubbing liquor falls into the recycle
tank and the clean gas exits at the top of the scrubber, passes
through a steam reheat section and an induced draft fan before
being vented through the stack.
Mist eliminators are washed by a separate recirculating
stream. The wash water is collected by a wash tray and recycled
through a wash tray recycle tank. A portion of the wash water
is pumped to the wash tray pond for solids settling. Clear
liquor is returned to the mist eliminator spray headers. Lime-
softened makeup water is added to replace the water lost through
evaporation and through occlusion with the solids.
A bleed stream is taken from the scrubber recycle
tank, diluted to about 670 solids with slurry pond recycle
liquor, and pumped to the pond system. At the present time,
scrubber solids are dredged and slurried to a disposal pond.
Bottom ash is sluiced to the bottom ash pond in a
recirculating system. Clear liquor from the bottom ash pond
clear well is used as sluice water. There are no aqueous dis-
charges from the plant. Water losses occur through cooling
tower evaporation and drift, scrubber evaporation, pond evapor-
ation, solids occlusion, and boiler losses.
B-9
-------
A total of nine aqueous samples were taken to char-
acterize the Colstrip water system and are listed below:
1) surge pond,
2) cooling tower makeup,
3) cooling tower blowdown,
4) bottom ash sluice water,
5) scrubber recycle slurry,
6) wash tray recycle slurry,
7) fly ash pond recycle,
8) effluent tank, and
9) Pond B overflow to Pond A.
Temperature and pH of each of the nine samples were
taken on location. A two-liter grab sample was taken. One
liter of this sample was filtered and acidified with nitric acid
for subsequent chemical analysis to determine the concentrations
of calcium, magnesium, sodium, potassium, chloride, sulfate,
phosphate, and silica. A portion of the filtered liquor was
taken prior to acidification for use in determining carbonate
concentration in each of the nine samples. The remaining liter
of sample was used to determine the nitrate concentration,
weight percent solids, and total dissolved solids. Sulfite con-
centrations for the scrubber recycle slurry, wash tray recycle,
fly ash pond recycle, and effluent tank samples were determined
by difference between the total sulfur concentration of the
unfiltered, nonacidified sample and the sulfate concentration of
the filtered, acidified sample.
Access to collecting a solid dry fly ash sample was
not possible, so a fly ash sample from the J. E. Corette Plant
in Billings, Montana was collected by Montana Power Co. person-
nel. The same coal is burned at both plants. The analytical
techniques used for these samples are explained in Section 3.0,
and the results are presented in Section 4.0.
B-10
-------
3.0 ANALYTICAL TECHNIQUES
The analytical techniques employed to obtain accurate,
reproducible analyses of water samples collected from the power'
plants are described in this section. Analytical techniques
were chosen consistent with the accuracy requirements of the
simulation model and the levels of concentration of the major
ions present in the water samples. Analytical techniques from
EPA's "Manual of Methods for Chemical Analysis of Water and
Wastes, Standard Methods for the Examination of Water and Waste-
water", 13th Ed. (1971), and techniques recommended in EPA Con-
tract CPA 70-143 were utilized in the characterization of the
water systems of the generating stations studied in this project.
3.1 Calcium, Magnesium, Sodium, Potassium, and Arsenic
Calcium, magnesium, sodium, potassium, and arsenic
ion concentrations were determined by atomic absorption utiliz-
ing a Perkin-Elmer, Model 403 spectrophotometer. Dilutions
were made with a 1% lanthanum chloride, 5% HC1 solution to
suppress interference from a number of other ions which occur
concurrently in the system. Certified atomic absorptions refer-
ence solutions have been used as standards to calibrate the
instrument. After dilution of each ion to the proper concentra-
tion range, the accuracy of the method is ±2% at the 95% confi-
dence level. The analytical procedures are reported in the
EPA's "Manual of Methods for Chemical Analysis of Water and
Wastes".
3.2 Chloride
Chloride was determined by specific ion electrode.
Samples and standards were run and a calibration curve was pre-
pared. Ionic strength adjusters were added to the samples to
give a constant background and eliminate possible interferences.
3.3 Total Sulfur and Sulfate
Total sulfur dissolved in the water samples was mea-
sured as sulfate. For total sulfur determinations, all sulfur
species are first oxidized to sulfate by hydrogen peroxide. The
sulfate was determined by an acid-base titration with standard
NaOH after converting all sulfate to sulfuric acid by passage
through a hydrogen-form cation-exchange column. This method can
be used for sulfate concentrations ranging from 0.001 to 0.5
molar with accuracies of ±2% at the 95% confidence level.
B-ll
-------
3.4 Carbonate
Liquid phase carbonate concentrations were determined
by chemical analyses of the aqueous C02 sample utilizing a. non-
dispersive infra-red analyzer. The aqueous C02 sample is
injected into an acid pool to liberate gaseous CQz, which is
then measured by the nondispersive infra-red analyzer. The
accuracy of the technique is +570 in the carbonate concentration
range of the water samples analyzed.
3.5 Nitrate
A specific ion electrode was calibrated and used to
determine nitrate ion concentrations after filtering the non-
acidified samples. The accuracy of this technique is +25% at
the concentration levels found.
3.6 Phosphate
Phosphate concentrations were determined colorimetric-
ally using the reference method in EPA's "Manual of Methods for
Chemical Analysis of Water and Wastes", page 249.
3.7 Silicate
Silica concentrations were determined by the molybdo-
silicate method in 14th Ed. of "Standard Methods for the Examina-
tion of Water and Wastewater". At a pH of about 1.2, ammonium
molybdate reacts with silica and phosphate to form heteropoly
acids. Oxalic acid is added to destroy the molybdophosphoric
acid. The intensity of the yellow color, measured by spectro-
photometric methods, is proportional to the concentration of
molybdate reactive silica.
3.8 Total Dissolved Solids
Total dissolved solids were determined gravimetrically
using the method reported in EPA's "Manual of Methods for Chem-
ical Analysis of Water and Wastes".
3.9 Sulfite
Sulfite ion concentrations were determined by iodo-
metric titration with sodium arsenite. The sample is added to
an excess of buffered iodine solution and the iodine remaining
B-12
-------
after the stoichiometric S02 oxidation is titrated with standard
sodium arsenite solution employing an amperometric dead-stop
method for end-point detection. The accuracy of the technique
is +2% above 0.5 mmole/S,.
B-13
-------
4.0 RESULTS
The analytical results of the five generating stations
studied will be discussed separately. An analysis of the over-
all accuracy of the analytical measurements will be made by
comparing the total ion concentrations measured with the total
dissolved solids determined, and by analyzing the computer cal-
culations of ion imbalances based on the pH, temperature, and
cation and anion input concentrations. Consistency comparisons
of the analytical results with plant design and operating data
will be made where possible. Potential problem areas associated
with scaling will also be identified.
4.1 Four Corners Generating Station
Results of the chemical analyses of the samples taken
at the Four Corners Plant are presented in Table 4-1. The pH,
temperature, and dissolved species concentrations are shown for
each sample.
The sum of the total ions is shown for comparison to
the total dissolved solids to allow a quantitative evaluation of
the accuracy of each sample analysis. The °L residual electro-
neutrality is also a measure of the sample consistency.
The analytical results of seven of the nine streams
monitored are consistent with the measured total dissolved
solids (TDS). The residual electroneutrality for these seven
streams as calculated by the equilibrium program reflect ion
imbalances of less than 1070. Inconsistencies were identified in
the analytical results of the effluent liquor of Venturi
Scrubber 3A and the thickener underflow. The solids content
(wt. % solids) of the slurry streams were consistent with
plant data.
The measured total ion concentration in the Scrubber
3A effluent liquor sample was 97o lower than that indicated by
the TDS measurement. The residual electroneutrality indicated
an ion imbalance of -22%. This was interpreted as a deficiency
of cations when compared to the low total ion concentration
indicated by the TDS measurement. A comparison of the chemical
analyses of the two scrubber liquors indicated that the measured
calcium concentration in Scrubber 3A is low. Duplicate analyses
for calcium and sodium confirmed the earlier analyses. Duplicate
analyses for total sulfur and total dissolved solids were also
B-14
-------
TABLE 4-1.
CHEMICAL ANALYSIS OF AQUEOUS SAMPLES OF FOUR CORNERS POWER PLANT
ARIZONA PUBLIC SERVICE
Dissolved Solids
pH
Temperature, *C
Calcium, mg/t
Magnesium, mg/t
Sodium, mg/t
Potassium, mg/t
Chloride, mg/t
Total Sulfur, mg/t
as S0»
Sulflje. mg/t
as SO!
CarboQate, mg/t
as CO,
Nitrate, mg/1
as NOl
Phosphate, mg/t
as PO,
Silicate, mg/t
as SiO,
1 Suspended Solids
Total Ions, mg/t
Total Dissolved
Solids, mg/t
T. Residual Electro-
neutrality**
Partial Pressure of
COi. atm
Makeup
Water
8.1
17
160
40
210
8
110
680
—
77
9
<0.3
100
<0.01
1390
1350
+8
4.8 x 10~»
Relative Saturations***
CaCO, 1.21
CaSO,-2HtO 0.21
Scrubber
1A
2.8
33
790
49
290
11
160
2740
8
30
25
2.2
270
2.17
4370
4370
-5
1.8 x ID'2
7.6 x 10"'
1.39
Liquor
3A
3.1
29
670
66
350
14
220
2930
60
27
47
3.5
360
8.80
4690
5150
-22
1.5 x lO'2
1.8 x 10"'
1.27
Thickener
Overflow
3.8
33
730
54
320
14
180
2540
18
16
24
<0.3
140
0.06
4140
4110
-5
9.6 x lO"1
3.8 x 10~T
1.28
Thickener
Underflow
8.3
21
330
38
280
8
110
1160
--
60
7
<0.3
150
10.0
2140
1760
+11
2.3 x 10-'
3.18
0.50
Bottom Ash
Sluice
6.9
26
160
35
210
8
140
690
--
74
—
<0.3
110
0.04*
1420
1420
+2
7.0 x 10-'
0.11
0.19
Sluice Tank
Effluent
6.7
26
700
46
290
11
160
2110
—
' 21
20
<0.3
780
2.19
4140
4070
+5
2.6 x 10"'
0.05
1.19
Ash Pond
Effluent
9.0
17
650
44
280
10
180
2040
—
29
23
<0,3
730
<0.01
3980
3880
-6
1.6 x 10"'
7.10
1.18
Ash Pond
Surface
8.9
17
620
43
270
10
170
1880
—
31
--
<0.3
560
0.04
3580
3600
+4
2.3 x 10"'
6.37
1.11
*Not representative
**Does not Include silica
***Critlcal values, above which scale potential exists, are 1,3 - 1.4 for CaSO»-2HjO and about 2,5 for CaCOs (see Appendix C)
-------
made. Initial analytical results were again confirmed. These
results indicate that there is a cation other than calcium or
sodium in the liquor which has not been accounted for or that
there were inconsistencies in the sampling.
Chemical analysis of the thickener underflow sample
at Four Corners revealed a 22% greater total ion concentration
than the TDS analysis showed. The ion imbalance calculated by
the equilibrium program was +1170, indicating an excess of cat-
ions. Repeat analyses of the calcium, sodium, total sulfur and
TDS confirmed earlier results. These inconsistencies were not
resolved. This problem could be related to the high suspended
solids in this sample (could have been solid-liquid reactions
after the sample was taken) or the pH measurement could have been
in error.
Several problem areas with respect to scaling have
been identified. Significant calcium carbonate supersaturation
was observed in three streams: 1) the thickener underflow (3-2),
2) the ash pond effluent (7.1), and 3) the ash pond surface
(6.4). The calcium carbonate precipitation rate studies (see
Appendix C) indicated that at relative saturations above 2.5,
the precipitation rate increases sharply, indicating a potential
for scaling. The calcium carbonate relative saturations for the
three streams listed above are all greater than 2.5.
Calcium sulfate dihydrate (gypsum) relative saturations
in the critical range (1.2-1.4) were observed in both Scrubber 1A
and 3A effluent streams and in the thickener underflow. Gypsum
relative saturations greater than 1.1 were also observed in the
sluice tank effluent and the ash pond. Drastic differences in
the pH of the thickener overflow and underflow were measured.
Analytical and computer results indicate that this may be due to
poor mixing of the lime with the scrubber effluent liquor in the
thickener.
4.2 Comanche Generating Station
Results of the chemical analyses presented in Table 4-2
are, in general, consistent with each other and with plant design
and operating data. The sum of all total ions measured were
within 6% of the total dissolved solids measured with the ex-
ception of the polishing pond effluent (±13%). The electro-
neutrality balances as calculated by the computer equilibrium
program closed within 10% for all streams monitored. None of
these samples contained high levels of suspended solids and
B-16
-------
TABLE 4-2. CHEMICAL ANALYSIS OF AQUEOUS SAMPLES OF COMANCHE
GENERATING STATION PUBLIC SERVICE OF COLORADO
Dissolved Cooling Tower
Species Makeup
PH
Temperature, °C
Calciun, mg/i.
Magnesium, mg/t
Sodium, ng/K
Potasslua, ng/Jl
Chloride, ng/ft
Total Sulfur, mg/fc
aa SO,*
Carbonate, mg/l
as COj"
Nitrate, ng/«
as N0j~
Phosphate, mg/J.
as P0,=
Silicate, ag/ft
aa SiOs~
X Suspended Solids
Total Ions, ng/H
Total Dissolved Solids, og/f.
Z Residual Elect roneutrality*
Partial Pressure of CO; , atn
Relative Saturations**
CaCOs
CaSO«,-2H20
6.20
14
36.5
10.2
19
1.7
9
163
6.0
9.0
0.10
56
<0.01
311
298
«
1.2 x 10"'
1 x 10"
0.028
Cooling Tower
Slowdown
6.30
22
205
65.5
89
13
53
965
2.7
16
3.5
280
<0.01
1690
1700
-10
5.6 x 10-"
5 x 10"
0.31
Makeup
Ash Sluice
8.55
12
53.4
14.2
19
2.3
7
134
101
13
0.10
11
<0.01
354
345
-5
2.2 x 10"*
1.56
0.031
Ash Pond
Inlet
7.45
20
115
18.3
29
3.6
12
260
111
19
0.1
48
0.29
616
573
+8
3.0 x 10"3
0.41
0.088
Ash Pond
Substrate
7.65
24
123
25.5
36
4.7
19
379
86
12
0.6
94
<0.01
779
773
-2.4
1.6 X 10"'
0.66
0.12
Ash Pond Polishing Pond
Effluent Effluent
7.25
16
105
25.4
44
5.8
16
355
80
17
0.8
110
<0.01
763
763
-2
3.1 x 10~3
0.12
0.11
7.70
23
149
33.4
39
6.9
27
528
67
13
2.1
130
<0.01
990
878
-7
1.1 x 10"'
0.60
,0.17
*Doea not include silica
**Critical values, above which scale potential exists, are 1.3 - 1.4 for CaSO^'2HzO and about 2.5 for CaCOj (see Appendix C)
-------
significant solid-liquid reaction after the sample was taken is
unlikely. This is confirmed by the small electroneutrality im-
balances. Concentrations of each ion were consistent with flow
rates and evaporation losses reported in plant operating data.
The bottom ash sluice water was the only stream which showed any
supersaturated species. Computer compilations predicted a cal-
cium carbonate relative saturation of 1.56. Calcium carbonate
precipitation rate studies indicate that this is below the crit-
ical value for scaling.
4.3 Bowen Generating Station
Chemical analyses of five of the seven streams pre-
sented in Table 4-3 were consistent with total dissolved solids,
computer equilibrium compilations, and plant operating data.
Chemical imbalances were observed in the cooling tower blowdown
and the ash pond inlet.
The cooling tower blowdown had a residual electroneu-
trality imbalance of -13% and an 8% greater total ion concentra-
tion than shown by the TDS determination. These values indicate
an error in the carbonate concentration measurement. A ±5% error
in the carbonate concentration would lower the inconsistencies
to within overall analytical error.
The major inconsistency in the analytical results was
observed in the TDS of the ash pond inlet stream. The sum of
the total ions measured was 35% lower than that indicated by the
TDS determination. Repeat analyses of the calcium, magnesium,
total sulfur, and TDS confirmed initial results. Silica may
account for some of the ion deficiency.
Extremely high calcium carbonate relative saturations
were identified in the ash pond inlet, ash-pond subsurface, and
ash pond effluent samples (38.8, 17.4, 17.1, respectively).
These relative saturations are substantially greater than the
critical level of about 2.5 as discussed in Appendix C.
4.4 Montour SES (PP&L)
Results of the chemical analyses of the samples taken
at the Montour Steam-Electric Station are presented in Table 4-4.
The sample showing the largest residual electroneutrality (-18.9%)
also shows a discrepancy between total ion and total dissolved
solids. The "ash basin at dike" sample analysis yielded a total
B-18
-------
TABLE 4-3. CHEMICAL ANALYSIS OF AQUEOUS SAMPLES OF BOWEN
GENERATING STATION GEORGIA POWER CO.
Dissolved
Species
pH
Temperature, *C
Calcium, Bg/i
Magneslua, Bg/l
SodiuB, Bg/l
PotassiuB, «g/l
Chloride, Bg/l
Total Sulfur, Bg/l
as SO,*
Carbonates, Bg/l
W as COa"
J^ Nitrate, Bg/l
as H03~
Phosphate, Bg/l
as PO,"
Silicate, Bg/l
as SiOj
Z Suspended Solids
Total Ions, mg/l
Total Dissolved Solids, Bg/l
Z Residual Electroneutrality*
Relative Saturations**
CaC03
CaSOS'2H20
Cooling Tower
Makeup
7.7
21
6.1
1.7
1.4
<0.4
2.1
1.9
20.4
4.0
<0.10
25
<0.01
65
57
-4.0
0.016
1.0 x 10~
Cooling Tower
Slowdown
7.9
23
16.1
2.1
0.2
<0.4
6.4
3.0
43
8.4
<0.10
28
<0.01
107
93
-12.8
0.15
3.0 x 10""
Bottom Ash
Sluice
6.5
36
21.6
2.3
1.5
1.5
3.5
38.4
39
5.2
<0.10
30
1.11
143
139
-5.0
0.01
5.0 x 10"*
Fly Ash
Sluice !
11.5
29
311
<0.10
9.4
19.8
3.9
514
22
9.5
<0.10
-
7.04
890
1370
-3.9
38.8
0.28
Ash Pond
Subsurface
10.4
19
89
1.7
18.7
5.9
7.7
168
24
8.9
<0.10
53
<0.01
377
364
+10.0
17.4
0.056
Ash Pond f
Effluent
10.4
19
89
1.7
19.6
5.4
8.2
182
24
11.2
<0.10
55
<0.01
396
374
+4.0
17.1
0.058
'lant Dralnoff
Water
8.4
21
18.8
2.0
1.9
3.5
18.3
28.8
31
10.1
<0.10
58
0.09
173
135
+6.0
0.31
3.0 x 10~3
*Does noc Include silica
"Critical values, above which scale potential exists, are 1.3 - 1.4 for CaSO<,'2H20 and about 2.5 for CaCOs (see Appendix C)
-------
TABLE 4-4. CHEMICAL ANALYSIS OF AQUEOUS SAMPLES OF MONTOUR SES,
PENNSYLVANIA POWER AND LIGHT COMPANY
Species
pH
Temperature, °C
Calcium, mg/t
Magnesium, ng/t
Sodium, Bg/2.
Potassium, mg/t
Chloride, mg/i
Total Sulfur, Bg/i
as SOi,=
Cd Carbonate, ng/i.
1 as C03=
° Nitrate, mg/£
as SO 3-
Phosphate, ng/i
as POi,"
Silica, mg/S.
as Si02
Arsenic, ng/t
as As
X Suspended Solids
Total Ion, mg/i
Total Dissolved Solids, «g/4
% Residual Electroneutrality*
Partial Pressure of C02> atB
Relative Saturations**
CaCOa
CaSOi»-2H20
Cooling Tower
Makeup
8.1
5.0
28.4
5.5
7.0
2.6
19
66
6.0
5.5
.029
0.9
—
.0008
110
100
+4.1
4.0 i 10"s
.013
.012
Cooling Tower
Slowdown
Unit 1
7.8
7.0
43.2
9.5
10.2
3.4
32
88
9.6
11.8
—
2.0
0.02
.0004
210
290
+7.6
2.8 x 10-"
.037
.019
Cooling Tower
Slowdown
Unit 2
7.3
30.0
49.7
12.6
10.2
2.6
33
131
9.0
10.2
—
3.1
0.02
—
261
250
+1.8
2.8 x 10-*
.025
.027
Ash Basin
at Dike
8.7
11.0
98.9
10.0
11.4
8.2
33
197
24
6.8
—
1.4
—
.004
391
470
-18.9
3.7 x 10"*
8.08
.066
Ash Basin
Overflow
7.7
6.5
98.9
10.0
11.8
7.4
34
245
9.6
9.9
1.02
2.0
—
.0012
430
460
+2.4
1.5 x 10"1
.024
. .093
Detention
Basin
Overflow
7.5
6.5
87.4
9.0
19.1
6.6
29
215
1.5
11.1
.056
1.4
—
.0016
380
290
+5.3
3.5 x 10-*
.021
.076
Hisc.
Wastes
7.7
10.5
28.4
6.0
7.4
0.9
18
66
10.8
5.5
--
0.6
—
.0016
144
170
+3.2
1.8 x 10"
.014
.011
Fly Ash
Slurry
Unit 2
8.9
29.5
142
10.4
12.6
9.9
38
267
3.8
13.3
.224
2.1
0.067
2.1
499
690
+11.3
"3.2 x 10-
6.83
.105
Butt OB Ash
Slurry
Unit 1
5.8
22.5
39.4
9.0
9.8
5.0
32
101
35.4
11.5
.046
2.0
—
.0532
245
200
+1.2
5 1.1 x 10'2
3.8 x 10" 2
.019
Mill
Reject
Slurry
6.9
17.5
39.9
11.7
9.4
4.2
33
78
7.8
11.8
—
0.5
—
.0056
196
190
+12.1
7.0 x lO""
.0029
.015
•Does not Include silica
**Critical values, above which scale potential exists, are 1.3 - 1.4 for CaSOii'2H20 and about 2.5 for CaCOj (see Appendix C)
-------
ion less than the total dissolved solids measurement which
indicates the presence of an unmeasured species. However, this
sample is not critical for performing the process simulations of
existing operations nor of alternatives.
No samples indicated CaSCK^HaO scaling tendencies but
the "ash basin at dike" and "Unit #2 fly ash slurry" samples
showed CaC03 relative saturations of 8.08 and 6.83, respectively.
However, these samples showed discrepancies in total species and
residual electroneutrality which may make the calculated scaling
tendencies questionable. No scaling problems have been encoun-
tered at Montour although erosive action of the fly ash slurry
may keep significant deposits from forming in the fly ash pipe-
line.
The low CaS04-2H20 relative saturations in the cooling
tower blowdown samples indicate that the cycles of concentration
in the cooling towers can be increased significantly if desired
without causing gypsum scale. Acid treatment may be necessary
for pH control (to prevent CaC03 scale) at higher cycles however,
depending on the CO2 equilibrium values in the circulating liquor.
This will be determined in the computer simulation studies of
existing operations and technical alternatives.
4.5 Colstrip SES (Montana Power Co.)
The results of the chemical analyses of the samples
taken at the Colstrip Steam-Electric Station are presented in
Table 4-5. As before, the sample location, pH, temperature, and
dissolved species concentrations are shown for each of the nine
samples. Total ion concentrations compare reasonably well
(within 15%) to the TDS determinations except for the surge pond
and wash tray recycle samples, which differ by about 20%. The
residual electroneutrality for the surge pond sample is +23.870
indicating that an additional anion which was not measured may
be present, or that one of the cationic species concentrations
is too high. Since the lime softened water results were used as
computer inputs, the discrepancy in the surge pond sample was
not resolved.
Only one of the nine samples, the bottom ash sluice
water, showed a CaC03 relative saturation greater than the crit-
ical value for scaling of about 2.5. The calculated value for
this sample is 4.55. The bottom ash sluice water sample also
was calculated to have a CaSO^'2E20 relative saturation of 1.67
which is above the critical level of 1.3-1.4. The erosive nature
B-21
-------
w
TABLE 4-5. CHEMICAL ANALYSIS OF AQUEOUS SAMPLES OF COLSTRIP SES,
MONTANA POWER COMPANY
Species
pH
Tenperature, °C
Calciua, mg/l
Magnesiun, Bg/1
Sodium, mg/i
Potassium, mg/l
Chloride, ng/l
Total Sulfur, «g/l
as SOi,=
Sulfi$e, mg/J.
as S03
Carbonate, ng/l
as C0j=
Nitrate, BgA
as N03~
Phosphate, mg/l
as P0»~
Silica, mg/l
as SiOa
Z Suspended Solids
Total Ion, ug/£
Total Dissolved Solids, «g/4
Z Residual Elect roneutrality *
Partial Pressure of COi , ata
Relative Saturations**
CaCOs
CaSO»-2H20
Surge Pond
6.7
10.5
57.9
19.5
53.5
4.2
22
174
—
17.3
1.7
—
1.8
.002
352
440
+23.8
1.9 x 10-3
.0026
.041
Cooling Tower
Makeup
10.3
6.0
39.9
10.7
53.1
4.2
17
188
—
6.0
1.4
—
1.3
.0016
322
360
+7.3
1.0 x 10~7
1.08
.034
Cooling Tower
Slowdown
6.7
27.5
533
193
710
50.3
266
3820
—
34.8
11.2
.26
5.0
.0014
5624
6000
-10.5
4.3 x 10"'
.051
1.11
Bottom Ash
Sluice
Water
• 10.4
9.5
722
70
295
13.1
79
2780
~
7.2
68
—
1.4
.0048
4036
4200
-7.9
2.6 x 10-'
4.55
1.67
Scrubber
Recycle
Slurry
3.9
50.5
504
5050
458
21.9
129
19,400
300
52.2
161
—
31
7.7
25,807
29,200
+11.0
5.1 x 10~'
1.87 x UT5
1.00
Wash Tray
Recycle
3.4
50.5
519
2925
153
11.5
67
10,600
1560
25.2
80.6
—
22
.88
14,403
16,300
+21.2
2.4 x 10~'
1.2 x 10~6
.82
Fly Ash
Pond
Recycle
5.5
5.0
484
1550
305
13.1
70
9000
400
9.6
130
.009
24
.0056
11,586
13,690
-10.6
2.4 x 10"'
2.2 x 10"s
1.38
Effluent
•Tank
4.4
17.6"
497
2075
315
15.5
74
11,800
100
31.2
118
.028
25
1.36
14,951
17,200
-13.9
1.3 x 10~2
1.5 x 10~6
1.31
Pond B
Overflow
to Pond A
4.8
10.5
464
1600
345
13.1
74
9521
—
9.0
136
—
21.4
.01
12,184
14,400
-13.7
3.0 x 10~s
1.5 x 10"'
1.30
*Does not Include silica
**Critical values, above which scale potential exists, are
1.3 - 1.4 for CaSOt,'2H?0 and about 2.5 for CaCCh (see Appendix C)
-------
of the slurry may keep deposits from forming if the water loop
is tightened since no scaling problems have been encountered in
these areas. The cooling tower blowdown sample shows that the
cooling towers cannot be operated at higher cycles of concen-
tration than presently done without producing a gypsum scale
potential in the system. To operate the towers at higher cycles
would require treatment for calcium removal, such as lime soft-
ening of a slipstream of the recirculating water.
B-23
-------
APPENDIX C
CaC03 AND Mg(OH)2 PRECIPITATION KINETICS
1.0 INTRODUCTION
In order to evaluate the technical feasibility of
several water recycle/reuse options it is necessary to establish
scaling potential criteria for CaC03 and Mg(OH)2. This appendix
describes the studies performed to establish precipitation kine-
tics data for these two species. Precipitation rates were de-
termined as a function of relative saturation and critical values
for scale formation were established at about 2.5 for CaC03 and
3.4 for Mg(OH)2.
First the results from the literature survey concern-
ing the precipitation kinetics of CaC03 and Mg(OH)2 are presented
Then the experimental apparatus and procedures are explained
followed by discussions of the results.
C-l
-------
2.0 LITERATURE SURVEY
The first step in evaluating the precipitation-kinetics
of CaC03 and Mg(OH)z is to search and retrieve physicochemical
and kinetic data from the literature. This section presents a
summary of the findings of the literature survey. Information
obtained in the areas of solubility, solubility product con-
stants, ion-pair formation, kinetic data, and their respective
temperature dependencies will be emphasized.
2.1 Basis of the Literature Survey
The following sources were consulted for this litera-
ture survey:
1) Chemical Abstracts - This source was consulted
for the period of May 1975 to January 1962.
2) Link, W. F., Solubilities of Inorganic and
Metal Organic Compounds - Volume II,
Washington, B.C., American ChemicaT Society.
3) W. L. Badger and Associates, Inc., "Critical
Review of Literature on Formation and Preven-
tion of Scale," July 1959, PG-161-399.
4) L. G. Sillen, Stability Constants of_ Metal-
Ion Complexes, London: The Chemical Society,
Burlington House, 1964.
2.2 Solubility Data
In this section, temperature-dependent solubility data
for the species CaCOs and Mg(OH)2 will be presented.
2.2.1 CaC03
Since the equilibrium reactions
C03 = + H20 £ HCOa- + OH- (2-1)
and HC03- + H20 £ H2C03 + OH- (2-2)
C-2
-------
and
H2C03 £ H20 + C0
(2-3)
are appreciable in solution, the solubility of CaCO3 is dependent
on the concentration of CO2 in solution, and therefore on the
partial pressure of COa in the atmosphere above the solution.
In view of this fact solubility data for CaC03 can be obtained
not only as a function of temperature, but al^o as a function
of partial ^pressure.of C02 and pH. Therefore, in Table 2-1,
temperature-dependent solubility data taken from Link (LI-001)
is presented for CaC03. In this case the water solution is in
contact with ordinary air containing approximately 3.15 parts
of C02 per 10,000. It is noticed that CaCO3 has an inverse
solubility; that is, one that decreases with increasing tempera-
ture.
TABLE 2-1.
SOLUBILITY OF CALCIUM CARBONATE IN WATER
IN CONTACT WITH ORDINARY AIR
Grams CaCO3 per Liter
0
10
20
25
30
40
50
0.081
0.070
0.065
0.056
0.052
0.044
0.038
Also, for comparison we present in Table 2-2 tempera-
ture-dependent solubility data for CaCO3 in water essentially
free of C02. This data is also taken from Link (LI-001). It is
clear from this data that the presence of C02 has a major effect
on the dissolution characteristics of CaC03.
In view of this fact, it is understandable that solu-
tion pH is also important in determining solubility characteris-
tics of this species. Solubility data for CaCO3 at 25°C as a
function of pH is presented in Table 2-3. This data is taken
from Jaulmes and Brun (JA-105).
C-3
-------
TABLE 2-2. SOLUBILITY OF CALCIUM CARBONATE IN WATER
ESSENTIALLY FREE OF CARBON DIOXIDE
Grains CaCO 3 per Liter
17
18
25
95
100
182
207
244
316
0.0145
0.0128
0.0132
0.024
0.0375
0.025
0.014
0.011
0.080
TABLE 2-3. SOLUBILITY OF CALCIUM CARBONATE
AT 25°C AS A FUNCTION OF pH
Grams CaC03 per Liter pH
1.87
0.35
0.10
0.032
0.012
0.0077
0.0071
0.0071
0.0070
6
7
8
9
10
11
12
13
14
2.2.2 Mg(OH)2
According to Gjalbaeck (GJ-001), Mg(OH)2 exists in
two well defined modifications of which the more soluble is the
labile phase and the less soluble is the stable phase. At a tem-
perature of 18°C, the following comparative results are reported
(LI-001).
C-4
-------
Solubility Solubility
(moles/liter) (moles/liter)
Method Stable Phase Labile Phase
Direct Determination 2.2 x 10" " 6.5 x 10~"
Mg+2 + 2NH4OH * Mg(OH)2 + NH,+ 1.9 x 10"" 5.5 x 10""
Conductivity Method 1.35 x 10~" 4.6 x 10""
Electrometric Method 1.6 x 10"" 7.0 x 10""
Temperature-dependent solubility data for Mg(OH)2,
taken from Link (LI-001), is presented in Table 2-4.
TABLE 2-4. SOLUBILITY OF MAGNESIUM HYDROXIDE IN WATER
T (°C)
18
25
35
45
75
100
110
142
150
158
Moles Mg(OH)2 per Liter
0.000200
0.000197
0.000169
0.000150
0.000118
0.000072
0.000074
0.000042
0.000037
0.000031
2.3 Solubility Product Data
For this program, precipitation rate data will be ex-
pressed as a function of the relative saturation of the particu-
lar precipitating species in solution. The relative saturation
is defined as the ratio of the solution activity product for the
precipitating species (in the case of CaC03, aca+2 x aCO ~2^
to the equilibrium solubility product, Ksp, for 3
the precipitating species at the temperature T. Therefore, in-
formation on the temperature dependence of the solubility prod-
ducts for CaCO3 and Mg(OH)2 is necessary for these studies.
C-5
-------
Data has been previously compiled concerning the
temperature dependence of solubility for these species for use
in the equilibrium program. However, in this section solubility
product data that is available from other independent sources
will be presented.
2.3.1
CaC03
Giringhelli and Bianucci (GH-001) have determined the
solubility product constant for CaC03 in the temperature range
0 to 50°C in water saturated with C02 at an equilibrium partial
pressure of 0.0004 atmospheres. Their data is presented in
Table 2-5.
TABLE 2-5.
SOLUBILITY PRODUCT CONSTANT FOR CaCO3
IN H20 SATURATED WITH CO?
Ksp
0
10
20
25
30
40
50
6.0 x 10_9
4.5 x 10 9
3.4 x 10~9
2.9 x 10 9
2.5 x 10 9
1.9 x 10 9
1.4 x 10 9
2.3.2
Feitknecht and Schindler (FE-105) have reported values
for the solubility product constant of Mg(OH)2 at 25°C. For the
labile phase of Mg(OH)2, they report a Ksp of 6.32 x 10~6; and
for the stable phase, a Ksp of 1.26 x 10 11.
No literature information was found that gave the
temperature dependence of the solubility product for Mg(OH)2.
However, temperature-dependent solubility product data for this
species has been previously determined for use in the equilibrium
computer program. From available thermodynamic data the tempera-
ture dependence of the solubility product constant was calculated
by using the integrated Gibbs-Helmholtz equation.
C-6
-------
2.4 Dissociation and Ion Pairing Data
No information was found in the literature that gave
the temperature dependence of the dissociation constants for the
ion-pairs CaCOa0 and Mg(OH)+. However, temperature-dependent
dissociation constants for CaCCh ° and Mg(OH)+ have been previ-
ously determined by calculations using the Fuoss equation
(FU-001) for use in the equilibrium program.
2.5 Formation Kinetic Data
In this section data available on the formation
kinetics of solid CaCOa and Mg(OH>2 will be presented.
2.5.1 CaC03
In 1965, R. Pytkowicz (PY-010) published an experimen-
tal study of the rate of CaCOs nucleation over a wide range of
carbonate concentrations.in natural and artificial sea waters.
This work was performed in order to study factors affecting the
induction period of nucleation at high supersaturations, and to
permit extrapolation of nucleation rate data to low supersatura-
tions.
In his experimental study, sodium carbonate solutions
were added to natural and artificial sea water, prepared by the
method of Lyman and Fleming (LY-001), with the original ionic
strength maintained by adding sodium chloride. In some cases,
sodium bicarbonate was also added to determine the effect of
changes in pH on the induction period. The time of nucleation
and solution pH were determined.
In Figure 2-1, the results of the addition of 8.25
mmoles/liter of Na2C03 to four natural sea water samples are
shown. The variation of solution pH is shown versus time.
Various amounts of NaaCOa were added to samples of
filtered and unfiltered natural sea water, and the time of
nucleation or induction period was determined. Na2COs and 2
mmoles/liter of NaHCO3 were added to some of the samples and
these results are shown in Figure 2-2.
C-7
-------
VISUAL
RANGE
NATURAL SEA WATER + 9.25 m moUa
Na2CO3/ LITER
• FIRST RUN
o SECOND
* THIRD
A FOURTH
_L
50 100 150
TIME (HOURS) AFTER ADDIND ADDING Na2C03
Figure 2-1. Visual and pH determination of the
time of nucleation.
10
DC
LJ
'o
£
O
00
oc
a
ID
O
O
0.1
FILTERED SEA WATER + CO3 =
FILTERED SEA WATER + CO3a -c HCO3~
UNFILTERED SEA WATER + CO3 =
J L
10 100 1000
TIME OF NUCLEATION (HOURS)
Figure 2-2. Time of nucleation versus
initial carbonate concentration.
C-8
-------
The effect of magnesium on the time of nucleation was
studied by adding Na2C03 to magnesium-free artificial sea water,
and to natural sea water that was enriched with magnesium. The
results are compared in Table 2-6 with selected values from
Figure 2-2.
Pytkowicz observed that at very high carbonate concen-
trations the nucleation was slowed by further addition of car-
bonate, as is shown in Figure 2-2. He also observed that mag-
nesium-free artificial sea water yielded much shorter times of
nucleation than did natural sea water, and that it did not
produce the minimum time of nucleation observed in Figure 2-2.
From Table 2-6 it can be seen that enrihhment with magnesium
decreased the rate of nucleation. Log-log plots of the time of
nucleation versus the carbonate concentration showed a second-
order decay of the nucleation time with increasing carbonate
concentration in magnesium-free artificial sea water, A sixth-
order decay was found in natural sea water.
TABLE 2-6. TIME OF NUCLEATION RESULTING FROM ADDED CARBONATE
Added
Carbonate
(mmole/1)
Time of
Nucleation
A. Magnesium-free artificial sea
water
B. Natural sea water (selected
values to match (A))
C. Natural sea water enriched in
magnesium to about twice the
original concentration
7.34
4.58
1.83
0.93
7.
4.
1.
35
58
83
0.93
4.75
3.85
1.0 min.
4. 7 min.
13 min.
20 min.
6.0-7.0 hr.
3.0-4.5 hr.
15-18 hr.
900-940 hr.
22-53 hr.
22-53 hr.
Pytkowicz stated that his results suggest that many
more collisions of carbonate ions are necessary to form CaC03
nuclei in the presence of magnesium. Also, it is apparent that
magnesium inhibits the formation of CaCOa nuclei and is the
predominant factor in determining the time of nucleation at high
carbonate concentrations.
C-9
-------
2.5.2 Mg(OH)2
In 1967, David Klein et al., (KL-052) performed an
experimental study of the homogeneous nucleation of Mg(OH)2.
They asserted that the rate of homogeneous nucleation of Mg(OH)2
precipitating from solution could be written as
dN/dt = k (IP)n (2-4)
where
N = number of nuclei formed
t = t ime
(IP) = ion activity product or a^ + 2 x SQ^-2
n = number of Mg(OH)2 units in the nucleus
(KL-053, KL-054)
It is known from the theory of homogeneous nucleation from dilute
solutions by Nielsen (NI-001) that the chief parameters which
determine nucleation rate are the salt-solution interfacial ten-
sion, the supersaturation of the solution, and the number of ions
in the nucleus. Therefore, determining the rate of nucleation
as a function of the ion product provides a conceptually simple
means for determining the size of the precipitation nucleus.
In their experimental study magnesium ion and hydroxide
ion were generated simultaneously in stoichiometric amounts by
electrolysis of a 0.1M solution of NaN03 with a magnesium anode
and a platinum cathode. Precautions were taken to remove most,
if not all, of the foreign particles from the solution to insure
homogeneous nucleation. This was accomplished by continuously
pumping the solution from the cell through a fiberglas filter
mat and back into the cell. At intervals during the electrolysis
a sample was removed and the number of particles present in it
was counted by using a Coulter electronic particle counter with
a 30 micron orifice. Also, the necessary analytical determina-
tions were made on each sample.
The results of one of their typical experiments is shown
in Figure 2-3. The solution pH and total number of particles
are plotted versus time duration of the experiment. The maximum
rate of formation of particles occurs at approximately the maxi-
mum pH of the solution.
C-10
-------
HOMOGENATION NUCLEATION OF MAGNESIUM HYDROXIDE
3.0
11.18
PH
11.14
11.10
2.5
2.0
u
o
20 40 60 80
TIME (MIN)
100
ft
1.5 CL
m
o
1.0
0.5
Figure 2-3. Variation with time, of pH (upper curve) and of
number of particules, during an experiment.
The results of four of their experiments are presented
in Figure 2-4. In this figure, the log of dN/dt is plotted ver-
sus the log of the ion activity product. The least-squares slope
for the points of Figure 2-4 is 33±4 and so the best value for
the number of Mg(OH)a units in the nucleus is 33.
3.0
2.5
I 2.0
»
E
0 1.5
1
10 1.0
0.5
-9.10 -9.08 -9.06
log (IP)
-9.04
Figure 2-4.
Dependence of rate of nucleation (particles/cm3 sec)
on solution concentration expressed as (IP)
C-ll
-------
These researchers also made a comparison of their ex-
perimental nucleation rate data with theoretical estimates. They
used the following theoretical expression for the rate of homo-
geneous nucleation (NI-001) :
exp (-AG*/kT) (2-5)
where
/ = rate of nucleation
/dt
D = diffusion coefficient of Mg(OH)2
d = mean diameter of one Mg and two OH ions
v = molecular volume of solvent
o
tj) = kT In S = kT In ((IP)/Ksp) ''3
n* = number of ions in the nucleus
AG* = free energy of nucleation
By using their experimental values for dN/dt at particular ion
activity products (IP) they were able to calculate a free energy
of nucleation (AG*) from Equation 2-5 for the Mg(OH)2 precipita-
tion process. This value based on experimental data was compared
with a theoretical thermodynamic value calculated from Gibb's
equation:
G* = n*/2
The two values agreed closely indicating that homogeneous nuclea-
tion of Mg(OH)2 was indeed observed and that the critical nuclei
size was approximately 33 Mg(OH)2 units.
C-12
-------
3.0 APPARATUS AND PROCEDURES
This section presents detailed descriptions of the
experimental apparatus and procedures used to correlate precipi-
tation rates with relative saturation for CaCOa and Mg(OH)2.
First the batch solid reactor is described along with the ancil-
liary equipment followed by a discussion of the experimental and
analytical procedures involved.
3.1 Description of Experimental Apparatus
The major experimental apparatus used in this precipi-
tation kinetics study is illustrated in Figure 3-1. The experi-
ments are centered around the batch-solid crystallizer shown in
Figure 3-2.
3.1.1 Inlet Feed Systems
Supersaturated reactor solutions are produced by intro-
ducing two separate feed solutions at constant flow rates into
the well stirred reactor. In the case of the CaCOa study, CaCl2
and NaaCOa of predetermined concentration are used as the stock
feed solutions. These solutions are made up quantitatively
prior to a run and stored respectively in two covered 16 gallon
Nalgene feed containers. Eastern Industries circulating pumps
are used to remove the feed liquor from the storage containers
through outlet ports at the bottom. From the pumps, a line is
returned to the storage containers in order to release excessive
head pressures from the pumps and to continuously circulate feed
solution (see Figure 3-1). Another line from the pump outlet is
directed to the Moore (Model 63-SD) constant differential pres-
sure regulators used to maintain a constant hydraulic pressure
drop across the Whitey control valves. This control system thus
maintains a constant predetermined feed liquor flow rate regard-
less of possible head pressure fluctuations.
The regulator system is followed by Matheson (Model
7641) flowmeter units, calibrated to 1% accuracy, and used to
measure the constant feed solution flow rates. A constant and
accurate flow rate is important since the steady state precipi-
tation rate calculations employ this experimental parameter.
From the flowmeter units the feed streams are directed
to stainless steel tubing coils immersed in the constant temper-
ature water bath before entering the reactor. By heat transfer,
the feed solutions are thus preheated to the desired reaction
temperature prior to entering the reactor.
C-13
-------
o
I
r
HEATING
COILS
CONSTANT PRESSURE DROP
REGULATOR
— EFFLUENT
I TO ANALYSES
I
TWO-LITER
STIRRED
REACTOR
_| vvvv?^x 1
CONSTANT TEMPERTURE BATH
Figure 3-1. Experimental system for liquid-phase reaction study.
-------
1/4 NPT-DRILL AND TAP
STIRRING ROD BEARING
(STIRRER NOT SHOWN)
FEED LIQUOR INLET
PORT
FOUR 7" X 1/8" X 7/6"
PLEXIGLASS BAFFLES
5 3/4 O. D. X 5 1/4 I. D.
PLEXIGLASS TUBE
DRILL AND TAP 10 - 32
/ 6 HOLES
1/4" MEMBRANE FILTER
SUBSTRATE
\0.
O-RING - FLANGE SEALS
Figure 3-2. Batch-solid reactor.
C-15
-------
Within the reactor, precipitation from the supersatu-
rated solution is initiated by seed crystals which are retained
within the reactor cavity by a filter membrane. The effluent
stream from the reactor is directed to a five-neck round bottom
flask supported in the constant temperature bath. This flask
contains pH and reference electrodes thus allowing continuous
monitoring of the effluent stream pH while maintaining reaction
temperature. From this point, the reactor effluent stream can be
sampled and the various analytical determinations performed.
3.1.2 Batch-Solid Reactor
The unique batch-solid crystallizer, designed and con-
structed for these kinetics experiments, is shown in Figure 3-2.
The reactor is based upon conventional 0-ring-flange design tech-
niques for ease of assembly and breakdown for cleaning. The
reactor is constructed primarily from inert Plexiglass material.
A 1/10 horsepower, A.C. variable-speed electric motor
(not shown in Figure 3-2) is mounted on the top plate of the
reactor and is connected to a Nalgene stirring rod via a 1/4"
universal joint. The stirring rod, in combination with the baf-
fled tube-reactor cavity, provide a continuously mixed suspension
of seed crystals in the supersaturated liquor. The stirring
speed can be controlled up to a maximum speed of 500 RPM.
Also mounted in the top plate of the reactor are the
two plexiglass inlet feed ports and the nylon Swagelok thermom-
eter port. The vacuum and pressure tight stirring rod bearing,
made from a 1/4" nylon Cajon male connector, is mounted on the
top plate along the longitudinal axis.
The solids are retained within the reactor cavity by a
Millipore filter membrane supported by a 1/4" thick perforated
Plexiglass substrate. The membrane-substrate combination is
sandwiched between the upper and lower portions of the reactor
by 0-ring-flange seals. The effluent stream thus exits from the
bottom portion of the reactor while seed and product crystals
remain in the upper portion. The volume of the reactor cavity
is approximately three liters.
3.2 Experimental Procedures
This section includes a descriptive presentation of
the preparation and experimental procedures developed specifi-
cally for the CaCOs and Mg(OH)a kinetics studies.
C-16
-------
3.2.1 Preparation of Feed Solutions
The CaC03 precipitation kinetics study uses two separate
feed solutions, CaCl2 and Wa2C03. For the Mg(OH)2 study, the
feed solutions are NaOH and M£C12 . The above feed solutions are
prepared and standardized for their respective concentrations
before starting each experimental run. The solids of each feed
compound are dissolved in a small volume (0.5 to 1.0 liter) of
deionized water. They are quantitatively transferred to the two
large polyethylene feed tanks, diluted with deionized water to
approximately 30 liters, and stirred to obtain a homogenous
solution. Contamination and oxidation are minimized by using
polyethylene floating lids and tank covers.
Analyses of calcium, magnesium and sodium are used to
determine the concentrations of their respective feed solutions.
A 200 ml sample from each feed solution is sufficient for all
analyses. Each sample is stored in a polyethylene bottle and
labeled with the necessary information such as: run number, date,
time, solution identification and operator's initials.
3.2.2 Preparation of Batch-Solid Crystallizer
The reactor vessel is washed and rinsed thoroughly
before assembly. A mild detergent and soft brush is used to
prevent excessive scratching of the Plexiglass reactor. Deionized
water is used to remove all traces of detergent and foreign
material.
The bottom portion of the reactor vessel is then filled
with deionized water to prevent the formation of air pockets
below the filter membrane. This will also insure a constant
outlet flow from the reactor. With the membrane support plate
positioned correctly on top of the bottom portion of the reactor,
a tared 142 mm Millipore filter membrane is then carefully placed
on the support plate. The top portion is placed in position and
reactor assembly is completed. A filter membrane with 0.8y pore
size is used for the CaC03 precipitation kinetics study; 0.45u
to 0.6u pore size is used for the Mg(OH)2 study.
Before placing the assembled reactor in the constant
temperature bath, the outlet stream line is connected. The
reactor is then filled to approximately one-half capacity with
deionized water and the remainder with feed solutions before
adding a predetermined amount of seed crystals with stirring.
C-17
-------
Inlet stream flowmeters are roughly adjusted by flush-
ing the system with approximately 1.0 liter of the feed solu-
tions. After flushing both inlet streams, samples for inlet
feed concentrations are taken. Inlet feed lines are attached
and the reactor is filled to capacity with equal volumes of feed
solutions, seed crystals are then added, and the system is then
placed in the bath. The stirrer blade is positioned approxi-
mately 3 cm above the filter membrane and is maintained at a
proper stirring speed by the variable control mounted on the
control panel.
While the reactor temperature is stabilizing, the 0.5
liter round-bottom five-neck flask is assembled and placed in
the constant temperature bath. This flask allows continuous
accurate monitoring of reactor effluent temperature and pH.
3.2.3 Experimental Run Procedure
When all preparations have been completed and the
charged reactor has reached constant temperature, the experiment
can proceed. The timing device and inlet feed stream pumps are
started with the inlet return valves opened completely. Inlet
stream flowmeters are adjusted as accurately as possible. Both
inlet streams and the reactor outlet stream are controlled by
identical flowmeters. The calibration graphs provided with the
Matheson Model 7641 flowmeters are referred to for the proper
settings. The Moore Model 63-SD constant differential pressure
flow controllers located in both inlet feed streams automati-
cally maintain the desired constant flow rate for the duration
of the experiment. However, continuous observation and manual
fine adjustments may also be necessary.
Initial samples are taken immediately after the system
has been properly adjusted. Refer to the sampling scheme for
the correct procedure. A three-way teflon valve located on the
control panel simplifies the actual sampling of the reactor
outlet stream. The determinations of (Ca"^"), (Mg++), pH and
temperature are performed during the actual experiment. The
remaining determinations of (Na+), (Cl~), alkalinity, and (C03=)
are performed upon completion of the experiment. The data
including technical observations are recorded in an experimental
log book. Samples are taken at predetermined intervals and
monitored until a steady-state precipitation rate is attained.
C-18
-------
Upon completion of an experimental run, the inlet feed
lines and stirring-motor assembly are disconnected from the
reactor. The remaining liquor above the filter membrane is
siphoned off by leaving the outlet reactor stream open. Deio-
nized water is used to wash down the stirrer blade, sides and
rim of the reactor in order to collect an accurate total of
product crystals. When the liquid level drops below the filter
membrane, the reactor is removed from the constant-temperature
bath and dissembled. The filter membrane plus product are dried
at 50-60°C, for 48 hours. Determination of the weight of pro-
duct crystals is performed by subtracting the weight of the
filter membrane and seed crystals from the total weight. Photo-
micrographs of the seed crystals and product crystals at compara-
tive magnifications are performed in order to determine crystal
size and growth characteristics.
3,2.4 Sampling Scheme for Experimental Run Procedure
The sample scheme for both precipitation kinetics
studies are identical with one exception. During the CaCCb
study, a separate sample for the (CDs") analysis is taken.
Polyethylene bottles for the collection of samples are
prepared in advance of an experimental run. Each bottle is
cleaned, dried, labeled and tarred before the addition of deio-
nized water which serves as the dilutant. It is then weighed
again to determine the exact amount of deionized water added in
order to accurately determine the dilution factor for that par-
ticular sample. Inlet feed samples are not diluted, but a
sufficient dilution factor must be approximated before inter-
mediate and steady-state samples are taken in order to maintain
a relative supersaturation <1.0 in the sample bottle.
3.2.4.1 Inlet Feed Solutions
Two clean 0.5 liter polyethylene bottles are filled
with the two feed solutions from the reactor inlet feed lines
after the system is adequately flushed. This amount of sample
insures that additional determinations can be made if necessary.
Each bottle is labeled with the run number, solution identifica-
tion, approximate concentration, date and operator's initials.
Determination of (Ca++) or (Mg++) and (Na+) is performed on
these samples after completion of the experiment.
A tarred 60 ml polyethylene bottle containing a known
amount of N1U-EDTA buffer solution is filled with the Na2C03 feed
solution from the reactor inlet feed line. The bottle is capped
C-19
-------
tightly and shaken. Before capping the bottle, it is reweighed
to determine the dilution factor. Each bottle is labeled with
the run number, sample identification, dilution factor, weights,
date and operator's initials.
3.2.4.2 Intermediate Diluted Filtrate Samples
At predetermined intervals during an actual experimen-
tal run, samples of the effluent are taken in order to determine
when steady-state is attained. A tarred 0.5 liter polyethylene
bottle containing a known amount of deionized water for quench-
ing purposes is filled to a specific total volume with sample.
The bottle is capped tightly and shaken immediately. The dilu-
tion factor is calculated and recorded on the bottle with the
run number, sample identification, elapsed time in minutes,
date, weights and operator's initials. The determination of
(Ca++) or (Mg"1"1") is performed on these samples using a direct
colorimetric titration method summarized in Section 3.3.
3.2,4.3 Steady-State Diluted Filtrate Samples
Upon reaching steady-state, diluted filtrate samples
are taken in the same manner as described in Section 3.2.4.2.
A minimum volume of 500 mis of diluted filtrate is sufficient
to perform the analyses of all major species in duplicate with
an adequate amount of sample left in reserve for additional
determinations if necessary. The methods of analyses for all
of the major species in both precipitation kinetics studies
are summarized in Section 3.3.
3.3 Analytical Procedures
This section describes the analytical procedures for
the determination of the principal ionic species encountered in
the precipitation kinetics studies on CaCOs and Mg(OH)2.
3.3.1 Determination of Sodium, Calcium and Magnesium
Samples taken from the inlet and outlet streams are
diluted with a lanthanum stock solution to control known inter-
ferences. With an accurately determined dilution step, the
samples are aspirated directly into the atomic absorption spec-
trophotometer for measurement of their respective concentrations.
C-20
-------
Intermediate samples taken during the experimental
runs are analyzed for calcium or magnesium to determine when
the steady-state condition is reached. This procedure utilizes
a colorimetric titration with diNa-EDTA and Calgamite indicator.
Analyses of these intermediate samples are repeated with atomic
absorption to verify the steady-state condition.
3.3.2 Determination of Chloride
Chloride present in the inlet and outlet streams is
determined by a manual potentiometric titration. The procedure
uses the millivolt scale of a pH meter to determine the end point
of the titration with 0.02M AgN03. A cup-type silver electrode
(Fisher No. 13-639-122) in conjunction with a silver-silver
chloride reference electrode with a sodium sulfate bridge (Fisher
No. 9-313-216) are used in this procedure.
3.3.3 Determination for Alkalinity
Hydroxyl ions present in the liquid samples by virtue
of the dissociation of solutes are neutralized by an electro-
metric titration with a standard acid. A Beckman Century SS-1
pH meter equipped with a calomel reference electrode and a
standard glass pH electrode are used in this procedure.
3.3.4 Determination of Total Carbonate
Carbon dioxide evolved from the reaction of a liquid
sample injected into a buffered acid pool is measured by a non-
dispersive infrared analyzer. The instrument allows samples to
be analyzed without introducing atmospheric carbon dioxide into
the system and removes harmful water vapor prior to reaching the
infrared cell. The amount of C02 in the injection ampule is
monitored by a recorder equipped with a disc-chart integrator to
measure the peak area. This peak area is used instead of the
peak height to accurately determine the amount of C02 present in
the sample because of pH fluctuations in the acid pool.
C-21
-------
4.0
EXPERIMENTAL RESULTS FOR CaC03
This section presents the results of the experiments
performed to determine CaCOa precipitation kinetics. First
the method used for correlating the data is described. Then the
actual data is presented in both tabular and graphical form and
discussed.
4.1
Kinetics Data Processing
For the batch-solid kinetics experiments, a steady-
state precipitation rate can be calculated from a straightforward
material balance. Steady-state material balances may be written
for the reactor in terms of either total calcium or total car-
bonate. That is:
inlet -
-------
the reactor composition approaches steady-state in a short period
of time compared to the total time duration of the run.
The CaC03 precipitation rate, R, determined for a par-
ticular run in the above manner, is expressed as a function of
the steady-state relative saturation. The relative saturation
is defined as the ratio of the reactor solution activity product
for the precipitating species, in this case, acot2 • aG032, to
the equilibrium solubility product, Ksp, for CaC03. Activities
for the particular solution species are calculated by inputting
pertinent reactor solution information, such as concentrations
of calcium, chloride, sodium, and carbonate, pH, and tempera-
ture, to the chemical equilibrium computer program. Reactor
solution relative saturations are controlled experimentally by
varying the mean reactor residence time or the inlet feed com-
positions from run to run.
A suitable rate expression for CaCOa solid precipita-
tion from supersaturated liquor may be written in the following
form:
R = k ' M ' <|> (4-2)
where: R = rate of solid precipitation,
k = rate constant, which may vary with liquor temperature,
composition, and transport parameters,
= driving force term related to the degree of
CaC03 supersaturation,
M = term dependent on the amount of solid phase
present.
The term, M, is usually assumed to be proportional to
the exposed surface area of the solid phase. This is obviously
difficult to quantify in experiments with suspensions of many
fine particles of seed crystals; therefore, no crystal surface
area measurements were attempted. Normally, as in this case of
CaC03 precipitation kinetics, the term is equated to the mass
of initial seed crystals and therefore has the units of "grams."
For dissolution and precipitation reactions, 4>, the
driving force term, is usually taken to be the difference be-
tween the actual and equilibrium quantities of the reacting
species, perhaps raised to some power. If one assumes a linear
C-23
-------
dependence of precipitation rate on the driving force function,
then equation 4-2 can be written as:
R(mMoles/tnin) = k(mMoles/gram-min) • M(grams) • (R.S.-l) (4-3)
where R. S. is the solution relative saturation as defined earlier.
In this study, the precipitation rate divided by the mass of seed
crystals (10.0 grains) is analyzed in terms of the solution rela-
tive saturation.
4.2 Results
Experimental results for the CaC03 system have been
summarized in Table 4-1. The reported precipitation rates for
these experiments are based on the total calcium material bal-
ance as derived in Section 4-1. The relative saturations were
calculated using the chemical equilibrium computer program as
described earlier. The precipitation rate for CaC03 (in mMoles/
gram-min) is plotted versus solution relative saturation in
Figure 4-1.
4.3 Discussion of Results
Previous investigators have developed the concept of
a metastable region of growth for many crystal systems. This
region is bounded by the equilibrium solubility product curve
for the particular precipitating species and a certain level of
supersaturation below which normal crystal growth will occur but
additional nuclei will not form. Since initiation of scaling
required nucleation, it is important to define the limits of
this metastable region for the CaC03 system.
From Figure 4-1, it is clear that the mechanism of
CaC03 crystal growth undergoes a radical change for solution
relative saturations greater than approximately 2.5. This abrupt
change in the CaC03 rate curve is attributed to an incipient
nucleation process in addition to normal crystal growth.
Photomicrographs comparing CaC03 seed crystals with
product crystals were used to detect the degree of formation of
new crystals or the onset of nuclei production during the kinetic
runs. These photographs suggested that the rapid rise in the
CaC03 precipitation rate at relative saturations greater than 2.5
is due to nucleation.
C-24
-------
TABLE 4-1. EXPERIMENTAL DATA - PRECIPITATION KINETICS OF CaCO
tun 1
1
2
3
4
5
O
1
to 6
7
a
9
10
11
Amount Feed Flow Rate
of Seed CaClj NajCOj
(e) (ml/rain) (ml/rain)
10.00 100 100
10.00 75 75
10.00 100 100
10.00 " i , ' "
10.00 " "
10.00 " "
10.00 " "
' v 10.00 " "
10.00 " " '
10.00 " "
10.00 H M
Peed Concentration
CaClz
(mKole/1)
0.750
1.400
1.400
1.390
1.480
1.560
2.930
2.000
0.767
2.605
0.213
NajC03
(mMole/1)
1.010
1.325
1.161
1.480
1.555
1.66S
2.815
2.010
0.795
2.280
0.205
Reactor
Temp.
30.00
30.00
30.00
30.00
30.00
30.00
30.30
30.10
30.20
30.10
30.00
Steady State
Effluent Cone.
Reactor
PH
9.61
9.60
9.58
9.84
9.84
9.85
9.84
10.42
11.11
9.65
11.17
Calcium
(nMole/1)
0.155 .
0.276
0.246
0.187
0.180
0.182
0.163
0.104
0.046
0.238
0.088
Carbonate
(mMole/1)
0.207
0.355
0.293
0.304
0.300
0.327
0.389
0.311
0.424
0.369
0.103
Relative
Saturation
Ksp
1.584
2.090
2.215
2.600
2.468
2.686
2.617
2.536
1.806
2.723
1.144
\ Precipitation
'* Rate
(mHole/g-aln)
.0044
.0064
.0091
.0102
.0112
.0120
.0260
.0179
.0067
.0213
.0004
-------
0.030
0.026
0.022
M
-------
For super saturations less than approximately 2.5, it
is clear that qualitatively the linear driving force function
appears to offer a good representation of the data. It is over
this linear region that the limits of normal crystal growth for
CaC03 are defined. That is, normal CaCO3 crystal growth without
nucleation is expected to occur in this linear metastable region.
4.4 Conclusions
From experimental measurements of CaCO3 precipitation
kinetics, we can state the following conclusions:
• A metastable crystal growth region bounded by a
solution relative saturation of approximately
2.5 times the solubility product is observed.
Below this level of supersaturation, precipitation
occurs primarily on existing seed crystals. For
solution supersaturations greater than approxi-
mately 2.5, nucleation begins to occur resulting
in rapid acceleration of growth rates.
The precipitation kinetics of CaC03 within the
limits of the metastable growth region may be
described by a rate expression of the form:
R = Rate (mMoles/min) = k • M • (R.S.-l) (4-4)
where K (mMoles/gram-min) is the rate constant
for the expression, M (grams) is the mass of
seed crystals, and R.S. is the relative saturation
defined by the ratio of the activity product to
the solubility product for the precipitating
species.
The dependence of the rate on the amount of seed
crystals present initially is not entirely clear.
Experimental attempts to demonstrate the effect
of the initial amount of seed on the rate did not
provide a complete description of the seed-dependent
term, M, in the rate expression.
C-27
-------
5.0 EXPERIMENTAL RESULTS FOR Mg(OH)2
This section presents the results of the experiments
performed to determine Mg(OH)2 precipitation kinetics. First,
the method used for correlating the data is described. Then the
actual data is presented in both tabular and graphical form and
discussed.
5.1 Kinetics Data Processing
For the batch-solid kinetics experiments, a steady-
state precipitation rate can be calculated from a straightfor-
ward material balance. Steady-state material balances may be
written for the reactor in terms of either total magnesium or
total hydroxide. That is:
(Fi ' Ci>inlet - outlet= R = Precipitation Rate (5-1)
where F- = flow rate in (Ji/min) ,
C- = concentration in (mMoles/5,) ,
R = rate of precipitation in (mMoles/min), and
i = magnesium or hydroxide.
Thus, equation 5-1 expresses the difference in the
rate of material entering and leaving the reactor, and this
difference is the rate at which solid material is produced in
the reactor, or the rate of precipitation. This precipitation
rate is calculated then by measuring the feed and effluent flow
rates and determining the concentrations by analytical means
for each kinetics experiment.
An additional check is provided by a total solids
material balance. That is, the amount of product solids for a
run is determined by emptying the reactor through the bottom
port following shutdown so that the solids are retained on the
filter membrane. After drying, the product cake is weighed and
compared to the amount of seed material introduced initially.
This mass difference is equal to the time integral of the ex-
perimental reaction rate multiplied by the Mg(OH)2 molecular
weight. This corresponds closely to the experimental steady-
state precipitation rate multiplied by the total run time, since
the reactor composition approaches steady-state in a short period
of time compared to the total time duration of the run.
C-28
-------
The Mg(OH)2 precipitation rate, R, determined for a
particular run in the above manner, is expressed as a function
of the steady-state solution relative saturation. The relative
saturation is defined as the ratio of the reactor solution
activity product for the precipitating species, in this case,
a^e+2. a2QH~» to the equilibrium solubility product, Ksp, for
MgtOH)2. Activities for the particular solution species are
calculated by inputting pertinent reactor solution information,
such as concentrations of magnesium, chloride and sodium, pH,
and temperature, to the chemical equilibrium computer program.
Reactor solution relative saturations are controlled experimen-
tally by varying the mean reactor residence time or the inlet
feed compositions from run to run.
A suitable rate expression for Mg(OH)2 solid precipita-
tion from supersaturated liquor may be written in the following
form:
R = k ' M ' <|> (5-2)
where R = rate of solid precipitation,
k = rate constant, which may vary with liquor
temperature, composition, and transport
parameters,
• = driving force term related to the degree of
Mg(OH)2 supersaturation,
M = term dependent on the amount of solid phase
present.
The term, M, is usually assumed to be proportional to
the exposed surface area of the solid phase. This is obviously
difficult to quantify in experiments with suspensions of many
fine particles of seed crystals; therefore, no crystal surface
area measurements were attempted.• Normally, as in this case of
Mg(OH)2 precipitation kinetics, the term is equated to the mass
of initial seed crystals and therefore has the units of "grams."
For dissolution and precipitation reactions, 4>, the
driving force team, is normally taken to be the difference
between the actual and equilibrium quantities of the reacting
species, perhaps raised to some power. If one assumes a linear
dependence of precipitation rate on the driving force function,
then equation 5-2 can be written as:
C-29
-------
R(mMoles/min) = k(mMoles/gram-min) • M(grams) • (R.S.-l) (5-3)
where R.S. is the solution relative saturation as defined
earlier. In this study, the precipitation rate divided by the
mass of seed crystals (10.0 grams) is analyzed in terms of the
solution relative saturation.
5.2 Results
Experimental results for the Mg(OH)2 system have been
summarized in Table 5-1. The reported precipitation rates for
these experiments are based on the average of the total magne-
sium and total hydroxide material balances as derived in Sec-
tion 3.1. The relative saturations were calculated using the
chemical equilibrium computer program as described earlier. The
precipitation rate for Mg(OH)2 (in mMoles/gram-min) is plotted
versus solution relative saturation in Figure 5-1.
5.3 Discussion of Results
Previous investigators have developed the concept of
a metastable region of growth for many crystal systems. This
region is bounded by the equilibrium solubility product curve
for the particular precipitating species and a certain level of
supersaturation below which normal crystal growth will occur but
additional nuclei will not form. Since initiation of scaling
requires nucleation, it is important to define the limits of
this metastable region for the Mg(OH)2 system.
From Figure 5-1, it is clear that the mechanism of
Mg(OH)2 crystal growth undergoes a radical change for solution
relative saturations greater than approximately 3.4. This abrupt
change in the Mg(OH)2 rate curve is attributed to an incipient
nucleation process in addition to normal crystal growth.
Photomicrographs of the Mg(OH)2 seed material wete
compared with product material from runs at low and high solu-
tion relative saturations. The occurrence of nucleation at high
solution relative saturation could not be-determined ftom these
representative photomicrographs due to the characteristic small
size (<_ l.Oy) of the Mg(OH)2 product crystals. Photomicrographs
of the crystals at higher resolution could not be obtained.
C-30
-------
TABLE 5-1. EXPERIMENTAL DATA-PRECIPITATION KINETICS OF Mg(OH)2
Run 1
1
2
3
4
i
4
7
8
Aaount of
Seed
(*•)
1O.O
10.00
10.00
10. OO
10.00
10.00
10.00
10.00
Feed Flow Kale
MgCl
(•1/aln)
75.0
75.0
75.0
75.0
75.0
75.0
75.0
75.0
Mai III
(•1/aln)
75.0
75.0
75.0
75.0
75.0
75.0
75.0
75.0
feed Com ent rat Ion
(aMole/l)
0.611
1.340
1. 740
2.215
0.896
0.6,;
0.711
0.452
NaOII
(•Mule/1)
1.190
2.940
2.8/5
1 665
1 II1!
1.485
1.425
0.871
Ked. iur
Temp.
10.11
29.9
10.0
lo.o
11). 0
l<>. (i
JO.O
10.0
Steady Stale
Ef fluent Cone.
Keactur
10.42
10.51
10 42
10.44
10.50
IO 50
10.50
10 45
Magltes lua
(•Mule/1)
0.217
0.257
0.1075
0.150
D.254
0.191
0.171
0.155
Hydroxide
(•Hole/1)
0.402
0.497
0.409
0.410
0.487
0.512
0.509
0.450
Relative
Saturat loit
Ksp
2.072
3.425
2.794
3.404
3.356
2.574
2.295
1.707
Precipitation
Kate
0.00128
0.00d45
U. 00818
0.01133
0.00292
0.00200
0.00215
0 . 00066
-------
0.016
0.014
0.012 •
0.010
_ oi
U. r-i
0.006
0.004
o.ooa
0.000
1.000 2.000
3.000 4.000
RELATIVE SATURATION
5.000 6.000
02-1783-1
Figure 5-1. Mg(OH)2 precipitation kinetics.
C-32
-------
For supersaturations less than approximately 3.4, it
is clear that qualitatively the linear driving force function
appears to offer a good representation of the data. It is over
this linear region that the limits of normal crystal growth for
Mg(OH)2 are defined. That is, normal Mg(OH)2 crystal growth
without nucleation is expected to occur in this linear metastable
region.
5.4 Conclusions
From experimental measurements of Mg(OH)2 precipita-
tion kinetics, we can state the following conclusions.
A metastable crystal growth region bounded by
a solution relative saturation of approximately
3.4 times the solubility product is observed.
Below this level of supersaturation, precipita-
tion occurs primarily on existing seed crystals.
For solution supersaturations greater than
approximately 3.4, nucleation begins to occur
resulting in rapid acceleration of growth rates.
The precipitation kinetics of Mg(OH)2 within the
limits of the metastable growth region may be
described by a rate of expression of the form:
R=Rate (mMoles/min)=k'M'(R.S.-1) (5-4)
where k (mtloles/gram-min) is the rate constant
for the expression, M (grams) is the mass of
seed crystals, and R.S. is the relative satura-
tion defined by the ratio of the activity
product to the solubility product for the pre-
cipitating species.
The dependence of the rate on the amount of seed
crystals present initially is not entirely clear.
Experimental attempts to demonstrate the effect
of the initial amount of seed on the rate did
not provide a complete description of the seed-
dependent term, M, in the rate expression.
C-33
-------
Appendix D. Ash Characterization for Colstrip and Montour Fly Ashes
1.0 INTRODUCTION
Five power plants have been selected for study repre-
senting typical situations in major geographical regions of the
United States. These five plants are: 1) Four Corners, Ari-
zona Public Service Co.; 2) Bowen, Georgia Power Co.; 3) Com-
anche, Public Service Co. of Colorado; 4) Montour, Pennsylvania
Power and Light Co.; and 5) Colstrip, Montana Power Co. This
appendix describes the results of bench-scale sluicing tests
performed with fly ash from the last two plants. The results
of similar studies performed on the ash from the first three
plants are presented in the final report for EPA Contract No.
68-02-1319, "Ash Characterization Studies", which was performed
in support of this program (see Appendix L).
1.1 Background
Recent emphasis on water recycle/reuse in the elec-
tric power industry has induced utilities to investigate the
feasibility of recycling water which has been used to sluice
coal ash. This system is known as a closed-loop ash sluicing
facility. The engineering involved in designing such a facility
necessitates the prediction of scaling potentials of CaCO 3,
Mg(OH)2 and CaSOif«2H20 so that the system can be designed to
control possible scaling problems. To predict scaling poten-
tials for these species, the dissolution characteristics of
the coal ash must be known. Therefore, it is important to
investigate the ash dissolution characteristics which will be
involved in such an ash handling facility. A bench-scale,
closed-loop ash sluicing facility was built to study the dis-
solution characteristics of the ash in a system of this type.
Measurements were made to determine the chemical composition
of the water at various locations in the system. The values
obtained will aid in the prediction of scaling potentials for
CaCO 3, Mg(OH)2 and CaSOi»-2H20 in closed-loop ash sluicing
facilities.
1. 2 Summary
Six experiments were performed using fly ash supplied
by PP&L and MFC. The first three were performed with ash from
the PP&L Montour Steam-Electric Station. The last three were
performed with ash from the MFC Corette Steam-Electric Station
as a substitute for fly ash from the Colstrip Steam-Electric
Station, which was not available because of the wet scrubbing
employed at Colstrip.
D-l
-------
The first experiment was performed with the fly ash
being sluiced near 1070 solids in a recycle system for about 50
hours. The pond recycle water comprised 84% of the sluice
water and simulated 2 cycle cooling tower blowdown was used as
makeup. The second experiment was performed with a 7% slurry
in the mix tank in a recycle system for about 30 hours. The
recycled water comprised 8870 of the sluice water and the makeup
water was of similar composition as was used in the first exper-
iment. The third experiment was performed under the same
conditions as those used in the second except the makeup water
simulated 8 cycle cooling tower blowdown from the Montour
cooling towers.
From these experiments it was learned that the reac-
tivity of this ash was less under these more realistic condi-
tions than was measured in batch dissolution studies using
deionized water. The reactivity of the ash is an important
parameter in determining the scaling potential of the slurry
in a wet ash sluicing system. These experiments also point out
that this ash does have the potential to be sluiced in a recir-
culating system without significant scale formation.
The last three experiments were performed with MFC ash,
and two cycle simulated cooling tower blowdown from the Montour
Station. Simulated Montour blowdown was used so that a compari-
son of ash reactivity between plants could be made and so the
effects of makeup water composition could be investigated. The
first two experiments performed with this ash were done for 30
hours with an 897o recycle, and about 770 solids in the slurry.
Carbon dioxide was bubbled into the ash pond in the second exper-
iment to simulate C02 transfer from the air to the pond. The
third experiment was performed on a once-through basis with the
same water quality used as makeup in the first two experiments.
These experiments showed that different ashes can
display great differences in reactitivity under similar condi-
tions. The MFC ash was much more alkaline producing pH's near
12 in the mix tank as compared to the more neutral pH's experi-
enced with the PP&L ash. The reactivity of the MFC ash was much
less under these experimental conditions than was found under
batch dissolution studies using deionized water.
These six experiments display more than anything that
ash sluicing system design must take into account a large number
of interacting factors. The reactivity of the ash is dependent
on the water quality of the sluice which is dependent upon the
percent recycle, the quality of the makeup and finally the reac-
tivity of the ash.
D-2
-------
2.0 EXPERIMENTAL
In this section a description of the six experiments
performed is presented. A description of the equipment used is
presented initially, along with the general approach used. This
is followed by a detailed description of the three experiments
performed with the fly ash from the Pennsylvania Power and Light
Montour Station. Finally, the three experiments performed with
the fly ash supplied by Montana Power Co. are described.
2.1 Technical Approach
A depiction of the laboratory scale ash sluicing
facility which was built to simulate a closed-loop ash handling
system is shown in Figure 2-1. Water from the settling pond
was pumped to the mixing tank, a 6-liter (1.6 gal) Plexiglass
cylinder where the coal ash was mixed with the sluice water.
The slurry formed was allowed to flow by gravity from the mixing
tank to the settling pond. The settling pond was constructed
of fiberglass and had a capacity of 454& (120 gal). The method
of gravity flow from the mixing tank to the settling pond was
adopted because dissolution occurs quickly in the mixing tank.
Therefore, the majority of dissolution occurs in the mixing tank
with only a minor fraction occurring in other portions of the
system. Batch dissolution studies indicate that the major por-
tion of the dissolution of the ash occurs within 15 minutes and
the mixing tank has a residence time of over 20 minutes.
Makeup water was fed into the mixing tank to replenish
water that was occluded with the sludge in the pond at 40 weight
percent solids. The chemical composition of this liquor varied
among runs to simulate the composition of actual makeup water
streams from the power plants studied (Montour and Colstrip).
The major portion of the sluice water was made up of pond water
recycled to the mixing tank.
The makeup water and the recycle were pumped with
peristaltic pumps. Rotometers were used to monitor flow rates.
The fly ash was fed into the mixing tank by a Model SCR-20
precision volumetric screw feeder manufactured by Vibra Screw,
Inc.
The liquor chemical compositions of the system must
be determined at steady state for values which can be effec-
tively used in a computer model of a closed-loop ash sluicing
facility. For the system to be at steady state, the chemical
D-3
-------
a
ASH STORAGE HOPPER
WORM SCREW FEEDER
i.
fr
TANK
-£X^
SLOWDOWN
SETTLING POND
OVERFLOW
Q - SAMPLE POINTS
Figure 2-1. Bench-scale simulation model of ash pond facilities
-------
composition of the liquor entering the mixing tank from the
settling pond must be constant, and the chemical composition of
the liquid flowing from the mixing tank to the settling pond
must be constant. The equation
7 =
-------
2.2 Montour SES
In Table 2-1 the operating conditions for the three
experiments performed with the ash supplied by Pennsylvania
Power & Light (PP&L) is presented. The feed rates of the
three streams that flow into the mixing tank are shown as well
as the residence time in both the tank and the pond at these
flow rates. The mixing tank has a volume of 6 liters, and the
average volume of the liquor above the sludge in the pond was
used to calculate the pond residence time.
The first experiment was carried out for about 50
hours to insure steady state operation. Because preliminary
results from this experiment indicated that a shorter period of
operation would also attain steady state, the second two experi-
ments were only performed over a 30 hour period. The volume of
the liquor in the pond did increase slightly over the time span
of the experiments. Since the first experiment was performed
over a much longer period the average volume of the pond liquor
was correspondingly larger.
The flow rate of the sluice water (makeup + recycle)
was the same for all three experiments. The ash feed rate and
the makeup water flow rate were larger in the first experiment
than in the other two. This caused the percent solids in the
slurry to be greater in the first experiment. The values re-
ported in Table 2-1 are the average measured values over the
length of the experiment. The makeup was calculated to be equal
to the amount of water that would be occluded with the sludge
in the pond at 40 weight percent solids.
Table 2-2 presents the composition of the makeup
waters used in these three experiments. These are measured
values that were obtained from sample analyses. The water
in the first two experiments was made to approximate cooling
tower blowdown from the Montour cooling towers at two cycles of
concentration. The water for the third experiment was made to
simulate eight cycle cooling tower blowdown from Montour.
The initial conditions of the pond water in each ex-
periment are shown in Table 2-3. The concentrations were
measured in the same manner that was used to measure the makeup
water. In the first two experiments the initial charge of pond
water was approximately the same as the makeup water. In the
third experiment the pond water from the second experiment was
used initially.
D-6
-------
TABLE 2-1. EXPERIMENTAL CONDITIONS (MONTOUR)
Experiment 1 Experiment 2 Experiment 3
Ash Feed Rate to
Mixing Tank, g/min 30 22 22
Makeup Water Rate to
Mixing Tank, mJl/min 45 35 35
Recycle from Pond to
Mixing Tank, mA/min 245 255 255
Percent Solids in
Slurry 10% 7% 7%
Mixing Tank
Residence Time, hrs .345 ,345 .345
Pond Residence
Time, hrs 6.8 6.1 6.1
Duration of Experiment,
Experiment, hrs 49 29.5 30
D-7
-------
TABLE 2-2. MAKEUP WATER COMPOSITION (MONTOUR)
Experiment 1 Experiment 2 Experiment 3
Chloride
(mg/£) 65.0 71.0 195.0
Sulfate
(mg/£) 129.0 131.0 416.0
Nitrate
(rag/A) 11.2 18.6 36.0
Sodium
(mg/£) 17.7 12.7 57.5
Calcium
(mg/O 51.3 46.9 174.0
Magnesium
(mg/£) 14.1 6.6 36.5
Carbonate
14.1 17.9 23.5
D-8
-------
TABLE 2-3. INITIAL POND WATER (MONTOUR)
Experiment 1 Experiment 2 Experiment 3
Chloride
Sulfate
(mg/£)
Nitrate
Sodium
(mg/JO
Calcium
Magnesium
(mg/A)
Carbonate
Volume,
liters
109.0
163.0
15.5
22.1
63.0
16.5
26.6
80.0
73.5
132.0
10.5
12.9
52.1
7.5
17.2
80.0
49.7
1680.0
6.8
85.1
613.0
24.2
12.6
80.0
D-9
-------
In summary, the first experiment was performed for 50
hours, with a larger percent solids in the slurry than was^used
in the other two experiments. The makeup water and the initial
pond water approximated the water quality found in two cycle
cooling tower blowdown at Montour. The second experiment was
performed with similar makeup and pond water compositions, but
with a lower ash feed rate, causing the slurry to have a lower
percent solids. The third experiment was performed with 8 cycle
cooling tower blowdown and the residual pond water from the
second experiment. The other operating conditions for the third
experiment were the same as those used in the second run.
2.3 Colstrip SES
In Table 2-4 the operating conditions for the three
experiments performed with the ash supplied by Montana Power
Company (MFC) is presented. The feed rates of the streams that
flow into the mixing tank are shown along with the residence
times of the pond and the tank at these flow rates. The mixing
tanks and the pond residence times were calculated in the same
manner that was done for the Montour experiments.
The first two experiments were performed under recycle
conditions for about 30 hours. The operating conditions were
exactly the same for both runs except that C02 was bubbled into
the pond in the second experiment in order to keep the pH of
the pond liquor near 7.5. This was done in order to study the
effect that C02 transfer in the pond had on the recirculating
system.
The third experiment was performed to simulate a
once-through ash sluicing operation. The percent solids in the
slurry was increased because the total water flow was decreased
and the ash feed rate was not changed. This was done in order
to see if a slurry of this ash near 107o solids could be sluiced
without scale formation. The pond water was contacted with C02
in order to maintain a pH near 7.7.
Table 2-5 presents the composition of the makeup water
used for these three experiments. These are measured values
that were obtained using sample analyses. The water used for
all three experiments simulated two cycle cooling tower blowdown
from the Montour station. This was done because ash sluicing is
performed as part of the S02 scrubbing at Colstrip and there
wasn't any representative stream composition from Colstrip to
use as makeup. The use of the Montour water allowed comparison
of the different ashes with similar makeup water composition.
D-10
-------
TABLE 2-4. EXPERIMENTAL CONDITIONS (COLSTRIP)
Experiment 1 Experiment 2 Experiment 3
Ash Feed Rate to
Mixing Tank, gm/min 15 15 15
Makeup Water Rate to
Mixing Tank, m&/min 30 30 150
Recycle from Pond to
Mixing Tank, mfc/min 255 255 0
Mixing Tank
Residence Time, hrs 0.35 0.35 0.66
Pond
Residence Time, hrs 6.1 6.1
Duration of
Experiment, hrs 30.5 30 8
D-ll
-------
TABLE 2-5. MAKEUP WATER COMPOSITION (COLSTRIP)
Experiment 1 Experiment 2 Experiment 3
Chloride
(mg/A) 32.0 32.0 28.8
Sulfate
(mg/£) 115.0 75.8 115.0
Nitrate
(mg/O 11.2 11.2 10.5
Sodium
15.2 11.0 11.5
Calcium
(mg/Z) 60.2 48.1 52.1
Magnesium
12.4 11.7 19.4
Carbonate
(mg/Z) 9.5 12.8 13.7
The initial pond water composition for each experiment
is presented in Table 2-6. For the first experiment the pond
was filled with 80 liters of the makeup water. For the second
and third experiments 80 liters of the pond water left from the
previous experiment were used as the initial pond water.
D-12
-------
TABLE 2-6. INITIAL POND WATER (COLSTRIP)
Experiment 1
Experiment 2
Experiment 3
Chloride
32.0
39.1
31.2
Sulfate
115.0
576.0
566.0
Nitrate
(mg/JO
11.2
10.5
12.4
Sodium
(mg/£)
15.2
21.9
52.9
Calcium
(rag/ A)
60:2
882.0
481.0
Magnesium
(rag/A)
12.4
0.0
0.0
Carbonate
(mg/£)
9.5
81.6
655.0
Volume,
liters
80.0
80.0
80.0
In summary, three experiments were performed with ash
similar to that produced at Colstrip. The first employed recycle
without CO 2 transfer in the pond. The second was operated under
identical conditions as were used in the first experiment except
CO2 was bubbled into the pond water. The third experiment was
a once-through operation using higher percent solids in the
slurry than was used in the previous two experiments.
D-13
-------
3.0 RESULTS
In this section the results of all six ash sluicing
experiments are presented. This includes information on the
water quality in the mix tank and the pond under steady-state
conditions. Mass balances were performed around the mix tank
and net dissolution rates of the leachable species were calcu-
lated. The results of the studies performed with the ash sup-
plied by PP&L are presented first followed by the results of
the studies with the MFC ash.
3.1 Montour
First the justification for steady-state operation
is presented using the results of Experiment 1. Then the final
samples are presented and the potential for scale formation is
looked at. The net dissolution rates in the mix tank and the
reactivities of the ash are compared for the three experiments.
Finally, the relevant conclusions concerning the results of
these experiments are drawn.
3.1.1 Steady-State Operation
The first experiment performed with the PP&L ash was
run for 50 hours to insure steady-state operation. Figures 3-1
through 3-7 present the concentration of the different key species
in the pond water as a function of time. The first four graphs
deal with the major species that were leached from the ash: so-
dium, calcium, sulfate, and magnesium. The other graphs deal
with three other species which are not normally leached from fly
ash to any significant degree: chloride, nitrate, and carbonate.
Figures 3-1, 3-2, and 3-3 include curves that were
fit to the data using Equation 3-1:
y - (y0 - yi)e"t/T + y± (3-1)
The.parameters y and y. were determined by plotting y versus
e 'T, where T = residence time of the pond (6.8 hours), and
calculating the slope and the y-intercept using a linear least
square regression. The results of these regressions are presen-
ted in Table 3-1. The values obtained for y and y- obtained
from the transformed plots (y versus e't/T) were then used to
calculate the curves presented in the original plots (y versus t).
D-14
-------
Ul
8.0
6.0
5
§ 4'°
m
2.0
10
20 30
TIME (MRS)
40
50
Figure 3-1. Sodium concentration in the pond vs. time,
-------
o
15.Oh
10.0H
2
2
o
_i
<
o
5.Ol-
I I
20 30
TIME (MRS)
40
50
Figure 3-2. Calcium concentration in the pond vs. time,
-------
30
20
m
_i
<
10
10
20
30
TIME (MRS)
40
50
Figure 3-3. Sulfur concentration in the pond -vs. time,
-------
o
I
oo
8.0
-------
3.0
a
O 2.5
o
a:
O
_i
x
0 2.0
1.5 •
10 20 30
TIME (MRS)
40
50
Figure 3-5. Chloride concentration in the pond vs. time
-------
o
I
.30
-------
to
0.4
J, 0.3
ill
<
e
< 0.2
0.1
10
20 30
TIME (MRS)
40
50
Figure 3-7. Carbonate concentration in the pond vs. time.
-------
TABLE 3-1. MONTOUR EXPERIMENT 1 REGRESSION RESULTS
, . ^ Coefficient of
slo?e 7- intercept ( } Determination
''
Figure 3-1
(Sodium) - 4.89 5.89 .996 .97
Figure 3-2
(Calcium) -10.9 12.7 1.85 .97
Figure 3-3
(Sulfate) -18.3 19.9 1.66 .99
These plots indicate that steady-state operation was
attained in the system with respect to sodium, calcium, and
sulfate concentration. They also point out that after 30 hours,
about 4% residence times, the system did not change significantly,
justifying the shorter (30 hours) operating period for the other
experiments.
Figure 3-4 is a plot of magnesium concentration in the
pond as a function of time. Although it is a leachable species,
the magnesium did not act as well behaved as the other leachable
species. The large, somewhat random variation in concentration
indicates that it did not reach steady state. Figures 3-5, 3-6,
and 3-7 are plots of the non-leachable species in the pond as a
function of time. The variation noted in these plots is most
probably due to sampling error caused by incomplete mixing in
the pond.
3.1.2 Chemical Analyses
For each experiment the results of the chemical analy-
ses from the final samples were input into the equilibrium pro-
gram. The program then calculated the relative saturations of
CaC03 and CaSOt. The relative saturation is a measurement of
the tendency of a particular solid to scale.
D-22
-------
Table 3-2 presents the chemical analyses as well as
the relative saturations calculated by the program. This table
includes the samples taken in both the mix tank and the ash pond.
These results show there was no tendency to form CaCO 3 scale in
any of these experiments. The low relative saturation of CaCO 3
in these samples was expected since all of the samples are acidic.
The relative saturation of CaSO i» was near one in all of the above
cases, but less than the critical scaling value of 1.3-1.4. The
concentrations of both calcium and sulfate did not change signi-
ficantly between the pond and the mix tank implying that CaSO^
precipitation was not occurring to any detectable extent.
3.1.3 Mass Balances
For each experiment, mass balances were performed
around the mix tank at different times in the experiment inclu-
ding the final samples. The difference between the amount of a
given species entering the tank and the amount leaving was assumed
to be leached from the ash. Table 3-3 presents the results of
these calculations for all three experiments.
These results confirm that in all three cases chloride,
nitrate, and carbonate were not leached from the ash. The size
and the fact that negative as well as positive rates are observed
indicate that the values represent measurement errors. The val-
ues reported for sodium, calcium, sulfur, and magnesium are sig-
nificantly larger and positive. This indicates that these spe-
cies were leached from the ash.
Using Equation 3-2, the weight percent of each species
leached from the ash was calculated:
W = -J x 100 (3-2)
r
where W = weight percent of species leached from the ash,
D = net dissolution rate of species (mmoles/min),
F = feed rate of the ash (mg/min), and
MW = the molecular weight of the species (g/mole).
The results of these calculations are presented in Table 3-4.
D-23
-------
TABLE 3-2. FINAL SAMPLE RESULTS (MONTOUR)
MIX TANK:
Sodium, mg/£
Calcium, mg/£
Sulfate, mg/£
Magne s ium , mg / £
Chloride, mg/£
Nitrate, mg/£
Carbonate, mg/£
Arsenic, mg/£
pH
Relative Saturation,
CaC03
if
Relative Saturation,
CaSCU-2H20
POND:
Sodium, mg/£
Calcium, mg/£
Sulfate, mg/£
Magnesium, mg/£
Chloride, mg/£
Nitrate, mg/£
Carbonate, mg/£
Arsenic, mg/£
pH
Relative Saturation,
CaC03
*
Relative Saturation,
Experiment 1
147.0
465.0
2054.0
44.2
83.8
17.4
3.0
0.148
5.6
7.7 x 10-"
.93
136.0
517.0
1950.0
24.3
88.8
15.5
14.4
0.118
6.0
2.0 x 10~3
1.0
Experiment 2
101.0
614.0
1843.0
30.6
76.7
15.5
8.4
0.120
6.9
2.5 x 10-2
1.10
85.1
589.0
1690.0
26.7
78.1
17.4
10.8
0.087
6.9
3.2 x 10~2
1.03
Experiment 3
145.0
690.0
1910.0
44.2
94.1
19.8
3.6
0.195
6.9
1.2 x 10'2
1.20
138.0
690.0
1900.0
43.7
85.2
20.5
3.6
0.171
6.9
1.3 x 10-2
1.19
Critical values, above which scale potential exists, are 1.3-1.
for CaSCK'2H20 and about 2.5 for CaC03 (see Appendix C).
D-24
-------
TABLE 3-3. NET DISSOLUTION RATES FROM MONTOUR FLY ASH
Species
Sodium, mmoles/min
Calcium, mmoles/min
Sulfate, mmoles/min
Magnesixim, mmoles/min
Chloride, mmoles/min
Nitrate, mmoles/min
Carbonate, mmoles/min
Experiment 1
(29.5 hrs) (49
.42
1.29
1.00 1
.51
.11
.02
-.02
hrs)
.38
.15
.17
.26
.01
.01
.05
Experiment 2
(29.5 hrs)
.31
.65
1.03
.09
.01
-.01
-.02
Experiment 3
(25 hrs)
.22
.33
.68
.01
-.01
-.01
-.01
(30 hrs)
.21
.45
.57
.01
-.03
-.01
-.01
-------
TABLE 3-4. MONTOUR FLY ASH REACTIVITY
Species Experiment 1 Experiment 2 Experiment 3
Sodium, wt. 7. 0.03 0.03 0.02
Calcium, wt. 7. 0.10 0.12 0.07
Sulfate, wt. 7o 0.35 0.45 0.27
Magnesium, wt. % 0.03 0.01 0.001
The differences in the values obtained for Experiment 1
and Experiment 2 can be attributed to the percent solids in the
slurry. In Section 2.1, it is shown that these two experiments
were operated very similarly except that the slurry was 107o sol-
ids in the first case and 7% solids in the second. The weight
percent of the calcium and sulfate leached from the ash was less
in the case with higher solids, but the same was not true for
sodium and magnesium.
The differences between Experiment 2 and Experiment 3
was the quality of the makeup water. In Experiment 3 where the
total dissolved solids were significantly higher in the makeup
water, and, therefore, in the slurry water, the ash was less
reactive. For all four species less was leached from the ash in
the third experiment. This implies that the reactivity of this
ash increases with the water quality of the leachate. This con-
clusion is further supported when the results of this study are
compared to the results of batch dissolution studies performed
with this ash, reported in Appendix K. In the batch
dissolution studies, the reactivity of the PP&L ash were examined
under deionized water at pH 6 and 8 and at 57o and 1070 slurries.
In every case, the reactivity of the:ash was greater under deion-
ized water than under the more realistic conditions employed in
this study.
D-26
-------
3.1.4 Conclusions
The results of the three experiments performed with
the PP&L ash lead to the following conclusions:
1) Steady-state behavior occurs in the
experimental equipment at the end
of three pond residence times.
2) Under the conditions studied, 84 to
88% recycle and 7 to 10% solids, this
ash can be sluiced without significant
scale formation. CaC03 scale does not
present a problem but the relative
saturation of gypsum is near the
critical range.
3) The reactivity of this ash is
dependent on the percent slurry and
the water quality of leachate.
Increased dissolved solids tends to
cause the reactivity of this ash
to decrease.
3.2 Colstrip
Three experiments were performed with the Colstrip
ash. The results of these three experiments are presented in
this section. This includes the chemical analyses of the final
samples, the relative saturations of CaC03 and CaSOi, for these
samples and the results of mass balances performed around the
mix tank to determine the reactivity of the ash. Finally,
general conclusions are drawn from the results presented.
3.2.1 Chemical Analyses
For each experiment the results of the chemical analyses
from the final sample were input into the equilibrium program.
The program then calculated the relative saturations of CaC03
and CaSOi,. The relative saturation is a measurement of the ten-
dency of a particular solid to precipitate.
D-27
-------
Table 3-5 presents the chemical analyses used as inputs
to the computer program and the calculated relative saturations.
Samples were taken from both the mix tank and the pond. These
results indicate that CaSO^ scale should not present a problem
under these operating conditions. However, CaCO3 scale seems
to represent a real danger. In all three cases in both the mix
tank and the pond, the relative saturation of CaCO3 exceeded its
critical scaling value of 2.5 (Appendix C).
The first experiment was performed without pH control
allowing the pH to rise to 12.6. This very high pH caused the
relative saturation of CaCO3 to reach a high value in the mix
tank where calcium was leached from the ash. In the pond, CaC03
precipitated causing the carbonate level to drop and decrease
the relative saturation from 7.8 to 3.7. In the second experi-
ment CO2 was bubbled into the pond maintaining a lower pH but
increasing the relative saturation in the mix tank and the pond
relative to the first experiment. In the second experiment, the
relative saturation in the pond fell relative to the tank even
though the calcium level remained the same and the carbonate
level increased, because the pH fell from 11.7 to 7.6.
In all three experiments, there was no detectable
level of magnesium in the liquid phase. This occurred because
the pH in the mix tank was always above 8. Therefore, any mag-
nesium leached from the ash or coming in with the makeup water
probably precipitated out in the form of Mg(OH)2- In the last
two experiments where the pond reached a pH of 7.6 no magnesium
was detected most probably because once the Mg(OH)2 solid was
formed in the tank, it settled with the sludge in the pond, and
did not redissolve.
3.2.2 Mass Balance
For each experiment mass balances were performed
around the mix tank at different times in the experiment inclu-
ding the final samples. The difference between the amount of
a given species entering the tank and the amount leaving was
assumed to be leached from the ash. Table 3-6 presents the
results of these calculations for all three experiments.
These results indicate that only calcium and sulfate
had significant net dissolution rates. Magnesium was expected
to be leached from this ash based on other studies (see Appen-
dix K) but as explained earlier the very high pH's occurring in
D-28
-------
TABLE 3-5. FINAL SAMPLE RESULTS (COLSTRIP)
Experiment 1
MIX TANK:
Sodium, mg/&
Calcium, mg/£
Sulfate, mg/&
Magnes ium , rag/ $,
Chloride, mg/&
Nitrate, mg/£
Carbonate, mg/£
PH
*
Relative Saturation,
CaCO,
3 *
Relative Saturation,
CaSO,, .2H20
POND:
Sodium, mg/£
Calcium, mg/£
Sulfate, mg/£
Magnesium, mg/n
Chloride, mg/s,
Nitrate, mg/£
Carbonate , mg/ £
PH
"V
Relative Saturation, '
CaC03
JL.
Relative Saturation,
CaSCU .2H20
27.6
1000.0
691.0
0.0
39.1
14.3
4.8
12.6
7.79
0.53
27.6
1240.0
595-0
0.0
39.1
14.3
2.4
12.6
3.67
0.51
Experiment 2
25.3
441.0
614.0
0.0
29.1
14.3
21.0
11.7
31.8
0.42
52.9
481.0
566.0
0.0
31.2
12.4
654.0
7.6
12.6
0.39
Experiment 3
3.9
922.0
451.0
0.0
29.5
10.5
10.8
12.8
16.9
0.30
4.6
377.0
413.0
0.0
29.5
9.9
1030.0
7.6
13.9
0.25
*
Critical values, above which scale potential exists, are 1.3-1.4
for CaSOif*2H20 and about 2.5 for CaC03 (see Appendix C) .
D-29
-------
TABLE 3-6. NET DISSOLUTION RATES FROM COLSTRIP FLY ASH
UJ
o
Species
Sodium, mmoles/tnin
Calcium, mmoles/min
Sulfate, mmoles/min
Magnesium, mmoles/min
Chloride, mmoles/min
Nitrate, mmoles/min
Carbonate, mmoles/min
Experiment
(15.5 hrs) (30.
-.11
1.40
.31
.03
.01
.03 0
0
1
5 hrs)
.02
.83
.44
.02
.01
.01
Experiment
(25 hrs) (30
.40
6.13
.14
-.01
0
0
-.98
2
hrs)
.29
.04
.30
.01
.02
.01
2.69
Experiment
(6 hrs) (8
-.05
2.96 3.
.50
-.12
0 0
0 0
-.01
3
hrs)
05
26
53
12
01
-------
the mix tank did not allow any magnesium to remain in solution
Chloride, nitrate and sodium were not leached to any significant
degree in these experiments. In both the first and the third
experiments, no significant change in the total carbonate level
was measured. In the second experiment, a substantial amount
of carbonate dropped out of solution in the tank. Even though
calcium carbonate was dropping out of solution in all three
experiments, the total amount of carbonate species was much
larger in the second experiment. The carbonate level in the
tank was higher in Experiment 2 because C02 was bubbled into
the pond and the pond water was recycled into the tank. In
Experiment 1, C02 was not bubbled into the pond and in Experi-
ment 3 the pond water was not recycled.
Using Equation 3-2, the weight percent of each species
leached from the ash was calculated.
W =
The results of these calculations are presented in Table 3-7.
TABLE 3-7. COLSTRIP FLY ASH REACTIVITY
Experiment 1 Experiment 2 Experiment 3
Calcium, wt. % 0.08 0.82 0.83
Sulfate, wt. % ' 0.24 0.14 0.33
Magnesium, wt. "L 0.0 0.0 0.0
In the first experiment the calcium concentration
remained very high because calcium carbonate precipitation was
controlled by the low carbonate concentration in the tank and
the pond (.08-. 04 mmoles/5.) . The high calcium concentration
inhibited the degree to which calcium was leached from the ash.
The amount of calcium leached in the second experiment was much
greater because the calcium concentration in the leachate was
D-31
-------
lower due to the supply of carbonate ion in the pond allowing
greater precipitation of calcium as CaCOs. In Experiment 3,
the calcium concentration was low due to the fact that there
was no recycle from the pond.
The sulfate concentration did not vary to the same
degree that calcium did. The total amount of sulfate that was
leached from the ash in this experiment was less than that which
was leached from the ash with deionized water. From other ex-
periments (see Appendix K) performed at pH's 4-8 with deionized
water, the weight percent of leachable sulfate ranged from .55
to .60 which is about twice as much as the values presented in
Table 3-7.
3.2.3 Conclusions
The results of the three experiments performed with
the MFC ash lead to the following conclusions:
1) This ash is very alkaline and therefore
causes CaCOs and Mg(OH)2 scale problems
that would not be seen with a less
alkaline or an acidic ash.
2) Increased C02 transfer in the pond
causes the pH to drop in the pond
and the tank with recycle, and
increase the amount of calcium
leached from the ash.
3) The reactivity of this ash decreases
with decreasing water quality.
D-32
-------
APPENDIX E
COMPUTER MODELS
1.0 INTRODUCTION
In all five power plant studies computer models of
the large water consumers at the plants were used to simulate
existing and alternative modes of operation. These models were
used to predict temperatures, flow rates, and compositions of
the important streams in the cooling tower, ash sluicing and
SOa scrubbing situations found at these power plants.
The purpose of this appendix is to discuss these
models in greater detail than is presented in the individual
plant studies (Appendices F-J). In Section 2.0 flowsheets and
descriptions of the individual models used in this study are
presented. Section 3.0 presents descriptions of the individual
subroutines used in these models and Section 4.0 describes the
basis of the chemical equilibrium program used to predict the
compositions of the liquid streams in the models.
E-l
-------
2.0 MODEL DESCRIPTIONS
The process simulation used in this study is a group
of computer programs for simulating aqueous inorganic chemical
processes. The programs include an executive system and a set
of equipment subroutines. The function of the executive system
is to interconnect the various units in the appropriate fashion
and control the sequencing of the computer operations.
The process units are interconnected by means of a
process matrix during an initialization phase of computer opera-
tions. In this phase, model input data are read into the machine,
the process matrix is used to define the processing scheme, and
each equipment box is initialized. Each processing unit is la-
beled by a number called an equipment number and the subroutine
designation. Each process stream is labeled by a stream number.
The process matrix used to define each processing
scheme is given on the first page of computer printout for each
simulation case. Each process unit is listed in the process ma-
trix. Input and output stream vectors are assigned to each unit
so that the interconnections specified in the process matrix
correspond to the interconnections of the process flow diagram.
The executive system also must be given the order in
which the process calculations are to be made. This order is
indicated immediately under the process matrix in the printout.
To execute the process calculations, the executive system takes
each subsequent equipment number from the order of calculations
and determines the subroutine name and the input and output
streams from the process matrix.
In this section descriptions of the models used to
simulate the ash sluicing, cooling tower and SOa scrubbing sys-
tems are presented. This includes flowsheets of the models, an
explanation of what each subroutine represents in the flowsheet
and descriptions of the actions required of the executive system.
In many cases the same model was used to simulate simi-
lar systems at different plants. In those cases only one des-
cription is presented since the simulations only differed in the
inputs to the system and not the structure of the system. The
headings for each subroutine include the names of the plants
where these models were employed.
E-2
-------
2.1 Ash Sluicing (Bowen, Montour, Comanche)
Two of the power plants studied (Bowen and Montour)
employed wet ash sluicing to dispose of the ash produced from
the combustion of the coal. One of the plants (Comanche)
sluiced bottom ash but used dry methods to dispose of the fly
ash collected in the electrostatic precipitator. The model
used to simulate the ash sluicing operations at these three
plants is discussed in this section.
Figure 2-1 presents the flow sheet of the ash sluic-
ing simulation used to simulate the ash sluicing operations
at Bowen, Montour, and Comanche. This flowsheet identifies
the input subroutines, used to calculate the initial streams,
the equipment subroutines, used to model individual pieces of
equipment, the overall system balance subroutine, and the
order of calculations. The input subroutines are shown as
circles, the equipment subroutines are represented by rec-
tangles , and the overall system balance is placed in the upper-
right hand corner. Descriptions of the individual subroutines
are presented in Section 3.0.
The order of process calculations is presented at
the bottom of Figure 2-1. The numbers are presented in the
order in which the corresponding subroutines are called by the
executive system. The order of calculations indicates that
the inputs are initialized by calling the input subroutines.
Then the equipment and system balance subroutines are called.
These subroutines specify the composition of all the streams
numbered in Figure 2-1.
Several assumptions were made in modeling ash sluic-
ing systems with this simulation. These include:
1) Ionic reactions taking place in the
liquid phase are rapid and thus at
equilibrium.
2) Solid-liquid equilibrium is achieved
in the ash pond, with the exception
of CaSCK which is allowed to remain
supersaturated.
3) Ash dissolution is essentially complete
before the slurry reaches the pond.
E-3
-------
M
FLY ASH 4
BOTTOM ASH15
FLY ASH SLUICE WATER 2
BOTTOM ASH SLUICE WATER 3
-^- 7 EVAPORATION
-—- 8 SLUDGE
-^ 16 EFFLUENT
-^ 9 FLY ASH SLUICE EVAPORATION
-^ 10 BOTTOM ASH SLUICE EVAPORATION
IILDTK3
0
|6>>
RATHDI
10
11
14
r
17
^-7 SLUICE WATER
VAPORIZED
»~8 SLUDGE
POND *— *
(NONE) |
-9 FLY ASH SLUICE
VAPORIZATION
| _J—*-10BTM ASH SLUICE
VAPORIZATION
ORDER OF PROCESS CALCULATION: 1. 2, 3, 4. 5. 6. 7. 0. 9. 10 *
Figure 2-1. Ash sluicing simulation flow scheme.
02- 1529- I
-------
4) All solids precipitation occurs in
reaction vessels or the pond. RATHD1
calculates nucleation amounts and
then precipitation rates based on
kinetic expressions.
5) Subroutine RATHD1 models nucleation
as an instanteous rate if the species'
relative saturation exceeds the critical
value. Nucleation is allowed such that
the various species' relative saturations
are returned to their respective critical
levels. At this point, no further nuclea-
tion is allowed.
2.2 Cooling Tower Model
Four of the power plants studied employed cooling
towers to dispose of waste heat from the condensers. Three of
the plants (Bowen, Montour and Colstrip) employed cooling systems
with hot side blowdown. The other plant (Comanche) took its blow-
down before the condenser and therefore had a much cooler blow-
down stream. The models of these two types of cooling systems
differed slightly and will be discussed separately.
2.2.1 Hot Side Blowdown at Bowen, Montour and Colstrip
Figure 2-2 presents the flow sheet of the cooling tower
simulation used to simulate the cooling systems at Bowen, Montour
and Colstrip. This flow sheet identifies the input subroutines,
used to calculate the initial streams, the equipment subroutines,
used to model the individual pieces of equipment, the overall
system balance subroutine, and the order of calculations. The
input subroutines are shown as circles, the equipment subroutines
are represented by rectangles, and the overall system balance is
placed in the lower-right hand corner. Descriptions of the indi-
vidual subroutines are presented in Section 3.0.
The order of process calculations is presented in the
lower-left hand corner. The numbers are presented in the order
in which the corresponding subroutines are called by the execu-
tive system. The order of calculations in Figure 2-1 indicates
that the inputs are initialized and the first approximation of
the cooling tower inlet water is calculated by CTGES. CLGTR1
then computes the outlet air rate and composition, the amount
of water evaporated, and the outlet water and drift compositions.
E-5
-------
w
I
/FLU
1 (Al
\^
asi\
R> r"
/CTGEiK
out i
•
1
/i
I GUESS) / —
,^—
V 2
-^
y
-/
CLGTR1
(COOLING
TOWER)
2
6
B ,^ OUTLET AIR
7 ^ DRIFT
a
^*
HLDTK3
(SUMMER)
/WTRINP\
JIRIII c.inir j 4 ^
\ ACID) /
\^4 S
^*- -^
ALKINP\
LIME OR
ODA ASHW
*vjL-/
5 ^
,-^M.
15
HLDTK3
(SUMMER)
CHMTRT
(SOFTENER)
1
I
0
10
11
— 13
^_^
14
DIVDR6
(TEE)
9
— 12
DIVDR5
(TEE)
8
9
' ••>
^ 11
16
CLRHTR
(CONDENSER!
13
AIR
CHEMICAL
WASTE
SLOWDOWN
MAKEUP WATER
ACID
ORDER OF PROCESS CALCULATIONS:
1, 2, 3, 4, 5. 6. 7, (12. 13, 8. 9, 10. 9. 1 1. 6. 7), 8
SOFTENING CHEMICALS
CTBAL1
(OVERALL
SYSTEM
BALANCE)
7
6
7
I
a
i
10
OUTLET AIR
• DRIFT
SLOWDOWN
• CHEMICAL WASTE
02-1270-1
Figure 2-2. Hot side blowdown cooling tower simulation flow scheme.
-------
Next, CTBAL1 computes the blowdown composition and flow The
convergence loop is entered through DIVDR5 which calculates
the circulating water composition. Calculations are performed
around the loop to CLGTR1 which calculates a new air rate and
drift composition. CTBAL1 then calculates a new blowdown stream.
At this point, the species concentrations of the blow-
down stream are compared to the previous values . If the differ-
ences in concentration of each species for consecutive iterations
are within the specified convergence criteria, the convergence
scheme is completed. If not, the cycle is repeated.
This model calculates the temperature, flows and com-
positions of all the streams which are numbered in Figure 2-2.
This model determines the amount of acid, if any, necessary
to control CaC03 scale. This model further determines if soft-
ening is required to control CaSOi»'2H20 scale and how much lime
is necessary.
This model has four major assumptions associated with
it. These include:
1) Equilibrium exists with respect to C02
and H20 in the atmosphere and cooling
tower exit water.
2) Ionic reactions taking place in the
liquid phase are rapid and thus at
equilibrium.
3) The temperature of the cooled water
stream approaches the wet bulb tem-
perature of ambient air within a
predictable range.
4) The compositions and temperatures of
the cooled water and drift streams are
equal.
The assumption involving the temperature of the cooled
water stream is a recognized design parameter in cooling tower
evaluation and gives a good approximation. The assumption con-
cerning the temperature and composition of the drift stream
should be very close to actuality as is the assumption in regard
to H20 gas-liquid equilibrium. The assumption with regard to
E-7
-------
equilibrium is conservative since the partial pressure of C02
in actual cooling towers tends to be greater than the equilibrium
value. The lower equilibrium concentration of carbonate species,
assumed in the model, causes the pH to be slightly higher in the
model than in actual operation. The higher pH causes the rela-
tive saturation of CaCO3 to increase more than the lowered car-
bonate species concentration causes it to decrease.
2.2.2 Cold Side Slowdown at Comanche
Figure 2-3 presents the flowsheet of the cooling tower
simulation used to simulate the cooling system at Comanche. This
flowsheet identifies the input subroutines, used to calculate the
initial stream, the equipment subroutines, used to model the indi-
vidual pieces of equipment, the overall system balance subroutine
and the order of calculations. The input subroutines are shown
as circles, the equipment subroutines are represented by rectan-
gles, and the overall system balance is placed in the lower-right
hand corner. Descriptions of the individual subroutines are pre-
sented in Section 3.0.
The order of process calculations is presented in the
lower left hand corner of Figure 2-3. The numbers are presented
in the order in which the corresponding subroutines are called
by the executive system. The order presented in Figure 2-3 in-
dicates that the inputs are initialized and the first approxima-
tion of the cooling tower inlet water is input by WTRINP. CLGTR1
then computes the outlet air rate and composition, the amount of
water evaporated, and the outlet water and drift composition.
Next CTBAL1 computes the blowdown composition and flow. The con-
vergence loop is entered through DIVDR5 which calculates the
circulating water composition. Calculations are performed around
the loop to CLGTR1 which calculates a new air rate and drift com-
position. CTBAL1 then calculates a new blowdown stream.
At this point, the species concentrations of the blow-
down stream are compared to the previous values. If the differ-
ences in concentration of each species for consecutive iterations
are within the specified convergence criteria, the convergence
scheme is completed. If not, the cycle is repeated.
This model calculates the temperatures, flows and com-
positions of all the streams which are numbered in Figure 2-3.
This model determines the amount of acid, if any, necessary
to control CaCO3 scale. This model further determines if soft-
ening is required to control CaSOi»«2H20 scale and how much lime
is necessary.
E-8
-------
16
5 OUTLET AIR
6 DRIFT
13 CHEMICAL WASTE
ORDER OF PROCESS CALCULATIONS:
I, 2, 3, 4, 5. 6, 7, (8, 9, 10, II, 12, 13, 14, 6, 7)
AIR 4
MAKEUP ..
WATEB
AGIO 3
8 SLOWDOWN
SOFTENING CHEMICALS 12
CTBAL
(OVERALL
BALANCE)
7
5 AIR
6 DRIFT
a SLOWDOWN
13 CHEMICAL WASTE
Figure 2-3.
02-1527-1
Process simulation scheme for Comanche
cooling tower system.
E-9
-------
The assumptions associated with this model are identical
to those presented in Section 2.2.1 for the cooling tower model
with hot side blowdown.
2.3 S02-Particulate Scrubbing
Two of the power plants studied (Four Corners and Col-
strip) employed wet scrubbing to remove S02 and particulates
from the flue gas. The design of the systems differ to a certain
degree and will be discussed separately.
2.3.1 S02-Particulate Scrubbing at Four Corners
Figure 2-4 presents the flowsheet of the scrubbing sim-
ulation used to simulate the scrubbing system at Four Corners.
The flowsheet identifies the input subroutines, used to calculate
initial streams, the equipment subroutines, used to model indi-
vidual pieces of equipment, the overall system balance subroutine
and the order of calculations. The input subroutines are shown
as circles, the equipment subroutines are represented as rectan-
gles and the overall system balance is placed to the side. Des-
criptions of the individual subroutines are presented in Section
3.0.
The order of process calculations is presented at the
bottom of Figure 2-4. The numbers are presented in the order in
which the corresponding subroutines are called by the executive
system. Once the inputs are initialized and the first approxi-
mation for the thickener overflow (Stream 15) is made, SYSTB4
computes the compositions and flow rates for stack gas and scrub-
ber effluent streams. Then calculations are performed in Boxes
8, 9, and 10. At this point, the composition of Stream 15 is
compared with the previously calculated composition for this
stream. If the differences in composition of each species for
consecutive iterations are within the specified convergence cri-
teria, this convergence scheme is completed. If not, then new
values for Stream 15 components are assigned and the cycle is
repeated.
Once this convergence is finished, the remainder of
the stream computations are performed. First, the scrubber cal-
culations (Equipment Boxes 11 and 12) are made. Next, the thick-
ener underflow composition is determined in Equipment'Box 13,
and the sluicing operation simulated in Equipment Box 14. The
composition of the scrubber recycle loop makeup (Stream 19) is
computed in Equipment Box 15. Finally, the fan and reheat re-
quirements are computed in Boxes 16 and 17.
E-10
-------
STACK 6AS
MAKEUP
WATER
(WTRMKP)
5
WATER
RECYCLE
IN
(WTRINP)
6
FLUE OAS
"I
TO POND
MAKEUP WATER
LIME
THICKENER OVERFLOW
1
18
30
1B
MATERIAL
BALANCE
(8Y8TB4)
7
STACK GAS
SCRUBBER SLOWDOWN
ORDER OF CALCULATIONS
I,2.3,4,6,e<7,8,e,l0)ll.l2.l3,l4,l6,l6,l7
Figure 2-4.
Four Corners scrubbing simulation
scheme (existing operations).
E-ll
-------
Several assumptions are inherent in performing this
simulation with the subroutines outlined above. These assump-
tions are enumerated below:
1) The stack gas is saturated with respect
to HgQ.
2) Equilibrium exists between COa in the stack
gas and liquor in the scrubber bottoms.
3) The scrubber bottoms and stack gas temperatures
are the adiabatic saturation temperature of
the flue gas .
4) The scrubber was modeled without allowing
solids precipitation to occur. However, dis-
solution of Mg(OH)2, Ca(OH)2, and CaS03-%H20
solids entering the scrubber was allowed.
This dissolution pertains to particulates
removed as well as slurry solids entering the
scrubber. The fraction of each solid species
that will dissolve in the scrubber is specified
by the user.
5) All oxidation was assumed to occur in the
scrubber.
6) In Subroutine SYSTB4, no CaSO^HzO, CaS03-%H20,
or CaC03 solids are allowed to form. This was
done to model the scrubber blowdown stream as
accurately as possible. Realistically, actual
conditions are somewhere between no precipitation
and solid-liquid equilibrium. The short resi-
dence time in the scrubbing loop and the low in-
ventory of precipitating solid crystals indicate
that the assumption of no solids formation in the
loop is adequate.
7) All solids precipitation occurs in reaction
vessels (Subroutines HLDTK3 or RATHD1) .
HLDTK3 assumes sol id- liquid equilibrium is
achieved. RATHD1 calculates nucleation
amounts and then precipitation rates based
on kinetic expressions.
E-12
-------
8) Subroutine RATHD1 models nucleation as an
instantaneous rate if the species' relative
saturation exceeds the critical value.
Nucleation is allowed such that the various
species' relative saturations are returned
to_their respective critical levels. At
this point, no further nucleation is allowed.
9) Ionic reactions taking place in the liquid
phase are rapid and thus in equilibrium.
A different model was used to simulate the alternatives
at Four Corners. In Figure 2-5 a flowsheet of the model used
for Alternative 3 is presented. A description of this model
only is presented because it is the most complicated alternative
model and contains most of the features of the other models.
The order of process calculations is somewhat altered
in this simulation compared to the preceding one. The three
input routines, FLUGS1, ALKINP, and WTRMKP, again initiate the
computations. At this point, Subroutine WTRINP provides an ini-
tial estimate of the flow rate and composition of the ash pond
overflow stream (Stream Number 13).
Next is the beginning of the first convergence routine.
Subroutine SYSTB4 performs overall material and energy balances
in computing the FILTER underflow stream (Stream Number 5).
Then, EVAPND and FILTR2 are used to model the ash pond. These
three routines (SYSTB4, EVAPND, and FILTR2) are repeated until
the compositions in Stream Number 13 are consistent for consec-
utive iterations. After the second iteration, a convergence
scheme is implemented to facilitate this convergence.
Once the ash pond overflow stream flow rate and compo-
sition have been determined, the overall material balances are
correct. Then the next convergence loop is entered. Computa-
tions are repeated in the slurry recycle loop (Subroutines
SCRUBS, RATHD1, and DIVDR2) until the scrubber feed stream com-
position has converged.
The assumptions inherent in the use of this model are
identical to those presented for the existing operations model
used for the Four Corners system.
E-13
-------
STACK GAS
/ PLUGS
llFLUE Q
LKINP \
1 1
14
1 \ 2
1
SCRUBS
7
OIVDR2
B
SUMMER
12
*-
13
— 10
WTRINP
iS
* 1
8
i i
\ <
RATHDI
e
B
DIVDER
6
\
7
FILTER
6
1
5
EVAPND
1 0
I
11
K
FILTR2
11
— 6
/ WTRMKP \
9 I (MAKEUP I
V WATER) /
EVAPORATIC
15^
POND 18 .^
RECYCLE ~*
8YSTB4
STACK QAS
CLARIFIER UNDERFLOW
12
SETTLED
SOLIDS
ORDER OF PROCESS CALCULATIONS
1,2.3.13(4,10,II.16,12,«t7,8,0)
Figure 2-5. Process model for Four Corners Alternative Three
E-14
-------
2.3.2 SQ2 •» Particulate Scrubbing at Colstrip
Figure 2-6 presents the flowsheet of the scrubbing
simulation used to simulate the scrubbing system at Colstrip.
The flowsheet identifies the input subroutines, used to calcu-
late initial streams, the equipment subroutines, used to model
individual pieces of equipment, the overall system balance sub-
routine, and the order of calculations. The input subroutines
are shown as circles, the equipment subroutines are represented
by rectangles and the overall system balance is placed in the
lower right hand corner. Descriptions of the individual sub-
routines are presented in Section 3.0.
The order of calculations is presented at the bottom
of Figure 2-6. The numbers are presented in the order in which
the corresponding subroutines are called by the executive system.
The order indicates that the inputs are initialized and overall
balance calculations are performed first. Then, iterative cal-
culations are performed around the scrubbing loop (Boxes 10,
11, and 12) until calculated rates are satisfied for the input
recycle tank volume. Once this convergence is achieved, calcu-
lations around the effluent tank are performed and reheat and
fan requirements are calculated.
The assumptions inherent in using this model are iden-
tical to those listed in Section 2.3.1 for the simulation of the
Four Corners scrubbing model.
E-15
-------
STACK
QAS
)
8
DIVDR3
7
14 •»
^ 13
T
DIVDER
9
i10
RATHLD
13
(HOLD TANK)
1"
ASPND1
6
(POND)
-
~T
18
-POND EVAPORATION
12
SETTLED I
SOLIDS
ADDITIONAL 5
MAKEUP •
(NOT USED)
STACK QAS
SETTLED
SOLIDS
POND
EVAPORATION
ORDER OF PROCESS CALCULATIONS:
' 1, 2, 3. 4. 5. 6. 7, 8. 9. 10. 5. 6. 7. 8. 9, 16, (10, 11. 12.) 13. 14, 15 *
02-1265-1
Figure 2-6. Colstrip scrubbing simulation flow scheme
E-16
-------
3.0 SUBROUTINE DESCRIPTIONS
In this section descriptions of all the subroutines
used to model the water systems in this study are presented.
These include three separate classes which will be described
separately. Section 3.1 is devoted to input subroutines, Sec-
tion 3.2 discusses the equipment subroutines, and the system
balance subroutines are covered in Section 3.3.
3.1 Input Subroutines
The ALKINP subroutine enters any of the following
species into the simulation system: CaO, MgO, CaC03, MgC03,
CaS03> MgS03, CaSO^, MgS04> Ca(OH)2, Mg(OH)2, and inerts.
ALKINP is used in conjunction with a hold tank or chemical
treatment routine to add a solid stream of these compounds to
the simulation flow scheme. In order to specify this solid
stream, ALKINP requires the total flow rate and the weight frac-
tions of the solid species. ALKINP was used in the cooling
tower models as a source of chemicals to the system in cases
where softening was required to prevent CaSO^ZHaO scale. It
was also used in the scrubbing model for the Four Corners plant
as a source of alkali additive.
The subroutine ASHINP originates a solid ash stream
with specified flow rate and composition. Composition is input
as soluble weight fractions of CaO, CaSO^, MgO, MgSO^, MgCl2,
Na20, and NaCl. The insolubles in the ash are all input as
inerts. In the ash sluicing simulations, ASHINP is used to
originate both the bottom ash and fly ash inputs.
The CTGES subroutine is a subroutine, used only in
cooling tower simulations, which generates an initial guess of
the composition and temperature of the process water entering
the cooling tower. This is done by multiplying the concentra-
tion of the ionic species in=the makgup water by the cycles of
concentration, except for C03 and SO^. The C03 is specified as
that which is in equilibrium with the atmosphere. The SO., is
the value necessary to attain an input pH, thereby allowing
the subroutine to take acid addition into account.
The FLUGS1 subroutine is a general flue gas stream
simulation routine. It provides an input stream based on spe-
cified values of gas properties and fly ash properties The
gas properties are temperature, pressure, flow rate, and mole
E-17
-------
fraction composition. Fly ash properties include flow rate and
weight fraction composition with respect to CaO, CaSOi,, MgO,
Na20, NaCl, and inerts. FLUGS1 was used in all of the scrubbing
models studied. This subroutine was also used in the cooling
tower models to initiate the inlet air stream. No fly ash was
included in the cooling tower simulations.
There are two water input subroutines, WTRMKP and
WTRINP. The concentration of the major ionic components which
normally occur in water are specified by feoth routines. jThe _^
ions in solution may include S03, SO,,, C03, NO^, Cl~, Ca , Mg ,
and Na . The difference is that the liquid flow rate of the out-
put stream from WTRINP is input by the user, whereas the flow
rate associated with the WTRMKP output stream is computed by the
material balance within the simulation. WTRINP is used in all
of the models used in this study. It was used to specify the
makeup water in the ash sluicing model, the acid stream in the
cooling tower model, and to estimate the recycle streams in the
scrubbing models. WTRMKP is only used in the cooling tower and
scrubbing models. In both cases it is used to specify the com-
position of the makeup water to the system.
3.2 Equipment Subroutines
The pond system at Colstrip was modeled by ASPND1.
The solid waste weight fraction solids and the pond evaporation
rate are specified as inputs to this subroutine. This subrou-
tine calculates the composition of the recycle from the ash pond.
The CHMTRT subroutine is used only in the cooling
tower models^ This subroutine is used in conjunction with CTBAL1.
to remove Ca from the cooling cycle by means of chemical treat-
ment. The subroutine simulates the operation,of a number of
treatment options which can remove either Ca , or Mg and Ca"*"1".
CHMTRT is set up to model side-stream treatment of the recircu-
lating cooling water and requires an initial guess of how much
water is diverted for treatment. Using the input of the required
calcium removal rate determined in CTBAL1, this initial treat-
ment water flow rate is adjusted to give the desired calcium re-
moval rate for the specified treatment option. The treatment
options include the following:
1) Lime treatment for Ca and Mg removal.
i I
2) Lime treatment for Ca removal.
E-18
-------
3) Lime-soda ash treatment for Ca4"4" and Kg*4" removal.
4) Lime-soda ash treatment for Ca4"4" removal.
5) Caustic soda treatment for Ca4"4" and Kg4"4" removal.
6) Caustic soda treatment for Ca4"^ removal.
7) Sodium zeolite treatment for Ca4""1" and Kg4"4" removal,
Stoichiometric factors for chemical addition must be
specified. The weight fraction solids in the chemical addition
stream and the waste stream must also be specified.
Output information from CHMTRT includes the rate and
composition of the chemical addition stream, the adjusted flow
rate for entering water, the composition and flow rate for
treated water, and the composition and flow rate of the waste
stream.
The CLGTR1 subroutine simulates the operation of a
wet cooling tower. The following data are required as input
information for this routine: complete specification of ambient
air, specification of drift rate, complete specification of cool-
ing water entering the cooling tower, specification of ambient
wet bulb temperature, specification of the temperature approach
of the cooling water leaving the cooling tower to the ambient
wet bulb temperature, and specification of the relative satura-
tion of the exit air with respect to water. CLGTR1 uses this
information in heat and material balance calculations to deter-
mine the amount of water evaporated. Knowing this in turn allows
the cooling tower exit air to be completely determined. The
composition of the water stream leaving the cooling tower, which
is also assumed to be the composition of the drift, is then
found.
The CLRHTR subroutine serves the function of a heat
exchanger in the cooling tower and scrubbing models. It changes
the temperature of liquid and gaseous streams and performs en-
thalpy calculations to determine the heat duty. In the cooling
tower models CLRHTR is used to simulate the condenser. In the
scrubbing models it is used to simulate the reheat required to
send the flue gas up the stack.
E-19
-------
Four subroutines DIVDER, DIVDR3, DIVDR5, and DIVDR6
simulate process "tees" and are used to split streams. DIVDER
requires complete information about one effluent stream and the
flow of the other effluent stream to calculate the flow rate
and composition of the feed stream. DIVDR3 calculates the flow
and composition of one effluent stream from complete information
about the feed stream and the flow of the other effluent stream.
DIVDR5 requires complete information about one effluent stream
and the flow rate for the feed stream in order to entirely spe-
cify the feed stream and the second effluent stream. DIVDR6 re-
quires the input of a completely specified feed stream along with
the flow rate of the first effluent stream in order to completely
specify both effluent streams.
Subroutine EVAPND models an evaporation pond. Based
on the area of the pond and certain climatological data, this
routine will compute the evaporation rate and the equilibrium
composition of the material remaining in the pond. EVAPND is
used in the scrubbing models.
Subroutine FILTER models a solid-liquid separator
(i.e., a clarifier or vacuum filter). The filter bottoms stream
must be completely specified and the filtration efficiency must
be provided. With these inputs FILTER computes the flow rate
and compositions of both the filter feed and the filter overflow
streams. An assumption that complete solid-liquid equilibrium
is reached in the separation device is made when Subroutine FILTER
is used. FILTER is used in the scrubbing models.
The HLDTK3 subroutine simulates an equilibrium reaction
vessel. All of the input streams must be specified with respect
to composition and flow rate. HLDTK3 will then use this infor-
mation to generate an output stream. Mass balances and the equi-
librium program are used to calculate an output stream with spe-
cies in solid-liquid equilibrium. HLDTK3 is used in the cooling
tower and ash sluicing models.
PMPFAN is a subroutine which calculates the fan require-
ments to blow the stack gas up the stack. This subroutine is used
in the scrubber models.
RATHLD was used in the scrubbing model at the Colstrip
Plant. Process tanks in the system were modeled by RATHLD.
Both the scrubber recycle tank and the effluent tank where the
E-20
-------
scrubber blowdown is diluted from 12% to 6% solids were included
Solid precipitation rates were calculated based on the tank
volume, slurry flow rate, and precipitation kinetics data for
CaS03*%H20 and CaS04«2H20. Detailed documentation of the sub-
routine RATHDl is presented at the end of this section. RATHD1
was used to model reaction vessels and was developed specific-
ally for this study.
The SCRUBS subroutine is used to model a concurrent
contactor. SCRUBS is used in the scrubbing models to simulate
the venturi scrubber. It performs the appropriate vapor-liquid
mass transfer calculations based on specific S02, C02 , and par-
ticulate sorption efficiencies and H20 vaporization rate. The
scrubber gas and slurry effluent streams are computed, by this
subroutine.
The subroutine SUMMER simply sums the input streams
to calculate an output stream. Flow rates of all species are
computed in this routine. SUMMRl calculates one of the input
streams using complete composition and flow information about
the other input and output streams. Both of these subroutines
are used in the scrubber simulations.
E-21
-------
RATHD1: Rate Hold Tank 1
Function;
This routine simulates a process hold tank with up to
four input streams. Complete information about the input streams
and hold tank volume must be provided. A surface area- dependent
solid-liquid mass transfer rate determination is employed and
CaS03'%H20, CaSO^HaO, and CaC03 solid precipitation rates are
calculated.
Input Information:
Input stream compositions and flow rates as well as
the tank volume must be supplied. The inlet crystal sizes for
CaC03> CaS03«%H20, and CaSO^HaO may also be specified. If
these areas are not specified they will be calculated.
Output Information :
The output stream will be completely specified. The
average solid crystal area and the nucleation and precipitation
rates for the appropriate species are also computed.
Description:
First, the input parameters are converted from input
units to program units.
The next step in these calculations is to determine
the molar rates of the key species in the effluent stream (n
jt = all key species, gmole/sec) . This calculation is per-
QS
'
formed by subroutine ADDER, which adds the molar rates of the
key species in the input streams (nq . for 8 = 1-, through I/)
o , j t , j. 4-
nOS,jt
ns,jt
S-
for jt = all key species. Subroutine ADDER also determines the
solids flow rates into the rate hold tank.
E-22
-------
nS,js (2)
S- IX
for js = CaS03-%H20, CaS04'2H20, and CaC03.
ADDER also calculates an average inlet particle size
for each of the precipitating species. (This calculation is
based on a mass average). If the inlet crystal area is zero,
an area of 7500 cm2/gram is set. If an inlet crystal area is
specified in RATHD1 this value overrides the calculated value.
The subroutine next calculates the nucleation rates.
First, to determine which species may nucleate, subroutine
EQUILB is called allowing supersaturation. Evaluation of the
relative saturations of CaC03, CaS03*%H20, and CaSOit'ZHzO
determines if nucleation will occur. Specifics concerning
these evaluations are provided in the subsequent discussion.
The parameter APD (Activity Product Divisor) is set
according to the relative saturation results. The Activity
Product Divisor is used in the equilibrium program in calcula-
ting equilibrium compositions. For example, one of the simul-
taneous equations that is solved by Subroutine EQUILB is shown
in Equation 3.
= o2
a HfeO T,
1 ^SPtCaSOu^HzO)
In this equation the a's represent the activities of the various
species and KCD/r on ou nN is the solubility product constant
for calcium SPCCabO^-/H2O) sulfate dihydrate. Similar expres-
sions can be written for each precipitating species. The para-
meter APD had different values depending on whether a particular
species is subsaturated or supersaturated.
The tests performed in checking for nucleation are
as follows :
E-23
-------
1. If the relative saturation of a species is
less than or equal to one, the APD^ parameter
is not set in RATHD1 and remains equal to one.
This makes that species an equilibrium species.
2. If the species' relative saturation is between
one and the critical value for the onset of
nucleation, APD. is specified to be a large
number (i.e., l600 times Kgp). This allows
the relative saturation of a metastable solid
to be unrestricted and no nucleation occurs.
3. If the species' relative saturation exceeds the
critical value, APD. is set equal to the criti-
cal relative saturation. This allows the solids
to precipitate (nucleate) when subroutine EQUILB
is called.
Should conditions for nucleation exist, subroutine
EQUILB is employed to compute the rate of nucleation for each
nucleating species and the solid and liquid species distribu-
tion. (Nucleation is modeled as an instantaneous rate and, as
such, is independent of hold tank volume.) At this point, an
average surface area per gram of solid is computed by the fol-
lowing equation.
(SAPG.s)(XNN.g) 4- (SAGS.g)(XNS.s)
SAGjg = (3)
where
js = the nucleating species, (CaC03, CaS03«%H20
and CaS04'2H20)
SAG = surface area per gram of solid after nuclea-
tion has occurred
SAPG = surface area per gram of nuclei
SAGS = surface area per gram of solid entering the
rate hold tank
E-24
-------
XNN = mass of the nuclei
XNS = mass of the solid seed entering the hold tank
Once nucleation has been computed, precipitation of
solids from the metastable supersaturated liquor is calculated
The form of the rate expression is shown in Equation 4.
Rjs = kjs SAGjs Wjs mOS,js PH*0'1)
' V (IIajs - KSP,js> / 1000-°
The quantities used in Equation 4 are defined as follows.
R. = precipitation rate (gmole/sec) of solid js,
i.e., nos>js - nljs.
k. = precipitation rate constant (gmole/sec cm3)
-'s for solid js.
MW. = molecular weight (g/gmole) of solid js.
J s
mnc, . = molality of solid js in the output stream
Ub'JS (gmole/kg liq H20) .
pH20(l) (Tnq) = density of liquid water (g/cm3) at the
temperature of the output stream (TQS, °K) .
V = hold tank volume (cm3) .
Ila, . = activity product of the ions which form
Js solid js, for example, fCaSOlt.2H20 = aOS,Ca
* aOS,SO= ' UOS,H20(1); '
K . = solubility product constant for solid js,
or , J S
1000.0 = conversion factor (g/kg) .
The parameter VRK (VRK = k. • V • SAG.g/1000), a
volumetric rate constant, is set in JfaSubroutineJ RATHD1 and
rates are calculated in Subroutine RATE. The actual mechanism
for determining the precipitation rates is to calculate the
E-25
-------
total amount of solids leaving the rate hold tank (by solving
a set of simultaneous equations) and subtracting the quantity
of solid seed crystals which enter the hold tank plus the amount
of solids formed by nucleation.
Prior to the calculation of rates , two checks are
performed. First, any species which is subsaturated is treated
as an equilibrium species. Secondly, if no solid seeds are
available for crystallization, no precipitation is allowed
(APD. is set to a high value).
Once precipitation has been calculated, solid crystal
area is computed according to Equation 5.
(XNNjg + PNMjs)(ANOjg) + (XNSjg + PSM. g) (ASOj g)
SAG. = - (5)
J (XNI.s + PNMjs + PSM.s)
where
js = solid species
XNN = mass of nuclei
PNM = mass of solid precipitating on the nuclei
XNS = mass of seed
PSM = mass of solid precipitating on the seed crystal
SAG = area of solid exiting the hold tank
ANO = area per gram of the nuclei exiting the hold tank
ASO = area per gram of the seed exiting the hold tank
Subroutine RATHD1 models a well-mixed hold tank. The
outlet parameters (compositions, area, etc.) are assumed to be
uniform throughout the vessel. Since outlet areas are required
to compute the rates, the rate calculations must be repeated
until the outlet surface area (for each precipitating species)
agrees to within 0.0170 for consecutive iterations. When the
areas are converged, the calculations in Subroutine RATHD1 are
complete.
E-26
-------
^COMMON DATA FOR ALL SUBROUTINES
@STREAM DATA
@STREAM CONNECTION ARRAY
^EQUIPMENT NAME
@ORDER OF PROCESS CALCULATIONS
@EQUIP,PARAMETERS
@FLAGS FOR THE RECYCLE LOOPS
^CURRENT INDEX
(§START AND END OF RECYCLE LOOPS
,DNAT(9),XA(4),LABEL(13),AH20,XO
CKS(IO),NHY(10),ESK(10,10),ISK(10)
,ANO(11)
,CLC(11),SAPG(11),PNM(11)
40
SUBROUTINE RATHD1(P,$)
INCLUDE CMMN, LIST
COMMON
*SV(30,140),
*ISTM(25,10),
*IDEQP(25),
*ISEQ(30),
*PA(25,24),
*L(10),
*NL,
*LSRL(10),LERL(10),
*WV,XLSU(2),XSO(8),XLD(8)
DIMENSION P(24)
END
COMMON/PAGE/LINE
COMMON/SOLIDS/LOCSDS,
*APS(10)
COMMON/APDRVR/APD(11)
DIMENSION XNS(ll),XNN(11),SAG(11),SAGT(11),Z(11)
*IFLAG(11)
DIMENSION RS(2,11),C
*XNI(11)
DIMENSION ZSI(ll),PSM(11),ZS(11),ASO(11)
DATA CV(3)/2.5/CV(5)/3./CV(7)/1.3/
DATA VI(3)/11.808/VI(5)/20.048/VI(7)/13.793/
DATA CLC(3)/1.150E-4/CLC(5)/1.591E-4/CLC(7)/1.159E-4/
DATA SAPG(3)/1.02E5/SAPG(5)/1.26E5/SAPG(7)/1.19E5/
DIMENSION PP(2), VRK(IO)
LOCSL=ISEQ(NL)
LIS1=ISTM(LOCSL,1)
LIS2=ISTM(LOCSL,2)
LIS3=ISTM(LOCSL,3)
LIS4=ISTM(LOCSL,4)
LOS=ISTM(LOCSL,6)
IF(LISl.LE.O) LIS1=30
IF(LIS2.LE.O) LIS2=30
IF(LIS3.LE.O) LIS3=30
IF(LIS4.LE.O) LIS4=30
IF(SV(LOS,2).GT.1.5) RETURN 2
IF (SV(LOS,2).GT.0.5) GO TO 1
WNES=0.0
LINE=LINE + 5
IF (LINE .LT. 43) GO TO 40
LINE = 8
PRINT 98
CALL DATIME
PRINT 97,LABEL
PRINT 96
PRINT 99, ISEQ(NL)
E-27
-------
GALS=P(1)*7.480519
PRINT 101, P(l), GALS
PRINT 102, P(2),P(3),P(4),P(8)
96 FORMAT(IX, 'SYSTEM AND EQUIPMENT PARAMETERS')
97 FORMAT(1H+,34X.13A6)
98 FORMAT(lHl)
99 FORMAT(/5X,'RATHD1 EQUIPMENT NUMBER ', 12)
101 FORMAT(10X,'VOLUME = '1PE10.4,' CU FT ='1PE10.4,' GAL')
102 FORMAT(10X, 'RATE CONSTANTS (GMOLE/SEC CM SQ)+', /, 15X,
*'CAC03=1,1PE11.4,' CAS04=',1PE11.4,' CAS03=',1PE11.4,'
*LIMESTONE=',1PE11.4)
IF(P(9).LE.O.) GO TO 50
WRITE(6,110) P(10), P(ll), P(12)
110 FORMAT(36X,'CRYSTAL AREA SPECIFIED'/10X,'CAC03=',1PE10.4,
*5X,'CAS03*1/2H20=',1PE10.4.5X,'CAS04*2H20=',1PE10.4)
50 CONTINUE
RETURN 2
1 CONTINUE
CALL ADDER (LIS1,LIS2,LIS3,LIS4,LOS)
IF(SV(LOS,49).LT.7.) SV(LOS,57)=1.E-10
CALL TOLISP(LOS)
SV(LOS,35)=SV(LOS,35)+SV(LOS,102)
SV(LOS,117)=0.0
SV(LOS,36)=SV(LOS,36)+SV(LOS,109)
SV(LOS,39)=SV(LOS,39)+SV(LOS,102)+SV(LOS,109)
DO 5 1=3,7,2
XNN(I)=0.
IL=(I-3)/2+2
IF(P(9);GT.O.) SV(LOS,120+I)=P(8+IL)
IF(SV(LOS,120+1).LE.O.) SV(LOS,120+I)=7500.
ZSI(I)=VI(I)/SV(LOS,120+1)
5 XNS(I)=SV(LOS,100+I)
IOPT=2**20
NS=1
T=SV(LOS,5)
CALL EQUILB(LOS,IOPT,T,PP,WNES,NS)
NS=0
IFLS=0
DO 10 1=3,7,2
IL=(I-3)/2+2
RS(1,1)=EXP(APS(IL)-CKS (IL))
IFLAG(I)=0
IF(RS(1,I).LE.l.) GO TO 10
APD(IL)=1000.*RS(1,1) @SUBSATURATE METASTABLE SOLIDS
IF(RS(1,I).LE.1.05*CV(I)) TO TO 10
IFLAG(I)=1
IFLS=1
APD(IL)=CV(I)
10 CONTINUE
E-28
-------
IF(IFLS.EQ.O) GO TO 11
IF(RS(1,3).GT.CV(3)) SV(LOS,103)=SV(LOS,32)
IF(RS(1,5).GT.CV(5)) SV(LOS,105)=SV(LOS,31)
IF(RS(1,7).GT.CV(7)) SV(LOS,107)=SV(LOS,33)
CALL EQUILB(LOS,IOPT,T,PP,WNES,NS)
DO 12 1=3,7,2
IL=(I-3)/2+2
RS(2,I)=EXP(APS(IL)-CKS(IL))
APD(IL)=1.
IF(IFLAG(I).NE.l) GO TO 12
XNN(I)=SV(LOS,100+I)
WRITE(6,106) I,XNN(I)
106 FORMAT(10X,'SOLID NO. ',12,' GMOLES/S NUCLEATING',G12 . 6)
12 CONTINUE
11 CONTINUE
DO 28 1=3,7,2
28 SAG(I) = (SAPG(I)*XNN(I)+SV(LOS,120+I)*XNS(I))/(XNN(I)+XNS(I))
CALL ADDER(LIS1,LIS2,LIS3,LIS4,LOS)
IF(SV(LOS,49).LT.7.) SV(LOS,57)=1.E-10
SV(LOS,35)=SV(LOS,35)-SV(LOS,101)
SV(LOS,117)=0.0
SV(LOS,36)=SV(LOS,36)-SV(LOS,108)
DO 23 1=3,7,2
XNI(I)=XNN(I)+XNS(I)
23 CONTINUE
V=28316.85*P(1)
20 VRK(1)=P(2)*V*SAG(3)/1000.
VRK(2)=P(4)*V*SAG(5)/1000.
VRK(3)=P(3)*V*SAG(7)/1000.
VRK(7)=P(8)*V*7.5
JO
DO 21 1=3,7,2
SV(LOS,100+I)=XNI(I)
21 J=J+IFLAG(I)
NRS=1
IF(J.NE.O) NRS=2
IF(RS(NRS;3).LE.l.) VRK(1)=0.
IF(RS(NRS,5).LE.l.) VRK(2)=0.
IR(RS(NRS,7).LE.l.) VRK(3)=0.
DO 22 1=3,7,2
IL=(I-3)/2+2
IF(IFLAG(I).GT.1.0R.XNS(I).GT.O.) GO TO 22
IF(I.EQ.3) VRK(1)=0.
IF(I.EQ.5) VRK(2)=0.
APD(iL?=1000R*RS(l?i) QSUBSATURATE METASTABLE SOLIDS
22 CONTINUE
T = SV(LOS,5)
CALL RATE (LOS,IOPT,T,PP,WNES,NS,VRK)
DO 25 1=3,7,2
E-29
-------
IL=(I-3)/2+2
ANO(I)=0.
APD(IL)=1.
IF(IFLAG(I).EQ.O) GO TO 29
AI=((XNI(I)-XNN(I))*SV(LOS,120+I)+SAPG(I)*XNN(I))/XNI(I)
PNM(I)=(SV(LOS,100+I)-XNI(I))*(XNN(I)*SAPG(I))/(AI*XNI(I))
Z(I)=CLC(I)*((1.+PNM(I)/XNN(I))**.333)
ANO(I)=VI(I)/Z(I)
29 CONTINUE
PSM(I)=SV(LOS,100+I)-XNI(I)-PNM(I)
IF(PSM(I).LE.O.) PSM(I)=0.
ZS(I)=ZSI(I)*((XNS(I)+PSM(I))/XNS(I))**.333
ASO(I)=VI(I)/ZS(I)
25 CONTINUE
DO 26 1=3,7,2
26 SAGT(I)=((XNN(I)+PNM(I))*ANO(I)+ASO(I)*(XNS(I)+PSM(I)))
*/SV(LOS,100+I)
IFLS=0
DO 27 1=3,7,2
IF(ABS((SAGT(I)-SAG(I))/SAGT(I)).GT..0001) IFLS=1
27 SAG(I)=SAGT(I)
IF(IFLS.NE.O) GO TO 20
WRITE(6.115)
115 FORMAT(1H1)
DO 30 1=3,7,2
SV (LOS,120+I)=SAG(I)
WRITE(6,105) I,ANO(I),SAG(I)
105 FORMAT(/IX,'SOLID NO. ',12,' OUTLET NUCLEI SURFACE AREA ',
*G12.6,'(CM2/GM)'/2X,'AVERAGE PARTICLE SURFACE AREA ',
*G12.6, '(CM2/GM)')
30 CONTINUE
CALL BOXCHK(0,0)
RETURN 2
END
E-30
-------
3.3 System Balance Subroutines
The CTBAL1 subroutine is an overall mass balance routine
used^to assess cooling system water treatment requirements. The
required inputs to the routine include complete specification of
the ambient air, the air leaving the cooling tower, the drift,
and the acid stream used for pH control. (The first three inputs
are determined by CLGTR1.) Other inputs are the number of cycles
of concentration, temperature change across the condenser, makeup
water composition, and an acceptable range of CaCCL and CaSCL
relative saturations.
CTBAL1 uses this input information to check the CaC03
relative saturation in the recirculating cooling water. If the
CaC03 relative saturation is outside the acceptable range, the
acid rate is adjusted by an iterative procedure to bring the
CaC03 relative saturation within the range. Once the CaC03 con-
dition is satisfied, the CaSO^ relative saturation is checked.
If it exceeds the desired value, which was specified as an input,
chemical treatment is needed to lower the Ca^"1" concentration.
The amount of Ca"^ that must be removed is calculated, and this
number is placed in a computer memory location that is in common
with the chemical treatment routine CHMTRT, which is discussed
in Section 3.2.
The CTBAL1 outputs include a complete specification of
the blowdown stream leaving the cooling system, specification of
the makeup water rate, specification of a new acid rate, and
specification of the amount of calcium that must be removed in a
water treatment step to keep the relative saturations of CaC03
and CaSO^ in the desired ranges.
The PNDBAL subroutine is used in the ash sluicing model
to perform an overall mass balance. This routine determines the
plant effluent from the ash pond based on specified fly ash and
fly ash sluice water, bottom ash and bottom ash sluice water,
pond evaporation, the sludge solids content and the degree of
carbon dioxide mass transfer between the pond liquor and the at-
mosphere. For all species except CaSO^ solid-liquid equilibrium
is assumed in the pond. CaSO,, was allowed to remain supersatur-
ated because evidence exists which indicate that it can remain
supersaturated in an ash pond. For the recirculating ash sluicing
simulations, this represents a worse case than the assumption ot
equilibrium which would cause Ca++ and SOS to be removed from the
system.
E-31
-------
There are three overall system balance subroutines,
SYSTB1, SYSTB4, and SYSTB5, used in the scrubbing simulations
for Four Corners and Colstrip. All of these subroutines perform
overall material and energy balances. They calculate the amount
of water vaporized in the scrubber, the makeup water flow rate
and the flow rate and composition of the scrubber effluent
stream. These subroutines have the option to not allow solids
formation. These subroutines differ in the number of input and
output streams that they consider. SYSTBl has three input
streams and two output streams. SYSTB4 has four input streams
and two output streams. SYSTB5 has four input streams and
three output streams. The choice of the correct subroutine
is determined by the configuration of the system which is modeled.
E-32
-------
4.0 CHEMICAL EQUILIBRIUM PROGRAM
. . n The basis of the models used in this study is the
chemical equilibrium program. This section presents a descrip-
tion of the program and the assumptions used to calculate the
distribution of ionic species in aqueous systems. At the end
of this section the nomenclature used in this description is
presented.
4.1 Chemical Species
Inputs to the program are in the form of nine key
species which are listed in Table 4-1. Here, the prefix "t"
has been used to denote that these species are key or total
species as opposed to gas, liquid, or solid species. For exam-
ple, tC02 represents the total carbon dioxide species in the
system. The total carbon dioxide species would consist of_the
sum of actual molecular and ionic species such as HCO"i, C03,
H2C03(JO, CaC03(JO, etc.
TABLE 4-1. EQUILIBRIUM PROGRAM KEY SPECIES
1.
2.
3.
4.
5.
tS02
tC02
tS03
tN205
tCaO
6.
7.
8.
9.
tMgO
tNaaO
tHCl
tH20
The chemical species which are considered sig-
nificant in an aqueous system of these total species are listed
in Table 4-2. These species are grouped according to liquid and
solid species. Some species may exist as a liquid and as a solid,
e.g., CaC03. Where ambiguity may arise, these species will be
denoted using standard chemical notation. For example, calcium
carbonate as a solid will be denoted by CaC03(s). In writing
algebraic equations (as opposed to chemical reaction equations),
an abbreviated notation will frequently be used to avoid the use
of numerous parentheses and brackets. In this abbreviated nota-
tion, solid calcium carbonate will be denoted by sCaC03 and dis-
solved or liquid calcium carbonate will be denoted by £CaC03.
E-33
-------
TABLE 4-2. EQUILIBRIUM SPECIES
A: Liquid Species
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
H20
H+
OH~
HSO"3
sol
so"
HCO~3
col
NO!
HSOZ
H2S03
1. Ca(OH)
2 . CaCO 3
3. CaS03
4. CaS03'
5. CaSOi,
6. CaSO.,*
7. Mg(OH)
12. H2G03
13. Ca4"4"
14. CaOH+
15. CaS03
16 . CaCO 3
17 . CaHCot
18. CaSOif
19 . CaNot
20. Mg"1"1"
21 . MgOH+
22. MgS03
B: Solid Species
2 8.
9.
10.
%H20 11.
12.
2H20 13.
2
23. MgHCot
24. MgSO*
25. MgC03
26. Na+
27. NaOH
28. NaCO~3
29. NaHCOs
30. NaSO^
31 . NaNO 3
32. Cl"
MgC03
MgC03-3H20
MgC03-5H20
MgS03
MgS03'3H20
MgS03«6H20
E-34
-------
All possible chemical species are not considered. For
example, no solid nitrates are listed. Nitrates as a class are
so soluble that within the range of interest of this program solid
nitrates will not occur, so that the solubility relations for ni-
trates may be ignored. Nitrates need not be included in the solid
equilibrium species.
4.2 System Equilibria
There are three types of equilibria which are considered
in this program, liquid-liquid, gas-liquid, and solid-liquid.
Liquid-liquid equilibria relate the activities of species within
the liquid phase. Gas-liquid equilibria relate the activity of a
liquid species to the equilibrium partial pressure of the corres-
ponding gas species, i.e., the gas phase is assumed to behave
ideally. Solid-liquid equilibria relate the activities of liquid
species to the solubility product constant of the corresponding
solid, i.e., the activities of the solid species are taken to be
one.
Liquid-liquid and gas-liquid equilibria are different
in that all terms in the liquid-liquid equilibria must enter
into the mass balance relations given in Section 4.3, whereas
the partial pressures of gases calculated with gas-liquid equil-
ibria have not been related to the mass balances. In the case
of solid-liquid equilibria, the quantity of solid does enter into
the mass balance, but it does not enter into the solubility pro-
duct expression. The presence of a solid depends on whether its
solubility product is exceeded. If so, the quantity depends on
the mass balance.
The chemical reactions which are considered and their
associated equilibrium relations are given in Table 4-3. In the
case of the solid-liquid equilibria, water of hydration (v) must
be assigned to some solids. The values of v are functions^of
temperature according to the relative thermodynamic stability
of the hydrates of each solid. Temperature ranges and hydrate
numbers for each solid so treated are given in Table 4-4.
4.3 Material Balance
As stated in Section 4.2, the mass balances involve
only the liquid and solid phases. The mass balance equations
require that the actual molecular and ionic species in the
liquid and solid phases be equal to the input total "key"
species (see Section 4.1).
E-35
-------
TABLE 4-3. EQUILIBRIUM RELATIONS
f .
Reaction
H20 * H+ + OK"
H2S03 j H* + HSO"3
HSol * H* + SO*
HSO^ J H* + SO,
H2CO 3 i H* + HCOl
HCOl ~l H + COl
CaOH* * Ca4* + OH~
CaSO 3 J Ca4* + Sol
CaC03 J Ca + CO 3
CaHCot X Ca4* + HCOl
CaSOi, X Ca + SO*
CaNot I Ca4* + NO!
MgOH4" * Kg4* + OH"
MgS03 X Mg4* +• Sol
MgHC03 J Kg4"*" + HO)!
MgSOt * Mg*4" -t- SOZ
MgC03 J Mg4* + COl
NaOH J Na+ + OH"
NaCOl * Na+ + Col
NaHC03 X Na+ + HCO^
NaSo" * Na4" + S0»
4. .
Reaction
S02(g) + H20(i) * H2£
C02(g) + H:0(£) I H2(
C:
Reaction
CaC03(s) * Ca4* + Col
CaSO,-2H20(s) X Ca4"1" + SO^ +
CaS03 -JH20(s) X Ca4* + Sol +
Ca(OH)2(s) X Ca4* + 20H~
Mg(OH)2(s) t Mg4* + 20H"
MgS03'ViH20(s) * Mg4* + Sol +
MgC03-v2H20(J.) * Mg + COl +
Liquid-Liquid Equilibria
i-qui 1 ' nr ium Reiacion
KI Y(H2U) - JIH ) a(OH~)
K2 a(H2S03) - a(H+) a(HSOl)
K3 a(HSOl) - ad)4") a(Sol)
Ei, a(HSO») • a(H+) a(SOO
Ks a(H2C03) - a(H ) a(HCO^)
Ks a(HCOl) - a(H+) a(Col)
K, a(CaOH*) • aCCa4*) a(OH")
K9 a(CaS03) - a^a4*) a(sol)
K9 a(CaC03) • a(Ca4*) a(Col)
K, „ a(CaHCot) - a(Ca4*) a(HCOl)
Ku a(CaSO») - a(Ca4*) a(SoZ)
K[2 a(CaHot) " aCCa4*) a(HOl)
Kir a (MgOH*) « a(Mg**) a (OH")
K18 a(MgS03) - a(Mg**) a(sol)
Ki9 a(MgHCot) - a(Mg4*) a(HCOl)
K2 o .(HgSOO - a(Mg4*) .(so;)
K2j a(MgCO,) • a(Mg**) a(CO^)
K25 a(NaOH) - a(Na*) a(OH")
K26 a(NaCOl) - a(Na+) a(Co")
K27 a(NaHC03) - a(Na+) a(HCOl)
K29 aCHaSOi,} » a(Na ) a(SO*)
K29 a(NaN03} - a(Na ) a(NOO
Gas-Liquid Equilibria
Equilibrium Relation
iO, (*) P(S02) Y(H20) K2(S02) - a(H2S03)
»3(4) P(C02j Y(HzO) K2(C02) - a(H2C02)
Solid-Liquid Equilibria
Equilibrium Relation
K (CaCOa) >_ a(Ca+*) a(COl)
2H20(8,) KS (CaSO») >_ aJCa"1*) a(SOZ) Y(H20)2
%H20(£) K=m(CaS03) > afCa4*) a(SOl) v (H20)
sp
Ksp[Ca(OH)2) >_ a(Ca4*) a(OH~)2
Kgp[Mg(OH)2] i a(Mg4*) a(OH~)2
V2(H20)(l) Kgp(MgS03) 1 aCMg4*) a(Sol) Y (HzO)Vl
V2(H20)(J.) K (MgC03) > atMg4*) a(Col) Y(HjO)V2
ar
(4.2-1)
(4.2-2)
(4.2-3)
(4.2-4)
(4.2-5)
(4.2-6)
(4.2-7)
(4.2-8)
(4.2-9)
(4.2-10)
(4.2-11)
(4.2-12)
(4.2-13)
(4.2-14)
(4.2-15)
(4.2-16)
(4.2-17)
(4.2-18)
(4.2-19)
(4.2-20)
(4.2-21)
(4.2-22)
(4.2-23)
(4.2-24)
(4.2-25)
(4.2-26)
(4.2-27)
(4.2-28)
(4.2-29)
(4.2-30)
(4.2-31)
E-36
-------
TABLE 4-4. HYDRATE NUMBERS
Solid Species Hydrate Number Temperature Range
MgS03-ViH20 3 Greater than 38°C
6 Less than 38°C
MgC03-v2H20 0 Greater than 100°C
3 20°C to 100°C
5 Less than 20°C
The mass balance equations for each total species are
given in Table 4-5. The molar quantity (in gmoles) of each
species [n(j)] is computed by multiplying the molality [m(j)] by
the amount of liquid water (L ) measured in kilograms. Here the
index j is used to denote thexactual molecular and ionic species
in the liquid and solid phases.
Actually the mass balance relation for water is a hydro-
gen balance (Equation 4.3-9a). This balance equation is not used
in the equilibrium program. A balance of the electrical charges,
or an electroneutrality balance, is used instead. This balance
is given in Table 4-5 by Equation 4.3-9b. This balance, like the
water balance, reflects the fact that some hydrogen enters the
system as HC1.
4.4 Constants
Two sets of constants are needed to implement this
program. These are equilibrium constants and the constants for
the activity coefficient correlation. Equilibrium constants
are discussed in Section 4.4.1 and activity coefficients in
Section 4.4.2.
4.4.1 Equilibrium Constants
The relation for calculating equilibrium constants is
given in Equation 4.4-1.
E-37
-------
TABLE 4-5. MASS BALANCES
n(tS02) = n(HSCh) + n(S03) + n(£H2S03) + n(£CaS03)
+ n(£MgS03) + n(sCaS03-%H20) + n(sMgS03 -v !H20) (4.3-1)
n(tC02) = n(HC03) + n(C07) + n(£H2C03) + n(£CaC03)
+ n(CaHCot) + n(MgHCot) + n(£MgC03) + n(NaC03)
) + n(sCaC03) + n(sMgC03 • V2H20) (4.3-2)
n(tS03) = n(S(K) + n(HSCK) + n(.HCaSOO + n(NaSO^)
+ n(sCaSO.t-2H20) (4.3-3)
n(tN205) = %[n(N03) + n(CaNot) + n(£NaN03)] (4.3-4)
n(tCaO) = n(Ca++) + n(CaOH+) + n(^CaS03) + n(&CaC03)
) + n(sCaC03)
+ n(sCaSO,t-2H20) + n(sCaS03-%H20) +n[sCa(OH)2J (4.3-5)
n(tMgO) = n(Mg++) + n(MgOH+) + n(£MgS03) + n(MgHCot)
+ n(AMgC03) + n(^MgSO^) + ntsMg(OH)2]
+ n(sMgC03-v2H20) + n(sMgS03 -v iH20) (4.3-6)
n(tNa20) = %[n(Na+) + n(JlNaOH) + n(NaC03> + n(fi-NaHC03)
+ n(NaSCK) + n(^NaN03)] (4.3-7)
E-38
-------
TABLE 4-5. MASS BALANCES (Continued)
n(tHCl) = n(Cl ) (4.3-8)
n(tH20) = n(£H20) + %[n(H) + n(OH~) + n(HS03) + n(HSCU)
+ n(HCC-3) + n(CaOH+) + n(CaHCot) + n(MgOH+)
) + n(£NaOH) + n(£NaHCot) 1
+ n(£H2C03) + n(£H2S03) + n[sCa(OH)2 + n sMg(OH)2]
+ % n(sCaS03'%H20) + 2 n(sCaSCK -2H20)
+ vi n(sMgS03«ViH20) + v2 n(sMgC03 'V2H20)
- % n(tHCl) (4.3-9a)
n(H+) + 2 n(Ca++) + n(CaOH+) + n(CaHCot) + n(CaNot) + 2 n(Mg)
+ n(MgOH+) + n(MgHCot) + n(Na+) = n(OH~) +
+ 2 n(SO^) + 2 n(SO^) + n(HCO^) + 2 N(CO") + n(N03)
) + nCNaCOs) + n(NaSOZ) + n(Cl") (4.3-9b)
logic(K) = -Ax/I - BxlogioT - CXT + Dx (4.4-1)
where T is the Kelvin temperature and A , Bx> C x, and Dx are
constants. Constants for these chemical reactions are stored in
the chemical equilibrium program.
4.4.2 Activity CoeJffiLcients
The activity coefficients are correlated with the ionic
strength (I) of the solution. This quantity is related to the
molality (m) and charge (z) of all the species in solution by
Equation 4.4.-2.
E-39
-------
I = % tn(jO z(j£)2 (4.4-2)
For ionic species, the logarithm of the activity coef-
ficients is correlated with ionic strength by Equation 4.4-3.
= A z(j£)2 - - p + MjJOl (4.4-3)
1 4- Ba°(jA)I*
Here, A and B are temperature- dependent constants for the mixture.
A= 1.8248 • 103/(DT)3/2 (4.4-4)
B = 50.292/(DT)% (4.4-5)
where D is the dielectric constant of water in cgs units and T is
the Kelvin temperature. The other constants in Equation 3.2-2
are parameters for each ionic species . Values for these are
stored within the equilibrium program.
In the case of uncharged species, the activity coeffi-
cient is determined by Equation 4.4-6.
logio[Y(j*)] = 0.076 I (4.4-6)
4. 5 Makeup Water Adjustment
Grab samples of many of the streams studied were taken
and analyzed for calcium, magnesium, sodium, chloride, carbonates,
nitrate, sulfate, and sulfite. This subsection describes how the
data for the makeup water were modified for use in the cooling
tower and scrubbing models.
Prior to the initiation of the cooling system simula-
tions, the makeup water compositions were altered slightly from
the sample values. This was done in an effort to reduce the re-
sidual electroneutrality which may result from analytical inac-
curacies. The residual electroneutrality is a parameter computed
in the aqueous ionic equilibrium program. It is defined as the
E-40
-------
total positive charges in solution minus the total negative char-
ges . When using the equilibrium program to evaluate analytical
accuracy, the measured pH is specified and a residual electroneu-
trality is computed. If this charge imbalance is significant
when compared to the total charge (sum of the absolute values) in
the solution, then analytical inaccuracies are indicated.
When the equilibrium program is employed in a process
simulation, the residual electroneutrality is minimized and the
stream pH is calculated. When the equilibrium program is used
in this way, analytical errors can have significant impact on
the pH of an unbuffered liquor such as the makeup water stream.
The composition of the makeup water was altered slightly to
generate a solution pH similar to the measured pH when the resi-
dual electroneutrality was minimized. That is why there is some
deviation between the compositions measured at the plants and
the compositions used in the simulations. It should also be noted
that silica was not included in the electroneutrality balances.
The changes made are presented in Tables 4-6 through
4-10. These tables present the composition changes required for
the five plants studied.
TABLE 4-6. MONTOUR MAKEUP WATER COMPOSITION ADJUSTMENTS
Sample Balanced
ing/«<
Calcium (Ca7"1") 28.4 28.4
Magnesium (Mg^) 5.5 5.5
Sodium (Na+) 8.5 8.1
Chloride (Cl~) 19.0 22.0
Carbonates (CO^) 6.0 6.0
Nitrate (N0~3) 5.5 5.5
Sulfate (SO") 67.0 67.0
Residual Electroneutrality 1.7 x 10"4 -1.5 x 10"9
pH 8.1 8.1
Relative Saturation
CaC03 0 013 0.013
CaS(V2H20 0.012 0.012
E-41
-------
TABLE 4-7. COLSTRIP MAKEUP WATER COMPOSITION ADJUSTMENTS
Calcium
Magnesium
Sodium
Chloride
Carbonates (as CO?)
Nitrate (as NO^)
Sulfate (as SOT)
Residual Electroneutrality
PH
Relative Saturation
CaC03
Sample
mg/£
39.9
10.7
57.3
17.0
6.0
1.4
188.0
6.6 x 10~"
10.3
1.08
0.034
Balanced
mg/Jl
39.9
10.7
40.3
17.0
6.0
1.4
188.0
-7.6 x 10~10
10.5
0.91
0.039
TABLE 4-8. FOUR CORNERS MAKEUP WATER COMPOSITION ADJUSTMENTS
Calcium (Ca"1"1")
j 1
Magnesium (Mg )
Sodium (Na+)
Chloride (Cl~)
Carbonates (GOT)
Nitrate (NO^)
Sulfate (SOT)
Residual Electroneutrality
PH
Relative Saturation
CaC03
CaSO^-2H20
Sample
mg/£
160.
40.
218.
110.
77.
9.
680.
1.7 x 10"3
8.1
1.2
0.21
Balanced
mg/£
160.
40.
189.
135.
77.
9.
680.
3.0 x 10"8
8.4
2.6
0.21
E-42
-------
TABLE 4-9. COMANCHE MAKEUP WATER COMPOSITION ADJUSTMENTS
Calcium (Ca++)
Magnesium (Mg )
Sodium (Na )
Chloride (Cl~)
Carbonates (COl)
Nitrate (N0~3)
Sulfate (SO^)
Residual Electroneutrality
pH
Relative Saturation
CaC03
CaS04-2H20
TABLE 4-10. BOWEN MAKEUP
_l i
Calcium (Ca )
Magnesium (Mg )
Sodium (Na+)
Chloride (Cl~)
Carbonates (C07)
Nitrate (N0~3)
Sulfate (SO")
Residual Electroneutrality
PH
Relative Saturation
CaC03
CaS04'2H20
Sample
mg/£-
36.5
10.2
19.0
9.0
6.0
9.0
163.0
-3.6 x 10'1*
6.2
1 x 10'"
0.028
WATER COMPOSITION
Sample
mg/£
6.1
1.7
1.4
2.1
20.4
4.0
1.9
-3.3 x 10"5
7.7
0.016
1.0 x 10"*
Balanced
mg/£
36.5
10.2
26.2
5.3
5.4
12.4
163.2
2.7 x 10~9
6.9
9.4 x 10"11
0.027
ADJUSTMENTS
Balanced
mg/£
6.0
1.7
1.4
2.1
20.4
4.3
1.9
-1.1 x 10~8
7.8
0.02
9.4 x 10"5
E-43
-------
NOMENCLATURE
activity
a° activity coefficient correlation parameter
A activity coefficient correlation parameter
A equilibrium constant correlation parameter
X
b activity coefficient correlation parameter
B activity coefficient correlation parameter
Bx equilibrium constant correlation parameter
Cx equilibrium constant correlation parameter
D dielectric constant of water
Dx equilibrium constant correlation parameter
g designation for gas species
I ionic strength
K liquid-liquid equilibrium constant, where n is the
reaction index number
K gas-liquid equilibrium constant
K solubility product constant
sp
H designation for liquid species
m molality
n molar flow rate
P partial pressure
s designation for solid species
t designation for total species
T Kelvin temperature
z species charge
Greek
Y activity coefficient
v hydration number
E-44
-------
Appendix F. Recycle/Reuse Options at Four Corners (Arizona Public Service)
1.0 INTRODUCTION
This appendix describes the analysis of the scrubbing
system at the Arizona Public Service Four Corners Station under
EPA Contract No. 68-03-2339, Water Recycle/Reuse Alternatives
in Coal-Fired Steam-Electric Power Plants.The results o£ the
computer modeling performed for existing operations and for
the recycle/reuse alternatives, with rough cost estimates for
the technically feasible options are discussed.
1.1 Summary
Three major topics are discussed in this appendix:
1) Existing Operations Modeling,
2) Alternatives Modeling, and
3) Economics.
The results of the existing operations simulations
compare well to the sample data obtained at the plant. Poten-
tial scaling conditions were found at several points in the
scrubbing system for low (270) solids operation. The scrubber
effluent, thickener overflow, and thickener underflow all were
identified as showing CaSOil*2H20 relative saturations above
1.3, indicating a tendency to form gypsum scale. No calcium
carbonate scaling was noted. A simulation of high (9%) solids
operation showed increased scaling potential in the system.
Four alternatives were investigated for the particulate
scrubbing system at Four Corners. Table 1-1 presents a summary
of these four alternatives compared to existing operations.
The results of the first alternative simulation indi-
cate that the present system tankage capacity is not sufficient
to allow ample gypsum precipitation to prevent scaling.
In the second alternative, a tank capacity of 37,500
cubic meters (1.33 x 106 cubic feet) was simulated. Gypsum rel-
ative saturations were reduced to levels below the critical level
required for the onset of scaling. Two cases were studied with
different scrubber liquid-to-gas ratios (L/G). The existing L/G
of 4.7 £/m3 @ STP (35.2 gal/lOOOscf) gave a scrubber bottoms pH
of 2.9 and an L/G of 10.0 £/m3 @ STP (74.8 gal/1000 scf) gave a
pH of 3.9 (assuming 50% S02 removal), indicating that higher
L/G's are desirable for corrosion control.
F-l
-------
TABLE 1-1. SUMMARY OF RECYCLE/REUSE OPTIONS AT FOUR CORNERS1
Weight Percent Solids
in Thickener Bottoms
Hold Tank Volume,
m3 (ft3)
Liquid to Gas Ratio,
£/ra3 @ STP (gal/scf)
7. Recycle from the
Ash Pond
^ SO a Removal, 7.
NJ
Oxidation, 7.
Particulate Removal
prior to scrubber, %
Scrubber Makeup Rate,
I/sec (GPM)
Costs :
Capital, 1976 $
Operating, 1976 $ 3
(mils/kWh)
Existing Alternative
Condition Two
Case 1 Case 2 Case 1 Case 2
10 30 30 30
0 0 37,500 37.500
(1.33 x 10s) (1. 33 x 106)
4.7 4.7 4.7 10.0
(35.2) (35.2) (35.2) (74.8)
000 0
30 30 50 50
98.6 98.6 98.6 98.6
None None None None
223 70.7 70.7 70.7
(3540) (1730) (1120) (1120)
3,334.000 4,275,000
628,000 1,101,000
( 128) (.225)
Alternative
Three
Case 1
30
37,500
(1.33 x 106)
10.0
(74.8)
28
50
98.6
None
50.8
(805)
4,328,000
1,109,000
(.226)
Case 2
30
21,200
(0.75 x 106)
10.0
(74.8)
28
50
98 6
None
50.8
(805)
3,317,000
958,000
(.195)
Alternative
Four
Case 1
30
8900
(0.31 x 106)
10.0
(74.8)
0
50
98.6
60
41.0
(650)
3,385,000
968,000
(.198)
'Alternative One is not included because it was not deemed technically feasible due to high CaSO,, relative saturations
in the SOj scrubber.
2These roup.h cost estimates were trade to compare technically feasible options and do not include a "difficulty to retrofit'
factor.
'Includes capital amortization at 157. per year
-------
The third alternative simulation, recycling the ash
pond overflow to the scrubbing system, indicated that the pond
overflow has no major impact on the gypsum relative saturations
in the system but reduces the water makeup requirements from
70.7 £/sec (1122 GPM) for Alternative 2 to about 50.8 5,/sec
(807 GPM). Also, a simulation with ash pond overflow recycle
using a reaction tank volume of 21,200 nr (7.5 x 10s ft3)
showed that a more reasonable reaction tank volume can be
utilized. This simulation showed a gypsum relative saturation
of 1.19 in the scrubber effluent slurry.
The fourth alternative shows that reaction tank
volume may be decreased further by removing a portion of the
fly ash by dry methods prior to the scrubbing system. A
volume of 8900 m3 (3.14 x 105 ft3) was used to obtain a gypsum
relative saturation of 1.19 in the scrubber effluent (607o of
fly ash removed prior to scrubber). Water makeup requirements
were also reduced to 41.0 £/sec (650 GPM).
All of these alternatives assumed 98.6% oxidation of
the S02 sorbed in the scrubbers. Process modification may cause
the oxidation to decrease, thereby decreasing the reaction tank
volumes necessary to prevent scale. Since less CaSOn*2H20 is
formed, the reaction time required to form gypsum is also
decreased. Modifications made after this study was completed
decreased the oxidation at Four Corners so that all of the
sulfate formed coprecipitated with calcium sulfite (<15% oxida-
tion) . Lime was added to the sump below the venturi to increase
S02 removal. The resulting higher pH liquors apparently reduced
the oxidation.
The rough-cost estimates of the technically feasible
options (Alternatives 2-4) indicate that three to four million
dollars would be required to upgrade the particulate scrubbing
system so that scale potential is eliminated and water require-
ments reduced. The least expensive alternative was Alternative
3, case two (recycle ash pond overflow, reduced reaction tank
volume), followed closely by the fourth option. Alternative^
2, cases one and two, and Alternative 3, case one showed similar
installed costs. Energy consumption did not vary radically
among alternatives, although Alternative 2, case one (increased
tank volume, low L/G) indicated a lower energy requirement was
necessary and therefore less operating costs.
Detailed discussions of the existing operations
simulations, the alternative simulations, and thorough cost
estimates constitute the main body of this appendix.
F-3
-------
2.0 PLANT CHARACTERISTICS
The Arizona Public Service Company (APS) Four Corners
Plant is a five-unit 2,150 Mw coal-fired electric generating
station located near Farmington, New Mexico. The coal utilized
at Four Corners is approximately 20% ash and 0.5 - 1.0% sulfur
with a heating value of about 9,300 Btu/lb. The plant uses a
cooling pond and bottom ash wet sluicing for all units, partic-
late wet scrubbing for Units 1-3, and electrostatic precipi-
tators for Units 4 and 5 (dry ash disposal).
This section of the appendix describes the
characterization of the Four Corners plant's water system
including the cooling, ash sluicing, and wet scrubbing systems.
First, an overall water balance for the plant is presented
which shows the major in-plant water flows and chemical analyses
for the streams which were sampled. Then a detailed description
of each of the major water consumers in the plant is given.
This is followed by a discussion of the simulation basis for
modeling operations at the Four Corners plant. Finally, the
computer simulation results are presented and discussed.
This discussion will include a comparison of the simulation
results and the chemical analyses of the samples taken. Areas
exhibiting scale potential will be identified.
2.1 Water Balance
A schematic of the Four Corners plant water system
is shown in Figure 2-1. The major streams are shown for the
particulate scrubbing system and bottom ash sluicing systems.
Makeup water for the plant (Stream 18) is taken from the San
Juan River and stored in Morgan Lake, which serves as the
source for all water used in the system. A periodic blowdown
(Stream 17) is taken from Morgan Lake to control the total
dissolved solids concentration. This blowdown is discharged
to the Chaco River which flows into the San Juan River.
Cooling water, bottom ash sluicing water, boiler
makeup water, and makeup water for the particulate scrubbing
system are taken from Morgan Lake. In addition to the blowdown
stream from Morgan Lake, water leaves the plant through evapo-
ration from Morgan Lake, evaporation from the ash pond,
evaporation in the scrubbers, and ash pond overflow. Some
vaporization also occurs in bottom ash sluicing operations
due to the high ash temperature.
F-4
-------
Ul
I A
Figure 2-1. Arizona Public Service Four Corners Station water balance
(Sheet 1 of 3)
-------
I
Cn
8
\\\
1
'
*
1
i
§
STSM* UNITS
TKtIH
TJ
arc*** tw/rs
Figure 2-1. Arizona Public Service Four Corners Station water balance. (Sheet 2 of 3)
-------
Scream Number
Stream Kane
Flo*,:
English
PH
Calcium
Magnesium
Sodium
'oca a slum
Chloride
CnrbDnate{A« COi)
Sulfale(Ai SO,)
Sulflt*
(74-1,000)
346,000 acfm
t43B,000)
(675,000)
407,000 acf»*
(515,000)
<3>
43« lit*
(550)
(8720)
(3.1)
790
_. <"0)
(66J_
(»0)
11
(14)
160
("0)
30
(27)
27to
(2930)
8
(60)
2!
(47J
2.2
(3.S)
270
(360)
2.2
(8-8)
'i vn
-l*«oj_
"1 £/eee
(162)
2070 Bps«
(2565)
_
3>
Feed
716 fc/sec
11,830 gpn
2.8
790
49
290
11
160
30
2740
8
25
2.2
270
2.2
4370
<£>
Feed
380 I/sec
6029 Bpm
2.8
790
49
290
11
160
30
2740
a
25
2.2
270
1.2
4370
Overflow
375 l/Bec
S9SO gpa
3.8
730
54
320
14
1BD
16
2540
la
14
<0.3
140
D.Ob
4110
<»>
Thickem-v
Underflow
5 t/aec
80 gpm
S.3
330
38
280
B
110
50
1160
7
<0.3
no
10. 0
1760
-$-
Sluice
Tank
Feed
64 t/aec
1010 gpm
2.8
790
49
290
U
160
30
27^,0
8
25
Z.2
2>0
2.2
4370
"$
' Asli
Pond
69 t/sec
1100 gpm
9.0
SSO
44
280
10
180
29
2040
23
129 fc/sec
2040 gpm
8.1
160
49
210
8
110
77
680
9
<0.3
100
1305
<£
851 i/aec
13A90 gpn
<6>
40 f,/sec
640 gpm
6.9
160
35
210
H
140
74
690
<0.3
110
1120
<0>
95 i/aec
1500 gpm
8.1
160
40
210
8
110
77
680
9
<0.3
100
<0.01
1350
o
95 1/eec
1500 gpm
O
115 e/sec
1620 gpm
8.1
160
40
210
e
110
77
680
9
<0.3
100
<0.01
1350
<§>
LO
940 t/sec
15000 gpn
7.7
55
10
37
2
a. s
116.4
117.6
1
139
350
o
Sluice to
25 H/aec
396 gp™
6.9
160
35
210
B
140
14
fi90
'0. 1
110
1420
Figure 2-1. Arizona Public Service Four Corners Station water balance.
(Sheet 3 of 3)
F-5b
-------
This study deals primarily with the particulate wet
scrubbing system and subsequent ash disposal for Units 1-3.
Since water is recycled in the cooling and bottom ash sluicing
systems there is little potential for water recycle-reuse
alternatives in these systems. Also, scaling problems have
been encountered in the scrubbing system and the study of
recycle/reuse alternatives at Four Corners dictates addressing
the causes and potential solutions to these problems.
Streams which are not shown in Figure 2-1 include
the cooling water which circulates between Morgan Lake and the
condensers for each unit, boiler makeup water, and water treat-
ment wastes. Boiler makeup water is taken from Morgan Lake
after it is passed through water treatment and evaporators.
Water-treating wastes recycled to Morgan Lake for October,
1975, totaled about 1.8 x 109 S. (4.8 x 108 gal), including
evaporator and demineralizer wastes.
Since Morgan Lake has been deemed a navigable water-
way, these water treatment wastes along with general plant
drainage present a problem. However, these wastes may be dis-
charged to the ash pond, preventing any contamination of Morgan
Lake through the addition of dissolved solids. The impact of
this process change on the results of this study which primar-
ily concerns the scrubbing system will be minimal. The results
of the simulations which involve ash pond overflow recycle
will not be adversely affected since the ash pond will already
be saturated with respect to calcium sulfite, calcium carbonate,
and calcium sulfate and any addition of these ions will only
cause increased precipitation in the pond. In addition, the
ash pond overflow only provides 28% of the total scrubbing
system makeup water. The operation of the scrubbing system
with respect to scaling will therefore not be adversely
affected by the addition of water treating wastes to the ash
pond.
The level of suspended solids in Morgan Lake due
mainly to bottom ash sluicing operations is not addressed in
this study. The focus of this study is on the chemical aspects
of water recycle/reuse alternatives at Four Corners.
Additional streams not shown in Figure 2-1 are the
evaporation from Morgan Lake and the ash pond. During October
1975, about 1.6 x 10* £ (4.2 x 108 gal) was evaporated from
Morgan Lake and 2.1 x 107 «, (5.5 x 106 gal) was evaporated from
F-6
-------
the ash pond. Also, seepage from Morgan Lake during October
1975, was about 4.8 x 1CP £ (1.3 x 1(T gal). The influent '
rate from the San Juan River to Morgan Lake was 2.9 x 109 £
(7.7 x 108 gal) while the blowdown from the lake was 7 3 x
108 £ (1.9 x 108 gal).
All of the flows for the particulate scrubbing are
design values reported by APS. The remaining stream flow rates
are average values calculated from data supplied by APS over
the period January to December 1975. Table 2-1 presents the
values reported on a monthly basis. In most cases the flows
measured in November are very close to the average for the
year. The two streams with the largest variation are the
makeup and blowdown streams from Morgan Lake. This is because
Morgan Lake acts as a large surge tank for the Four Corners
Plant and makeup and blowdown requirements are determined by
a combination of factors. The TDS and species concentrations
for each stream were taken from sample data taken at the plant
in November of 1975. A more detailed description of the sam-
ples taken and analytical procedures used is presented in
Appendix B.
Calculated parameters for the sampled streams are
presented in Table 2-2. Included are the relative saturations
of CaC03, CaSO^ and Mg(OH)2 as well as the partial pressure of
C02 and the 70 residual electroneutrality. These parameters
are useful for characterization of the individual streams.
The relative saturation is a parameter which indi-
cates the potential of a stream to produce scale. When the
relative saturation is greater than the critical value, solids
formation can be expected. The critical values for the three
species reported in Table 2-2 are 2.5 for CaC03, 3.4 for
Mg(OH)2, and 1.3 - 1.4 for CaSO^HzO. The relative saturations
of CaC03 only show a tendency for solid formation in the
thickener underflow and the ash pond. CaS04'2H20 relative
saturations are all near the critical value with the exception
of the makeup water, the thickener underflow and the bottom ash
sluice. This is not surprising since gypsum scale is a serious
problem at Four Corners.
The equilibrium partial pressure of C02 above the
streams sampled at Four Corners is also presented in Table^2-2.
The partial pressure of C02 in the atmosphere is 3.3 x 10
atm. Most of the streams have partial pressures of C02>near
atmospheric. The scrubber liquor seems to be high in dissolved
F-7
-------
TABLE 2-1. MONTHLY VARIATION OF FLOWS FROM SELECTED STREAMS AT FOUR CORNERS*
Stream Name
Sluice Tank Feed
Ash Pond Overflow
Makeup to Units
1, 2 and 3
Makeup to Transfer
Tank
Bottom Ash Sluice
from Units 1, 2 and 3
Makeup to Units
4 and 5
Bottom Ash Sluice
from Units 4 and 5
Blowdoun from Morgan
Lake
Makup to Morgan Lake
Bottom Ash Sluice
to Ash Pond
Stream
Number
9
10
11
12
14
15
16
17
18
19
Ave
63.9
(1013)
69.1
(1095)
197
(3123)
129
(2045)
40.4
(640)
95.3
(1511)
94.7
(1501)
114.5
(1815)
938
(14870)
25.0
(396)
Jan
65.5
(1038)
15.3
(243)
214
(3392)
132
(2092)
82.2
(1303)
119.4
(1893)
118.2
(1874)
0
(0)
639
(10130)
0
Feb
53.2
(843)
51.0
(808)
—
—
76.6
(1214)
100.7
(1596)
99.7
(1580)
0
(0)
778
(12330)
0
(0)
Mar
65.7
(1041)
83.6
(1325)
—
—
21.9
(347)
79.8
(1265)
79.0
(1252)
0
(0)
485
(7690)
21.4
(339)
April
64.4
(1021)
91.9
(1457)
—
—
0
(0)
—
—
0
(0)
—
31.4
(498)
May
64.1
(1016)
lit. 2
(1176)
—
—
59.2
(938)
—
—
0
(0)
—
0
(0)
June
64.8
(1027)
62.2
(986)
—
—
42.4
(672)
46.1
(731)
44.2
(701)
0
(0)
—
0
(0)
July
64.8
(1027)
96.5
(1530)
—
—
60.0
(951)
73.9
(1171)
73.9
(1171)
0
(0)
934
(14810)
0
(0)
Aug
65.1
(1032)
70.7
(1121)
—
—
61.8
(980)
133.2
(2111)
134.5
(2132)
271.8
(4308)
1343
(21290)
75.0
(1189)
Sept
64.6
(1024)
66.0
(1046)
—
—
0
(0)
103.8
(1645)
102.7
(1628)
278.7
(4418)
1578
(25010)
68.5
(1086)
Oct
64.2
(1018)
61.0
(967)
180
(2853)
127
(2013)
53.0
(840)
87.6
(1389)
86.7
(1374)
272.0
(4311)
1080
(17120)
0
(0)
Nov
64.4
(1021)
98.1
(1555)
—
—
28.1
(445)
92.8
(1471)
88.9
(1409)
272.0
(4311)
890
(14110)
32.9
(521)
Dec
65.5
(1038)
55.2
(875)
—
—
0
(0)
114.4
(1813)
113.3
(1796)
272.0
(4311)
717
(11370)
69.6
(1103)
*Flows are reported in i/sec with GPM in parentheses.
-------
TABLE 2-2. PARAMETERS CALCULATED BY THE EQUILIBRIUM PROGRAM
FOR FOUR CORNERS SAMPLES*
Relative
Stream Name
Makeup Water
Scrubber Liquor (1A)
Scrubber Liquor (3A)
Thickener Overflow
Thickener Underflow
Bottom Ash Sluice
Sluice Tank Effluent
Ash Pond Effluent
Ash Pond Surface
CaC03
1.21
7.6 x 10~9
1.8 x 10~8
3.8 x 10~7
3.18
0.11
0.05
7.10
6.37
Saturations**
Mg(OH)2 CaS0lt«2H20
2.2
5.3
1.5
5.9
8.8
3.4
1.2
6.9
1.1
x 10 5
x l(f15
x Hf1*
x 10~13
x 10~5
x 10~7
x 10~7
x 10~ *
x 10~3
0.21
1.39
1.27
1.28
0.50
0.19
1.19
1.18
1.11
Equilibrium Partial
Pressure of C02
atm
4.8 x 10~"
1.78 x 10~2
1.51 x 10"2
9.55 x 10~3
2.3 x 10~"
6.96 x 10"3
2.57 x 10~3
1.6 x 10~5
2.3 x 10~5
% Residual
Elect roneutrality
8.0
-5.0
-22.0
-5.0
11.0
2.0
5.0
-6.0
4.0
* These values were calculated using raw analytical data with all species concentrations and pH
specified. The percent residual electroneutrality is the difference between the positive ions
and the negative ions divided by the total charge times 100. This value gives an indication of
analytical error as well as indicating the possible existence of an unaccounted for species in
solution.
**Critical values, above which scale potential exists, are 1.3-1.4 for CaSOit«2H20, about 2.5 for
CaC03, and about 3.4 for Mg(OH)2 (see Appendix C).
-------
carbon dioxide because of the large amount of COz in the flue
gas. The carbonate level in the ash pond is depressed from the
atmospheric value because CaC03 precipitation is occurring as
indicated by the high relative saturations of CaC03.
The percent residual electroneutrality is the differ-
ence between the total positive charge and the total negative
charge as a percent of the total charge. It is an indication
of how accurately the actual stream is represented by the com-
puter model. More information on the residual electroneutrality
is presented in Appendix E. The values reported in Table 2-2
are quite good for the most part and tend to confirm the accuracy
of the analysis.
2.1.1 Cooling and Bottom Ash Sluicing Systems
Water from Morgan Lake is used for both cooling and
bottom ash sluicing. Water is brought into Morgan Lake from
the San Juan River at a point 15 miles west of Farmington, New
Mexico. This intake includes coarse trash racks, closure gates,
traveling bar racks, sand traps and pumps. When the Four Corners
plant was built, a permit to withdraw water from the San Juan
River was obtained. However, this water could only be diverted
at certain times of the year so Morgan Lake was constructed as
a reservoir from which water could be continuously withdrawn
for circulating water makeup, boiler water makeup, service water,
cooling water, and ash sluicing water. Thus, Morgan Lake serves
two purposes at the plant:
1) makeup water reservoir
2) cooling pond.
Morgan Lake has a surface area of about 5.16 km2
(1,275 acres) with an average depth of 8.8 m (25 ft), and is
located three miles south of the San Juan River. The circulat-
ing cooling water is withdrawn from the lake at the west end
(deepest end) on the south shore. The condenser discharge is
sent 1,520 m (5,000 ft) along the south shore through canals.
Winds which are primarily from the southwest carry the warm
water across the lake away from the plant intake (DE-165) .
2.1.2 Particulate Scrubbing System
The potential for reduction of water requirements
appears to be greatest in the wet scrubbing system. For this
F-10
-------
reason, the wet scrubbing system will be described in more
detail than the cooling and ash sluicing systems. The particu-
late scrubbing system at Four Corners consists of six venturi
scrubbers (two each for Units 1, 2, and 3), two thickeners,
two thickener transfer tanks for return of thickener overflow
and a sluice tank for combining thickener underflows and the
scrubbing loop bleed stream. The scrubbing system is designed
to clean 2.18 x 106 m3 @ STP/hr (1.28 x 10? scfm) of gas at base
load. The design slurry flow through the six scrubbers is
2880 S,/sec (45,700 GPM) , resulting in a liquid-to-gas ratio of
about 4.8 £/m3 @ STP (35.7 gal/1000 scf).
The flue gas and liquor undergo intimate contact as
they flow through the venturi throat. The particulate removal
efficiency is in excess of 99%, giving a scrubber outlet grain
loading of about 92 mg/m3 @ STP (.04 gr/scf) . About 30% of the
600 ppm SOa in the flue gas is also transferred to the liquid
phase, over 987o of which is oxidized to form sulfate in the
scrubbing liquor.
The gas-liquid stream leaving the venturi throat
passes through a disengagement zone where the liquid is sepa-
rated from the gas and falls into a reservoir at the bottom of
the scrubber. The gas passes through a demister to minimize
entrainment and then to the stack. The liquor collected in the
reservoir is recycled for further gas-liquid contacting after
a bleed stream is removed. The major portion of this bleed
stream is routed to the thickener and the remainder is sent to
the sluice tank to reslurry the thickener underflow. The
solids concentration in the recycle slurry is generally con-
trolled between 1-2% by this bleed stream. Thickener transfer
tank clear liquor, a combination of thickener overflow and
makeup water, replaces the slurry which is removed from the
scrubber recycle loop.
The scrubbing system is piped so that the two scrub-
ber trains from each unit are connected to separate thickeners.
In this manner, the system does not rely completely on one
thickener train, so that if one thickener develops operating
problems, the units can still operate at half load. The
thickener underflow solids concentration can vary between 10k
and 60% solids but typically is on the lower end of this range.
Lime is added in the center of the thickeners to maintain the
bottoms stream pH at approximately eight. This stream is
sluiced with the bleed stream from the scrubber recycle loop
to produce a waste stream of about 8% solids which is pumped
to the ash pond for disposal. As mentioned previously, the
thickener overflow streams are pumped to the thickener transrer
F-ll
-------
tanks where makeup water is added. The clear liquor from
these tanks is pumped to the scrubber reservoir on demand.
Water requirements for this scrubbing system depend
primarily on two factors:
1) the evaporation rate in the scrubbers
2) the amount of water associated with the solid
waste stream
The scrubber evaporation rate is approximately 30 £/sec (470
GPM) when all scrubbers are operating at full load. When the
slurry discharge stream sent to the pond is 8% solids, the
water associated with this stream is about 190 H/sec (3,000
GPM) . The total water makeup requirements for the scrubbing
system for October 1975 were about 3.4 x 108 £ (9.0 x 107 gal).
This total includes the ash pond overflow liquor which is
routed to the Chaco River and evaporation from both the ash
pond and the scrubbers.
2.2 Existing Operations
The most severe operating problems related to water
usage and disposal exist in the scrubbing system. Some SOa
removal is achieved in the particulate scrubbers and gypsum
scaling conditions have been reported. Also, a substantial
portion of the water usage at Four Corners is related to the
wet scrubbing system.
The following section presents an analysis of the
design scrubber operating conditions based on sample analyses
and operating data for the Four Corners plant. First, the
simulation basis is presented, including a brief model descrip-
tion and a discussion of the input data used to simulate design
conditions at Four Corners. Then the results of the simula-
tions are compared to the sample results.
2.2.1 Simulation Basis
A process simulation of the Four Corners scrubbing
system operating at design conditions was performed to charac-
terize the system and to determine if a potential for water
recycle/reuse exists with the present configuration. This
section first briefly discusses the model, followed by a
F-12
-------
description of the operating parameters used as inputs to the
model. A detailed discussion of the process model is included
in Appendix E.
The process simulation flow scheme shown in Figure 2-2
was used to model the scrubbing system at Four Corners. This
model calculates all stream compositions and flow rates using
precipitation rate kinetics for CaSQ^*2R20 and CaS03'%H20 (see
Appendix C) , which are the solids formed in lime/limes tone
scrubbing systems, and various input parameters. These para-
meters characterize the operating conditions for a particular
scrubbing system and include flue gas flow and composition, fly
ash rate and composition, makeup water composition, lime addi-
tion rate, tank volumes, scrubber feed flow rate and percent
suspended solids, percent oxidation in the system, and percent
solids in the sludge.
As shown by the order of process calculations in
Figure 2-2, once the inputs are initialized and the first
approximation for the thickener overflow (Stream 15) is made,
SYSTB4 computes the compositions and flow rates for stack gas
and scrubber effluent streams. Then iterative calculations are
performed in Boxes 8, 9, and 10, completing the convergence
loop until the composition of the thickener overflow remains
with the specified convergence criterion. Then the remaining
calculations are performed in Boxes 11 through 17.
Several assumptions are inherent in performing this
simulation with the model outlined above. These are enumerated
below:
1) The stack gas is saturated with respect
to water.
2) Equilibrium exists between C02 in the stack
gas and liquor in the scrubber bottoms.
3) The scrubber bottoms and stack gas temperatures
are the adiabatic saturation temperature of
the flue gas.
4) The scrubber was modeled without allowing
solids precipitation to occur. However,
dissolution of Mg(OH)2, Ca(OH)2, and CaS03-%H20
solids entering the scrubber was allowed.
This dissolution pertains to particulates
F-13
-------
STACK GAS
BCRUBBER
RESER-
VOIR
(NONE)
MAKEUP
WATER
(WTRMKP)
S
THICKENER OVERFLOW
1 ^
i a
?°
IB
MATERIAL
BALANCE
C8Y8TB4)
7
STACK OAS
SCRUBBER SLOWDOWN
ORDER OF CALCULATIONS
1,2,3.4.6,6(7,6,9,10)11.12,13.14,16.16,17
Figure 2-2.
Four Corners scrubber simulation
scheme (existing operations).
F-14
-------
removed as well as slurry solids entering the
scrubber. The fraction of each solid species
that will dissolve in the scrubber was specified.
5) All oxidation was assumed to occur in the
scrubber.
6) No CaSOi^HzO, CaS03-%H20, or CaC03 solids formed
in the scrubbing loop. This was done to model
the scrubber blowdown stream as accurately as
possible. Realistically, actual conditions are
somewhere between no precipitation and solid-
liquid equilibrium. The short residence time
in the scrubbing loop and the low inventory of
precipitating solid crystals indicate that the
assumption of no solids formation in the loop
is adequate.
7) All solids precipitation occurs in reaction
vessels (Subroutines HLDTK3 or RATHDl) .
8) Ionic reactions taking place in the liquid
phase are rapid and thus at equilibrium.
A summary of the input stream data employed in this
simulation is provided in Table 2-3. The flue gas composition
was determined by a combustion calculation from a coal analysis
supplied by APS. The fly ash composition was provided by APS.
The lime and water makeup compositions were measured by chemical
analyses and adjusted to minimize the residual electroneutral-
ity (Appendix E).
The system and equipment parameters are also listed
in Table 2-3. The S02 sorption efficiency, S02 oxidation, gas
phase pressure drop, particulate removal efficiency, scrubber
blowdown pH and solids concentration, were either supplied by
APS or computed from data obtained from APS. The fractions of
CaO and MgO from the fly ash which hydrate in the system were
computed from the results of ash characterization studies per-
formed in support of this project (FU-R-61). The amount of
hy drat ion was determined from the leaching measured at low pH
since this most closely approximates the existing scrubber
operation at Four Corners.
F-15
-------
TABLE 2-3.
INPUT DATA FOR FOUR CORNERS SCRUBBING
SIMULATION*
Flue Gas
Flow, m3/hr
(ACFM)
Temperature, °C
(°F)
Composition, mole %
SO 2
CO 2
02
N2
H20
Fly Ash Rate, kg/min
(Ib/min)
System Parameters
S02 Removal Efficiency, 7o
Oxidation, %
Particulate Removal Efficiency, 70
Liquid-to-Gas Ratio, £/m3 @ STP
(gal/1000 scf)
Scrubber Slurry Solids, wt %
Thickener Underflow, wt 70 solids
Sludge, wt % solids
Makeup Water Composition, mg/&
Calcium
Magnesium
Sodium
Chloride
Carbonates (as C07)
Sulfates (as SOT)
Nitrates (as N07)
4.1 x 106
(2.41 x 106)
129
265
0.0643
13.1
4.65
74.58
7.62
975
(2144)
30
98.6
99.7
4.8
35.7
2.0
10.0
50
160.3
40.1
188.6
134.7
77.4
682
9.3
*A11 flows are for all six scrubber modules
F-16
-------
2.2.2 Simulation Results
This section describes the results from the simulation
of design scrubber operations at Four Corners. As Table 2-3
shows, some differences between the simulation and the sample
data are to be noted, but overall the two compare favorably.
Comparison of some sample data with simulated results
indicates that parts of the system may not have been at steady
state during sampling. For instance, the thickener underflow
stream's measured CaSOif-2H20 relative saturation was 0.5. Since
gypsum precipitation was noted in the thickener and to some
extent in the scrubber, it is not likely that the relative
saturation would be much less than one. A second indication
of unsteady state operation is the low (2.2%) solids concentra-
tion in the sluice tank effluent. Since the thickener underflow
(10% solids) was sluiced by the scrubber effluent (2.2%, solids),
a concentration of 8%, solids was simulated. Some of these
discrepancies may be due to nonhomogeneous sampling and/or
analytical errors as well as unsteady-state operation.
An examination of the existing operations at low
slurry solids concentration (Table 2-4) reveals potential
chemical scaling conditions at several points in the scrubbing
loop. A section describing causes of chemical scaling is pre-
sented in Section 3.0. For discussion purposes here, it is
noted that streams with relative saturations above 1.3 for
CaSCK'2H20 and 2.5 for CaC03 may exhibit scale formation. The
simulated scrubber effluent, thickener overflow, and thickener
underflow all show CaSOit'2H20 relative saturations in the
scaling region. Operation under these conditions for extended
periods may necessitate system shutdown for cleaning. No
carbonate scaling problems are indicated.
One method of reducing the amount of water consumed
by the scrubbing system at the Four Corners plant is increasing
the solids concentration of the solid waste stream. Since less
water exits the system, less water makeup is required. A
simulation of system operation at 30% solids in the thickener
underflow and approximately 17% solids in the sluice tank
effluent was performed. The results of this simulation are
presented in Table 2-5. From this table it can be seen that
the scaling potential is somewhat higher in the scrubber than
at lower solids levels. It is obvious that the solution to the
water recycle/reuse problems cannot be achieved simply by
raising the solids content of the solid waste stream. Possible
F-17
-------
TABLE 2-4. FOUR CORNERS SCRUBBING SIMULATION RESULTS FOR DESIGN CONDITIONS
i
i—1
CO
Stream
Flow Rate,** I/sec
(GPM)
pH
Suspended Solids, wt. %
Relative Saturations ***
CaS
-------
TABLE 2-5. FOUR CORNERS SCRUBBING SIMULATION RESULTS WITH 30% SOLID WASTE OPERATION
Stream
Flow Rate, t/sec
(GPM)
pH
Suspended Solids, wt. %
Relative Saturations*
CaSOk.2H20
CaCO,
Composition, ing/ 1
Calcium
,-jj Magnesium
I Sodium
\O Chloride
Total Sulfur (as S0=)
Sulfite (as S(T)
Carbonate (as C0=)
Nitrate (as N0~)
Scrubber Liquor Slowdown
160
(2,540)
2.3
9.0
2.52
1.2 x 10"'
1,480
77
322
188
4,954
72
126
12.4
Thickener Overflow
81.5
(1,290)
3.8
0
1.30
7.7 x 10"«
794
77
322
188
2,678
72
126
12.4
Thickener Underflow
26
(410)
7.1
30.0
1.32
2.22
846
77
322
188
2,664
72
114
12.4
Sluice Tank Effluent
79.7
(1,260)
2.5
17.3
1.0
1.3 x 10" "
642
77
322
188
2,668
72
126
12.4
* Critical values, above which scale potential exists, are 1.3-1.4 for CaSOi-2H20 and about 2.5 for CaC03 (see Appendix C).
-------
alternatives for decreasing water use levels are presented in
Section 3.0. Any attempt at decreasing the water makeup
requirements to the scrubbing system must be accompanied by an
effort to reduce the scaling potential which results.
F-20
-------
3.0 TECHNICAL ALTERNATIVES
A key to reducing the water requirements for the
particulate scrubbing system at the Four Corners Plant is pro-
ducing a concentrated solid waste stream. An increase in solids
concentration from 8% (existing operations) to 30% in the ash
pond feed stream will reduce water requirements roughly by a
factor of three. However, as was indicated in the plant char-
acterization section, a scaling problem already complicates
scrubber system operation. Further reduction of water makeup
into the system which would result by increased waste solids
concentration will compound these existing problems.
The Four Corners scrubbing system was originally
designed for particulate scrubbing only, and several aspects
of this system contribute to the scaling conditions which are
currently in evidence. These aspects were considered in
deciding upon practicable technical alternatives. The alterna-
tives investigated include: (1) employing existing thickener
transfer tanks as solid-liquid reaction vessels, (2) substan-
tially increasing reaction tank capacity with two different
liquid-to-gas ratios, (3) recycling ash pond overflow back to
the scrubbing system utilizing two different hold-tank volumes,
and (4) reducing the flue gas fly ash content into the scrubbers
All alternatives considered in this study were simu-
lated with an additional change from existing operating pro-
cedure. The recirculating slurry was specified to be 10 weight
percent solids rather than 2 percent. This increase in solids
concentration will tend to lower the relative saturation
required for a specified set of system operating parameters.
This will also assist with the elimination of chemical scaling
conditions.
The reasons for considering each alternative are
explained in this section. Flowsheets are provided in order
to point out the differences between the models used for the
alternative simulations and the one used for existing opera-
tions The results from each simulation are discussed from a
standpoint of technical feasibility. Finally, conclusions
drawn from these simulations are presented in the last subsec-
tion.
F-21
-------
3.1 Alternative One
One major problem with the Four Corners scrubbing
system is lack of reaction time (hold tank volume) for solid-
liquid mass transfer. Alternative 1 proposes a means of
doubling the solid-liquid reaction time by utilizing existing
tank capacity available in the present system. In this section
the simulation basis for this alternative is presented followed
by the results of the simulation.
3.1.1 Simulation Basis
The present system configuration (Figure 2-1) provides
only minimal solid-liquid contact (approximately one minute) in
the scrubber recirculation loop. This holding time is insuffi-
cient as evidenced by the present scaling conditions. If the
solid concentration in the solid waste stream is increased and
the water make-up requirements reduced, these scaling problems
will worsen.
One method of providing additional reaction time would
be to use the existing thickener transfer tanks as solid-liquid
reaction tanks. This was the first alternative considered. A
schematic flow diagram of Alternative 1 is shown in Figure 3-1.
(The same flow scheme would be used for the 'B' scrubber train
but is not pictured.) The system modifications required to
implement this alternative are largely piping changes.
The major system alteration is that the slurry
recirculation loop has been changed to encompass the transfer
tank and the scrubber reservoir rather than only the scrubber
reservoir. In the existing operational flow scheme, the trans-
fer tank contains only clear liquor from the thickener overflow
and precipitation takes place only by nucleation. With
Alternative 1, approximately one minute of additional solid-
liquid reaction time is provided.
A small tank contained within the transfer tank has
also been proposed. An opening in the small tank at the base
would allow liquor to flow through the small tank to the trans-
fer tank. All streams enter the transfer tank through the
proposed new tank. The function of this tank would be primar-
ily one of mixing the scrubber effluent stream with the lime
slurry additive. This would produce high relative saturations
resulting in controlled nucleation in this tank. It was
reasoned that by providing a small volume tank with high
F-22
-------
FLUE QAS.
INLET
LIME
t
STACK QAS
SCRUBBERS
SCRUBBER
FEED
MAKEUP WATER
NUCLEATION
TANK
'A'
TRANSFER
TANK
THICKNER OVERFLOW
THICKNER
ASH POND OVERFLOW
(DISCHARGED TO
CHACO RIVER)
ASH POND
Figure 3-1.
Schematic flow diagram for
Four Comers Alternative One
F-23
-------
precipitation driving forces sufficient nucleation and precipi-
tation rates might occur in the tanks such that scaling could
be avoided in the scrubber.
Employing this system configuration, the thickener
feed stream would be a slipstream from the transfer tank effluent
(scrubber feed). As such, no lime addition to the thickener
would be required since the thickener feed pH should be between
6 and 8.
In order to simulate Alternative 1 a new model was made
of the scrubbing system. A flow sheet of this model is presented
in Figure 3-2. This model is somewhat simpler than the model
used for existing operations, but is sufficient to illustrate the
effect that a larger hold tank volume has on the CaSOit'ZHaO
scale potential in the scrubbing loop.
3.1.2 Simulation Results
The results of this simulation are presented in Table
3-1. The key to these results is the relative saturation of
CaS(K«2H20 in the scrubber effluent. The value of 1.33 indicates
that scrubber operation under these conditions would be at the
risk of gypsum scale formation. It should be noted here that
calcium carbonate and gypsum precipitation is over 90% nuclea-
tion in this system configuration. Considering the results of
this simulation case, it does not appear that this nucleation
can be controlled and this alternative was judged not to be
technically feasible.
3.2 Alternative Two
Since elimination of scaling potential in the scrubber
could not be achieved by implementing Alternative 1, increasing
reaction tank capacity further was the next alternative consider-
ed. For Alternative 2 to become an operational system, tanks
would have to be installed. Two separate liquid-to-gas ratios
were considered with this alternative. This section presents the
simulation basis and the results of the simulation of Alternative
2.
3.2.1 Simulation Basis
Implementation of Alternative 2 would provide 37,500 m3
(1.33 x 10s ft3) of combined reaction volume for the entire scrub-
bing system. A process flow diagram for this alternative is shown
in Figure 3-3.
F-24
-------
STACK GAS
DIVDR2
10
DIVDER
6
FILTER
5
FLUE QAS
MAKEUP WATER
LIME
SYSTBI
4
•STACK QAS
-5-^-CLARIFIER UNDERFLOW
CLARIFIER
UNDERFLOW
ORDER OFCALCULATIONS
1.2,3,4,5,6(7,8.9,10)
Figure 3-2. Process model for Four Corners Alternative One
F-25
-------
TABLE 3-1.
FOUR CORNERS SCRUBBING SIMULATION
RESULTS FOR ALTERNATIVE ONE*
Flow
Stream
Rate, ll sec
(GPM)
Scrubber
Effluent
160
(2,550)
Nucleation
Tank Effluent
2,840
(45,000)
Transfer
Tank Effluent
2,960
(46,900)
Filter
Bottoms
42
(670)
2.. 8
7.2
7.1
6.9
Suspended Solids, wt
10
10
10
30
Relative Saturations**
CaSO^HjjO
CaC03
Composition, mg/£
Calcium
Magnesium
Sodium
Chloride
Sulfate (as S0~)
Sulfite (as S0~)
Carbonate (as CO )
Nitrate (as N0~)
3
1.33
4.8 x 10~<»
793
82
442
233
2,659
72
116
16
1.3
3.5
802
83
447
236
3,042
76
121
16
1.26
2.3
816
83
449
236
2,722
74
120
16
1.0
1.0
630
82
442
233
2,330
28
111
16
* Makeup water flow rate is 71.5 &/sec (1,130 GPM).
** Critical values, above which scale potential exists, are 1.3-1.4 for
CaS04'2H20, and about 2.5 for CaC03 (see Appendix C).
F-26
-------
STACK GAS
^SCRUBBER FEED
FLUE QAS
WATER MAKEUP
LIME
REACTION TANKS
CLARIFIER OVERFLOW
CLARIHERS
30LID WASTE
TO ASH POND
Figure 3-3. Process flow diagram for
Four Corners Alternative Two
F-27
-------
Basically, the philosophy represented by this alterna-
tive is one of allowing sufficient solid-liquid reaction time so
that nucleation does not occur anywhere within the system. This
can be done by providing sufficient reaction tank volume and sup-
plying the scrubber feed slurry from the reaction tank effluent.
Also, two different L/G's were simulated. The first was the
present design L/G of 4.7 £/m3 @ STP (35.2 gal/lOOOscf). The
second L/G which was simulated was approximately 10 Jl/m3 @ STP
(74.8 gal/1000 scf). The second case was performed because the
simulated scrubber effluent pH was low, about 2.9.
The process model used to simulate Alternative 2 is
presented in Figure 3-4. The principal differences between the
simulation of Alternative 1 and Alternative 2 are (1) the reac-
tion tank volume is much larger in Alternative 2 and (2) only
one reaction tank is simulated in Alternative 2 instead of two
smaller tanks in series as in Alternative 1.
3.2.2 Simulation Results
The results from the two cases considered in Alterna-
tive 2 are presented in Table 3-2. Both cases model a system
which could effectively remove 50% of the S02 from the flue gas
without gypsum scaling. The gypsum relative saturations for
the scrubber effluent liquor are 1.16 and 1.14 for Cases 1 and
2, respectively. These are well below the level required for
the onset of scaling. In fact, these cases represent systems
with very conservative sized reaction tanks. Hold tank sizing
will be addressed in subsequent sections.
A simulated scrubber bottoms pH of 2.9 resulted when
the design L/G, 4.7 £/m3 @ STP (35.2 gal/1000 scf) was employed.
This pH could cause corrosion and possibly other operating prob-
lems if the system operated in this manner for extended periods
of time. With this in mind, the system was simulated using an
L/G of 10.0 £/m3 @ STP (74.8 gal/1000 scf). An increase in L/G
will cause a smaller pH drop across the scrubber since less 362
is absorbed per liter of liquor. Since the increase in acidic
species concentration across the scrubber is smaller with the
higher L/G, the scrubber bottoms pH should rise.
A scrubber bottom pH of 3.9 was calculated for this
case. This operating condition is still not ideal from the
standpoint of corrosion control; however, it is somewhat better
than 2.9 pH scrubber liquor. The proposed L/G is certainly
within the normal operating range for most venturi scrubbers.
F-28
-------
STACK QA3
FLUE QAS
MAKEUP WATER
LIME
CLARIFIER
UNDERFLOW
ORDER OF CALCULATIONS
1,2,3,4,5,6(7,8.8.)
Figure 3-4. Process model for Four Corners Alternative Two
F-29
-------
TABLE 3-2. FOUR CORNERS SCRUBBING SIMULATION RESULTS FOR ALTERNATIVE TWO*
i
u>
o
Stream
Flow Rate, l/sec
(GPM)
pH
Suspended Solids
Relative Saturations**
CaSO-.'2HiO
CaC03
Composition, mg/l
Calcium
Magnesium
Sodium
Chloride
Sulfate (as SOT)
Sulfite (as S0°)
Carbonate (as C0°)
Nitrate (as NO",)
Scrubber
Case 1
2827
(44,800)
2.9
10
1.16
1.0 x 10"'
674
84
447
235
2,432
34
116
15
Effluent
Case 2
6048
(95,800)
3.9
10
1.14
6.7 x 10"6
673
84
446
235
2,431
33
116
16
Reaction
Tank Effluent
Case 1 Case 2
3,016 6
(47,800) (98
6.9
10
1.07 1
1.03 1
674
84
447
235
2,432 2
34
116
16
,235
,800)
6.9
10
.07
.03
673
84
446
235
,431
33
116
16
Filter
Case 1
40.9
(650)
6.9
30
1.0
1.0
631
84
447
235
2,336
28
116
16
Bottoms
Case 2
40.9
(650)
6.9
30
1.0
1.0
631
84
447
235
2,336
28
116
16
* Makeup water flow rate is 70.7 I/sec (1,120 GPM)
Case 1 has a L/G of 4.7 */m3 @ STP (35.2 gal/1,000 scf)
Case 2 has a L/g of 10 */m3 @ STP (74.8 gal/1,000 scf)
**Critical values, above which scale potential exists, are 1.3-1.4 for CaSO,,'2HJ0, and about 2.5 for CaCO, (see Appendix C) .
-------
It is noted that the higher L/G does reduce further the gypsum
relative saturation in the scrubber effluent stream. For these
reasons the remaining simulations were conducted using the
10.0 £/m @ STP liquid-to-gas ratio. Operation of the Four
Corners scrubbing system at this L/G probably will require
increased pumping capacity.
3.3 Alternative Three
Further reduction of water requirements could be
achieved by^recycling the ash pond overflow which is currently
discharged into the Chaco River. This is desirable from a water
use standpoint as well as from an emissions viewpoint. Simula-
tion of this alternative assists in evaluating the impact of
this recycle on the scrubbing system. This section presents the
simulation basis and the results of the simulation of Alterna-
tive 3.
3.3.1 Simulation Basis
Figure 3-5 indicates the flow scheme for this alterna-
tive. This is the same system as was modeled in Alternative 2
with the exception that the ash pond overflow is returned to
the scrubbing system reaction tank. This water has a much
higher total dissolved solids level than does the makeup water
taken from Morgan Lake. The objective of simulating this alter-
native was to measure the impact that the poorer quality water
might have on the operation of the scrubber system.
Two cases were considered in this alternative. The
first case simulated the system with a 37,500 m3 (1.33 x 106
ft) reaction tank. This is the same volume tank which was
modeled in Alternative 2. The second case simulated a system
with perhaps a more realistically sized reaction tank (21,200
m3 or 7.5 x 105 ft3). The process model used to simulate this
alternative is presented in Figure 3-6.
3.3.2 Simulation Results
The results from Alternative 3 are summarized in Table
3-3. The recycle of ash pond overflow to the scrubbing system
has no major impact on the simulated scrubber bottoms gypsum
relative saturation. The water requirements would be reduced
by 19.9 £/sec (315 gpm) with this system configuration.
F-31
-------
I .STACK QAS
QAS
WATER MAKEUP
SCRUBBERS
SCRUBBER FEED
LIME
1
REACTION TANKS
CLARIFIER OVERFLOW
ASH POND OVERFLOW
CLARIFIER
UNDERFLOW
(30% SOLID)
ASH POND
Figure 3-5.
Process flow diagram for Four Corners
Alternative Three.
F-32
-------
STACK QA3
13
SYST84
ORDER OF PROCESS CALCULATIONS
1,2,3,13(4,10, II.) 6.12,6(7,8,0)
Figure 3-6. Process model for Four Corners
Alternative Three.
F-33
-------
TABLE 3-3. FOUR CORNERS SCRUBBING SIMULATION RESULTS FOR ALTERNATIVE THREE*
Stream
Flow Rate, if sec
(GPM)
PH
Suspended Solids , Wt. %
Relative Saturations **
CaSO»-2H20
CaC03
Composition, mg/t
Calcium
Magnesium
Sodium
Chloride
Sulfate (as S0°)
Sulfite (as S0°)
Carbonate (as C0~)
Nitrate (as NO*)
Scrubber
Case 1
6048
(95,800)
3.8
10
1.14
4.5 x 10"6
641
145
793
386
3,017
35
105
25
Effluent
Case 2
6048
(95,800)
3.7
10
1.19
2.7 x 10"6
673
138
757
361
: 3,118
40
93
23
Reaction
Tank Effluent
Case 1 Case 2
6,230 6
(98,700) (98
7.0
10
1.04 1
1.03 1
641
145
793
386
3,107 3
35
105
25
,230
,700)
7.0
10
.12
.05
673
138
757
361
,118
40
93
23
Filter
Case 1
40.8
(650)
7.0
30
1.0
1.0
599
145
793
386
3,014
30
105
25
Bottoms
Case 2
40.8
(650)
7.0
30
1.0
1.0
599
138
757
361
2,952
29
93
23
*Makeup water flow rate is 50.8 Vsec (805 GPM)
Case 1 has a reaction tank volume of 37,500 m3
Case 2 has a reaction tank volume of 21,200 m3
**Critical values, above which scale potential exists, are 1.3-1.4 for CaSCK^HzO and about 2.5 for CaC03 (see Appendix C).
-------
Case 2 represents a reduction in reaction tank
capacity from 37,500 m3 (1.33 x 106 ft3) to 21,200 m3 (7.5 x
10 ft3). The gypsum relative saturation in the scrubber
effluent increased from 1.14 to 1.19. A well designed and well
controlled system can function adequately at gypsum relative
saturations up to 1.25 in the scrubber slurry stream. Above
this control point fluctuations in the operation of a system
would make scale control in the scrubber difficult.
Case 2 represents a realistically sized reaction tank
based on the information available. Further size reduction
simulations were not considered to be cost effective for several
reasons. Proper design of a scrubbing system would require
further testing and data gathering to be performed.
3.4 Alternative Four
One final alternative, a system design where a 6070
efficient particulate control device (such as a mechanical
collector) is placed upstream of the scrubbing system, was
considered. The discussion of this alternative will be brief
since very few changes from Alternative 3 were necessary to
model Alternative 4.
3.4.1 Simulation Basis
Figure 3-7 is a process flow diagram for Alternative
4. A 60% efficient mechanical collector has been located prior
to the venturi scrubbers. Otherwise this diagram is identical
to the flow sheet presented for Alternative 3. Therefore, the
process model used for Alternative 4 is the same as that used
for Alternative 3. The difference between these two alterna-
tives is the fly ash concentration of flue gas which enters the
scrubber. In Alternative 4, this concentration is only 4070 of
that specified for Alternative 3. This, in effect, simulates
a 60% efficient mechanical collector. From the standpoint of
water recycle/reuse, it is assumed that the fly ash removed by
the mechanical collector would be disposed of by dry methods.
With less fly ash solids being removed, the reaction tank volume
required to maintain non-scaling conditions is reduced. Since
more of the recirculated solids will be gypsum, more sites are
provided for precipitation, which reduces the reaction time
required for gypsum formation.
F-35
-------
, STACK QA8
SCRUBBER FEED
WATER MAKEUP
LIME
REACTION TANKS
CLARIFIER OVERFLOW
ASH POND OVERFLOW
CLARIFIER
UNDERFLOW
(30% SOLID)
Figure 3-7. Process flow diagram for
Four Corners Alternative Four
F-36
-------
TABLE 3-4.
FOUR CORNERS SCRUBBING SIMULATION
RESULTS FOR ALTERNATIVE FOUR*
Stream
Flow Rate,
i/sec
(GPM)
Scrubber
Effluent
6,039
(95,700)
Reaction
Tank Effluent
6,142
(97,300)
Filter
Bottoms
18.7
(296)
PH
3.6
7.0
7.0
Suspended Solids
10.1
10
30
Relative Saturations**
CaSO^HzO
CaC03
Composition, mg/&
Calcium
Magnesium
Sodium
Chloride
Sulfate (as S0~)
Sulfite (as So")
Carbonate (as CO )
Nitrate (as N0~)
3
1.19
2.0 x 10~6
692
157
825
447
3,377
44
108
29
1.14
1.05
685
156
821
445
3,225
42
106
29
1.0
1.0
604
156
822
445
3,044
30
107
29
* Makeup water flow rate is 41 £/sec (650 GPM).
**Critical values, above which scale potential exists, are 1.3-1.4 for
CaS04'2H20 and about 2.5 for CaC03 (see Appendix C).
F-37
-------
3.4.2 Simulation Results
Table 3-4 presents a summary of the Alternative 4
results. Two results of a special note should be mentioned
here. First, the reaction tank volume required to achieve non-
scaling conditions is less than that needed in any of the other
technically feasible alternatives considered in this study.
This volume for Alternative 4 is 8,900 m3 (3.14 x 105 ft3) com-
pared to 21,200 m3 (7.5 x 105 ft3) for Case 2 of Alternative 3.
This reduction is directly attributable to the reduction in the
fly ash removal by the scrubbing system. Less of the circulated
solids are inert fly ash and more of the solids are gypsum. The
increase in precipitation sites lowers the reaction time required
for gypsum precipitation.
A second result which is also a consequence of removing
a portion of the fly ash prior to the scrubbing system is the
reduction in water makeup requirements. A 207o decrease (from
Case 2 of Alternative 3) in the water makeup flow is noted.
Again, this result is based on the assumption that dry methods
would be employed in the disposal of the fly ash removed.
It should also be mentioned that lime requirements
are increased in this alternative. Since less fly ash is picked
up by the scrubbing system, less alkalinity is derived from the
sorbed fly ash. The increase in lime flow was from 30 to 40
kg/min.
3.5 Conclusions
This section discusses the conclusions which can be
drawn from the simulation of the various alternatives. A sum-
mary of the simulation results is provided in Table 3-5. The
conclusions are listed below:
1) Alternatives 2, 3 and 4 are technically
feasible from the standpoint of scale
control. It appears that present system
tankage capacity (Alternative 1) is not
sufficient to achieve scale-free operation.
2) Based on data available, it is difficult
to model the Four Corners System from the
standpoint of accurately predicting S02
removal and sulfite oxidation. Further
testing is recommended before detailed
F-38
-------
TABLE 3-5. FOUR CORNERS WATER MANAGEMENT SIMULATIONS SUMMARY
Existing Operations Alternative 1 Alternative 2 Alternative 3 Alternative 4
Hold Tank Volume, m3
L/G, SL/m3 @ STP
Scrubber Bottoms
pH
CaSO,,.H20 R.S. **
Thickener Bottoms
% Solids
PH
Water Makeup
Requirements, a/sec
Case 1
--
4.7
2.6
1.38
10.0
7.7
223
Case 2
--
4.7
2.3
2.52
30.0
7.1
109
Case 1
214*
4.7
2.8
1.33
30.0
6.9
71.5
Case 1
37.500
4.7
2.9
1.16
30.0
6.9
70.7
Case 2
37,500
10.0
3.9
1.14
30.0
6.9
70.7
Case 1
37,500
10.0
3.8
1.14
30.0
7.0
50.8
Case 2
21,200
10.0
3.7
1.19
30.0
7.0
50.8
Case 1
8,900
10.0
3.6
1.19
30.0
7.0
41.0
•^Combined volume of proposed nucleated hold tank and existing transfer tank.
**Based on 98.67. oxidation.
Comments:
Alternative 1:
Alternative 2:
Alternative 3:
Alternative 4:
This alternative would utilize existing thickener transfer tanks as reaction
tanks. Small nucleating tanks within these transfer tanks were also modeled.
The resulting scrubber effluent gypsum relative saturation would cause continued
scaling problems in the scrubber.
To reduce the scaling potential in the scrubber, a much larger reaction tank was
specified in Alternative 2. This lowered the gypsum relative saturation.
However, a 2.9 pH may cause additional operating problems so that L/G was
increased. This resulted in a more reasonable pH.
Alternative 3 involves modeling recycle of ash pond overflow to reduce water
makeup requirements. Case 2 represents a more realistically sized hold tank. A
gypsum relative saturation between 1.2 and 1.25 in the scrubber bottoms stream
is generally acceptable for nonscaling scrubber operation.
Alternative 4 models a system which has a 60% efficient particulate collection
device prior to the wet scrubbing system. A marked decrease in required hold
tank volume is noted. Water makeup requirements would also be reduced if the
fly ash was disposed of by dry methods.
-------
hold tank sizing and any scrubbing system
alterations are attempted. It is felt that
the 5070 S02 removal and 98% oxidation
levels which were specified are adequate
to indicate trends and evaluate alternatives.
This 507o SOa removal corresponds to roughly
0.7 pounds of emitted 862 per million Btu
which is well below the existing Federal
new source standard of 1.2 Ib S02/MM Btu.
3) A large increase in reaction tank volume
will be necessary to eliminate scaling
problems in the scrubber with the existing
oxidation level. More detailed information
on SOz removal rate, venturi contactor
efficiency, effect of fly ash erosion of
scale, and effects of process modifications
on oxidation is essential in correctly sizing
these tanks.
4) The present pump capacity may produce a
scrubber effluent slurry which could cause
operating problems due to low pH. Doubling
the pump capacity (increasing the liquid-to-
gas ratio from 4.7 to 10 £/m3 @ STP or from
35.2 to 74.8 gal/1000 scf) will increase the
scrubber bottoms pH and tend to lessen the
scaling tendency of this stream.
5) Recycle of ash pond overflow has little
impact on the operation of the scrubbing
system. Considerable reduction of water
makeup requirements would be achieved by
implementing this alternative. A major
reduction in makeup water requirements
can be achieved simply by increasing the
thickener underflow from 10% to 307. solids.
6) Separate equipment for fly ash removal
could decrease water requirements and
reduce the size of the reaction tanks
required. Decreased fly ash in the
system may reduce erosion and possibly
alleviate other potential operating diffi-
culties .
F-40
-------
hold tank sizing and any scrubbing system
alterations are attempted. It is felt that
the 50% S02 removal 'and 98% oxidation
levels which were specified are adequate
to indicate trends and evaluate alternatives.
This 50% S02 removal corresponds to roughly
0.7 pounds of emitted S02 per million Btu
which is well below the existing Federal
new source standard of 1.2 Ib SOz/MM Btu.
3) A large increase in reaction tank volume
will be necessary to eliminate scaling
problems in the scrubber with the existing
oxidation level. More detailed information
on SOz removal rate, venturi contactor
efficiency, effect of fly ash erosion of
scale, and effects of process modifications
on oxidation is essential in correctly sizing
these tanks.
4) The present pump capacity may produce a
scrubber effluent slurry which could cause
operating problems due to low pH. Doubling
the pump capacity (increasing the liquid-to-
gas ratio from 4.7 to 10 £/m3 @ STP or from
35.2 to 74.8 gal/1000 scf) will increase the
scrubber bottoms pH and tend to lessen the
scaling tendency of this stream.
5) Recycle of ash pond overflow has little
impact on the operation of the scrubbing
system. Considerable reduction of water
makeup requirements would be achieved by
implementing this alternative. A major
reduction in makeup water requirements
can be achieved simply by increasing the
thickener underflow from 10% to 30% solids.
6) Separate equipment for fly ash removal
could decrease water requirements and
reduce the size of the reaction tanks
required. Decreased fly ash in the
system may reduce erosion and possibly
alleviate other potential operating diffi-
culties .
F-40
-------
4.0 ECONOMICS
This section provides cost estimation for implementing
each^of the technically feasible alternatives discussed in
Section 3.0. Both rough capital costs and operating costs are
presented. The assumptions and techniques used in calculating
these costs are briefly outlined. It should be emphasized here
that these economics are only rough estimates for comparative
purposes.
A capital cost summary for the technically feasible
alternatives is provided in Table 4-1. All of these alterna-
tives involve the addition of six reaction tanks, two agitators
per tank to keep the slurries well mixed, and additional pump-
ing capacity. Alternatives 3 and 4 require piping to recycle
the ash pond water and Alternative 4 uses a cyclone for particu-
late removal before the scrubber. All values are in 1976
dollars.
The tank costs are given for field-erected tanks of
carbon steel construction. These costs include the addition of
a wear liner, mixer supports, baffles, nominal foundations and
plumbing. Engineering and labor costs were estimated for each
to be approximately 2470 of the material costs (GU-075). Terrain
and soil characteristics may require special site preparation
which will add to installation costs and the costs for inter-
connecting plumbing and pumping will also be a function of the
particular site. Agitator costs were determined for twelve
50 hp electrically driven agitators with rubber coated impellers
The additional pumps used to increase the L/G were
assumed to be 300 hp electrically driven centrifugal pumps.
The pumps are rubber-lined and have wear-resistant impellers.
Six pumps will be required with a capacity of about 8,500 GPM
each. The pump used to transport the 650 GPM 30% solids slurry
out to the ash pond will be a 25 hp reciprocal pump. To return
350 GPM of water from the ash pond, a 5 hp centrifugal pump
will be employed. A labor to material ratio of 0.36 was used
for installation costs and engineering was assumed to be 1070
of the combined labor and material costs (GU-075).
The piping costs are for the half mile of pipe
required to transport the 30% slurry out to the ash pond. Five
inch carbon steel pipe with average fittings, flanges, shop
coating, wrapping, and lined with rubber was assumed to extend
the full distance. A labor to material cost ratio of 0.8 was
used to determine the cost of underground installation.
F-41
-------
TABLE 4-1. CAPITAL COSTS FOR WATER RECYCLE/REUSE ALTERNATIVES AT FOUR CORNERS*
i
-p-
r-o
Item
Hold Tanks
Agitators
Pumps and Driver*
Piping
Cyclone
Contingency (201)
Contractural Fees (3D
Total
*Based on 98.6% oxidation in
Comments :
Alternative 2: To reduce
Alternative 2
Alternative 3
Case 1 Case 2
(1976 dollars) (1976 dollars)
2.387.000 2.
312,000
12.000
...
---
542,000
8J.PQ9
3,334,000 4,
the scrubbers.
the scaling potential in the
387,000
312.000
777,000
...
695.000
275,000
scrubber.
Case 1
(1976 dollars)
2,387,000
312,000
777,000
43.000
...
704,000
106.000
4,328,000
a much larger reaction
Case 2
(1976 dollars)
1.565.000
312.000
777,000
43,000
---
539,000
81,000
3.317.000
tank was specified
Alternative 4
Case 1
(1976 dollars)
1.060,000
312.000
777.000
43.000
560.000
550,000
__8_3iOOO
3.385.000
in Alternative 2.
This lowered the gypsum relative saturation. However, a 2.9 pH may cause additional operating problems so
that L/G was increased. This resulted in a more reasonable pH.
Alternative 3: Alternative 3 involves modeling recycle of ash pond overflow to reduce water makeup requirements. Case 2
represents a more realistically sized hold tank. A gypsum relative saturation between 1.2 and 1.25 in the
scrubber bottoms stream is generally acceptable for nonscaling scrubber operation.
Alternative 4: Alternative 4 models a system which has a 60% efficient particulate collection device prior to the wet
scrubbing system. A marked decrease in required hold tank volume is noted. Water makeup requirements would
also be reduced if the fly ash was disposed of by dry methods.
-------
Engineering costs (direct and indirect) were assumed to be 7.2%
of the combined labor and material cost (GU-075).
Table 4-2 presents the operating costs associated
with the different alternatives presented for Four Corners.
These values are reported in 1976 dollars per year. Power
costs were based on an 80% load factor and a wholesale price of
2£/kw-hr for electricity. Capital cost amortization is also
included using 15% per year for a 30 year lifetime.
A comparison of Alternative 3, case 2 with Alternative
4 shows that the reduction in capital cost due to a reduction
in the required reaction volume when cyclones are employed is
roughly offset by the cost of the cyclones. These costs are
based on installation of new dust collectors. Existing col-
lectors are presently inoperable on Units 1, 2, and 3 at Four
Corners, but it may be possible to place these in working order
at less expense than supplying new cyclones.
An alternate method of reducing the water makeup
requirements from Morgan Lake would be to purify the ash pond
overflow and return it to the system rather than routing the
overflow to the Chaco River. (It is not possible to recycle
the ash pond overflow with the existing system configuration
because of scaling problems). However, this stream flow is
about 173 liters/sec (2760 GPM). A brine concentrator/reverse
osmosis unit designed to handle this flow would be about $8.5
million capital investment (RE-211). Furthermore, this method
would not significantly improve the existing scaling problems.
It is emphasized here that these costs are based on a
system which achieves 507o S02 removal. The reaction tanks were
roughly sized accordingly. The economics and system design
could change somewhat in the event that greater than 50%
removal is feasible under the proposed venturi scrubber operat-
ing conditions. With this limitation in mind and without fur-
ther testing, either Alternative 3, case 2 or Alternative 4
would be recommended. Ash pond overflow recycle appears to be
feasible and the higher L/G is recommended. The size of the
reaction tanks depends on whether dust collectors are employed
upstream of wet scrubbers.
An additional 10 liters/sec (160 GPM) of water makeup
can be eliminated by employing dust collectors. Implementation
of Alternative 4 necessitates specific plant information such
F-43
-------
TABLE 4-2. OPERATING COSTS FOR WATER RECYCLE/REUSE ALTERNATIVES AT FOUR CORNERS"
Alternative 2 Alternative 3 Alternatiye4
Case 1 Case 2Case 1Case 2Case 1
Item (1976 dollars) (1976 dollars) (1976 dollars) (1976 dollars) (1976 dollars)
Power for the
Agitators 125,000 125,000 125,000 125,000 125,000
Power for the Pumps 3,000 335,000 335.000 335,000 335,000
Capital Charges
(157. per year) 500fOOP 641.000 649.000 498.000
Total 628,000 1,101,000 1,109.000 958.000
(mlls/VU-hr)** (.128) (.225) (.226) (.195) (.198)
i-rj *Based on 607. load factor
1 **Power production from Units 1, 2, and 3 (700 MM)
.p- Comments:
Alternative 2: To reduce the scaling potential in the scrubber, a much larger reaction tank was specified in Alternative
2. This lowered the gypsum relative saturation. However, a 2.9 pH may cause additional operating problems
so the L/G was increased. This resulted in a more reasonable pH.
Alternative 3: Alternative 3 involves modeling recycle of ash pond overflow to reduce water makeup requirements. Case 2
represents a more realistically sized hold tank. A gypsum relative saturation between 1.2 and 1.25 in the
scrubber bottoms stream is generally acceptable for nonscaling scrubber operation,
Alternative It: Alternative 4 models a system which has a 607. efficient particulate collection device prior to the wet
scrubbing system. A marked decrease in required hold tank volume is noted. Water makeup requirements
would also be reduced if the fly ash was disposed of by dry methods.
-------
as erosion problems caused by fly ash slurry and ease of dry
disposal of the fly ash collected in the dust collection
devices.
F-45
-------
Appendix G. Recycle/Reuse Options at Bowen (Georgia Power Company)
1.0 INTRODUCTION
This appendix describes the analysis of the water
system at the Georgia Power Company's (GPC) Plant Bowen under
EPA Contract No. 68-03-2339, Water Recycle/Reuse Alternatives
in Coal-Fired Steam-Electric Power Plants"This section pre-
sents a summary of the important results of the study concer-
ning Bowen. Bowen was chosen with four other plants for evalu-
ation of the technical and economic feasibility of various
water recycle/reuse options. The major water systems at the
four-unit, 3180 Mw Bowen Plant are the cooling towers and fly
ash and bottom ash sluicing operations.
Three major task areas performed in this study in-
clude :
1) Existing Operations Modeling,
2) Alternatives Modeling, and
3) Economics.
The results of the existing operations simulations
of the cooling towers compare well to the sample data obtained
at the plant. The calculated CaC03 and CaS04-2H20 relative _^
saturations in the cooling tower water (0.1-0.3 and 2.5 x 10
respectively) indicate that the cycles of concentration may be
significantly increased without calcium sulfate (gypsum) scale.
However, an increase in cycles of concentration will probably
require treatment such as acid addition to control calcium car-
bonate scale.
Nine cooling tower simulations were performed to de-
termine the degree of acid treatment necessary for increased
cycles of concentration in the towers (to reduce tower blowdown
quantity) and the effects of increased calcium levels in the
cooling tower makeup water (operational effects of poorer
quality makeup water). No scale potential for CaS0^2H20 was
identified in any of the cases. Sulfunc acid treatment was
required for CaC03 scale control of all cases.
Table 1-1 presents a summary of the technically
feasible options for the Bowen water system as compared to
existing operations and the relative costs of each of these
alternatives. Two process alternatives were studied for the
G-l
-------
TABLE 1-1. SUMMARY OF TECHNICALLY FEASIBLE OPTIONS AT BOWEN
o
Cooling Tower Makeup Source
Cycles of Concentration in
Towers
Cooling System Treatment
Acid Addition Rate, kg/day1
Existing Condition
Makeup Pond ,
Service Water
1.7
None
0 (0)
Alternative One
Makeup 1'ond ,
Service Water
5.7
HjSOi,
481 (1060)
Alternative Two
Makeup Pond ,
Service Water
15.
H2SO,.
608 (1340)
Alternative Three
Makeup Pond, Service Water,
Brine Concentrator Distillate
15.
H2SOu
608 (1340)
(Ib/day)
Ash Sluice Makeup Source
% Recycle in Fly Ash
System
Cooling Tower Blowdown
0
Cooling Tower Blowdown
0
Cooling Tower Blowdown
60
Cooling Tower Blowdown
60
% Recycle in Bottom Ash 0
System
Ash System Treatment None
Plant Makeup Requirements, 3250 (51,500)
I/sec (GPM)
Plant Discharge Rate, 1600 (25,000)
Jt/sec (GPM)
Costs 2
Capital, 1976 $
Operating, 1976 $/yr 3
(rails/kw-hr)
0
None
1880 (29,800)
255 (4050)
100,000
52,900
(.002)
100
Recycle Softening
1670 (26,400)
41 (650)
1,223,000
402,000
(.018)
100
Recycle Softening, Brine
Concentration of Pond
Overflow
1630 (25,800)
0 (0)
6,380,000
1,735,000
(.078)
'AS 100% H2so...
2These rough cost estimates were made to compare technically feasible options and do not include a "difficulty to retrofit" factor.
'includes capital cost amortization at 15% per year.
-------
ash sluicing system at Bowen. The first case involved using
cooling tower blowdown from the towers operating at 5.7 cycles
of concentration to sluice both bottom and fly ash on a once-
through basis at about 10 wt. % solids (Alternative 1 in Table
1-1). The effects of C02 mass transfer in the ash pond and
sluice tank on the system operation were investigated. No gyp-
sum scale potential was identified in any of the cases with
once-through ash sluicing. It should be noted here that this
analysis was performed to study general water recycle/reuse al-
ternatives. Actual implementation of any of these alternatives
would require a more extensive investigation of process parameter
variability. More water quality data would be required along
with additional studies to fully characterize the ash reactivity
variations as a function of time.
Potential scaling of CaC03 is present in all cases
studied. However, the fly ash slurry line possibly can be kept
free of plugging by the addition of a fly ash slurry reaction
tank and by frequent flushing with a water stream of pH 6-7.
Pilot or bench scale testing is recommended to determine accur-
ately the size of reaction tank and frequency and quantity of
acid washing required or if other measures are necessary.
The second alternative for the ash sluicing system
involved using cooling tower blowdown from the towers operating
at 15.0 cycles of concentration as makeup water to a recircula-
ting ash sluice system (Alternative 2 in Table 1-1). If the
pond recycle water remains supersaturated with respect to gyp-
sum, scaling will occur in this system. However, this situation
may be remedied by chemical treatment. Sodium carbonate soft-
ening of approximately 80% of the pond recycle water will main-
tain a gypsum relative saturation of about 1.0 in the slurry
line and prevent calcium sulfate scaling. The calcium carbonate
sludge produced in the softening step may be disposed of in the
ash pond. Problems may also be encountered in the cooling towers
at 15 cycles with silica scale potential. Additional studies
to determine control limits should be conducted before implemen-
ting this alternative.
Zero discharge from the cooling and ash sluicing sys-
tems (Alternative 3 in Table 1-1) may be achieved by installing
a softening/reverse osmosis/brine concentration unit to treat
the ash pond overflow (41 £/sec or 650 GPM) and recycling approx-
imately 50% of the clean water as boiler makeup and the remainder
as cooling tower makeup.
G-3
-------
As with the once-through operations, CaCOs and Mg(OH)2
scale potential was noted but can probably be minimized by in-
stalling a reaction tank prior to the sluice line and frequent
flushing of the line with a pH 6-7 water stream.
The rough cost estimates presented for the alternatives
in Table 1-1 indicate that reducing the ash pond overflow to
225 £/sec(4050 GPM) by running the cooling towers at 5.7 cycles
of concentration and sluicing the ash on a once-through basis
using cooling tower blowdown is the less expensive option (about
$100,000 capital cost with about $53,000/yr operating costs).
This option necessitates acid treatment in the towers.
Reducing the ash pond overflow to about 41 £/sec (650
GPM) by operating the cooling towers at 15.0 cycles of concentra-
tion (with acid treatment) and using the tower blowdown as makeup
to a recirculating ash sluice system (with NaaCOa softening of
807o of the pond recycle) has an initial capital cost of about
$1,223,000 and operating costs including capital cost amortiza-
tion of about $402,000/yr. The inclusion of a softening/reverse
osmosis/brine concentrator unit to eliminate the ash pond over-
flow discharge (recycle to boiler and cooling tower makeup) for
this alternative would require a capital investment of about
$6.38 million total. The additional operating costs would be
about $l,333,000/yr, giving a total of approximately $1,735,0007
yr.
Detailed discussions of the existing operations simu-
lations, the alternative simulations, and the rough cost esti-
mates constitute the main body of this appendix.
G-4
-------
2.0 PLANT CHARACTERISTICS
Plant Bowen is a four-unit 3,180 Mw coal-fired electric
generating station located near Taylorsville, Georgia. The coal
utilized at Bowen is approximately 11% ash and 2.8% sulfur with a
heating value of about 11,500 Btu/lb. The plant employs cooling
towers and once-through bottom and fly ash wet sluicing for all
of the units.
This section of the appendix describes the analysis of
Plant Bowen's water system. First, an overall plant water bal-
ance is presented which shows the major in-plant flows and chemi-
cal analyses for the streams which were sampled. Then a detailed
description of each of the major water consumers in the plant is
given. This is followed by a brief discussion of the process
model and the input data used to simulate existing operations at
Plant Bowen. The computer simulation results are finally presen-
ted and discussed. This discussion will include a comparison of
the simulation results and the chemical analyses of the samples
taken. Areas which show a potential for water recycle/reuse will
be identified and discussed.
2.1 Water Balance
A flow schematic for the Bowen water system is shown in
Figure 2-1. The major streams in the plant, including the cooling
tower and ash handling systems, are shown in this diagram. The
numbers in the diamonds refer to the stream numbers shown below
the schematic where the design flows and results of the chemical
analyses of the spot samples taken at Bowen are presented. A
more detailed description of the samples taken and analytical
procedures used is presented in Appendix B.
Makeup water for the plant is taken from the Etowah
River and stored in a makeup pond. Water is removed from the
makeup pond at a design rate of 3,280 a/sec (52,000 GPM) and
used as general service water, boiler makeup, and cooling tower
makeup.
The general service water effluent is split so that
5% of the flow returns to the makeup pond and 95% is used as
cooling tower makeup. Water treatment wastes (not shown in
Figure 2-1) total about 9.5 s,/sec (150 GPM) and are pumped to
the ash pond. The major water consumers at the Bowen plant
are the cooling tower system and the ash handling systems,
which are discussed in the following sections.
G-5
-------
O
•WAJTfi/P fOHD
O-
-i A
-t B
Figure 2-1. Georgia Power Company
Plant Bowen water balance
-------
A
-------
Stream Number
Stream Name
Flow: Metrlc
English
PH
Calcium
Magnesium
Sodium
Potassium
Chloride
Carbonate (as CDs)
Sulface (as SOt)
Nitrate (as NOj)
Phosphate (as POn)
Silicates (as SlOj)
Suspended Solids
Dissolved Solids
o
Cooling
Tower
Makeup
3,230 Jl/sec.
51,000 gpm
7.7
6.1
1.7
1.4
<0.4
2.1
20.4
1.9
4.0
<0.1
25
<0.01
57
Cooling
Tower
Slowdown
1,900 J./sec.
30,000 gpm
7.9
16.1
2.1
0.2
<0.4
6.4
43.0
3.0
8.4
<0.1
28
<0.01
93
Discharged
Slowdown
320 «./sec.*
5,000 gpn*
7.9
16.1
2.1
0.2
<0.4
6.4
43.0
3.0
8.4
<0.1
28
<0.01
93
Slowdown
To Sluice
Ash
1,580 H/sec.*
25,000 gpn
7.9
16.1
2.1
0.2
<0.4
6.4
43.0
3.0
8.4
<0.1
28
<0.01
93
O
Ash
Pond
Overflow
1,580 I/sec.
25,000 gpm
10.4
89
1.7
19.6
5.4
8.2
24
182
11.2
<0.1
55
<0.01
374
Service
Water To
Cooling
Towers
670 fc/sec,
10,500 gpm
Service
Water
Slowdown
35 £/sec.
550 gpm
<•>
Bottom
Ash
Sluice
1,230 £/sec.*
19,500 gpm
6.5
21.6
2.3
1.5
1.5
3.5
39
38.4
5.2
<0.1
30
1.11
139
<£>
Fly
Ash
Sluice
350 £/sec.
5,500 gpm
11.5
311
<0.1
9.4
19.8
3.9
22
514
9.5
<0.1
7.0
1,370
HX*
Plant
Makeup
Water
3,280 H/sec.
52,000 gpm
o
I
*These stream flows vary due to the
and was obtained as the difference
blowdown, when not used for ash si
periodic sluicing of bottom ash. The value shown for Stream 8 assumes full load operation
between the cooling tower blowdown and the fly ash sluice water. Excess cooling tower
uicing, is discharged (Stream 3).
Figure 2-1. (Continued)
-------
The first step in characterizing the chemistry of the
Bowen water system is to examine the results of the spot samples
taken. The measured species concentrations were input to the
equilibrium program and several parameters were calculated which
determine the tendency of the liquor sampled to form chemical
scale and to absorb or desorb CO2 from the atmosphere. Another
parameter calculated checks the internal consistency of the
sample and is a measure of the analytical accuracy.
Table 2-1 presents a summary of the parameters calcu-
lated for each of the samples taken at Bowen. Relative satura-
tions for CaC03, Mg(OH) 2, and CaSOlt-2H20 are given in the first
three columns. These parameters indicate the tendency of the
stream to form scale. Critical values for relative saturation
of each species, above which scale formation is likely, are
2.5 for CaC03, 3.4 for Mg(OH)2, and 1.3-1.4 for CaS04.2H20 (see
Appendix C).
None of the streams sampled showed a tendency to form
Mg(OH)2 or CaSOi^HaO scale. The highest gypsum relative sat-
uration was 0.28 in the fly ash slurry water, well below the
critical range of 1.3-1.4. Three of the seven streams sampled
showed CaCO3 relative saturations greater than the critical
value of 2.5. The fly ash sluice and pond samples showed CaC03
relative saturations of 17.1-38.8. The decrease in relative
saturation from the fly ash sluice to the pond is most probably
a result of CaC03 precipitation and/or C02 absorption from the
atmosphere.
The equilibrium partial pressures shown in Table 2-1
are an indication of the tendency of a liquor to absorb or
desorb C02 when in contact with the atmosphere. A value less
than 3 x 10" \ the equilibrium partial pressure of C02 in air,
indicates a tendency to absorb C02 and a value greater indicates
a tendency to desorb C02. The value for the fly ash sluice
water is 7 x 10" * ° atm indicating a strong tendency to absorb
C02. The value for C02 partial pressure of the pond water,
2 x 10"7, is larger than the fly ash sluice but still less than
atmospheric which indicates that some C02 transfer is occurring
in the pond but that complete equilibrium is not achieved. The
value for the cooling tower blowdown sample is about 5 x 10-1*
indicating that the cooling tower blowdown C02 concentration is
very near the equilibrium value.
G-7
-------
TABLE 2-1. PARAMETERS CALCULATED BY EQUILIBRIUM PROGRAM FOR BOWEN SAMPLES
a
00
Stream
Stream Name Number
Cooling Tower Makeup 1
Cooling Tower Slowdown 2
Bottom Ash Sluice 8
Fly Ash Sluice 9
Ash Pond Subsurface
Ash Pond Effluent 5
Plant Drainoff Water
Relative Saturations *
CaC03 Mg(OH)2
0.
0
0.
38.
17.
17.
0.
,016 6.3 x 10~7
.15 2.8 x 10~6
,01 3.2 x 10"e
,8
4 0.065
1 0.065
31 1.7 x 10"5
CaSCV2H20
1.0 x 10""
3.0 x 10'"
5.0 x 10""
0.28
0.056
0.058
0.003
Equilibrium partial
pressure of C02, atm
3.
5
9.
7
2,
2.
1,
,95 x 10""
.33 x ID"1*
.5 x 10"3
.0 x 10"10
.0 x 10"7
.0 x 10"7
.15 x 10""
% Residual
Electroneutrality
-4.
-12,
-5.
-3.
+10
+4
+6
0
.8
.0
.9
.0
.0
.0
* Critical values, above which scale potential exists, are 1.3-1.4 for CaS04-2H20, about 2.5 for CaC03> and about 3.4 for Mg(OH)2 (see Appendix C).
-------
The last parameter shown in Table 2-1, % residual
electroneutrality, is calculated to determine the internal con-
sistency of each sample with pH specified. A value of ±15%
is considered acceptable. All of the Bowen samples were within
this range. A more detailed description of how this parameter
is calculated is presented in Appendix E.
2.2 Cooling Tower System
Each of the four units at Plant Bowen have independent
cooling systems with one cooling tower for each unit. Units 1
and 2 are identical (700 Mw) and have identical cooling towers.
Units 3 and 4 are each rated at 880 Mw and also have identical
cooling towers. Water circulates between the condenser and
the cooling tower of each unit at a rate of 16,280 £/sec
(258,000 GPM) for Units 1 and 2 and 19,530 £/sec (310,000 GPM)
for Units 3 and 4. A blowdown stream is removed from the cir-
culating water after the condenser. The water removed as blow-
down is replenished with fresh makeup water.
The blowdown rate is maintained so that the dissolved
species concentrations remain low enough to prevent scaling in
the condenser. The relationship between the blowdown rate, the
cooling tower evaporation rate, the drift rate, and the amount
of concentration that dissolved species undergo is expressed
below:
C -
where
C = cycles of concentration
E = evaporation rate
B = blowdown rate
D = drift rate
Present operation of the Bowen cooling system maintains the
blowdown rate so that the makeup water is concentrated about
1.7 times (i.e., C ^ 1.7). This allows the towers to operate
scale-free without acid addition to control pH. Excess blow-
down is discharged.
G-9
-------
2.2.1 Simulation Basis
Existing cooling tower operations at Plant Bowen were
simulated to verify the model validity so that water recycle/
reuse alternatives can be evaluated. These simulations will
also help identify potential areas for recycle/reuse of water
at Plant Bowen. This section presents the basis for simulating
existing operations for the cooling towers at Bowen. First,
the process model used is briefly described followed by a pre-
sentation of the input data. A detailed description of the pro-
cess model is included in Appendix E.
The existing operations of the Bowen water system
were simulated by means of the computer model shown in Figure
2-2. This is a generalized cooling tower model with capabil-
ities of simulating sulfuric acid addition and slipstream soft-
ening for calcium removal. Neither option was used for existing
operations.
Given the inputs of air flow, temperature and compo-
sition, makeup water composition, flow and temperatures of the
circulating water, drift rate, and cycles of concentration, the
model performs iterative calculations around the cooling loop
to determine the blowdown, evaporation and makeup rates, and
compositions for all water streams. An acid addition rate
(if required) is determined to keep the CaC03 relative satura-
tion within a specified range. If slipstream softening is re-
quired (determined by model) the slipstream and chemical addi-
tion rates are calculated.
Several assumptions are inherent in performing this
simulation with the subroutines outlined above. These assump-
tions are enumerated below:
1) Equilibrium exists between C02 and
H20 in the atmosphere and cooling
tower exit water.
2) The temperature of the cooled water
stream approaches the wet bulb tem-
perature of ambient air within a
predictable range.
3) The compositions and temperatures of
the cooled water and drift streams
are equal.
G-10
-------
o
I AIR I ' ^
1 (PLUGS l>) *"
X. j ^r^ 2 ^^V
^ -^ /WATER \ „
/ RE- \ 2 fc.
ICYCLE INI ^
MWTRINP)/ ' '
X^4 ^X
/ ACID \ 4
yWTRINPjy II
\s^_^^ ^ (HLDTK 0
15
CHEMICALS 5
ADDITION 1 • te- lo
k(ALKINP)/ WATER '3
^ w/ TREAT- ^
^^^-^ j<< 10 MENT
WASTE (CHMTRT)
A in ,, »^
3
MAkPtlP UUATPQ _.. ._ ,^^
4
n 1. 1 U — ^"
5
SOP 1 LNINu CyML.MIL.ALo ^*
6
COOLING
TOWER
(CLGTR 1)
14
TEE
(DIVDR3)
7
OVERALL
SYSTEM
BALANCE
-------
4) Ionic reactions taking place in the
liquid phase are rapid and thus at
equilibrium.
The assumption involving the temperature of the cooled
water stream is a recognized design parameter in cooling tower
evaluation and gives a good approximation. The assumption con-
cerning the temperature and composition of the drift stream
should be very close to actuality as is the assumption in regard
to H20 gas-liquid equilibrium. The assumption with regard to
CO2 equilibrium is conservative since the partial pressure of
CO2 in actual cooling towers tends to be greater than the equi-
librium value. The assumption in the model causes the pH to be
slightly higher in the model than in actual operation. The
higher pH causes the relative saturation of CaCO3 to increase
more than the lowered carbonate species concentration causes it
to decrease.
A summary of the input stream data employed in the
existing operations simulations is presented in Table 2-2.
The cooling tower design air flows were obtained from GPC and
adjusted to the temperature and water content shown. The air
temperature and composition were calculated using local clima-
tological data for Atlanta between December, 1974, and November,
1975. The makeup water composition was obtained from chemical
analyses.
The cooling tower drift rate, approach, cycles of
concentration, and circulating water flow were obtained direct-
ly from GPC or calculated from data obtained from GPC. The
condenser temperature change was also obtained from GPC. The
ambient air wet bulb temperatures were derived from Atlanta
climatological data for December, 1974, to November, 1975.
2.2.2 Simulation Results
This section describes the results from the simulation
of existing cooling tower operations at Bowen. Three simula-
tions were performed:
1) Cooling Towers 1 & 2, Summer Operation
2) Cooling Towers 1 & 2, Winter Operation
3) Cooling Towers 3 & 4, Summer Operation
G-12
-------
TABLE 2-2. INPUT DATA FOR BOWEN COOLING TOWER SIMULATIONS
FLOWS
Air, m3/hr
(ACFM)
Drift, A/sec
(GPM)
Circulating Water, £/sec
(GPM)
TEMPERATURES
Ambient Air, °C
Approach, °C
Condenser AT, °C
Wet Bulb, °C
Condenser Outlet, °C
ADDITIONAL DATA
Relative Humidity, %
Cycles of Concentration
Makeup Water Composition, ing/ 5,
Calcium
Magnesium
Sodium
Chloride =
Carbonate, as CQ^
Sulfate, as SOij
Nitrate, as NOs
Units
Summer
2.46 x 10 7
(1.45 x 107)
3.3
(52)
16 , 300
(258,400)
23.9
(75)
10.6
(19)
14.2
(25.6)
21.1
(70)
46.1
(115)
78.0
1.7
6.0
1.7
1.4
2.1
20.4
1.9
4.3
1 & 2
Winter
3.6 x 107
(2.12 x 10 7)
3.3
(52)
16,300
(258,400)
7.8
(46)
10.6
(19)
14.2
(25.6)
5.6
(42)
30.6
(87)
73.0
1.7
6.0
1.7
1.4
2.1
20.4
1.9
4.3
Units 3 & 4
Summer
3.52 x 107
(2.07 x 10 7)
3.9
(62)
19,500
(310,000)
23.9
(75)
10.0
(18)
15.6
(28)
21.1
(70)
46.7
(116)
78.0
1.7
6.0
1.7
1.4
2.1
20.4
1.9
4.3
ff-13
-------
The simulation results will be compared to the chemi-
cal analyses results from the samples gathered at Bowen in order
to evaluate the validity of the process model. These simula-
tions will be used to discuss possible water recycle/reuse al-
ternatives at Bowen, and potential problems which could be
caused by implementing alternatives will be discussed.
A summary of the simulation results for existing
cooling tower operations at Bowen is presented in Table 2-3.
The first column in Table 2-3 shows a summary of actual plant
data concerning the cooling tower blowdown characteristics for
cooling tower number 3. Process simulations were performed for
the summer and winter operation of towers 1 and 2 and the summer
operation of towers 3 and 4.
The cooling tower blowdown pH value of 7.93 for the
simulated operation (Case 3) compares very well to the measured
value of 7.9.
Comparison of some of the sample data with simulation
results indicates that the system may not have been at steady
state during sampling. For example, measured calcium and chlor-
ide concentrations are higher than the computed values whereas
magnesium and sodium measured concentrations are lower than the
simulation results. These discrepancies may also be due to
nonhomogeneous sampling and/or analytical errors as well as un-
steady-state operation.
The sulfate and nitrate concentrations measured com-
pare favorably with the simulated values. The calcium carbon-
ate relative saturations are consistent in that the system is
operating well below the critical level for scaling of 2.5
(see Appendix C).
Calcium carbonate relative saturation is very depen-
dent on pH due to the carbonate-bicarbonate-carbonic acid equi-
librium in solution and accounts for the differences shown.
This dependence on pH is illustrated by comparing the values
for Case 2 to the values for Case 3. The blowdown concentra-
tions are approximately the same, yet the respective relative
saturations are 0.11 and 0.30. The lower pH of 7.87 as opposed
to 7.93 is enough of a difference to lower the relative satur-
ation from 0.30 to 0.11, even with a slightly higher carbonate
concentration in Case 2.
G-14
-------
TABLE 2-3. BOWEN EXISTING COOLING TOWER OPERATIONS
o
i
Cooling Tower Slowdown
Flow, £/sec per tower
(GPM)
pH
Composition, mg/£
Calcium
Magnesium
Sodium
Chloride
Carbonate, as CO 3
Sulfate, as SOit
Nitrate, as N03
Relative Saturations *
CaCOs
CaSOit'ZHaO
Partial Pressure COa , atm
Plant Data
(Tower No. 3)
442
(7,000)
7.9
16.1
2.1
0.2
6.4
43.0
3.0
8.4
0.15
3.0 x 10"*
5.33 x Hf*
Case 1
Towers 1&2 Summer
467
(7,400)
7.94
10.3
2.9
2.3
3.6
33.8
3.3
7.4
.30
2.5 x 10~"
5.63 x 10"1*
i
Simulations
Case 2
Towers 1&2 Winter
391
(6,200)
7.87
10.2
2.9
2.3
3.6
34.3
3.3
7.4
.11
2.4 x lO"1*
5.11 x 10'*
Case 3
Towers 3&4 Summer
625
(9,900)
7.93
10.3
2.9
2.3
3.6
33.8
3.3
7.4
.30
2.5 x 10'"
5.89 x 10~k
* Critical values, above which scale potential exists, are 1.3-1.4 for CaSO^«2H20 and about 2.5 for
CaC03 (see Appendix C).
-------
The low measured calcium sulfate relative saturations
are confirmed by the simulation results. The critical value
of relative saturation for scaling of calcium sulfate to occur
is 1.3-1.4.
These simulations indicate a potential for reducing
water requirements and discharges for the cooling towers by
increasing the cycles of concentration but not so much as to
reach the critical scaling level for calcium carbonate. If
acid treatment for calcium carbonate scale control is instituted
in the Bowen cooling tower system, the cycles of concentration
may be increased until calcium sulfate or some other species
such as CaHPOi,(s) reaches the respective critical value for
scaling. The decreased cooling tower blowdown resulting from
operation at higher cycles of concentration might then be used
for ash sluicing on a once-through basis or in a recirculating
system.
The effects of increased cycles of concentration on
cooling tower blowdown and the subsequent use of the cooling
tower blowdown in an ash sluicing system are discussed in Sec-
tion 3.0. The effects of increased calcium in the makeup water
are also investigated to determine the operational effects of
poorer quality makeup water.
2.3 Ash Handling Systems
Fly ash is collected by electrostatic precipitators
at a rate of about 24,200 kg/hr (53,300 Ib/hr) from Units 1
and 2 and about 22,500 kg/hr (49,500 Ib/hr) from Units 3 and 4.
The total rate of collection is therefore about 93,400 kg/hr
(205,600 Ib/hr). The collected fly ash is sluiced on a once-
through basis to the ash pond using cooling tower blowdown as
sluice water. Sluicing this amount of fly ash at about 7%
solids (see sample analyses in Figure 2-1) requires 350 a/sec
(5,500 GPM) of water. Detailed calculations showing how the
fly ash rates and water rates were obtained are presented in
Appendix K.
Bottom ash is periodically sluiced with cooling tower
blowdown to the ash pond also on a once-through basis. The re-
mainder of the cooling tower blowdown that is not used for
sluicing fly ash is used to sluice the bottom ash at about 1%
solids (see sample analyses in Figure 2-1). This water is
discharged when it is not used to sluice the bottom ash.
G-16
-------
2.3.1 Simulation Basis
Existing operations of the Bowen fly ash sluicing
system were simulated by means of the computer model shown in
Figure 2-3. This was done to verify the model and establish a
basis for comparison. The model uses information about the
composition and flows of the makeup water and the fly ash as
well as the percent solids in the sludge and pond evaporation
as inputs. From this information the flows and compositions of
all the streams are calculated. A detailed description of the
ash sluicing model is given in Appendix E.
Several assumptions were made in modeling the ash
sluicing system with this simulation. These include:
1) Solid-liquid equilibrium is achieved
in the ash pond, with the exception of
CaS04 which is allowed to remain super-
saturated.
2) Ash dissolution is essentially complete
before the slurry reaches the pond.
3) All solids precipitation occurs in
reaction vessels or the pond.
4) Ionic reactions taking place in the liquid
phase are rapid and thus at equilibrium.
The input data required to simulate the once-through
ash sluicing system at Bowen are presented in Table 2-4. Sluice
water rates were calculated based on a fly ash slurry solids
content of about 7 wt. "L and a bottom ash slurry solids content
of about 1 wt. 7o. The pond evaporation rate was calculated
based on average wind speed, ambient air composition, pond
surface area, and pond surface temperature. The sluice water
composition (cooling tower blowdown) was obtained from the
results of the existing operations cooling tower simulation
previously discussed.
The fly ash flow rate was obtained from precipitator
inlet and outlet grain loadings obtained from GPC. The bottom
ash flow was calculated as the difference between the total ash
from the coal and the fly ash. The soluble species data for the
fly ash were obtained from ash characterization studies performed
in support of this program (see Appendix L). Calculations per-
formed in obtaining this input data are explained in Appendix K.
G-17
-------
o
I-1
00
MAKEUP FOR FLY ASH SYSTEM
MAKEUP FOR BOTTOM ASH SYSTEM
*-7 SLUICE WATER
VAPORIZED
9 FLY ASH SLUICE
VAPORIZATION
10BTM ASH SLUICE
VAPORIZATION
BOTTOM ASH SLURRY
ORDER OF PROCESS CALCULATION: 1. 2. 3. 4. 5. 6. 7. 8. 9. 10*
Figure 2-3. Process simulation scheme for Bowen ash sluicing system.
02-1529- I
-------
TABLE 2-4. BOWEN EXISTING ASH SLUICING INPUT DATA
FLOWS (Unit 3)
Fly Ash, kg/min
(Ib/min)
Bottom Ash, kg/min
(Ib/min)
Fly Ash Sluice Water, &/sec
(GPM)
Bottom Ash Sluice Water, £/sec
(GPM)
Pond Evaporation*, if sec
(GPM)
375.1
(825.3)
205.3
(451.7)
83.3
(1320.0)
340.0
(5387.0)
1.1
(18)
SLUICE WATER COMPOSITION
Cooling Tower Slowdown @1.7 cycles, mg/&
Calcium
Magnesium
Sodium
Chloride
Carbonates, as CO3
Sulfates, as SOT
Nitrates, as NOl
10.2
2.9
2.4
3.6
33.8
3.3
7.4
POND DEPOSITS, wt. % solids
40.0
SOLUBLE FLY ASH SPECIES, wt. %
CaO
MgO
Na20
CaS04
0.21
0.0
0.15
1.25
* for entire plant.
G-19
-------
2.3.2 Simulation Results
Two simulations were performed for existing ash slurry
operations at Bowen. The first did not allow C02 transfer in
the ash pond and the second allowed the C02 in the pond to come
to equilibrium with the atmosphere.
Table 2-5 presents the results of these simulations.
The compositions of the pond liquor and the fly ash slurry sam-
pled at the plant are compared to those predicted by the model.
The results for the fly ash slurry stream show that
the model predicted higher calcium and sulfate values than the
sample showed resulting in a slightly higher gypsum relative
saturation (0.38 versus 0.28). A calcium carbonate relative
saturation about three times that of the plant data was pre-
dicted by the model due to the higher calcium and carbonate
values predicted. Since the sluice water carbonate concentra-
tion was 43 mg/& (plant data for cooling tower blowdown) some
calcium carbonate precipitation is indicated although no scale
buildup has been reported for Bowen. The erosive character of
the ash slurry may be preventing the buildup of scale by scrub-
bing the sluice pipe walls. Once the liquor reaches the pond,
there is little solid-liquid mixing, which may account for the
high CaC03 relative saturation calculated based on the pond
sample.
The pond liquor results show that actual operation
at Bowen is closer to no C02 transfer in the pond since the
sample pH is 10.4 and the predicted values are 10.8 for no C02
transfer and 8.2 for C02 equilibrium. The differences in com-
position may be attributed to non-steady-state operation. Due
to the long residence time in the pond, any process changes re-
sult in very slow system response. Thus, the pond liquor com-
position may vary in time with changes in the fly ash reactivity,
slurry solids content and load, but the variations will be
damped and will correspond more closely to a time averaged
composition than a spot sample.
G-20
-------
TABLE 2-5. BOWEN EXISTING ASH SLUICING OPERATIONS
o
NJ
Fly Ash Slurry
Plant Data
Composition, mg/&
Calcium 311.0
Magnesium <0.1
Sodium 21.2
Chloride 3.9
Carbonates, as COj 22.0
Sulfates, as SOV 514.0
Nitrates, as NO^ 9.5
£H 11.5
Relative Saturations*
CaC03 38.8
Mg(OH)2
CaS04'2H20 0.28
Equilibrium Partial
Pressure of C02 , 7 x 10
atm
Model
405.0
2.9
2.4
3.7
34.1
674.0
7.5
11.0
104.4
39.4
0.38
1 x 10~8
Plant Data
89.0
1.7
22.8
8.2
24.0
182.0
11.2
10.4
17.1
.065
0.058
2 x 10~7
Pond Liquor
Model
(No C02 Transfer)
65.6
2.6
2.4
3.7
1.0
135.0
7.5
10.8
1.0
1.0
0.038
2.1 x 10~9
Model
(C02 Equilibrium)
72.2
2.9
2.4
3.7
54.3
135.0
7.5
8.2
1.0
7.3 x 10~8
0.041
3 x 10~4
* Critical values, above which scale potential exists, are 1.3-1.4 for CaSOi»-2H20, about 2.5 for CaC03,
and about 3.4 for Mg(OH)2 (see Appendix C).
-------
3.0 TECHNICAL ALTERNATIVES
A modular approach to studying water recycle/reuse
alternatives at Bowen was used in that the major plant water
systems were divided into two subsystems to form separate pro-
cess simulations. One subsystem consists of the cooling towers
with associated treatment facilities (where necessary), hold
tanks, and condensers. The other sybsystem consists of the ash
handling systems. The studies for each subsystem will be dis-
cussed separately. The effects of increasing the cycles of con-
centration in the cooling towers and of poorer quality makeup
water (increased calcium levels) are presented first. Then the
use of cooling tower blowdown in a once-through and a recircu-
lating ash sluice system is evaluated. The effects of carbon
dioxide mass transfer between the atmosphere and the pond
liquor are also investigated.
3.1 Cooling Tower System
The existing operations simulations indicated that
the cycles of concentration may be greatly increased in the
cooling towers without scaling with respect to calcium sulfate,
but only limited increases in the cycles of concentration may
be implemented before calcium carbonate reaches saturation.
However, calcium carbonate scaling potential can be controlled
with acid treatment of the circulating water. This section
first presents a description of the simulation bases used,
then a discussion of the results with respect to increased
cycles of concentration and calcium in the makeup water.
3.1.1 Simulation Basis
The process model used to simulate alternatives for
cooling tower operation is identical to that used for existing
operations (see Appendix E). Acid treatment for calcium
carbonate scale control was implemented to keep the CaC03
relative saturation between 0.5 and 1.0.
A total of nine simulations were performed for alter-
native cooling tower operations. Three simulations were per-
formed with the existing makeup water quality and cycles of
concentration of 5.7, 10.0, and 15.0. Three simulations were
conducted with the calcium concentration in the makeup water
doubled, and three additional cases were run with the calcium
level tripled. Cycles of concentration values used in these
simulations were also 5.7, 10.0, and 15.U so that correlations
could be made using all nine runs. Since calcium carbonate
G-22
-------
relative saturation is the limiting factor of cycles of concen-
tration and the carbonates are essentially fixed by being in
equilibrium with the atmosphere, the acid requirements will
correlate to the calcium levels. Any changes in the calcium
concentration of the makeup water will necessitate changes in
the acid addition rate. The increased calcium runs were per-
formed to determine the magnitude of those changes.
All of the alternative cooling tower simulations were
performed for summer operation of towers three and four since
these conditions represent the case of maximum blowdown rates.
Increased evaporation rates realized during the summer months
necessitates an increase in blowdown rate over that required
during the winter months to maintain a constant value for cycles
of concentration.
The only changes in the input data for the first
three alternative simulations are the values for cycles of
concentration. The makeup water compositions used in the last
six cooling tower simulations were changed from existing data
by increasing the calcium concentration. It should be noted
that the chloride concentrations were adjusted in addition to
the calcium levels. This was done to maintain a solution pH in
the same range as the existing makeup water.
3.1.2 Effect of Increased Cycles of Concentration
The simulation results from the first three alterna-
tive cooling tower operation runs are presented in Table 3-1
along with the results for existing operations. These three
alternate simulations represent tower operation at 5.7, 10.0,
and 15.0 cycles of concentration. Sulfuric acid treatment was
required to control calcium carbonate scale potential for all
three cases, confirming the indications of the existing opera-
tions simulations that only limited increases in cycles of
concentration could be achieved without treatment.
The first alternate operating run (Case A) was made
with 5.7 cycles of concentration. Increasing the cycles from
1.7 to 5.7 will require acid treatment as shown in Table 3-1.
However, no calcium sulfate scale potential will be realized
since the relative saturation is 0.012, well below the critical
scaling value of 1.3 - 1.4. Although the critical scaling value
for relative saturation of CaC03 is about 2.5 (Appendix C),
acid addition requirements were calculated based on keeping
CaC03 subsaturated. This will minimize the effects of upsets in
G-23
-------
TABLE 3-1.
o
i
EFFECT OF INCREASED CYCLES OF CONCENTRATION IN
BOWEN COOLING TOWERS*
Case No.
Cycles of Cencentration
Makeup Water Rate, £/sec
(GPM)
Acid Addition Rate, kg/day**
(Ib/day)
Slowdown
Flow, £/sec
(GPM)
pH
Composition, mg/£
Calcium
Magnesium
Sodium
Chloride =
Carbonates (as CO 3)
Sulfates (as SOT)
Nitrates (as NOJ)
Temperature, °C
(°F)
Relative Saturations***
CaCOs
CaSO^HaO
Existing Operation
3
1.7
1060
(16800)
0.
(0.)
625
(9900)
7.9
10.3
2.9
2.3
3.6
33.8
3.3
7.4
46.7
(116)
0.30
2.5 x 10~"
Increased
4
5.7
530
(8400)
481
(1060)
89.8
(1420)
8.0
34.3
9.7
7.9
12.2
39.2
69.7
24.8
46.7
(116)
0.97
0.012
Cycles of Concentration
6
10
485
(7690)
567
(1250)
45.0
(714)
7.9
60.2
17.0
13.8
21.3
32.1
152
43.4
46.7
(116)
0.94
0.034
5
15
468
(7420)
608
(1340)
27.5
(436)
7.7
90.2
25.5
20.7
32.0
20.9
250
65.0
46.7
(116)
0.51
0.066
* All flows for Unit 3 or 4; the existing makeup water quality for summer months was used as a basis.
**As 100% EzSOk .
***Critical values, above which scale potential exists, are 1-3-1.4 for CaSOit-2H20 and about 2.5
for CaC03 (see Appendix C).
-------
the system which might cause calcium and carbonate levels to
reach the critical limit, such as increases in calcium content
of the makeup water.
A cycles of concentration value of 5.7 for all towers
will produce a blowdown rate for the plant of 312 £/sec (4950
GPM) which will provide a once-through ash sluicing system at
Bowen_with enough water for a 10% solids ash slurry. The impact
of using the blowdown water from the towers operating at 5.7
cycles of concentration will be discussed in Section 3.2.2.
Case 5 represents tower operation at 15.0 cycles of
concentration. Even when the towers are operated at this high
level of concentration, no scaling problems are noted as long
as acid treatment is used for calcium carbonate scale control.
The calcium sulfate relative saturation for Case 5 was only
0.066, still well below the critical level of 1.3 - 1.4. Acid
requirements were increased by 127 kg/day (280 Ib/day) from
the calculated requirements of 481 kg/day (1060 Ib/day) for
operation at 5.7 cycles of concentration.
Case 6 was run to provide an additional data point
for determining the effects of cycles of concentration on acid
requirements for a given makeup water quality. The results
from this run and Cases 4 and 5 are consolidated with the data
obtained from the increased calcium cases in Section 3.1.3.
Graphs depicting the effects of cycles of concentration and
calcium in the makeup water on the acid requirements are pre-
sented.
Operating the cooling towers at higher cycles of
concentration may cause species other than gypsum or calcium
carbonate to become supersaturated and possibly form scale.
Table 3-2 shows relative saturations for silica and phosphate
solids in addition to the species already considered in the
cooling tower blowdown at 15 cycles. None of the phosphate
solids are above saturation but two silica solids are super-
saturated. The Si02 relative saturation is 1.28 and the
Mg(Si02)3(OH)2 (sepiolite) relative saturation is 14.0. These
solids may cause problems at Bowen at 15 cycles of concentra-
tion but the respective critical values for these species are
not known. The kinetics of the solid precipitation will
determine if these solids will cause problems. Additional
testing should be performed to determine the control limits for
silica solids before implementing water recycle/reuse alterna-
tives which require increased cycles of concentration in the
G-25
-------
TABLE 3-2. RELATIVE SATURATIONS OF SCALE-FORMING
SPECIES AT 15 CYCLES IN BOWEN COOLING
TOWERS*
Species Relative Saturation
Ca(OH)2 1.5 x 10~9
CaC03 0.52
CaSO^HjO 0.066
CaHPO., 0.031
Ca3(P04)2 2.2 x 10"3
Mg(OH)2 2.7 x 10""
MgC03 4.0 x 10"5
Si02 1.28
Mg(Si02)3(OH)2 14.0
MgaSizOsCOH), 0.36
CaH2SiO,, 3.0 x 10~ 3
Ca(H3SiOO2 0.075
*This simulation required an acid addition rate of 608 kg/day
(1340 Ib/day)
G-26
-------
towers. If necessary the silica concentration can be lowered
by lime-soda ash or magnesium bicarbonate softening of either
the makeup water or a slipstream from the circulating water
(RO-266, TH-192). ft
3.1.3 Effect of Calcium Concentration in the Makeup Water
Six additional cooling tower simulations were per-
formed. Three runs were made at cycles of concentration of
5.7, 10.0, and 15.0 with the calcium level in the makeup water
tripled. Three runs with double calcium in the makeup water
were also performed at cycles of 5.7, 10.0, and 15.0. The
results from these six simulations are presented in Table 3-3.
As in the cases presented in the previous section, acid
addition rates were calculated to produce calcium carbonate
relative saturations between 0.5 and 1.0, as shown in Table
3-3.
As can be seen from the relative saturations given in
Table 3-3, even in the worst case (triple calcium, 15.0 cycles),
scaling potential of calcium sulfate is nonexistent. The high-
est relative saturation was 0.13 (Case 9) which is significantly
below the critical scaling level of 1.3 - 1.4.
Also shown in Table 3-3 is the acid addition rate
expressed as a ratio of acid to calcium in the circulating
water stream on a molar basis. This ratio is plotted versus
cycles of concentration and calcium level in the makeup water
in Figures 3-1 and 3-2. Data from Cases 4-12 were used to
produce these graphs. As can be seen from these two graphs,
the acid/ calcium ratio decreases with increasing calcium in
the makeup water and also with increasing cycles of concentra-
tion. This means that as the calcium level in the recirculat-
ing water increases, the acid/calcium ratio decreases, as shown
in Figure 3-3. As calcium continues to increase, the slope of
the curve in Figure 3-3 decreases sharply and the acid/ calcium
ratio approaches a constant value. This can be explained in
the following way:
The relative saturation of calcium carbonate is
defined by Equation 3-1:
arn =)
R.s = _| - C0s_ (3-D
KsPCaC03
G-27
-------
TABLE 3-3. EFFECTS OF MAKEUP WATER CALCIUM CONCENTRATION
ON BOWEN COOLING TOWER OPERATION
Case No.
Cycles of Concentration
Makeup Water Calcium, mg/£
Slowdown
Temperature, °C
(°F)
Composition, mg/£
Calcium
Magnesium
Sodium
Chloride =
Carbonates (as CO 3)
Sulfates (as SOT)
Nitrates (as NOT)
pH
Relative Saturations*
CaCOa
CaSC%-2H20
Acid Addition Rate, kg/day**
(Ib/day)
Acid /Calcium Ratio x 10 3***
7
5.7
18
46.7
(116)
103
9.7
7.9
134
17.9
86.5
24.8
7.6
0.52
0.03
619
(1365)
1.47
8
10
18
46.7
(116)
180
17.0
13.8
235
15.5
164
43.4
7.5
0.54
0.07
621
(1370)
0.84
9
15
18
46.7
(116)
270
25.5
20.7
352
18.9
249
65.0
7.6
0.98
0.13
607
(1340)
0.55
10
5.7
12
46.7
(116)
68.4
9.7
7.9
73.0
22.6
83.0
24.8
7.7
0.60
0.02
590
(1300)
2.11
11
10
12
46,7
(116)
120
17.0
13.8
128
24.1
158
43.4
7.7
0.95
0.06
592
(1305)
1.21
12
15
12
46.7
(116)
180
25.5
20.7
192
21.2
248
65.0
7.7
0.92
0.10
604
(1330)
0.82
*Critical values, above which scale potential exists, are 1.3-1.4 for
GaSOit-2H20 and about 2.5 for CaC03 (see Appendix C) .
**100% H2S04.
***Acid addition/calcium rate in circulating water (molar basis).
G-28
-------
4.0
O
I
to
(D
<
CO
DC
d
or
2
o
o
o
o
3.0
2.0
X
1.0
6.0
© 5.7 CYCLES OF CONCENTRATION
H 10.0 CYCLES OF CONCENTRATION
A 15.0 CYCLES OF CONCENTRATION
12-° 18.0
CALCIUM IN MAKE-UP WATER, MG/L
Figure 3-1. Acid requirements as a function of makeup water calcium.
-------
O
I
U)
O
4.0
CO
w
CD
5 3.0
_i
O
5
g
^-
-------
4.0
O
u>
5>
co
<
CQ
IT
<
O
5
_>
o
Q
5
CO
o
3.0
2.0
1.0
CASE 10
50
CASE 5
CASE 12
100 150 200 250
CALCIUM IN CIRCULATING WATER. MG/L
300
Figure 3-3. Acid requirements as a function of calcium in circulating water
-------
where R. S0 ™ = relative saturation of calcium carbonate
CaCO 3
ar ++ = calcium ion activity
L>a
3,,^= = carbonate ion activity
cu 3
K = solubility product constant for calcium
spCaC03 carbonate
As cycles of concentration increase, the ionic
strength of the circulating liquor increases, resulting in
increased ion pair formation which lowers the activity of the
calcium ion. To keep a constant calcium carbonate relative
saturation, the carbonate ion activity must be decreased at
higher calcium levels but the lowering of the calcium ion
activity at higher ionic strengths results in a decrease in
acid addition per unit of total calcium.
For increased levels of calcium, the carbonate ion
activity must decrease to keep the relative saturation constant
and less than one. At these lower levels, the carbonate ion
activity is roughly proportional to the solution pH and there-
fore roughly inversely proportional to the acid rate. If the
calcium ion activities were directly proportional to the cal-
cium concentrations, then the acid/calcium ratio would be a
constant for high calcium levels. As calcium concentrations
increase, however, the nonlinearities between the activity and
the concentration become more substantial, resulting in the
slight curvature observed in Figure 3-3. This slight curvature
is also due to slight nonlinearities between the carbonate ion
activity and the acid rate.
On the other hand, as calcium levels decrease, the
carbonate ion activity must increase to keep the relative satu-
ration constant. At higher carbonate activities, the acid rate
is no longer linearly inversely proportional to the activity,
resulting in the steep curvature observed at low calcium
levels in Figure 3-3.
This curve may be used to determine theoretical acid
addition rates for cooling tower operation at Bowen provided
the circulating water rate and calcium concentration are known.
For example, if the cooling tower is operating with 15,800
IIsec (250,000 GPM) of circulating water with a calcium con-
centration of 150 mg/£, the acid/calcium ratio determined from
G-32
-------
Figure 3-3 is approximately 10"3. The acid addition rate in
gmoles/sec of 100% H2SC\ is then determined as follows:
ACID = 10"3 gmple/sec HzSO^ 150 mg Ca 15,800£
gmole/sec Calcium £ sec
mmole Ca gmole
40.1 mg Ca A 1000 mmole
0.059 gmoles/sec
It should be noted here that this curve applies only
to the quality of makeup water sampled at Bowen with the calcium
and chloride levels varied. Changes in magnesium, sulfate, or
any other species that will have an effect on the level of
chemical complexes formed in the system will affect the relative
saturation of CaCOs . The acid rate required may therefore
depend on the concentrations of other species than calcium in
the makeup water.
3.1.4 Summary of Cooling Tower Alternatives
The first set of simulations concerning increased
cycles of concentration showed that with respect to CaCOa and
CaSCU^HaO scale control the cycles may be increased to 15
easily which results in a 95.670 reduction in the cooling tower
blowdown rate. However, at 15 cycles with the makeup water
quality sampled, two silica solids are supersaturated, Si02 and
Mg(Si02) 3 (OH) 2 (sepiolite) . The kinetics of these solids are
not known so that additional testing should be performed to
determine the control limits for these species.
The effects on cooling tower operation of the makeup
water calcium concentration were determined and acid addition
rates were correlated from the results of the remaining simula-
tions performed. Even at 15 cycles with three times the cal-
cium level of the makeup water sampled, no gypsum scale
potential was identified.
3. 2 Ash Handling Systems
For a system using cooling tower blowdown water
exclusively as ash sluice water, the cycles of concentration
in the tower system is determined by the water requirements of
the ash sluice system. Two alternatives for ash handling at
Bowen were studied:
G-33
-------
1) once through sluice (10% solids)
2) recirculating sluice system
For each of these two alternatives, the effects of
CO2 transfer in the pond and in the sluice tank are examined as
well as the scaling potential of the systems. This section
first discusses the simulation basis for these simulations,
including the process model and input data. Then the results
of the simulations are examined.
3.2.1 Simulation Basis
The process model used to simulate alternatives for
ash sluicing operations is identical to that used for existing
operations (see Appendix E). The ash flow rates and character-
istics were the same as the values used for existing operations
simulations. Sluice water rates were determined so that both
bottom ash and fly ash were slurried at 1070 solids .
Table 3-4 presents the input data that was used for
the alternative sluicing operations. The makeup water for
once-through sluicing of bottom ash and fly ash is 5.7 cycle
cooling tower blowdown and for recirculating sluicing is 15
cycle blowdown.
3.2.2 Once-Through Ash Sluicing System
The simulation results for once-through ash sluicing
at Bowen using cooling tower blowdown, with the towers operat-
ing at 5.7 cycles of concentration, as the sluice water are
shown in Table 3-5. For this simulation, no transfer of carbon
dioxide was allowed between the process liquor and the atmos-
phere at any point in the system. The effects of C02 mass
transfer in the pond and in the sluice tank are examined in
Section 3.2.4.
The relative saturation of calcium sulfate in the fly
ash slurry indicates that gypsum scaling will not present any
problem in the once-through ash sluice system. The calculated
relative saturation of 0.66 is well below the critical scaling
level of 1.3 - 1.4. However, scaling potential for both cal-
cium carbonate and magnesium hydroxide is noted in the fly ash
slurry. The calculated values for relative saturation of CaC03
G-34
-------
TABLE 3-4. BOWEN ALTERNATIVE ASH SLUICING INPUT DATA
Once-Through Recirculating
Flows (Total Plant)
Fly Ash, kg/min
(Ib/min)
Fly Ash Sluice Maekup, £/sec
(GPM)
Fly Ash Sluice Recycle, £/sec
(GPM)
Bottom Ash, kg/min
(Ib/min)
Bottom Ash Sluice Makeup, I/ sec
(GPM)
Bottom Ash Sluice Recycle, £/sec
(GPM)
Makeup Water Composition, mg/1*
Calcium
Magnesium
Sodium
Chloride =
Carbonates (as C03)
Sulfates (as SffO
Nitrates (as NOT)
Pond Deposits, wt 7, solids
Soluble Ash Species (fly ash) , wt
1554
(3430)
234
(3700)
0
(0)
515
(1135)
77
(1220)
0
(0)
34.3
9.7
7.9
12.2
39
69
2
7
24.8
40
1554
(3430)
94.6
(1500)
140.4
(2200)
515
(1135)
0
(0)
77
(1220)
2
5
7
90,
25
20
32.0
20.9
249
65.0
40
CaO
MgO
Na20
CaSO,
0.33
0.001
0.08
1.25
0.33
0.001
0.08
1.25
*0nce-through sluicing makeup water is 5.7 cycle cooling tower
blowdown. Recirculating sluicing makeup water is 15 cycle
cooling tower blowdown.
G-35
-------
TABLE 3-5.
BOWEN ONCE-THROUGH ASH SLUICING
AT 10% SOLIDS*
Makeup Fly Ash
Sluice Water** Slurry
Flow, 5,/sec
(GPM)
Composition, mg/£
Calcium
Magnesium
Sodium
Chloride
Carbonates (as C03)
Sulfates (as SOT)
Nitrates (as NOT)
Temperature, °C
(°F)
PH
Relative Saturations***
CaC03
Mg(OH)2
CaSOlt-2H20
310
(4920)
34.3
.9.7
7.9
12.2
39.2
69.7
24.8
46.7
(116)
8.0
0.98
4.7 x 10""
0.012
234
(3700)
712
10.5
74.5
12.2
39.2
1060
24.8
46.7
(116)
11.4
127
547
0.66
Ash Pond
Overflow
255
(4050)
520
.03
58.3
12.4
0.7
817
25.2
21.1
(70)
12.0
1.0
1.0
0.55
*No CO2 transfer in the system.
**5.7 cycle cooling tower blowdown.
***Critical values, above which scale potential exists are
1.3-1.4 for CaSO^HzO, about 2.5 for CaC03 , and about 3.4
for Mg(OH)2 (see Appendix C).
G-36
-------
and Mg(OH)2 are far in excess of the respective critical
values of 2.5 and 3.4 (Appendix C).
One possible solution to the precipitation of calcium
carbonate and magnesium hydroxide as scale in the sluice line
is to install a reaction tank prior to the sluice line, whereby
a significant portion of the solids may be formed in the tank
as opposed to the line. One method to aid fly ash-sluice water
mixing and precipitation of solids in the fly ash slurry tanks
is to utilize two smaller tanks as opposed to one large tank
to avoid channeling of streams in the slurry tank. Although
the tank may be designed to minimize fouling in the slurry
line, there is a high probability that some scaling of CaCOs
and Mg(OH)2 will occur and eventually foul the line.
A potential remedy to this situation is to period-
ically flush the line with a low pH water stream. Flush water
in the pH range of 6 to 7 should be adequate to remove solid
CaCOs and Mg(OH)2 since they easily dissociate in this pH
range. A possible source of flush water is acidified cooling
tower blowdown. Frequent flushing of the fly ash slurry line
with low pH water should maintain the line free of solid CaC03
and Mg(OH)2. However, pilot or bench scale studies should be
performed prior to implementing this alternative to accurately
size the reaction tank and determine the quantity and frequency
of flush water required or if other measures are necessary.
3.2.3 Recirculating Ash Sluicing System
The simulation results for a recirculating ash
sluicing system at Bowen using cooling tower blowdown, with the
towers operating at 15.0 cycles, as the makeup sluice water
are shown in Table 3-6. As with the simulation discussed in
the previous section, no C02 transfer between the atmosphere
and the process liquor was allowed at any point in the system.
The effects of C02 mass transfer will be discussed in Section
3.2.4.
The degree of recycle achievable in the ash sluicing
system will depend upon the CaSOif«2H20 relative saturation
since gypsum scale is of greater concern than that of CaC03 or
Mg(OH)2. Gypsum scale is very difficult to remove from process
vessels and equipment once it is formed but CaC03 and Mg(OH)2
scale most likely can be dissolved by acid washing.
G-37
-------
TABLE 3-6. BOWEN RECIRCULATING ASH SLUICING*
Flow,
£/sec
(GPM)
Makeup **
Sluice Water
93,3
(1480)
Pond
Recycle
138.9
(2200)
Fly Ash
S lurry
232.2
(3680)
Ash Pond
Overflow
40.9
(650)
Composition, mg/&
Calcium 90.2
Magnesium 25.5
Sodium 20.7
Chloride = 32.0
Carbonates (as C03) 20.9
Sulfates (as SOV) 250
Nitrates (as NOT) 65.0
Temperature, °C 46.7
<°F) (116)
pH . 7.9
Relative Saturations ***
1170
187.6
32.8
0.7
1280
66.6
21.1
(70)
12.5
1410
11.0
187
32.6
8.9
1850
66.2
46.7
(116)
11.7
1170
187.6
32.8
0.7
1280
66.6
21.1
(70)
12.5
CaC03
Mg(OH)2
CaSOij -2H20
0.79
6.3 x 10~*
0.067
1.0
1.0
1.0
29.3
1458
1.28
1.0
1.0
1.0
*No C02 transfer allowed, gypsum precipitation in pond
allowed, 60% of sluice water recycled.
**15-cycle cooling tower blowdown.
***Critical values, above which scale potential exists, are
1.3-1.4 for CaSO^HzO, about 2.5 for CaC03 , and about 3.4
for Mg(OH)2 (see Appendix C).
G-38
-------
The amount of ash pond liquor recycled to the sluice
system should be the maximum amount possible without exceeding
the critical scaling level of 1.3 - 1.4 for the relative
saturation of CaSOlf«2H20 in the fly ash sluice system. Exceed-
ing this critical value could cause plugging of the slurry
line. Maximizing the amount of recycle in the overall system
will minimize the quantity of ash pond effluent to be treated
and therefore minimize the cost of treatment. These simula-
tions were performed for the exclusive use of pond water for
sluicing bottom ash and 60% recycle in the fly ash system.
The level of supersaturation in the pond recycle
water with respect to gypsum will affect the scaling potential
in the slurry line and should be considered. The case shown in
Table 3-6 assumes that the gypsum relative saturation of the
pond recycle is 1.0. A discussion of the effects of super-
saturation in the pond recycle liquor is given in Section
3.2.5.
The relative saturation of gypsum in the fly ash
slurry reaches 1.28 for the recirculating configuration as
shown in Table 3-6, indicating an approach to the critical
range for scaling. This level of recycle in the ash sluice
system therefore represents the maximum operating limit
(based on CaSCU^HaO) without chemical treatment, since the
critical range for scale formation is 1.3 - 1.4. For this
case about 60% of the fly ash sluice water is recycled from the
pond.
As in the once-through sluicing case, the relative
saturations of calcium carbonate and magnesium hydroxide are
significantly above the critical levels. The values shown in
Table 3-6 are 29.3 for CaC03 and 1,458. for Mg(OH)2, whereas
the respective critical values are 2.5 and 3.4. Again, CaC03
and Mg(OH)2 scale formation in the slurry line may be reduced
by installing a reaction tank prior to the sluice line. Sizing
this tank is critical to the successful operation of this ash
sluicing configuration. Additional data should be taken on a
pilot scale so that the reaction tanks may be accurately sized
before implementing this alternative. Flush water may possibly
be used (at pH 6-7} to clean any CaC03 or Mg(OH)2 solid
deposits at periodic intervals, as suggested for once-through
sluicing. Pilot or bench scale testing to determine the level
of acid washing that is sufficient to prevent the line from
plugging should be conducted before this alternative is
implemented.
G-39
-------
The ash pond overflow is reduced from 255 &/sec (4050
GPM) in the once-through simulation case to 40.9 £/sec (650
GPM) for this case. Treatment of this stream to achieve zero
discharge could be accomplished by a combination softening/
reverse osmosis/brine concentration treatment unit. A portion
of the clean water resulting from treatment (50%) could be
recycled to the system as boiler makeup water. In the event
S02 scrubbers are installed at Bowen, the ash pond overflow
could be used as makeup water to the scrubbers to utilize the
available ash alkalinity.
3.2.4 Effects of Carbon Dioxide Mass Transfer
Five additional cases were studied to determine the
effects on the operation of the ash sluicing system of carbon
dioxide mass transfer between the process liquor and the atmos-
phere. The results from these additional cases along with the
two base cases previously discussed are shown in Table 3-7.
Two additional cases for once-through sluicing opera-
tion were run: 1) allowing the process liquor in the pond to
be in equilibrium with the atmosphere with respect to COa and
2) allowing C02 equilibrium with the atmosphere in the sluice
tank.
Allowing CO2 equilibrium in the ash pond has no
effect on the fly ash slurry, but reduces the ash pond overflow
pH to 8.0 from the value of 12.0 for the base case. Carbon
dioxide equilibration in the sluice tank indicates an increase
in scale potential for CaC03 but completely eliminates Mg(OH)2
scale potential. The decrease in the pH of the ash slurry
causing a carbonate shift away from the carbonate ion (C0=)
towards bicarbonate (HCOj), is more than offset by the increase
in total COa in the liquid phase, resulting in an increase in
CaC03 scaling potential (relative saturation changed from 126.7
for the base case to 965.6).
The drop in pH from 11.4 to 9.1 for the ash slurry
is the reason for the Mg(OH)2 relative saturation decrease from
547.4 to 0.035. Gypsum relative saturation decreased slightly
between the base case and the additional case. The net result
is that achieving C02 equilibrium between the atmosphere and
the process liquor in the tank is beneficial from the stand-
point of reducing Mg(OH)2 scale potential. In actual practice
it is not likely that equilibrium would be completely achieved
in the tank but the relatively high values of pH (11.4) enhance
G-40
-------
TABLE 3-7. EFFECTS OF C02 TRANSFER ON BOWEN ASH SLUICING OPERATIONS
Q
Once-Through Sluicing Recirculating Sluicing
C02 Equilibrium
in Pond
C02 Equilibrium
in Tank
Fly Ash Slurry
Relative Satu-
rations*
CaC03
CaSO^HaO
Mg(OH)2
pH
Pond Overflow pH
Base Case
No
No
126.7
0.66
547.4
11.4
12.0
Case 2
Yes
No
126.7
0.66
547.4
11.4
8.0
Case 3 Base Case
No
Yes No
956.6 29.3
0.62 1.28
0.035 1,458.
9.1 11.7
12.5
Case 2
Yes
No
89.1
1.35
1,198.
11.3
7.9
Case 3
No
Yes
2,766.
1.19
0.051
9.3
—
Case 4
Yes
Yes
956.6
1.28
0.062
9.1
—
^Critical values, above which scale potential exists, are 1.3-1.4 for CaSOit-21^0,
about 2.5 for CaC03, and about 3.4 for Mg(OH)2 (see Appendix C).
-------
the sorption of C02 along with agitation of the liquid phase.
However, the short residence time in the sluice tank would
minimize the C02 transfer (see Appendix L).
Three additional cases were run for the recirculating
ash sluicing system: 1) C02 equilibrium in the pond only,
2) C02 equilibrium in the tank only, and 3) C02 equilibrium in
both the tank and the pond. Case 2 in Table 3-7 shows that C02
equilibration between the atmosphere and the liquor in the pond
increases the CaCOa scale potential (relative saturation
increased from 29.3 to 89.1) but decreases the Mg(OH)2 scale
potential (relative saturation decreased from 1,458. to 1,198.)
in the fly ash slurry. The calcium sulfate relative saturation
increased from 1.28 to 1.35.
Case 3, representing C02 equilibrium with the atmos-
phere in the sluice tank but no C02 transfer in the pond, indi-
cates that the CaC03 scale potential is greatly increased, but
the CaSOt,42H20 and Mg(OH)2 relative saturations are decreased
to values below the critical scaling levels. The gypsum rela-
tive saturation decreased from 1.28 in the base case to 1.19
due to the increased calcium associated with carbonate ions in
solution, which lowers the calcium ion activity. The Mg(OH)2
relative saturation decrease from 1,458. in the base case to
0.051 is due to the decrease in the pH of the ash slurry from
11.7 to 9.3.
The last case which represents operation-where C02
equilibrium is achieved both in the pond and in the sluice tank
showed an increase in CaC03 scale potential, no change in gyp-
sum relative saturation, and elimination of Mg(OH)2 scale
potential.
Since gypsum scale is the most important factor to
control in the system due to the difficulty of removing the
scale, Case 3 represents the most favorable case. Although
complete C02 equilibration in the tank and no C02 transfer in
the pond may not be achieved in actual practice, the system may
be operated to maximize the C02 transfer in the tank and mini-
mize C02 transfer in the pond. The transfer in the pond may be
minimized by taking the recycle liquor from a point near the
discharge of the slurry, such that enough residence time has
been allowed for solids settling. Although C02 transfer has
not been quantified in this study, pilot scale studies to
determine the optimum ash sluicing recycle configuration may
G-42
-------
provide data to allow a more accurate account of the level of
C02 transfer in actual operations.
3.2.5 Effect of CaSCK^HzO Supersaturation in the Pond
Recycle Water
If the pond recycle water in a recirculating ash
sluice system at Bowen remains supersaturated with respect to
gypsum, scaling may occur in the fly ash sluice line. The
degree of supers aturation in the pond recycle water cannot be
accurately quantified but will depend on the degree of turbu-
lence in the pond and on the residence time in the pond. The
greater the degree of mixing in the pond due to thermal
gradients or wind turbulence, the more desupersaturated the
liquor will become. Longer residence times will also encourage
precipitation.
However, the lack of CaSQ^'2K20 crystals in the pond
will discourage any precipitation and therefore, limit the
degree of desupersaturation. Since ponds are generally not
very well mixed, the pond will most likely remain supersatu-
rated with respect to gypsum as long as no chemical treatment
is used, and scaling may occur. Pilot or bench scale testing
may provide information to more accurately determine the degree
of desupersaturation.
The magnitude of chemical treatment to remove calcium
from the system was calculated using the chemical equilibrium
program. Sodium carbonate softening of 8070 of the pond recycle
water is necessary assuming that no CaSOn'2R20 precipitation
occurs in the pond and that all of the cooling tower blowdown
(towers operating at 15 cycles) is used as makeup water to the
fly ash system. Bottom ash was assumed to be sluiced exclus-
ively with pond water.
Treatment of 807o of the recycle liquor corresponds
to removing 2.7 gmole/sec calcium from a 206 £/sec (3270 GPM)
stream. Treatment inefficiencies were taken into account by
assuming that the treated stream contains 50 mg/£ calcium. The
equilibrium value with stoichiometric addition of Na2C03 is
22.4 mg/& as calculated by the chemical equilibrium program.
G-43
-------
3.2.6 Summary of Ash Sluicing Operations
Two sluicing configurations were studied: once-
through sluicing of both bottom ash and fly ash at 10% solids
and recirculating sluicing with 60% recycle in the fly ash
system and 100% recycle in the bottom ash system. The once-
through calculations show that no gypsum scale potential is
present but CaC03 and Mg(OH)2 are highly supersaturated.
The recirculating system using 15 cycle cooling tower
blowdown as makeup requires softening 80%, of the pond recycle
assuming the pond will not desupersaturate with respect to
gypsum. A pond overflow of 41.0 H/sec (650 GPM) is produced
from the recirculating system as opposed to 255 &/sec (4050
GPM) from the once-through system.
The net result of the studies concerning C02 transfer
indicate that the level of COZ transfer has only a small effect
on gypsum scale potential but does alter the CaCO3 and Mg(OH)2
relative saturations. In general C02 absorption raises the
CaC03 relative saturation and lowers the Mg(OH)2 relative
saturation.
3.3 Conclusions
From the results of the cooling tower and ash sluice
system simulations discussed in the previous sections, two
alternatives for reducing plant discharges are considered
technically feasible. These are:
1) Cooling tower operation at 5.7 cycles with
acid treatment and once-through ash sluicing
with discharge of ash pond overflow after pH
adj us tment, and
2) Cooling tower operation at 15.0 cycles with acid
treatment and recirculating ash sluice (Na2C03
softening of 80%, of pond recycle) with either
discharge of the ash pond overflow after pH
adjustment or treatment of the overflow with a
softening/reverse osmosis/brine concentration
unit and recycle of the clean water as boiler
makeup and cooling tower makeup.
G-44
-------
The first alternative will require the addition of
acid treatment in the cooling towers and reaction tanks prior
to the fly ash sluice line to minimize CaC03 and Mg(OH)2 scale
formation in the line. Adjustment of the pH of the ash pond
overflow may be required, depending on the amount of carbon
dioxide mass transfer occurring in the pond. The calculated
pH for equilibrium with respect to C02 between the pond liquor
and the atmosphere is 8.0 whereas the value for no C02 transfer
is 12.0. This alternative would not allow Bowen to achieve
zero-discharge without expensive treatment of the ash pond
overflow (255 £/sec or 4050 GPM) , but would reduce the plant
makeup water and discharge rates significantly. The existing
ash pond overflow rate for Bowen is about 1600 £/sec (25,000
GPM) and could be reduced to about 255 £/sec (4050 GPM) by this
alternative.
Treatment of the ash pond overflow by the lime-soda
ash process would reduce the calcium, magnesium, and silica
levels but sulfate concentrations would reach a high enough
level for gypsum scaling to occur. Effective treatment could
be achieved by softening/reverse osmosis/brine concentration
but only at a severe economic penalty due to the magnitude of
the stream flow.
If in the future, S02 scrubbers are installed at
Bowen, the ash pond overflow could be used as makeup water to
the scrubbing system to make use of the available alkalinity
from the ash. However, this study does not include the addi-
tion of scrubbers at Bowen, but considers only the cooling and
ash handling systems.
The second alternative will require the addition of
acid treatment in the cooling towers and reaction tanks in the
fly ash sluice system as in the first alternative. In addition,
recycle lines and pumps to return a portion of the ash pond
liquor for sluicing and sodium carbonate softening of 8070 of
the pond recycle water are required. Zero-discharge may be
achieved with this alternative by treatment of the ash pond
overflow by a softening/reverse osmosis/brine concentration
unit and returning the cleaned water to the boilers' and cool-
ing towers' makeup systems. Discharge of the ash pond overflow
may require pH adjustment as in the first alternative depending
on the level of C02 transfer in the pond. Also, silica removal
may be required at 15 cycles of concentration in the cooling
tower system.
G-45
-------
It should be emphasized here that neither alternative
should be implemented before more information is gathered from
a bench or pilot scale test program to determine 1) the actual
size of reaction tank required in the sluice system, 2) the
quantity and frequency of acid wash water required to minimize
CaC03 and Mg(OH)2 scale formation, 3) the level of gypsum
desupersaturation in the pond, and 4) the scaling control limits
for silica solids.
An economic analysis based on rough cost estimates for
these two alternatives is presented in the next section.
G-46
-------
4.0 ECONOMICS
This section provides rough cost estimations for
implementing each of the technically feasible alternatives
discussed in Section 3,0. Both rough capital costs and operat-
ing costs are presented. The assumptions used in calculating
these costs are briefly outlined. It is emphasized that these
values are only rough estimates for comparative purposes.
A capital cost summary for the two technically feasi-
ble alternatives is presented in Table 4-1. The fly ash slurry
tanks were sized based on a five minute residence time of the
slurry to allow most of the ash soluble species to be leached
in the tank. These tanks were assumed to be general storage
tanks equipped with wear liners for costing purposes. One
tank was used for the fly ash slurry from each unit and was
assumed to have one agitator to keep the slurry well mixed.
Pond overflow recycle pumps and piping were sized
based on the flows calculated in the simulations discussed in
Section 3.0. Twelve-inch carbon steel buried pipe with average
fittings, flanges, shop coating, and wrapping was assumed for
pond return lines to the fly ash systems. Eight and ten-inch
pipe was assumed for the bottom ash systems. A labor to mate-
rial ratio of 0.8 was used to determine installation costs.
Engineering costs (direct and indirect) were assumed to be 7.27o
of the combined labor and material cost (GU-075).
Cast steel pumps with electric motor drivers were
used for all streams. A labor to material ratio of 0.36 was
used for installation costs. Engineering was assumed to be 10%
of the combined labor and material cost (GU-075). All pump and
piping costs were upgraded from 1970 dollars to 1976 dollars
using a factor of 1.56 (based on Chemical Engineering Index).
Since both alternatives involve sluicing the ash at
10 wt% solids, the tank and agitator costs are identical. The
difference in capital cost is due to the installation of pumps
and piping for recycling a portion of the ash pond liquor and
Na2C03 softening of the pond recycle liquor for alternative
two.
A summary of the operating costs for the two alterna-
tives is shown in Table 4-2. Four major breakdowns are shown:
acid treatment, power consumption, softening, and capital cost
G-47
-------
TABLE 4-1. CAPITAL COSTS* FOR WATER RECYCLE/REUSE ALTERNATIVES AT BOWEN
i
j>
oo
Alternative One
(Once-through ash sluice)
Alternative Two
(Recirculating ash sluice)
Fly Ash Slurry Tanks**
Agitators
Pond Overflow Recycle Pumps
Pond Overflow Recycle Piping
Sodium Carbonate Softening
Contingency (20%)
Contractual Fees (3%)
TOTAL
61,000
21,000
—
—
—
16,000
2,000
100,000
61,000
21,000
105,000
507,000
300,000
199,000
30,000
1,223,000
*1976 dollars
**Includes wear liner and agitator supports
***$91.7/GPM (1976 dollars) or $75/GPM (1974 dollars)
References: GU-075, MC-136, NE-107
-------
TABLE 4-2. OPERATING COSTS1 FOR WATER RECYCLE/REUSE
ALTERNATIVES AT BOWEN
Alternative One Alternative Two
Cooling Tower Acid Treatment2
Power Consumption3
Agitators
Recycle Pumps
Softening Chemicals'*
Capital Charges5
TOTAL
(mils/kw-hr)
35,800
2,100
—
—
15 ,000
52,900
(.002)
45,400
2,100
51,300
120,000
183,000
401,800
(.018)
X1976 dollars/yr based on 80% load factor
2$60/ton for sulfuric acid
32c/kw-hr
"$69/106 gal (NE-107)
515% per year based on 30-year lifetime
G-49
-------
amortization. The acid treatment costs were based on $60/ton
for sulfuric acid and were calculated based on the simulation
results in Section 3.0. Operating the towers at 15.0 cycles of
concentration (alternative 2) will require 27% more acid than
operation at 5.7 cycles (alternative 1). The difference in
power consumption for the two alternatives is due to the recycle
pumps employed in the second alternative. A cost of 2c/kw-hr
was used to determine power costs.
The results shown in Tables 4-1 and 4-2 indicate that
the first alternative is significantly less expensive than the
second. However, to achieve zero-discharge by eliminating the
ash pond overflow discharge is not practical for the first
alternative (once-through ash sluicing) due to the magnitude of
the flow (255 a/sec or 4,050 GPM).
Additional capital and operating costs for treating
the ash pond overflow (41 £/sec or 650 GPM) from the second
alternative are presented in Table 4-3. The overflow can be
treated by a combination of softening, reverse osmosis, and
brine concentration and the clean water recycled to the plant
boiler makeup system. The additional capital cost is about
$5.16 million giving a total capital cost of about $6.38 million
for achieving zero-discharge with a recirculating ash sluice
system. The additional operating costs total approximately
$l,333,000/yr to give a total operating cost of about $1.74
million/yr for achieving zero discharge.
G-50
-------
TABLE 4-3. CAPITAL AND OPERATING COSTS FOR ELIMINATING
ASH POND OVERFLOW AT BOWEN
Capital Cost1 Operating Cost'
Softening/ Reverse Osmosis/
Brine Concentrator3
Additional Pump
Additional Piping
Additional Capital Charges4
Total Additional Costs
Costs from Tables 4-1, 4-2
(to nearest $1000)
TOTAL
(mils/kw-hr)
5,040,000
22,000
95,000
--
5,157,000
1,223,000
6,380,000
546,000
13,000
--
774,000
1,333,000
402,000
1,735,000
(.078)
'1976 dollars
21976 dollars per year
3capital cost = $7,750/GPM feed (LE-239)
operating costs = $2/1,000 gal not including capital
cost amortization
" 15% per year for 30 year lifetime
G-51
-------
Appendix H. Recycle/Reuse Options at Comanche (Public Service of Colorado)
1.0 INTRODUCTION
This appendix describes the analysis of the water sys-
tem at the Public Service of Colorado's (PSC) Comanche Plant.
The work was done under EPA Contract No. 68-03-2339, Water Re-
cycle/Reuse Alternatives in Coal-Fired Steam-Electric Power
Plants. In this section a summary of the important results is
presented. Comanche was chosen along with four other plants for
evaluation of technical and economic feasibility of various water
recycle/reuse options. The major water systems at the two-unit,
700 Mw Comanche plant are the cooling tower and bottom ash
sluicing systems. Fly ash is disposed of in a dry form.
The results of the existing operations simulations'for
the Comanche cooling system compare well with the sample data ob-
tained at the plant. The calculated CaC03 and CaSO<,'2H20 rela-
tive saturations in the recirculating cooling water (8.2 x 10~4
and 0.252, respectively) indicate that the cycles of concentra-
tion may be significantly increased without forming calcium
carbonate or calcium sulfate (gypsum) scale. However, the high
level of the silica concentration in the makeup water may re-
quire some form of silica removal, such as lime-soda ash or
magnesium bicarbonate treatment, in order to prevent silica
scaling at higher cycles of concentration. Pilot or bench-scale
studies to more accurately quantify silica scaling potentials
are recommended before increasing the cycles of concentrations
in the Comanche cooling system.
Cooling system simulations were carried out to deter-
mine the effects of operating at increased cycles of concentra-
tion in the towers. In addition, system sensitivity to composi-
tion changes in the makeup water were investigated by simulations
using a makeup water with twice the sulfate concentration found
in the sample data.
Table 1-1 presents a summary of the three alternatives
which were examined for Comanche. It should be noted here that
this analysis was performed to study general water recycle/reuse
alternatives. Actual implementation of any of the alternatives
would require a more extensive investigation of process parameter
variability. More water quality data would be required along
with additional studies to fully characterize the ash reactivity
variations as a function of time. The first one involved using
cooling system blowdown from the towers designed to operate at
five cycles of concentration to sluice both fly ash and bottom
ash on a once-through basis. The effects of C02 mass transfer
H-l
-------
TABLE 1-1. SUMMARY OF WATER RECYCLE/REUSE OPTIONS AT COMANCHE
Existing
Conditions
Alternative
One
Alternative
Two
Alternative
Three
ffi
ro
Softened River Water Softened River Water Softened River Water Softened River Water
5.0 5.0 7.6 8.4
(Sulfuric acid and zinc polyphosphate used for all conditions)
Cooling Tower Makeup Source
Cycles of Concentration in
Cooling Towers
Cooling System Treatment
Fly Ash Disposal Method
Type, 7. solids
Bottom Ash Disposal Method
Type, 7. solids
Recycle in Fly Ash
System, 7.
Recycle in Bottom Ash
System, 7.
Treatment in Ash Systems
Plant Makeup Requirements
I/sec (GPM)
Plant Discharge
*/sec (GPM)
Costs '
Capital Investment, 1976 $
Operating Expenditures, 1976 $/yr:
(mils/kW-hr)
Additional Cost to Treat Pond
Overflow for Zero Discharge
Capital, 1976 $
Operating, 1976 S/yr '
(mils/kW-hr)
Total Cost for Zero Discharge
Capital, 1976 $
Operating, 1976 $/yr 2
(mils/kW-hr)
1 These rough cost estimates were made to compare technically feasible options and do not include a "difficulty to retrofit" factor:
* Includes capital amortization at 157, per year
'About 813 GPU of pond water is recycled in this alternative and represents about 76/<, of the total sluice water required for
17, solids in the bottom ash slurry
Dry Wet, 107.
Wet, 17. Wet, 47.
0
0 0
None None
590 (9350) 520 (8250)
156 (2470) 65.4 (1040)
342,000
90,000
(0.02)
8,280,000
2,136,000
(0.43)
8,622,000
2,226,000
(0.45)
Wet, 107,
Wet, 47.
107.
1007.
Brine Concentration
of Makeup (507.)
455 (7210)
28.8 (460)
3,662,000
863,000
(0.18)
3,706,000
944,000
(0.19)
7,368,000
1,807,000
(0.37)
Dry
Wet, 1%
---
767.3
None
450 (7120)
30.2 (480)
222,000
38,000
(0.008)
3,853,000
989,000
(0.20)
4, 105,000
1 027,000
(0.21)
-------
in the ash pond and the sluice tank were examined for this
system. No gypsum scale potential was identified in any of the
once-through sluicing cases, but potential scaling of CaC03 and
Mg(OH)2 was present.
It is possible that the fly ash slurry line can be
kept free of plugging by the addition of a fly ash slurry reac-
tion tank and by frequent flushing with water of pH 6 to 7.
Pilot or bench scale studies are recommended to size the reaction
tank and determine the quantity of wash water required or if
other measures are required before implementing fly ash sluicing
at Comanche. This alternative will result in an ash pond over-
flow of about 32.7 £/sec (518 GPM) for each unit as compared to
the existing configuration bottom ash pond overflow rate of about
78 S,/sec (1230 GPM) per unit.
The second alternative involves sluicing the fly ash
at about 10 wt.70 solids using 90% cooling tower blowdown and 1070
ash pond recycle water. Bottom ash is sluiced at about 4 wt.%
solids using only pond recycle water. Gypsum relative satura-
tions in the fly ash sluice line were calculated to be 1.54 -
1.74 depending on the level of COa transfer in the pond. This
range exceeds the critical relative saturation range for scaling
of CaS04'2H20 of 1.3-1.4. Therefore, some form of treatment
would be required such as brine concentration of a portion of
the tower blowdown. Lime treatment of the blowdown for calcium
removal was found to be insufficient for scale prevention due to
the sulfate concentrations in the system. Desupersaturation of
gypsum in the ash pond will also not prevent scaling since only
a small portion (10%) of the ash pond liquor is recycled to the
fly ash system.
As with the once-through simulations, potential for
CaC03 and Mg(OH)2 scaling was identified. But, as before, this
possibly can be minimized by installing a reaction tank prior
to the sluice line and by flushing the line frequently with a
pH 6-7 water stream. Again, further testing is suggested. This
alternative will produce an ash pond overflow of about 14.4 £/sec
(230 GPM) for each unit.
The third alternative is to continue to dispose of fly
ash in a dry form and sluice the bottom ash on a recirculating
basis using cooling tower blowdown and pond recycle with the
towers operating at 8.4 cycles of concentration. This will pro-
vide 16.0 £/sec (260 GPM) of cooling tower blowdown per unit and
will not alter the boiler refractory cooling systems. For this
H-3
-------
alternative about 15.1 &/sec (240 GPM) of ash pond overflow per
unit will be obtained. This water may be discharged or recycled
to the boiler and cooling tower makeup systems after appropriate
treatment.
Rough cost estimates were made for the once-through
sluice system and the recirculating system using cooling tower
blowdown to sluice fly ash with 50% of the blowdown treated by
brine concentration. Operating the cooling system at 5 cycles
of concentration and sluicing the fly ash and bottom ash on a
once-through basis is the less expensive alternative ($342,000
for capital cost and about $90,000/yr operating cost, including
capital amortization at 15% per year). The third alternative is
the least expensive with $222,000 capital costs and $38,000/yr
operating costs.
In order to reduce the ash pond overflow to 14.4 £/sec
(229 GPM) for each unit by operating the cooling systems at 7.6
cycles of concentration with the cooling system blowdown as
sluicing makeup, the entire plant ash sluice system will require
an initial capital cost of about $3.7 million and an operating
cost of about $863,000/yr, including capital amortization at 15%
per year. These costs do not include the possible necessity of
silica removal.
If zero discharge of ash pond overflow is desired, the
once-through system becomes more expensive due to the greater
amount of ash pond overflow to be treated. A softening/reverse
osmosis/brine concentration system to eliminate ash pond overflow
would require an additional operating cost of approximately
$2,136,000/yr. The total overall costs would be about $8,622,000
for capital costs and $2,226,000/yr for operating costs (including
capital cost amortization at 1570 per year) .
The additional costs for obtaining zero discharge with
the recirculating system would be about $3.7 million for capital
costs and $944,000/yr for operating costs, giving total overall
costs of about $7.4 million for capital costs and $1.8 million/yr
for operating costs including capital amortization at 15% per
year.
The costs associated with achieving zero discharge with
dry fly ash disposal (third alternative) are about $4.1 million
for capital costs and $1,027,000/yr for operating costs. These
costs include brine concentration, additional piping, additional
pumping costs, and capital amortization at 15% per year.
H-4
-------
Detailed discussions of the existing operations simu-
lations, the alternative simulations, and the rough cost esti-
mates make up the main body of this appendix.
H-5
-------
2.0 PLANT CHARACTERISTICS
The Public Service of Colorado Comanche generating sta-
tion is a coal-fired system composed of two units, each having a
350 Mw capacity, and is located near Pueblo, Colorado. The coal
burned at Comanche is about 7.3% ash and 0.4% sulfur with a heat-
ing value of about 8300 Btu/lb. The basic flow schemes are the
same for both units as described in this section.
The Comanche cooling system uses wet cooling towers to
discharge heat. The ash removal system consists of (1) wet
sluicing for bottom ash, and (2) electrostatic precipitation and
subsequent dry disposal for fly ash. The bottom ash slurry is
sent to ash ponds for disposal.
These major features of the Comanche operation are dis-
cussed in detail in the following sections. The overall water
balance will be described first followed by a more detailed de-
scription of the cooling and bottom ash sluicing systems. Then
a brief description of the computer simulation model that was
used to characterize the existing operations of the cooling sys-
tem will be presented. A more detailed description of the models
used can be found in Appendix E. The results of the existing
operations simulations are compared to actual plant data and
examined for potential water recycle/reuse alternatives.
2.1 Overall Water Balance
A flow schematic for the Comanche water system is given
in Figure 2-1. Both the cooling system and the ash sluicing sys-
tem are shown with the design flow rates and chemical analyses of
the streams. The chemical analyses were performed on spot sam-
ples collected at Comanche. A detailed description of the sam-
pling and analytical procedures used is presented in Appendix B.
Under existing operations the water input to the over-
all system is first taken from the Arkansas River and stored in
a raw water reservoir. From here a small portion of the raw
water (about 7 a/sec or 105 GPM) is sent to the coal handling
facilities to suppress dust generation. Another portion of the
flow from the reservoir is sent to the ash removal system to
sluice bottom ash into the ash ponds. The remainder of the raw
water leaving the reservoir is sent to the Comanche lime treat-
ment facility to reduce the calcium hardness. The lime sludge
produced during the softening process is sent to a special ash
H-6
-------
OtfANSAS
Figure 2-1. Public Service of Colorado Comanche Plant water balance
-------
sc
-J
3&NO
f>LTe*
OMAN*
fUTSA.
PCT/tOLf
sroxtat
CVAPOKATiOf*
TO
f ST CHMU.I3
"I'M
Figure 2-1. (Continued)
-------
Stream Number
Stream Name
Flow
pH
Summer: ""'J0.
English
Winter: Me"ich
English
Calcium
Magnesium
Sodium
Potassium
Chloride
Carbonate (as C03)
Sulfate (as SO,,)
Nitrate (as N03)
Phosphates (aa PCM
Silicates (as SlOa)
Suspended Solids
Dissolved Solids
^
Plant
Makeup
590 i/sec.
9,348 gpm
488 I/ sec.
7,731 gpm
8.55
53.4
14.2
19
2.3
7
101
134
13
<0.1
11
<0.01
345
^>
Makeup
Ash
Sluice
69 11 sec.
1,100 gpm
89 £/sec.
1,410 gpm
8.55
53.4
14.2
19
2.3
7
101
134
13
<0.1 "
11
<0.01
345
3>
Coal
Dust
Suppression
7 fc/sec.
105 gpm
7 I/sec.
105 gpm
8.55
53.4
14.2
19
2.3
7
101
134
13
<0.1
11
<0.01
345
Lime
Softener
Feed
514 £/sec.
8,143 gpm
392 I/sec.
6,216 gpm
8.55
53.4
14.2
19
2.3
7
101
134
13
<0.1
11
<0.01
345
^
Softened
Water
510 i/sec.
8,078 gpm
389 £/sec.
6,166 gpm
Potable
Makeup
0.4 H/sec.
6 gpm
0.4 t/sec.
6 gpm
Potable
Water
0.2 A/sec.
3 gpm
0.2 A/sec.
3 gpm
4>
Sewage
Treatment
Feed
0.2 t/sec.
3 gpm
0.2 £/sec.
3 gpm
<$>
Softening
Wastes
4 «./sec.
65 gpm
3 £/sec.
50 gpm
Figure 2-1. (Continued)
-------
a
o
Stream Number
Stream Name
Flow
pH
Metric
Su"aer: English
Winter: ^"K.
English
Calcium
Magnesium
Sodlun
Potassium
Chloride
Carbonate (as C03)
Sulfate (as SO,,)
Nitrate (as H03)
Phosphates (as P0»)
Silicates (as S103)
Suspended Solids
Dissolved Solids
Softened
Water
509 i/sec.
8,072 gpm
389 J./sec.
6,160 gpm
XX
Boiler
Makeup
5 a/sec.
76 gpm
5 A/sec.
76 gpm
<5>
Service
Water
11 H/aec.
172 gpm
8 Jt/sec.
132 gpm
<£>
Service
Water to
Towers
3.3 I/sec.
52 gpm
2.7 Jl/sec.
42 gpm
<»
Cooling
Tower
Makeup
241 I/sec.
3.826 gpa
184 I/ sec.
2,910 gpm
6.2
36.5
10.2
19
1.7
9
6.0
163
9
<0.1
54
<0.01
298
<£>
Cooling
Tower
Slowdown
41 Jl/sec.
655 gpm
32 Jl/sec.
500 gpm
6.3
205
65.5
89
13
53
2.7
965
16
3.5
280
<0.01
1,700
<5>
Boiler
Cooling
16 i/sec.
260 gpm
16 I/sec.
260 gpm
6.3
205
65.5
89
13
53
2.7
965
16
3.5
280
<0.01
1,700
Boiler
Cooling
14 i/sec.
225 gpm
14 4/sec.
225 gpm
<4x
Makeup
Ash
Sluice
35 H/sec.
550 gpm
44 Jl/sec.
705 gpm
8.55
53.4
14.2
19
2.3
7
101
134
13
<0.1
11
<0.01
345
Figure 2-1. (Continued)
-------
ffi
I
Scream Number
Scream Name
Flow
Metric
Sumer: English
Winter: *'^c.
English
pH
Calcium
Magnesium
Sodium
Potassium
Chloride
Carbonate (as C03)
Sulfate (as SO*)
Nitrate (as N03)
Phosphates (as POH)
Silicates (as SIOj)
Suspended Solids
Dissolved Solids
<$>
Bottom
Ash
Slurry
63 ^/sec.
1.000 gpm
63 11 sec.
1,000 gpm
XX
Bottom
Ash
Slurry
74 «./sec.
1,170 gpm
74 I/sec.
1,170 gpm
<8>
Clean
Boiler
Makeup
2 Jl/sec.
35 gpm
2 I/sec.
35 gpm
^p
Demineralize
Waste
0.4 Jl/sec.
6 gpm
0.4 i/sec.
6 gpm
XX
Ash
Pond
Influent
74 i/sec.
1,170 gpm
74 Jl/sec.
1,170 gpm
7.45
115
18.3
29
3.6
12
111
260
19
0.1
48
0.29
573
<^>
Ash
Pond
Effluent
74 H/sec.
1,170 gpm
74 11 sec.
1,170 gpm
7.25
105
24.4
44
5.8
16
80
355
17
0.8
110
<0.01
763
<$>
Polishing
Pond
Influent
152 S,/sec.
2,405 gpm
151 H/sec.
2,390 gpm
<^>
Boiler
Slowdown
2 Jl/sec.
30 gpm
2 A/sec.
30 gpm
<£r>
Polishing
Pond
Effluent
156 Jl/sec.
2,465 gpm
155 Jl/sec.
2,450 gpm
7.7
149
33.4
39
6.9
27
67
528
13
2.1
130
<0.01
878
Figure 2-1. (Continued)
-------
pond which is kept separate from the ponds receiving bottom ash
slurries. The softened water is used for service water and for
makeup water to the two cooling systems.
The water effluent from the overall operation comes
from the overflow from the final polishing pond which is fed by
the two boiler blowdown streams, the lime sludge disposal pond
overflow, and the two bottom ash disposal pond overflows. The
final polishing pond effluent is sent to the St. Charles River.
The remaining system water losses are cooling tower evaporation
and drift and other evaporative losses.
The first step in characterizing the chemistry of the
Comanche water system is to examine the results of the spot sam-
ples taken. The measured species concentrations were input to
the equilibrium program and several parameters were calculated
which determine the tendency of the liquor sampled to form chem-
ical scale and to absorb or desorb C02 from the atmosphere.
Another parameter calculated checks the internal consistency of
the sample and is a measure of the analytical accuracy.
One apparent inconsistency is that the softener feed
has a silica concentration of 11 mg/£ whereas the cooling tower
makeup (softener effluent) has a silica concentration of 56 mg/£.
The 11 mg/& is consistent with Public Service of Colorado data
but the 56 mg/£ in the cooling tower makeup is consistent with
the 280 mg/£ in the cooling tower bottoms (about 5 cycles of con-
centration) . This discrepancy was not resolved. The larger value
of 56 mg/£ was used in this study for the makeup water to repre-
sent a worst case.
If silica problems are encountered at Comanche, hot or
warm lime-soda softening should reduce silica. Another possibil-
ity is to use the magnesium bicarbonate process to remove silica.
Silica removals of 48-8470 were reported for this process in EPA
report 600/2-76-285 entitled "Recovery of Lime and Magnesium in
Potable Water Treatment" (TH-192).
Table 2-1 presents a summary of the parameters calcu-
lated for each of the samples taken at Comanche. Relative sat-
urations for CaC03, Mg(OH)2, and CaS04«2H20 are given in the
first three columns. These parameters indicate the tendency of
the stream to form scale. Critical values for relative saturation
of each species, above which scale formation is likely, are 2.5
for CaC03, 3.4 for Mg(OH)2, and 1.3-1.4 for CaSO^•2H20'(see
Appendix C).
H-8
-------
TABLE 2-1. PARAMETERS CALCULATED BY EQUILIBRIUM PROGRAM FOR COMANCHE SAMPLES
Stream
Stream Name
Cooling Tower Makeup
Cooling Tower Slowdown
* Ash Sluice Makeup
Ash Pond Inlet
Ash Pond Subsurface
Ash Pond Effluent
Polishing Pond Effluent
No.
14
15
1
23
—
24
27
0
0
1
0
0
0
0
Equilibrium Partial
Relative Saturations* Pressure of C02, % Residual
CaC03
.0001
.0005
.56
.41
.66
.12
.60
Mg(OH)2 CaSOij
7.
1.
3.
1.
7.
2.
8.
9
9
8
1
0
8
9
x lO'10
x 10~8
x 10~5
x 10~6
x 10~6
x 10~7
x 10~6
0.
0.
0.
0.
0.
0.
0.
•2H20
028
31
031
088
12
11
17
atm x 10^
12.
5.
2.
30.
15.
31.
10.
2
6
2
0
7
4
7
Electroneutrality
+9
-10
-5
+8
-2.4
-2
-7
*Critical values, above which scale potential exists, are 1.3-1.4 for CaSOi,'2H20, about 2.5 for CaCOs,
and about 3.4 for Mg(OH)a (see Appendix C)
-------
None of the streams sampled show a tendency to form
CaC03) Mg(OH)2, or CaS0lf-2H20 scale. The highest CaC03 relative
saturation was found in the ash sluice makeup water (1.56) but
even that value was below the critical level. The relative sat-
urations for the cooling tower blowdown indicate that the cycles
of concentration may be increased somewhat since the gypsum rela-
tive saturation is only 0.31.
Equilibrium partial pressures of C02 above the liquor
sampled were calculated by the equilibrium program and show the
tendency of a stream to absorb or desorb COa when in contact
with the atmosphere. A value less than 3 x 10"" atm, the equi-
librium partial pressure of COa in air, indicates a tendency
to absorb COa and a value greater indicates a tendency to desorb
C02. The value for the cooling tower blowdown sample is very
near 3 x 10"^ indicating that COa equilibrium was essentially
achieved in the cooling towers.
Percent residual electroneutrality is a parameter cal-
culated to determine the internal consistency of each sample with
pH specified. A value of ±15% is considered acceptable. All of
the Comanche samples had a residual electroneutrality within ±10%.
A more detailed description of how this parameter is calculated
is presented in Appendix E.
2.2 Cooling System
The Comanche generating station has two cooling systems,
one for each unit. The systems are identical and employ wet
cooling towers for evaporative cooling. Water circulates at a
design rate of 9240 2,/sec (146,400 GPM) between the condenser
and cooling tower for each unit.
The circulating cooling water characteristics are con-
trolled by the makeup water composition and by the amount of
chemical additives introduced into the system. Sulfuric acid
is added for pH control, and zinc polyphosphate is added to
inhibit scaling.
The blowdown stream is maintained at a rate sufficient
to keep dissolved species from concentrating to the point of
saturation. Otherwise, scaling of the lines and equipment could
result. The relationship between the blowdown rate, the cooling
tower evaporation rate, the drift rate, and the amount of con-
centration that dissolved species undergo is expressed below:
H-10
-------
r - E + B + D
L B + D
C = cycles of concentration (number of
times that dissolved species in the
makeup water are concentrated in the
circulating water)
E = evaporation rate
B = blowdown rate
D = drift rate (rate at which water is
entrained in the vapor leaving the
cooling tower)
Present operation of the Comanche cooling system maintains the
blowdown rate so that the makeup water is concentrated about 5
times (i.e., C = 5). This level is maintained to supply boiler
refractory cooling and ash sluicing water.
2.2.1 Simulation Basis
Existing operations simulations were performed for the
Comanche cooling tower system to verify the validity of the model
in predicting scaling tendencies in the tower and/or condenser
and to determine any potential for increased recycle/reuse. This
section first briefly describes the computer model used to simu-
late the present Comanche water system. Then the inputs to this
computer model (such items as process flowrates, concentrations,
temperatures, etc.) are examined. A detailed description of the
process model is included in Appendix E.
The process simulation flow scheme shown in Figure 2-2
was used to model cooling tower operations at Comanche. This is
a generalized cooling tower model with capabilities of simulating
sulfuric acid addition and slipstream softening for calcium re-
moval. Only acid addition was used for existing operations.
Given the inputs of air flow, temperature and compo-
sition, makeup water composition, flow and temperature of the
circulating water, drift rate, and cycles of concentration, the
model performs iterative calculations around the cooling loop
to determine the blowdown, evaporation and makeup rates, and
compositions for all water streams. An acid addition rate is
H-ll
-------
16
5 OUTLET AIR
6 DRIFT
ALKINP
SOFTENIN
HEMICAL3
13 CHEMICAL WASTE
ORDER Of PROCESS CALCULATIONS:
I, 2, 3. 4. 5. 6. 7. (8. 9, 10, II, 12, 13. 14, 6, 7) *
AIR 4
MAKEUP
WATER
17
ACID 3
SOFTENING CHEMICALS 12
3 SLOWDOWN
^
CTBAL
(OVERALL
BALANCE)
7
^»
_ ^,
..... —f.
-_^
5 AIR
6 DRIFT
*• 8 SLOWDOWN
13 CHEMICAL WASTE
Figure 2-2.
02- 1527- I
Process simulation scheme for Comanche
cooling tower system.
H-12
-------
determined to keep the CaC03 relative saturation within a
specified range. If slipstream softening is required (determined
by model) the slipstream and chemical addition rates are calcu-
lated.
Several assumptions are inherent in performing this
simulation with the subroutines shown in Figure 2-2. These
assumptions are enumerated below:
1) Equilibrium exists between C02 and
HaO in the atmosphere and cooling
tower exit water.
2) The temperature of the cooled water
stream approaches the wet bulb tem-
perature of ambient air within a
predictable range.
3) The compositions and temperatures of
the cooled water and drift streams
are equal.
4) Ionic reactions taking place in the
liquid phase are rapid and thus at
equilibrium.
The assumption involving the temperature of the cooled
water stream is a recognized design parameter in cooling tower
evaluation and gives a good approximation. The assumption con-
cerning the temperature and composition of the drift stream
should be very close to actuality, as is the assumption in re-
gard to H20 gas-liquid equilibrium. The assumption with regard
to CO2 equilibrium is conservative since the partial pressure of
CO2 in actual cooling towers tends to be greater than the equi-
librium value. The lower equilibrium concentration of carbonate
species assumed in the model causes the pH to be slightly higher
in the model than in actual operation. The higher pH causes the
relative saturation of CaC03 to increase more than the lowered
carbonate species concentration causes it to decrease.
The data used as input to this model is presented in
Table 2-2. Some of this information was obtained directly from
PSC while other inputs were calculated from PSC data, local
meteorological data, and sample analyses. The air flows were ob-
tained from PSC and adjusted to a representative temperature and
composition by means of local climatological data. The ambient
H-13
-------
TABLE 2-2. INPUT DATA FOR COMANCHE
COOLING TOWER SIMULATIONS
FLOWS
^
Air, m /hr
(ACFM)
Drift, &/sec
(GPM)
Circulating Water, £/sec
(GPM)
TEMPERATURES
Ambient Air, °C
(°F)
Approach, °C
(°F)
Condenser AT, °C
(°F)
Wet Bulb, °C
(°F)
Condenser Outlet, °C
(°F)
ADDITIONAL DATA
Relative Humidity, %
Cycles of Concentration
Makeup Water Composition, mg/£.
Calcium
Magnesium
Sodium
Chloride _
Carbonate (as C03)
Sulfate (as SO^
Nitrate (as N03)
Winter
2.7 x 107
(1.6 x 107)
9.0
(142)
9,240
(146,400)
0
(32)
11.1
(20)
14.1
(26)
-2.8
(27)
22.8
(73)
53
5
36.5
10.2
26,2
5.3
5.4
163
12.4
Summer
2.7 x 107
(1.6 x 107)
9.0
(142)
9,240
(146,400)
22.2
(72)
8.3
(15)
14.4
(26)
17.8
(64)
40.6
(105)
68
5
36.5
10.2
26.2
5.3
5 4
-/ • *T
163
12.4
H-14
-------
air wet bulb temperatures were derived from climatological data
from the National Oceanic and Atmospheric Administration averages
(NA-166). The water makeup composition was obtained from the
spot sample and adjusted to minimize residual electroneutrality
(see Appendix E).
2.2.2 Simulation Results
This section gives the results of simulating the exist-
ing operations at Comanche. The simulations are based on the
heavy-load summer conditions and heavy-load winter conditions,
whose operating parameters have been described in the previous
sections. The simulation results are compared with the results
coming from the chemical analyses in order to evaluate the per-
formance of the computer process model in giving a reasonable
approximation of typical Comanche operating conditions.
Table 2-3 is a summary of the most important simulation
results for the existing operations of the Comanche cooling sys-
tem along with the plant data for actual operation.
The blowdown flow for summer operation is within the
data range reported by PSC. The winter blowdown is slightly
lower probably due to differences in climatological data. The
blowdown stream pH values compare well (6.3 versus 6.4). When
the sample blowdown composition is compared with simulation com-
positions for summer and winter operations, however, it appears
that the measured concentrations are for the most part greater
than the simulation concentrations. The difference between
measured and simulation concentrations for calcium, magnesium,
chloride, sulfate, and carbonate indicates that the Comanche
cooling system was operating at a concentration factor in excess
of the value of 5.0 used in the simulations. Using the measured
concentrations, an actual concentration factor of about 5.8 is
indicated. This difference is well within the range of typical
operation, especially since the measurements may not have been
taken at the heavy-load conditions assumed for the simulations.
The major discrepancy in the blowdown composition re-
sults involves the nitrate ion, whose measured concentration does
not agree very well with its simulation concentrations. However,
the measured value of the makeup concentration (12.4 mg/i) is not
consistent with the measured blowdown composition (16 mg/&)
either, if a concentration factor of 5.0 or above is assumed.
Therefore, analytical uncertainties or errors are indicated in
the case of the nitrate species. The problem with the nitrate
H-15
-------
TABLE 2-3. COMAHCHE EXISTING COOLING TOWER OPERATIONS
Simulations (5 cycles of concentrations)
Cooling Tower Slowdown
Flow, &/sec per tower
(GPM)
PH
Composition, mg/£
Calcium
Magnesium
Sodium
Chloride =
Sulfate (as SOi^
Carbonate (as CO 3)
Nitrate (as N03)
Relative Saturations*
CaC03
CaSOk'2E20
Partial Pressure COa , atm
Plant Data
31.5 - 41.3
(500 - 655)
6.3
205
65.5
89
53
965
2.7
16
5.0 x Hf*
0.31
5.6 x 10'1*
Case 1
Winter Operation
24.6
(390)
6.42
181
50.9
131
26.5
824
2.2
61.8
2.3 x lO'1"
0.30
3.3 x 10~4
Case 2
Summer Operation
37.8
(600)
6.45
181
50.8
131
26.5
824
1.85
61.8
8.2 x 10""
0.25
3.6 x 10"1*
^Critical values, above which scale potential exists, are 1.3-1.4 for CaSOit'2H20 and
about 2.5 for CaCOs (see Appendix C)
-------
concentration serves to explain the lesser discrepancies with
the chloride and sodium ions, because these ions were adjusted
in the makeup composition used as a simulation input. Any errors
in the nitrate concentration, therefore, influence the simulation
blowdown values of sodium and chloride.
Although the simulation results of the sodium, chloride,
and nitrate ions do not compare precisely with their measured
values, the discrepancies do not severely limit the usefulness
of the process model in simulating the cooling system operation.
These ions do not form ionic pair bonds comparable in strength
to the CaSCU and CaC03 bonds. Therefore, any perturbations in
the ions respective concentrations (assuming that the pH is kept
constant) should not greatly influence the Ca , SCU, and COT
activities and resultant scaling potentials. This is borne out
when the relative saturations of CaSO^'2E20 and CaC03 are
examined.
The relative saturation of CaC03 with the measured blow-
down composition (5 x 10"1*) is in good agreement with the summer
and winter simulation values (8.2 x 10"" and 2.3 x 10""*). The
relative saturation of CaSO^^HaO also agrees well. The measured
composition's relative saturation for CaSO^^HaO is 0.31, the
simulation values are 0.25 for the summer case and 0.30 for the
winter case. The comparison becomes even closer if, as mentioned
above, the measured composition corresponds to a concentration
factor slightly greater than that used for the simulation cases.
Judging from the agreement between the measured compo-
sition and the simulation results, the computer process model can
be used to adequately simulate typical cooling system behavior.
The simulation results indicate a potential for reducing cooling
water requirements by increasing the cycles of concentration.
The effects of increasing the cycles of concentration with re-
spect to cooling tower operation and the subsequent use of blow-
down water for ash sluicing are the subject of the following
section. The sensitivity of the cooling system to increases of
SO^ in the makeup water is also investigated to determine the
operational effects of poorer quality makeup water.
2.3 Ash Disposal System
The ash disposal system at the Comanche generating
station consists of two ash ponds with a surface area of about
5060 m2 (54,000 ft2) each. Only bottom ash is being sluiced
into the ponds at present. The fly ash is trucked away in a
dry form.
H-17
-------
The bottom ash is sluiced to the ash ponds at about
1 wt. 70 solids. The sluicing is intermittent; bottom ash is
pulled from the ash hopper about six hours per day. The sluice
water comes from two sources: blowdown from the cooling system
and untreated water from the plant raw water reservoir. The
flow rates for these streams were averaged over a 24-hour period
to give representative flow rates for simulating a continuous
system.
Part of the cooling system blowdown stream is diverted
for use as boiler refractory cooling water. This water subse-
quently flows down into the bottom ash hoppers where an overflow
stream is sent to the ash ponds. But when bottom ash is being
pulled, the refractory cooling water goes out in the bottom ash
sluice stream.
The bottom ash sluice stream travels to the ash pond
at about 210 cm/sec (7 ft/sec). The sluice lines for the two
boiler units are 760 m and 590 m (2,500 and 1,930 ft) long,
giving line retention times of about 5.9 and 4.6 minutes,
respectively.
No simulations of the existing ash handling systems
were performed but in the next section the alternatives looked
at do include wet sluicing of the ash and the appropriate simu-
lations were used to study the alternatives.
H-18
-------
3.0 TECHNICAL ALTERNATIVES
Water recycle/reuse alternatives for the Comanche
plant are presented in this section. The Comanche water system
was divided into two subsystems for purposes of simulation.
One subsystem consists of the cooling towers, with associated
treatment facilities and condensers. The other subsystem con-
sists of the ash disposal operations.
The operating characteristics of the cooling system
are examined first. Then the ash handling system is analyzed
in the context of the possible range of cooling system blowdown
streams available for sluicing. Water recycle/reuse alterna-
tives, therefore, will be based on an overall picture of the
Comanche water system.
3.1 Cooling Towers
Before a water management strategy can be made for
the Comanche ash sluicing system, the cooling system must be
fully characterized with respect to the effects of increasing
cycles of concentration. This is necessary to provide informa-
tion about the cooling tower blowdown, which is used as sluice
water. Also, information is needed regarding the cooling sys-
tem sensitivity, i.e., how changes in makeup water composition
affect the system. The scaling potential of CaSC\«2H20 and
CaC03 are the most crucial areas concerning the operation of
the towers at higher cycles of concentration.
In order to characterize the cooling system, simula-
tions were run which varied the cycles of concentration of the
recirculating cooling water by adjusting the blowdown flow
rate. Additional simulations were subsequently run in which
the concentration of S04 in the makeup water was doubled. The
differences in system behavior between the first set of simula-
tions using the makeup water compositions determined by sample
analysis and the second set of simulations using the makeup
water with increased sulfate serve as a measure of the sensiti-
vity of the Comanche cooling system to changes in river water
quality with respect to sulfate.
Section 3.1.1 is concerned with the simulation basis
used with the process model in making the computer characteri-
zation of the Comanche cooling towers. Section 3.1.2 and 3.1.3
are discussions of the results of these simulations.
H-19
-------
Section 3.1.2 addresses the changes in the cooling
system arising from increasing the cycles of concentration.
The changes of particular interest are the changes in the ap-
proach toward scaling of the chemical species in the recircula-
ting cooling water and the change in the acid rate required to
maintain pH control.
Section 3.1.3 discusses the changes which result from
using makeup water with twice the sulfate concentration used in
the previous simulations. Simulations were run at different
cycles of concentration as before in order to determine how the
sulfate concentration increase affects the approach to scaling.
3.1.1 Simulation Basis
The basis for the simulations to characterize the
Comanche cooling towers is the same as discussed earlier in
Section 2.2.1. The process model is unchanged. The sole dif-
ferences between the existing operations simulation discussed
earlier and these alternative operations simulations lie in
altering the computer inputs to the process model.
The first series of calculations were directed at
determining the effects of increasing the cycles of concentra-
tion. The second series of simulations were conducted in a
similar manner, except that the S07+concentration was doubled
in the water makeup stream. The Na ion was also increased in
order to maintain a constant pH. (The net effect was to in-
crease sulfate concentration by addition of Na2S04 to the make-
up water.
All of the alternative cooling tower simulations were
performed for summer operation of the cooling towers because
summer conditions represent the case of maximum blowdown rates.
Increased evaporation needed during the summer months requires
an increase in blowdown rate to maintain a constant concentra-
tion factor. Therefore, summer operation represents a conser-
vative or limiting case. For a given makeup water composition
and blowdown rate, the summer operation results in a higher
concentration factor than does the winter operation.
3.1.2 Effect of Increased Cycles of Concentration
Simulation results of cooling tower operations at
5.0, 7.6, and 15.0 cycles of concentration are presented in
H-20
-------
Table 3-1. Sulfuric acid treatment was used to maintain a
slightly acidic blowdown pH, which represents typical Comanche
practice. Additional acid treatment was not needed to control
CaCO3 scale, because CaCO3 relative saturation remained well
below the critical scaling value of 2.5 (Appendix C) for all
three cases.
Although CaCO3 scaling is not a problem for operation
at 15.0 cycles of concentration, CaSO^«2H20 does approach the
critical scaling value range for relative saturation of 1.3-1.4
at this level of concentration. The relative saturation of
CaS0lt«2H20 versus cycles of concentration is plotted in Figure
3-1. As might be expected for such a relatively dilute aqueous
system, the relative saturation behaves quite linearly with re-
spect to increasing the cycles of concentration.
Operating the cooling towers at higher cycles of con-
centration may cause species other than gypsum or calcium car-
bonate to become supersaturated and possibly form scale. Table
3-2 presents the calculated relative saturations at 5.0, 7.6,
and 15 cycles of concentration for the important phosphate and
silica species.
None of the phosphate species were supersaturated at
15 cycles of concentration, the highest value being about 0.5
for CaHPO^. All of the silica solids are subsaturated for all
cases except for Si02 and Mg(Si02)3(OH)2. Existing operations
(5 cycles) shows a relative saturation of 1.36 for Si02 but no
evidence of scaling exists. This suggests that a critical value
greater than one exists for Si02 as well as CaC03, Mg(OH)2, and
CaSOit*2H20. The magnitude of this value is unknown, so that
increasing the cycles of concentration in the towers at Comanche
could cause scaling problems.
The relative saturation of Mg(Si02)3(OH)2 (sepiolite)
is 0.54 for existing operations but rises to about 1.8 for 7.6
cycles and 31. for 15 cycles. The critical value for this spe-
cies also is not known, so that the scaling limit for cycles
of concentration is also unknown.
In light of the above discussion, additional testing
should be performed to more accurately determine the control
limits for magnesium-silica solids before implementing water
recycle/reuse alternatives requiring increased cycles of con-
centration in the cooling towers. Any significant increase in
H-21
-------
TABLE 3-1. EFFECTS OF INCREASED CYCLES IN
COMANCHE COOLING TOWERS*
Cycles of Concentration
Makeup Water Rate, £/sec
(GPM)
Acid Addition Rate, kg/day**
(Ib/day)
Slowdown
Flow, H/ sec
(GPM)
PH
Composition, mg/£
Calcium
Magnesium
Sodium
Chloride _
Carbonate (as C03)
Sulfate (as SO~O
Nitrate (as NOj)
Temperature, °C
(°F)
Relative Saturations***
CaCOa
CaSO,'2H20
Existing
Operations
5,0
235
(3720)
48.1
(106)
38
(600)
6.5
181
50.8
131
26.5
1.9
824
61.8
26.1
(79)
8.8 x KT"
0.25
Increased Cycles
of Concentration
7.6
216
(3430)
41.3
(91,0)
20
(310)
6.9
275
77.1
198
40.2
3,8
1250
93.6
26.1
(79)
8.2 x 10"1*
0.43
15.0
200
(3190)
40.5
(89.3)
4.5
(70)
7.0
540
151
388
78.8
4.6
2450
184
26.1
(79)
0.018
0.98
All flows are for one unit.
***
As 100% E2SO^
Critical values, above which scaling potential exists, are 1.3-1.4
for CaSO^'2E20 and about 2.5 for CaC03 (see Appendix C)
-------
tu
i
ro
u>
o
CM
in
in
I.I •
1.0 -
0.8
0.8
o
m 0.7 -
O
O
0.6 -
0.6
0.4 -
111 O.3 •
IE
O.2-
0.1
6 7 « 9 IO II
CYCLES OF CONCENTRATION
13
Figure 3-1. Gypsum relative saturation as a function of cycles of
concentration at Comanche.
-------
TABLE 3-2. RELATIVE SATURATIONS OF PHOSPHATE AND SILICA
SOLIDS IN COMANCHE COOLING TOWERS
Cycles of Concentration 5 7.6 15.0
Relative Saturations
CaHPO .117 .192 .486
1+
Ca (PO ) .0011 .0027 .038
3 IT 2
Si02 1.36 2.06 4.07
Mg Si 0 (OH) 6.6 x 10~5 2.0 x 10"* 7.7 x 10~3
235 6
Mg Si 0 (OH) 6,3 x 10~5 1.1 x 10~4 5.1 x 10~3
d ^ Q 4
Mg(Si02)3(OH)2 0.54 1.79 31.0
CaH^iO^ 7.87 x 10~6 1.09 x 10~5 4.76 x 10~5
Ca(H3SiOit)2 5.0 x l(f* 1.05 x 10~ 3 9.12 x 10"3
H-24
-------
the cooling system cycles of concentration may necessitate
additional chemical treatment of the makeup water or a slip-
stream from the circulating water to reduce the silica concen-
tration. The makeup water concentration may be reduced by using
hot lime-soda ash treatment (RO-266) or the magnesium bicarbonate
process (TH-192) instead of the present lime treatment.
3.1.3 Effect of Sulfate Concentration in the Makeup Water
Three simulation runs were made at 5.0, 7.6, and 13.0
cycles of concentration for makeup water containing twice as
much sulfate as used for the existing operations simulations.
The results of these runs are presented in Table 3-3. As in
the simulations presented in the previous section, acid addition
rates were calculated to produce a slightly acidic blowdown
stream.
As can be seen from the relative saturations given in
Table 3-3, the CaCOs relative saturation is far removed from its
critical scaling value due to the fact that almost all of the
Ca^ ions are associated with SOT ions. The CaS0lf'2H20 relative
saturations for the simulations using the makeup water with the
doubled sulfate concentration are larger than the previous simu-
lations using the sample makeup water but are not larger by a
factor of two. Apparently, the S(K activity is not doubled by
increasing the sulfate concentrations by a factor of two. With
the makeup water containing twice as much sulfate, a relative
saturation of 1.0 is reached at about 13 cycles of concentration,
whereas this high a value for CaS04'2H20 relative saturation was
not reached until 15 cycles of concentration with the existing
makeup water quality.
In both sets of simulations the acid addition rate was
found to be insensitive to the number of cycles of concentration.
In every case an acid rate of about 43 kg/day (95 Ib/day) was
sufficient to maintain a neutral or slightly acidic blowdown
stream.
3.2 Ash Handling Systems
Three alternatives for ash handling at Comanche were
studied: (1) once-through sluicing of bottom ash and fly ash,
(2) the recirculation of ash pond liquor, and (3) bottom ash
sluicing and dry fly ash disposal. Fly ash sluicing is included
as well as the existing bottom ash sluicing for the first two
alternatives. For fly ash sluicing, the effects of C02 transfer
in the pond and in the sluice tank are examined as well as the
scaling potentials of the system.
H-25
-------
TABLE 3-3. EFFECTS OF MAKEUP WATER SULFATE CONCENTRATION*
ON COMANCHE COOLING TOWER OPERATION
Cycles of Concentration
Makeup Water Rate, £/sec
(GPM)
Acid Addition Rate, kg/day**
(Ib/day)
Slowdown
Flow, £/sec
(GPM)
PH
Composition, mg/Jl
Calcium
Magnesium
Sodium
Chloride _
Carbonate (as CO 3)
Sulfate (as SO^)
Nitrate (as NO^)
Temperature, °C
(°F)
Relative Saturations
CaCOa
CaSOif-2H20
5.0
235
(3720)
43,0
(94.8)
38
(600)
7.0
181
50.8
520
26.5
1.9
1640
61.8
26.1
(79)
0.0055
0.33
7.6
216
(3430)
43.7
(96.4)
20
(310)
6.9
275
77.1
789
40.2
3.8
2480
93.6
26.1
(79)
0.0056
0.55
13.0
203
(3220)
43.0
(94.8)
6.7
(105)
6.7
469
131
1340
68.5
4.7
4220
159
26.1
(79)
0.0058
1.01
**
***
Sulfate concentration of 326 mg/Jl as opposed to sampled level of 163 mg/£
As 100% H2SOi,
Critical values, above which scale potential exists, are 1.3-1.4 for
CaSOit-2H20 and about 2.5 for CaC03 (see Appendix C)
H-26
-------
This section first discusses the simulation basis for
these simulations, including a brief description of the process
model and the input data. The simulation results are then pre-
sented and discussed.
3.2.1 Simulation Basis
The basis for simulation of alternative ash sluicing
operations at Comanche is presented in this section. A descrip-
tion of the process model utilized and the important assumptions
made is first given. Then the input data used in these simula-
tions is summarized.
The process model used to simulate both once-through
and recirculating ash sluice systems at Comanche is shown in
Figure 3-2. The model uses information about the compositions
and flows of the makeup water and fly and bottom ash as well as
the percent solids in the sludge and pond evaporation as inputs.
From this information the flows and compositions of all the
streams in Figure 3-2 are calculated. A detailed description
of the ash sluicing model is given in Appendix E.
Several assumptions are inherent in performing ash
sluicing simulations with the subroutines outlined above. These
assumptions are listed below:
1) Ionic reactions taking place in the
liquid phase are rapid and thus at
equilibrium.
2) Solid-liquid equilibrium is achieved
in the ash pond, with the exception
of CaSCK which is allowed to remain
supersaturated.
3) Ash dissolution is essentially complete
by the time the slurry reaches the pond.
4) All solids precipitation occurs in
reaction vessels or the pond. RATHD1
calculates nucleation amounts and then
precipitation rates based on kinetic
expressions.
5) Subroutine RATHD1 models nucleation as
an instantaneous rate if the species'
H-27
-------
/ASHINPX
( FLY ASH 1
MAKEUP FOR FLY ASH SYSTEM V 2 /
\
S*~ "^^S. I
/WTR^NPy^ D(VDR3 J^
OLOWDOWNf "" 4 ^
"T"
1
*s
4
HLDTK3
^J_^-
5
3
ffi EFFLUENT — pJ
1 _ 16 DIVDER _ 12
00 6
8
6^
DIVDER
7
MAKEUP FOR BOTTOM ASH SYSTEM
/ASHINP\ 15
1 BOTTOM 1 _»^
V ASH 1
\ 3 /
RATHDI
10
r
„ 1
POND 1
OVFRFI rtuu i
13
HLDTK3
9
. __ .. |
L
14
*-
15~*" PNC
2— *-
3— •-
PI V
i
ASH
SLURRY
17
^^ 7
.___ ^^—^ n
)BAL -»~ 16
— ^ 9
5 -*-10
~~ L_ ^-7 SLUICE WATER
POND •-»-
(NONE) |
r*^
VAPORIZED
3 SLUDGE
9 FLY ASH SLUICE
VAPORIZATION
| — »-10BTM ASH SLUICE
i
VAPORIZATION
BOTTOM ASH SLURRY
ORDER OF PROCESS CALCULATION: 1. 2. 3, 4. 5. 6. 7, 8. 9. 10 *
Figure 3-2. Process simulation scheme for Comanche ash sluicing system.
02-1529- I
-------
relative saturation exceeds the
critical value. Nucleation is
allowed such that the various
species' relative saturations are
returned to their respective crit-
ical levels. At this point, no
further nucleation is allowed.
The input data used in simulating a once-through ash
sluicing system and a recirculating ash sluicing system at the
Comanche plant are given in Tables 3-4 and 3-5. The fly ash
flow rate and the bottom ash flow rate were taken to be 78% and
22%, respectively, of the total ash flow rate, which was calcu-
lated from Public Service of Colorado information.
Sluice water rates shown in Table 3-4 were calculated
based on (1) the use of the existing blowdown rate, taken at 5.0
cycles of concentration, and (2) a slurry solids content of about
10% for the fly ash stream. The sluice water composition corre-
sponds to the existing operations blowdown composition.
The soluble species data for the fly ash were obtained
from ash characterization studies performed in support of this
program (Appendix L). The results of the leaching studies per-
formed in this program were used. Calculations performed to
obtain the soluble species amounts are presented in Appendix L.
The sluice makeup and sluice recycle water flows given
in Table 3-5 are based on sluicing the fly ash at about 10 wt. %
solids using 90% cooling tower blowdown and 10% ash pond recycle
water. Bottom ash is sluiced at about 4 wt. % solids using only
pond recycle water. The boiler refractory cooling water flow rate
is unchanged from its design value of 16.4 Jl/sec (260 GPM). The
fly ash soluble species amounts were assumed to remain constant.
3.2.2 Once-Through Ash Sluicing System
The simulation results for once-through fly ash sluic-
ing at Comanche using cooling system blowdown as sluice water are
given in Table 3-6. The simulation is based on blowdown water
corresponding to cooling tower operation at 5.0 cycles of concen-
tration. No transfer of C02 was permitted between the atmosphere
and the sluice liquor at any point within the sluicing system.
The effects of CC-2 transfer will be discussed in Section 3.2.4.
H-29
-------
TABLE 3-4.
COMANCHE ONCE-THROUGH ASH
SLUICING INPUT DATA
Flow Rates (per unit)
Fly Ash, kg/min (Ib/min)
Bottom Ash, kg/min (Ib/min)
Fly Ash Sluice Water, £/sec (GPM)
Bottom Ash Sluice Water, SI/sec (GPM)
Pond Evaporation, £/sec (GPM)
Sluice Water Composition
(Cooling Tower Blowdown @ 5 Cycles)
Calcium
Magnesium
Sodium
Chloride
Carbonates, as C03
Nitrate, as NO3
Sulfate, as SOT
Pond Deposits (40 wt. 70 solids)
Soluble Ash Species
CaO
MgO
Na20
151.
45.
21.7
16.4
.15
mg/A
181.7
50.86
130.5
26.5
1.85
61.8
824.5
(333.7)
(99.)
(344)
(260)
(2.4)
(Fly Ash)
wt. 70
2.287
0.287
0.041
1.010
H-30
-------
TABLE 3-5. COMANCHE RECIRCULATING ASH
SLUICE INPUT DATA
(10% Recycle in Fly Ash System)
Flow Rates (per unit)
Fly Ash, kg/min (Ib/min) 151 (333.7)
Bottom Ash, kg/min (Ib/min) 45 (99)
Fly Ash Sluice Makeup, £/sec (GPM) 19.6 (310)
Bottom Ash Sluice Makeup, £/sec (GPM) 0.0 (0)
Fly Ash Sluice Recycle, £/sec (GPM) 2.2 (34)
Bottom Ash Sluice Recycle, £/sec (GPM) 19.6 (310)
Pond Evaporation, £/sec (GPM) .15 (2.4)
Sluice Water Composition mg/£
(Cooling Tower Slowdown @ 7.6 Cycles)
Calcium 275
Magnesium 77.1
Sodium 198
Chloride 40.2
Carbonates, as C07 3.8
Nitrate, as NO3 93.6
Sulfate, as SC^ 1250.
Pond Deposits (40 wt. 70 solids)
(Fly Ash)
Soluble Ash Species wt. %
CaO 2.287
Na20 0.041
MgO 0.077
1.010
H-31
-------
TABLE 3-6, COMANCHE ONCE-THROUGH
ASH SLUICING SIMULATION RESULTS*
Flow, &/sec
(GPM)
Composition, mg/&
Calcium
Magnesium
Sodium
Chloride
Sulfate. as SOT
Carbonate, as CO3
Nitrate, as N03
Fly Ash Slurry
Ash Pond Overflow
21.5
(356)
2,450.
106.
167.
27.
1,670.
1.9
62.
32.7
(518)
1290.
0.01
153.
27.
1300.
0.7
62.
Temperature, °C
~
Relative Saturations **
CaC03
40.6
(105)
Mg(OH)
5.37
1.24
35,360.
17.8
(64)
1.0
1.1
1.0
12.3
12.7
No C02 transfer in the system.
Critical values, above which scale potential exists are
1.3-1.4 for CaS04-2H20, about 2.5 for CaC03, and about 3 4
for Mg(OH)2 (see Appendix C)
H-32
-------
The relative saturation of CaSOIt«2H20 in the fly ash
slurry indicates that gypsum scaling will not occur in the once-
through ash sluice system. The calculated relative saturation
of 1.24 is below the critical scaling level of 1.3-1.4. But a
strong probability of scaling exists for calcium carbonate and
magnesium hydroxide in the fly ash slurry. The calculated val-
ues for relative saturation of CaCO3 and Mg(OH)2 are larger
than their respective critical values of 2.5 and 3,4 (see Ap-
pendix C) .
A reaction tank installed before the sluice line can
possibly be used to minimize the scaling of CaCO3 and Mg(OH)2
in the sluice line. This will permit a significant portion of
the solids to be formed within the reaction tank as opposed to
the line.
In order to facilitate fly ash sluice water mixing
and precipitation of solids in the fly ash slurry tanks two
small tanks can be used instead of a single large tank. This
will minimize channeling in the slurry tank. Although the
slurry tank may be designed to guard against fouling in the
slurry line, some scaling of CaCO3 and Mg(OH)2 can be expected,
A possible measure to prevent fouling would be to periodically
flush the line with a relatively low pH water stream. Water
with a pH of 6 to 7 should be adequate to remove solid CaC03
and Mg(OH)2 because these compounds readily dissociate in this
pH range. One possible source of this flush water is cooling
system blowdown, which is typically neutral or slightly acidic,
3.2.3 Recirculating Ash Sluicing System
The configuration examined to determine the effects
of recirculating ash pond water as sluice water uses the blow-
down from the cooling system to sluice fly ash with only 1070
of the sluice water recycled and uses ash pond water to sluice
bottom ash.
As in the once-through simulation discussed in the
previous section, no C02 transfer between the atmosphere and
the process liquor was permitted at any point in the system.
The effects of C02 transfer are discussed in Section 3.2.4.
The results of the simulation are presented in Table
3-7. As shown, strong potential for gypsum scaling exists
throughout the system. The fly ash slurry gypsum relative
H-33
-------
TABLE 3-7.
COMANCHE RECIRCULATING ASH
SLUICING SIMULATION RESULTS*
Flow, £/sec
(GPM)
Composition, mg/£
Calcium
Magnesium
Sodium
Chloride
Sulfate, as SOt,.
Carbonate, as CDs
Nitrate, as NO 3
Relative Saturations**
CaC03
CaSOt»'2H20
Mg(OH)2
Fly Ash Sluice
Water Recycle
2.1
(34)
1580.
0.01
240.
40.8
2190.
.73
94.
1.0
1.83
1.0
Fly Ash
Slurry
21.5
(342)
2680.0
124.0
240.0
40.4
2180.0
3.5
94.0
6.46
1.76
42,750.0
Ash Pond
Overflow
14.4
(229)
1580.0
0.01
240.0
40.8
2190.0
.73
94.0
1.0
1.83
1.0
12.7
12.7
12.7
Makeup water to sluice system is cooling tower blowdown at 7.6 cycles of
concentration.
Critical values, above which scale potential exists, are 1.3-1.4 for
about 2.5 for CaC03, and about 3.4 for Mg(OH)2 (see Appendix C)
H-34
-------
saturation of 1.76 exceeds the critical range for scale forma-
tion of 1.3-1.4. A potential solution to the problem of
CaSO(,-2H20 scaling with this configuration is to treat a por-
tion of the cooling tower blowdown prior to the ash sluicing
system. Two types of treatment were examined: (1) lime soft-
ening of the entire cooling tower blowdown for calcium removal
and (2) brine concentration of 50% of the tower blowdown. A
comparison of the effects of these treatment options is shown
in Table 3-8.
Lime treatment of the fly ash sluice makeup water is
not sufficient to prevent gypsum scaling in the sluice line.
The CaSO ^2^0 relative saturation for this case is 1.74, which
greatly exceeds the critical range for scale formation of 1.3-
1.4. However, brine concentration of 5070 of the makeup water
results in a gypsum relative saturation of 1.29 in the fly ash
slurry which is just below the critical range for scale forma-
tion. This lower scaling potential is due to the removal of
sulfate as well as calcium from the makeup water, whereas with
lime treatment, only calcium removal is realized.
Another possibility which may lower gypsum scale po-
tential in the slurry is gypsum desupersaturation in the pond.
The degree of desupersaturation that may occur in the pond can-
not be accurately quantified but will depend on the degree of
turbulence in the pond and on the residence time in the pond.
The greater the degree of mixing in the pond due to thermal
gradients or wind turbulence , the more desupersaturated the
liquor will become. Longer residence times will also encourage
gypsum precipitation.
However, the lack of suspended CaSO^^HzO crystals in
the pond will discourage any precipitation and therefore, limit
the degree of desupersaturation. An additional case allowing
CaSCU'2H20 precipitation to equilibrium (relative saturation
of 1.0) was run but since only 10% of the fly ash sluice water
is recycled, the gypsum relative saturation in the slurry was
only reduced to 1.68 from 1.76. Thus, the level of supersatur-
ation in the pond will have only a very small effect on the
slurry gypsum relative saturation with this configuration.
Since calcium and sulfate in the makeup water are the
limiting constituents, an alternative treatment is to remove cal-
cium and sulfate by controlled gypsum precipitation. Operation
of this treatment is very similar to the situation encountered
in lime /lime stone S02 scrubbing where gypsum is precipitated in
H-35
-------
TABLE 3-8.
COMANCHE FLY ASH SLUICE
MAKEUP WATER TREATMENT EFFECTS
Fly Ash Slurry
Treatment A*
Treatment B**
Composition, mg/£
Calcium
Magnesium
Sodium
Chloride
Sulfate, as SOT
Carbonate, as COT
Nitrate, as NO 3
Relative Saturations***
CaCOs
CaSO^-2H20
Mg(OH)2
2,550.0
137.0
239.0
40.0
2,180.0
3.8
94.0
6.97
1.74
45,070.0
2,660.0
99.0
140.0
20.0
1,560.0
1.9
47.0
3.61
1.29
37,500.0
12.7
12.8
Makeup water treated with lime for Ca removal.
5070 of makeup water treated with brine concentration.
Critical values, above which scale potential exists are
1.3-1.4 for CaS04-2H20, about 2.5 for CaC03, and about 3 4
for Mg(OH)2 (see Appendix C)
H-36
-------
a reaction tank with CaS0lt»2H20 solids recirculated. Calcium
and sulfate may be removed in this treatment alternative in two
steps: (1) addition of lime to precipitate gypsum in a tank
with recirculating CaSO^ZHjO seed crystals and (2) addition of
C02 to precipitate CaC03. However, this is not proven techno-
logy for water treatment and pilot studies would have to be
conducted to size process vessels before a full-sized instal-
lation could be considered.
Although treatment of the makeup water will prevent
gypsum scale, the relative saturations of CaC03 and Mg(OH)2
are still above the respective critical levels of 2.5 and
3.4. As in the once-through sluicing scheme, CaC03 and Mg(OH)2
scale formation in the slurry line may be reduced by installing
a reaction tank prior to the sluice line. The sizing of the
reaction tanks will be critical to the successful operation of
this ash sluicing configuration. Additional data taken on a
pilot scale should be gathered before implementing this tech-
nical alternative. Also, flush water may possibly be used to
clean CaC03 and Mg(OH)2 deposits at periodic intervals as sug-
gested for once-through sluicing. Pilot studies to determine
the level of acid washing necessary to prevent plugging should
be conducted before this alternative is implemented.
The recirculating ash sluice system with makeup water
treatment will produce an ash pond overflow of about 14.4 a/sec
(229 GPM) per unit which is reduced from the 32.7 £/sec (518 GPM}
flow per unit for the once-through alternative. Treatment of
the ash pond overflow to achieve zero discharge can be accomp-
lished by a brine concentrator or a brine concentrator/reverse
osmosis system. The clean water produced by such treatment
could be recycled as boiler makeup water and cooling tower
makeup water.
3.2.4 Effect of Carbon Dioxide Mass Transfer
Five additional cases were studied to determine the
effects on the operation of the ash sluicing system arising
from C02 transfer between the process liquor and the atmosphere.
The results from these cases plus the results of the two base
cases previously discussed are summarized in Table 3-9.
Two additional cases for once-through sluicing oper-
ation were run: (1) allowing the process liquor in the pond to
be in equilibrium with the atmosphere with respect to C02 and
(2) allowing C02 equilibrium with the atmosphere in the sluice
H-37
-------
TABLE 3-9. EFFECTS OF C02 MASS TRANSFER IN COMANCHE ASH SLUICING'
w
1
UJ
oo
Once-Through Sluicing Recirculating Sluicing
Base Case Case 2 Case 3 Base Case* Case 2 Case 3
C02 Equilibrium
in Pond No Yes No Yes No
COa Equilibrium
in Tank No No Yes No No Yes
Fly Ash Slurry
Relative Saturations **
CaC03 5.37 5.37 4,047.0 6.46 13.9 4,207.0
CaSOi,-2H20 1.24 1.24 1.25 1.68 1.74 1.54
Mg(OH)2 35,360.0 35,360.0 .045 43,010.0 45,520.0 .049
Slurry pH 12.3 12.3 9.5 12.7 12.7 9.5
Pond Overflow pH 12.7 12.7 12.7 8.0
Case 4
Yes
Yes
3,961.0
1.61
.056
9.5
Pond desupersaturation of CaSOi|'2H20 allowed.
Critical values, above which scale potential exists, are 1.3-1.4 for CaSOit*2H20,
about 2.5 for CaCOs, and about 3.4 for Mg(OH)2 (see Appendix C)
-------
tank. Allowing C02 equilibrium in the ash pond has no effect
on the fly ash slurry, but this reduces the ash pond overflow
pH to 7.9 from the base case value of 12.7. Carbon dioxide
equilibration in the sluice tank results in an increase in
scale potential for CaCO3 and completely eliminates Mg(OH)2
scale potential. In any event, the CaCO3 relative saturation
remains above the critical level of 2.5 which may cause scaling
problems. The gypsum relative saturation remained virtually
unchanged by the C02 equilibration.
Three additional cases were run for the recirculating
ash sluicing system using cooling tower blowdown to sluice fly
ash and ash pond recycle to sluice bottom ash. These were
(1) CO2 equilibrium in the pond only, (2) C02 equilibrium in
the tank only, and (3) C02 equilibrium for both the pond and
tank. Case 2 for this system in Table 3-9 indicates that C02
equilibration between the atmosphere and the pond increases
the CaCO3 scale potential (relative saturation increased from
6.46 to 13.9) and increases the Mg(OH)2 scale potential (rela-
tive saturation increased from 43,010 to 45,520). The CaS0lt-2H20
relative saturation increased slightly, from 1.68 to 1.74.
Case 3 for this system, representing C02 equilibrium
with the atmosphere in the sluice tank but no C02 transfer in
the pond, indicates that the CaC03 scaling potential is dras-
tically increased, but the Mg(OH)2 scaling potential is greatly
reduced. The gypsum scaling potential is decreased but still
remains above its critical scaling level. The decrease was
due to the increased amount of calcium associated with carbon-
ate ions in solution, which lowers the calcium ion activity.
The great decrease in Mg(OH)2 relative saturation is attribut-
able to the lowering of the pH between the base case and Case 3.
Case 4 represents operation where C02 equilibrium is
achieved in both the ash pond and the sluice tank. In this
case the CaCO3 relative saturation is greater than that of the
base case, while the Mg(OH)2 relative saturation is greatly
reduced. The gypsum relative saturation is only slightly
changed between Case 4 and the base case (1.61 versus 1.68).
Preventing gypsum scaling is the most important consid-
eration in the ash sluice system due to the difficulty of remov-
ing the scale, and Case 3 is the most favorable case on this
basis. However, all four cases have CaSOlt-2H20 relative satu-
rations exceeding the critical scaling level, and all will re-
quire some form of treatment to prevent gypsum scale potential.
H-39
-------
3.3 Conclusions
From the results of the cooling tower and ash sluice
system simulations discussed in the previous sections, two al-
ternatives for reducing plant discharges with fly ash sluicing
in the system are considered technically feasible. These are:
1) Cooling system operation at 5.0 cycles
of concentration (existing operations)
with once-through ash sluicing for both
fly ash and bottom ash. The ash pond
overflow can be discharged after pH
adjustment or can be treated with a
brine concentrator/reverse osmosis unit,
with the clean water recycled to the
boiler and cooling tower makeup systems.
2) Cooling system operation at 7.6 cycles
of concentration with the sluicing of
bottom ash accomplished by using re-
cycled ash pond water. The fly ash
will be sluiced with cooling tower blow-
down which has been treated to remove
calcium and sulfate, and recycled pond
water.
Another alternative exists in which fly ash disposal
is effected by dry methods, as is currently done at Comanche.
The cooling towers may be operated at 8.4 cycles of concentration
(with treatment for silica removal if necessary)
providing 16.4 I/sec (260 GPM) blowdown for boiler refractory
cooling for each unit. This water may then be used to sluice
bottom ash on a recirculating basis at about 1% solids, resulting
in an ash pond overflow of about 15.1 H/sec (240 GPM) per unit.
About 58.4 H/sec (930 GPM) of pond overflow is recycled in this
system. Zero discharge may be achieved with this alternative by
treating the overflow by brine concentration and recycling the
clean water to the boiler and cooling tower makeup systems.
The first alternative may necessitate the use of re-
action tanks before the fly ash sluice line to minimize CaCOs
and Mg(OH)2 scale formation in the line. Adjustment of the pH
of the ash pond overflow may be required, depending on the amount
of CO2 transfer occurring in the ash pond. The calculated pH for
equilibrium with respect to C02 between the pond and the atmos-
phere is 7.9, but the value for no CO2 transfer is 12.7.
H-40
-------
The second alternative will also include reaction
tanks in the fly ash sluice system, as in the first alternative.
In addition, recycle lines and pumps will be required to return
a portion of the ash pond liquor for sluicing. This alternative
involves operating the cooling towers at 7.6 cycles of concen-
tration, which may possibly result in scale in the condenser due
to silica. Chemical studies will need to be undertaken to in-
vestigate whether scaling will occur. If scaling will occur at
this increased level of concentration, it will be necessary to
lower silicate concentrations in the cooling tower makeup water.
This can be accomplished by using hot lime-soda ash softening
(RO-266) or the magnesium bicarbonate process (TH-192) for the
makeup water instead of just lime treatment. Approximately 50%
of the cooling tower blowdown must be treated to remove calcium
and sulfate to avoid gypsum scaling in the sluice line. The
cost estimates presented in the next section are based on brine
concentration used for this treatment step.
The second alternative can achieve zero-discharge by
treatment of the ash pond overflow with a brine concentrator/
reverse osmosis unit. Discharge of the ash pond overflow may
require pH adjustment as in the first alternative, depending
on the amount of CO2 transfer in the pond.
The third alternative should not require pH adjustment
of the ash pond overflow before discharge because of the insoluble
nature of the bottom ash.
It should be emphasized here that none of the alterna-
tives should be implemented before more information is gathered
from a bench or pilot scale test program to determine (1) the
actual size of reaction tank required in the sluice system, (2)
the quantity and frequency of acid wash water required to minimize
CaC03 and Mg(OH)2 scale formation, and (3) the solubility limits
for silica in the cooling tower system.
An economic analysis based on rough cost estimates
for these alternatives is given in the following section.
H-41
-------
4.0 ECONOMICS
This section provides rough cost estimates for the
technical alternatives discussed in Section 3.0. Both capital
costs and operating costs are given. These costs should be
considered to be rough estimates for comparative purposes.
The assumptions used in making these estimates are outlined
below.
A capital cost summary for the two technically feas-
ible alternatives employing wet fly ash disposal and the third
alternative (dry fly ash disposal) is presented in Table 4-1.
The fly ash slurry tanks were sized based on a five minute
residence time to permit most of the soluble ash species to be
leached in the tank. The tanks are general storage tanks
equipped with wear liners. One tank was used for each of the
two Comanche units and was assumed to have one agitator to
keep the slurry well mixed. Pumps and piping sizes were based
on the flows used in the simulations discussed in Section 3.0.
Eight-inch carbon steel buried pipe with average fittings,
flanges, shop coating, and wrapping was assumed for fly ash
sluice lines. Four-inch pipe was used for the fly ash recycle
line. A labor-to-material ratio was used for the fly ash re-
cycle installation costs. Engineering costs (direct and
indirect) were assumed to be 7.2% of the combined labor and
material cost (GU-075).
Cast steel pumps with electric motor drivers were
used for all streams. The fly ash slurry pumps were lined with
neoprene for wear resistance, a labor-to-material ratio of 0.36
was used for installation costs. Engineering was assumed to be
10% of the combined labor and material cost (GU-075). All pump
and piping costs were upgraded from 1970 dollars to 1976 dollars
using a factor of 1.56 (based on Chemical Engineering Index).
The operating cost for the once-through system, assum-
ing 2c/kW-hr and 15% per year for capital cost amortization,
totals $90,000/yr (1976 dollars). The recirculating sluice sys-
tem will cost about $863,000/yr, of which $549,000 is for capi-
tal cost amortization, $53,000 is for pump and agitator power
consumption, and the remainder is for brine concentration. If
additional treatment is necessary to prevent silica scale in
the condenser, the cost of this treatment will be an additional
operating cost. Also, if the pond overflow is discharged,
an acid cost may be involved, depending on the amount of C02
transfer in the ash pond and the resultant pond pH. The oper-
ating costs for alternative three are about $38,000/yr, $33,000/
yr being for capital amortization at 15% per year.
H-42
-------
TABLE 4-1. CAPITAL COSTS FOR WATER RECYCLE/
REUSE ALTERNATIVES AT COMANCHE
Alternative One Alternative Two Alternative Three
(1976 dollars) (1976 dollars) (1976 dollars)
Fly Ash Slurry Tanks* 29,000 29,000
Agitators 9,000 9,000
Ash Slurry Pumps 87,000 87,000
and Drivers
Ash Sluice Piping 157,000 157,000
Pond Recycle Pumps 39,000 24,000
and Drivers
Pond Recycle Piping 263,000 157,000
Brine Concentration** 2,400,000
Contingency (20%) 52,000 590,000 36,000
Contractural Fees (3%) 8,000 88,000 5.000
TOTAL 342,000 3,662,000 222,000
*
Includes wear liner and agitator supports.
**
$7750/GPM
References: GU-075, LE-239, MC-136
H-43
-------
TABLE 4-2. CAPITAL AND OPERATING COSTS FOR ATTAINING ZERO DISCHARGE AT COMANCHE
Sof tening/R.O. /Brine Concentration3
Additional Pumps
Additional Piping
•33 Total Additional Cost
1
•P" Costs from Table 4-1
Total Overall Cost for
Zero Discharge
(rails/kW-hr)
Alternative One Alternative Two
Capital Cost' Operating Cost2 Capital Cost1 Operating Cost2
8,107 2,096 3,550 917
17 17 12 5
156 23 144 22
8,280 2,136 3,706 944
342 90 3,662 863
8.622 2,226 7,368 1,807
(0.45) (0.37)
Alternative Three
Capital Cost1
3,720
13
150
3,883
222
4,105
Operating Cost2
962
5
22
989
38
1,027
(0.21)
1976 $ x 10
1976 5 x 10 3, including capital cost amortization at 15% for a 30-year lifetime, based on 80% load factor
Capital Cost = $7750/GPM feed (LE-239)
Operating Cost = $2/1000 gal (not including capital cost amortization)
-------
Additional capital and operating costs for treating
the ash pond overflow so that it may be recycled to achieve
zero discharge are presented in Table 4-2. The overflows can
be treated by a combination of softening, reverse osmosis, and
brine concentration. The clean water can then be recycled to
the plant boiler and cooling tower makeup systems. For the
once-through system, about 32.7 A/sec (518 GPM) must be treated
per unit. For the recirculating system, 14.4 H/sec (229 GPM)
must be treated per unit. For dry fly ash disposal, about 15.1
&/sec (240 GPM) of ash pond overflow from each unit must be
treated to achieve zero discharge at the plant.
The third alternative for achieving zero discharge
is the least expensive since no intermediate treatment to pre-
vent gypsum scaling is required. The second alternative (re-
circulating fly ash sluice system) produces an ash pond over-
flow of about the same magnitude as the third alternative but
requires treatment of the cooling tower blowdown to remove cal-
cium and sulfate.
The once-through sluicing of fly ash and bottom ash
is less expensive than the recirculating system if zero dis-
charge is not desired because no treatment is involved. How-
ever, to achieve zero discharge with the once-through system
requires treatment of a much larger stream than in the recir-
culating system. This makes the once-through alternative more
expensive.
H-45
-------
Appendix I. Recycle/Reuse Options at Montour (Pa. Power and Light Co.)
1.0 INTRODUCTION
This appendix describes the analysis of the water
system at Pennsylvania Power and Light Co.'s (PP&L) Montour
plant, under EPA Contract No. 68-03-2339, Water Recycle/Reuse
Alternatives in Coal-Fired Power Plants. This section presents
a summary of the important results of the study at Montour. The
major water systems at the two-unit, 1500 Mw Montour plant are
the natural draft cooling tower and once-through ash sluicing
systems. Montour was selected with four other plants for eval-
uation of the technical and economic feasibility of various
water recycle/reuse alternatives.
Three major topics are addressed in this study:
1) Existing Operations Modeling
2) Alternatives Modeling
3) Economics
The results of the existing operations simulations of
the cooling towers compare well to the sample data obtained at
the plant. The calculated CaC03 and CaSO^HaO relative satu-
rations in the cooling tower water (0.02-0.04 and 0.02-0.03,
respectively) indicate that the cycles of concentration may be
significantly increased without calcium sulfate (gypsum) scale.
However, a substantial increase in cycles of concentration will
probably require treatment such as acid addition to control
calcium carbonate scale.
Eight cooling tower simulations were performed to de-
termine the degree of acid treatment necessary for increased
cycles of concentration in the towers (to reduce cooling tower
blowdown) and the effects of different magnesium levels in the
makeup water. Magnesium was chosen to be studied since the
effects of calcium and sulfate were identified at other plants.
Magnesium is an important species due to the numerous complexes
formed and can have significant effects on operation. No scale
potential for CaSOit'2H20 was identified in any of the cases.
Sulfuric acid treatment was required for CaCOs scale control in
all cases at or above 8 cycles of concentration. Increased mag-
nesium levels reduced the scale potential of CaCOa and therefore
reduced the amount of sulfuric acid required to prevent CaCOa
scale formation. No scale potential was found for any phosphate
or silicate solids even at 20 cycles of concentration, which is
the highest cycles of concentration considered.
1-1
-------
Table 1-1 presents a summary of the technically feas-
ible options for the Montour water system as compared to exis-
ting operations and the relative costs of each of these alter-
natives. Four process alternatives were studied for the water
systems at Montour. All alternatives sluiced bottom ash and
fly ash at 5 wt. % solids. Mill rejects were sluiced at 0.5%
solids. In all cases ash pond liquor was recycled to the ash
sluicing operation. In one case, Alternative 4, a blowdown was
taken from the system to prevent CaS04'2H20 scale. The other
three alternatives did not discharge any liquid streams and con-
trolled CaS0lt-2H20 scale with softening of a portion of the pond
recycle water. It should be noted here that this analysis was
performed to study general water recycle/reuse alternatives.
Actual implementation of any of these alternatives would require
a more extensive investigation of process parameter variability.
More water quality data would be required along with additional
studies to fully characterize the ash leachability as a function
of time.
Potential scaling of CaC03 is present in all four
cases. However, the fly ash slurry line possibly can be kept
free of plugging by the addition of a fly ash slurry reaction
tank and/or by flushing with acidic water. Pilot or bench
scale testing is recommended to determine accurately the size
of reaction tank and frequency and quantity of acid washing
required or if other measures are necessary.
For each alternative the effect of C02 mass transfer
in both the pond and the sluice tank was studied. It was found
that C02 equilibrium in both the tank and the pond decreased
the pH and thus the scale potential for CaC03 and Mg(OH)2. C02
transfer did not significantly affect CaSO^HjO scale potential.
The first alternative assumes that the cooling tower
drift from one tower was equal to the design value of 32 £/sec
(500 GPM). Enough blowdown was drawn from the cooling towers
to serve as makeup to the recirculating sluicing operation.
Under this situation the cooling tower should be operating near
8 cycles of concentration.
The second alternative is identical to Alternative 1
except that the cooling tower drift was assumed to be negligi-
ble. This assumption increased the cycles of concentration from
8 to 20 even though the blowdown rate was not changed. This had
the effect of requiring more softening for the pond recycle
stream because of the poorer quality of makeup water to the ash
sluicing operations.
1-2
-------
TABLE 1-1. SUMMARY OF TECHNICALLY FEASIBLE OPTIONS AT MONTOUR
Existing Condition
Alternative 1
Alternative 2
Cycles of Concentration
in Cooling Towers
Assumed Drift Rate
in Cooling Towers
!l/sec (GPM)
Blowdown from Cooling
Towers
t/sec (GPM)
% Recycle in Fly Ash
1.5 - 2.0
62 (1,000)
725 (11,500)
20
62 (1,000)
48 (760)
40 (650)
Total Makeup Water Rate,
t/sec (GPM)
Ultimate Effluent Rate,
i/sec (GPM)
Treatment Required
Costs'
1,500 (24,000)
500 (7,900)
None
1,000 (16,000)
950 (15,000)
HaSO:, (Cooling Tower) '
Na2COa (Pond Recycle)
i, (Cooling Tower)'
NaiCOa (Pond Recycle)
Alternative 3
20
40 (650)
H
1
U>
Sluicing Syscea
Sluice Systea Makeup
Source
0
Cooling Tower
Slowdown
89.
Cooling Tower
Blowdown
89.
Cooling Tower
Blowdown
89.
River Water
985 (15,600)
O^ (Cooling Tower)'
Na2C03 (Pond Recycle)
Alternative 4
20
40 (650)
73.
River Water
1,035 (16,400)
50 (800)
H2SO* (Cooling Tower)'
.pital, 1976 $
'crating, 1976 S/yr
(mils/kW-hr)
640,000
173,000
(0.016)
668,000
187,000
(0.018)
622,000
169,000
(0.016)
485,000
103.0OO
(0.010)
*Sulfurlc acid treatment for CaCOj scale control
2Na2COj softening for Ca removal
i
These rough cost eatimates were made to compare technically feasible options anJ do not include a "difficulty to retrofit" factor.
Includes capital amortization of 15% per year.
-------
The other two alternatives assume that the cooling
towers can be operated at zero blowdown. This requires that
the drift be at least 65% of the design value. Under these
alternatives the makeup water to the ash sluicing operation is
obtained directly from the Susquehanna River or the plant makeup
pond. Alternative 3 employed softening and attained zero dis-
charge similar to the two previous alternatives. Alternative 4
controlled the CaSCU^HaO scaling potential by the use of a blow-
down stream of about 50 Si/sec (800 GPM) from the ash pond (both
units).
Rough cost estimates for the different alternatives
are also presented in Table 1-1. Alternative 4 is the least
expensive due to the fact that no softening was required. The
other three vary mostly in the degree of softening that was re-
quired for the recirculating ash sluicing system. It should be
emphasized that Alternatives 1 and 2 differ only in the assump-
tion concerning the drift rate in the cooling towers. If more
information could be obtained about the actual drift rate, a
more reliable cost estimate could be made.
Detailed discussions of the existing operations simu-
lations, the alternative simulations, and the rough cost esti-
mates constitute the main body of this appendix.
1-4
-------
2.0 PLANT CHARACTERISTICS
The Pennsylvania Power and Light (PP&L) Montour Sta-
tion consists of two 750 Mw coal-fired units located near
Washingtonville, Pennsylvania. The coal used at Montour is
supplied by three different mines: Oneida, Rushton, and Green-
wich. The Greenwich mine accounts for more than 80% of the coal
burned at Montour. Typical Greenwich coal is approximately 20%
ash and 2% sulfur with a heating value of about 12,000 Btu/lb.
The plant has two large natural draft cooling towers and uses
once-through sluicing for fly ash, bottom ash, and mill reject
disposal.
This section of the appendix describes the character-
ization of Montour's water system. First, an overall plant
water balance is presented which shows the major in-plant flows
and chemical analyses for the streams which were sampled. Then
a detailed description of each of the major water consumers in
the plant is given. This is followed by a discussion of the
process model and the input data used to simulate existing oper-
ations at Montour. The computer simulation results are finally
presented and discussed. This discussion includes a comparison
of the simulation results and the chemical analyses of the
samples taken. Areas which show a potential for water recycle/
reuse at Montour are identified and discussed.
2.1 Water Balance
A flow schematic for the Montour water system is shown
in Figure 2-1. The major streams in the plant, including the
cooling tower and ash handling systems, are shown in this dia-
gram. The numbers in the diamonds refer to the stream numbers
shown with the design flows and results of the chemical analyses
of the samples taken at Montour. A more detailed description
of the samples taken and.analytical procedures used is presen-
ted in Appendix B.
The composition of the makeup water is consistent with
data obtained from PP&L about the water quality of the Susque-
hanna River from January, 1974 until December, 1976. The con-
centrations of the key species used in this study (Ca, Mg, Na,
K, Cl, SC\ , and N03 as well as TDS and pH) were all in the
range of values found over this three year period. The sili-
cate and the phosphate concentrations found in the samples
were smaller than those found in the data supplied by PP&L.
1-5
-------
<^HILL'SCOJAC3Ut
B
Figure 2-1. Pennsylvania Power and Light Company
Montour Plant Water Balance.
-------
Of M tf/f»Ai Iff It
* f
tOJJ/J
ftf ASH
StvrCf
MILL XSJfCTS
C
-------
Stream Number
Stream Name
Flow: Metrlc
English
pH
Calcium
Magnesium
Sodium
Potassium
Chloride
Carbonate (as COj)
Sulfate (as SO,,)
Nitrate (as N03)
Phosphate (as POi,)
Silicates (as SiOj)
Arsenic
Suspended Solids
Dissolved Solids
<»
Cooling
Tower
Makeup
1,500 JL/sec.
24,000 gpm
8.1
28. A
5.5
7.0
2.6
19
6.0
66
5.5
.029
0.9
.0008
100
<£>
Cooling
Tower
Slowdown*
450 £/sec.
7,400 gpm
7.5
46.5
11.1
10.2
3.0
33
9.3
110
11.0
2.5
<.02
.0004
235
<»
Fly
Ash
Slurry
300 i/sec.
4,800 gpm
8.9
142.0
10.4
12.6
9.9
38
3.8
267
13.3
.224
2.1
.067
2.1
690
^
Mill
Reject
Slurry**
80 I/sec.
1,200 gpm
6.9
39.9
11.7
9.4
4.2
33
7.8
78
11.8
0.5
.0056
190
^>
Bottom
Ash
Slurry**
120 i/sec.
1,900 gpm
5.8
39.4
9.0
9.8
5.0
32
35.4
101
11.5
.040
2.0
.0532
200
<^
Misc.
Waste
90 Jl/sec.
1,400 gpm
7.7
28.4
6.0
7.4
0.9
18
10.8
66
5.5
0.6
.0016
170
Ash
Basin
500 if see.
7,900 gpm
8.7
98.9
10.0
11.4
8.2
33
24
197
6.8
1.4
.004
470
4>
Ash
Basin
Overflow
500 I/sec.
7,900 gpm
7.7
98.9
10.0
11.8
7.4
34
9.6
245
9.9
1.02
2.0
.0012
460
<^
Detention
Basin
Overflow
590 Z/sec.
9,300 gpm
7.5
87.4
9.0
19.1
6.6
29
1.5
215
11.1
.056
1.4
.0016
290
I
CT\
cr
*Average of the values found from each cooling tower.
••Normalized to continuous operation. Actual flows are intermittent.
Figure 2-1. (Continued)
-------
Makeup water for the plant is taken from the Susque-
hanna River and can be stored in a raw water reservoir. Water
is supplied to the system at a design rate of 1500 a/sec (24,000
GPM) and is used as general service water, boiler makeup, and
cooling tower makeup.
Miscellaneous plant wastes total about 90 I/SQC (1,400
GPM) and are sent to a detention basin before ultimate discharge.
The major water consumers at the Montour plant are the cooling
tower system and the ash handling systems, which are discussed
in the following sections.
The first step in analyzing the water systems at Mon-
tour is to examine the results of the grab samples obtained at
Montour. Parameters calculated by the equilibrium program for
the streams sampled are presented in Table 2-1. Included are
the relative saturations of CaC03, CaSO^, and Mg(OH)2 as well
as the equilibrium partial pressure of C02 and the "L residual
electroneutrality. These parameters are useful for characteri-
zation of the individual streams.
The relative saturation is a parameter which indicates
the potential of a stream to produce scale. When the relative
saturation is greater than the critical value, solids formation
is likely. The critical values for the three species reported
in Table 2-1 are 2.5 for CaC03, 3.4 for Mg(OH)2, and 1.3-1.4 for
CaS04.2H20. From the values in the table no scale potential is
calculated for any stream except for the fly ash sluice and the
ash pond, where CaC03 relative saturations are 6.8 and 8.1, re-
spectively.
The equilibrium partial pressure of C02 for each of
the streams sampled at Montour is also presented in Table 2-1.
The partial pressure of C02 in air is about 3.3 x 10"" atm.
Most of the streams have partial pressures of C02 near this
value. The cooling tower makeup and the fly ash sluice values
are significantly lower than atmsopheric indicating a tendency
to absorb C02 and the bottom ash sluice value is higher than
atmsopheric indicating a tendency to desorb C02.
The percent residual electroneutrality is the differ-
ence between the total positive charge and the total negative
charge as a percent of the total charge. It is an indication
of how accurately the actual stream is represented by the
computer model. More information on the residual electroneu-
trality is presented in Appendix E. The values reported in
1-7
-------
TABLE 2-1. PARAMETERS CALCULATED BY THE EQUILIBRIUM PROGRAM FOR MONTOUR SAMPLES
00
Relative Saturations*
Stream Name
Cooling Tower Makeup
Cooling Tower Slowdown
Fly Ash Sluice
Mill Rejects
Bottom Ash Sluice
Miscellaneous Wastes
Ash Basin
Ash Basin Overflow
(Acidified)
Detention Pond Overflow
Stream No.
1
2
3
4
5
6
7
8
9
CaC03
.013
.031
6.83
2.94 x 10"'
3.84 x 10""
.014
8.08
.024
.021
Mg(OH)2 CaSO,'2H20
5
2
2
4
5
2.
3.
1.
6.
.45
.34
.49
.79
.40
.82
,86
73
36
x 10' 7
x 10"'
x 10" 3
x 10' e
x 10-'°
x 10" '
x 10" s
x 10"'
x 10' '
.012
.023
.105
.015
.019
.011
.066
.093
.076
Equilibrium Partial
Pressure of COz
atm
3
2.
3,
6
1
1
3
1
3
,95
.84
.18
.96
.13
.78
.74
.46
.54
x 10' 5
x 10-*
x 10" 5
x 10~*
x 10~2
x 10" "
x 10"'
x lO'"
x 10" *
?„ Residual
Electroneutrality
4.
4.
11.
12.
1.
3
-18
2.
5,
1
7
3
1
2
.2
.9
.4
.3
*Critical values, above which scale potential exists, are 1.3-1.4 for CaSOi, • 2H20, about ?.. 5 for CaCOj ,
and about 3.4 for Mg(OH)2 (see Appendix C)
-------
Table 2-1 are generally quite good and tend to confirm the accu-
racy of the analyses.
2.2 Cooling Tower System
Each of the two 750 Mw units has an independent
cooling system. Water circulates between each tower and con-
denser at a rate of about 16,000 H/sec (250,000 GPM). The
blowdown is removed from the system before the condenser.
Water from either the river or the raw water reservoir or both
is added to the system as makeup to replace water lost through
evaporation, drift, and blowdown.
The towers normally operate at 1.5 - 2.0 cycles of
concentration, which may be defined as the ratio of blowdown
species concentrations to makeup species concentrations. In
terms of flow rates, cycles of concentration is defined as:
r _ E + B + D
L B + D
where C = cycles of concentration
E = evaporation rate
B = blowdown rate
D = drift rate
2.2.1 Simulation Basis
The existing operations of the Montour cooling towers
were simulated by means of the computer model shown in Figure
2-2. This was done in order to verify the validity of this
model and to establish a sound basis from which potential
recycle/reuse options could be compared. The model used infor-
mation concerning the inlet air flow and composition, the
makeup water composition, the recirculating water flow rate,
the drift rate, the desired cycles of concentration, and the
temperature change across the condenser as inputs. From these
inputs the heat load on the condenser, and the flows and compo-
sitions of all the streams in Figure 2-2 are calculated. A
detailed description of the cooling tower model including a
brief discussion of the subroutines in Figures 2-2 is presented
in Appendix E.
1-9
-------
I—1
o
SLOWDOWN
ORDER OF PROCESS CALCULATIONS
1,2,3,4,5,6,7,18,9,10,11,10,12,6.7)
SOFTENING
1
3
4
5
CHEMICALS ~~
OVERALL
SYSTEM
BALANCE
tCTBHLO
6
7
9
13
OUTLET AIR
DRIFT
^-SLOWDOWN
CHEMICAL WASTE
Figure 2-2. Process simulation scheme for Montour cooling tower system.
02-1169-1
-------
Several assumptions were made in modeling the cooling
l-i 4-T-i-io <-> i" mi -11 «a t- -1 rcn TVi d o o •inr>1iir1a.
towers with this simulation. These
1) Equilibrium exists between C02 and H20 in
the atmosphere and cooling tower exit
water.
2) The temperature of the cooled water
stream approaches the wet bulb tempera-
ture of ambient air within a predictable
range.
3) The compositions and temperatures of the
cooled water and drift streams are equal.
4) Ionic reactions taking place in the
liquid phase are rapid and thus at
equilibrium.
The assumption involving the temperature of the
cooled water stream is a recognized design parameter in cooling
tower evaluation and gives a good approximation. The assumption
concerning the temperature and composition of the drift stream
should be very close to actuality as is the assumption in regard
to H20 gas-liquid equilibrium. The assumption with regard to
C02 equilibrium is conservative since the partial pressure of C02
in actual cooling towers tends to be slightly greater than the
equilibrium value. The lower equilibrium concentration of
carbonate species, assumed in the model, causes the pH to be
slightly higher in the model than in actual operation. The
higher pH causes the relative saturation of CaC03 to increase
more than the lowered carbonate species concentration causes it
to decrease.
A summary of the input stream data employed in the
existing operations simulations is presented in Table 2-2.
The air temperature and composition were calculated using
local climatological data for Williamsport between December,
1975 and August, 1976. The makeup water composition was
obtained from chemical analyses of the spot sample taken in
November, 1976.
The cooling tower drift rate, approach, cycles of
concentration, and circulating water flow were obtained directly
from PP&L or calculated from data obtained from PP&L. The
condenser temperature change was also obtained from PP&L. The
1-11
-------
TABLE 2-2. INPUT DATA FOR MONTOUR COOLING TOWER SIMULATIONS
December 1975
August 1976
FLOWS
Air, rnVhr
(ACFM)
Drift, H/sec
(GPM)
Circulating Water, i/sec
(GPM)
3.5 x 107
(2.06 x 107)
31.5
(500)
14,600
(232,000)
3.5 x 107
(2.06 x 107)
31.5
(500)
18,700
(296,000)
TEMPERATURES
Ambient Air, °C
(°F)
Approach, °C
(°F)
Condenser AT, °C
(°F)
Wet Bulb, °C
(°F)
Condenser Outlet, °C
CF)
ADDITIONAL DATA
Relative Humidity, ?<>
Cycles of Concentration
Makeup Water Composition, mg/2,
Calcium
Magnesium
Sodium
Chloride
Carbonate (as C07)
Sulfate (as SOZ)
Nitrate (as
0
(32)
20
(36)
20
(36)
-2
(29)
36
(97)
72.0
2.0
28.4
5.5
8.1
22.0
6.0
67.0
5.5
21
(70)
10
(18.5)
16
(28)
18
(64)
43
(110)
72.0
1.5,2.0
28.4
5.5
8.1
22.0
6.0
67.0
5.5
1-12
-------
ambient air wet bulb temperatures were derived from Williamsport
climatological data for December, 1975 to August, 1976.
Examples of the calculations performed to obtain the data are
presented in Appendix K.
2.2.2 Simulation Results
This section describes the results from the simulation
of existing cooling tower operations at Montour. Simulations
were performed using weather data from December, 1975 and
August, 1976. The makeup water composition was based on that
which was sampled at the plant in November, 1976> with adjust-
ment to/.minimize the residual electroneutrality using the method
described in Appendix E.
Samples were taken from the blowdown of both cooling
towers. Table 2-3 presents the data obtained from both towers
as well as the results of the simulations performed. Two simu-
lations were done under summer conditions at 1.5 and 2.0 cycles,
and one under winter conditions at 2.0 cycles of concentration.
The three simulations agree reasonably well with the
sampled data. The pH in all three simulations is between the
values found at the plant. The simulations at 2 cycles show
concentrations which were slightly higher than those measured
for all species except magnesium and nitrate concentrations
which were between the sample values. The simulation at 1.5
cycles tended to have concentrations which were slightly lower
or in between sampled values except for carbonate which was
slightly higher. These discrepancies may be due to deviations
from steady state during sampling, nonhomogeneous sampling
and/or analytical errors.
The CaC03 relative saturations are consistent with
the input data. The larger value reported under summer condi-
tions is probably due to the small difference in pH which is
caused by temperature differences. The CaSOn relative satura-
tions agree very well with those found at the plant. In both
cases low values were found indicating that the cooling towers
were being run under extremely safe conditions.
In addition to relative saturations, the ACS Index
is reported. This is an index developed by the Asbestos Cement
Pipe Manufacturers Association and is obtained from the follow-
ing expression:
1-13
-------
TABLE 2-3. EXISTING COOLING TOWER SIMULATION RESULTS
Cooling Tower Blow down
FLOW, a/sec per tower
(GPM)
EM
COMPOSITION, mg/l
Calcium
Magnes ium
Sodium
Chloride
Sulface (as SOT)
Carbonate (as CO 3)
Nitrate (as NOi)
RELATIVE SATURATIONS***
CaCOj
CaSO.. -2H20
ACB Index
PARTIAL PRESSURE C02 , atm
Plant
Tower #1
820*
(13,000)
7.8
43.2
9.5
12.2
32.0
88.0
9.6
11.8
.037
.019
11.0
1.32 x 10""
Data
Tower //2
410**
(6,500)
7.3
49.7
12.6
11.7
33.0
131.0
9.0
10.2
.025
.027
10.6
4.36 x 10""
December '75
2.0 Cycles
290
(4,600)
7.5
56.7
11.0
16.1
44.0
135.0
13.1
11.0
.030
.033
11.0
3.3 x 10""
Simulations
August '76
1 . 5 Cycles
760
(12,000)
7.5
42.5
8.2
12.1
33.0
100.0
10.9
8.2
.042
.020
10.8
3.3 x 10""
August '76
2.0 Cycles
360
(5,700)
7.6
56.7
11.0
16.1
44.0
135.0
12.8
11.0
.073
.031
11.1
3.3 x 10""
*based on design evaporation and drift rates and 1.5 cycles of concentration
**based on design evaporation and drift rates and 2.0 cycles of concentration
**^critical values, above which scale potential exists, are 1.3-1.4 for CaSOi.-^HzO and
about 2.5 for CaC03 (see Appendix C)
-------
ACB = pH + log (Calcium x Alkalinity)
This index gives information concerning the potential for
corrosion of the type of fill used in the cooling towers at
Montour. An example calculation is presented in Appendix K.
The values calculated for the simulated runs compare very well
with those found in the plant.
These simulations indicate a potential for reducing
water requirements and discharges for cooling towers by increas-
ing the cycles of concentration. As the cycles are increased,
an increase in the relative saturations of CaC03 and CaSO^ is
expected indicating an increase in the potential for scaling
problems. CaCOa scale control can be achieved through acid
addition for pH adjustment. As cycles are increased the ACB
Index will increase which means there will be a decrease in the
corrosiveness of the water.
2. 3 Ash Handling Systems
Fly ash is collected by each electrostatic precipi-
tator at a rate of about 31,100 kg/hr (68,400 lb/hr) from each
unit. The fly ash is sluiced on a once-through basis to the ash
pond using cooling tower blowdown as sluice water. The ash is
slurried at about 5% solids using 145 A/sec (2400 GPM) of
cooling tower blowdown.
Bottom ash and mill rejects are periodically sluiced
on a once-through basis also. The amount of water needed for
sluicing determines the rate at which blowdown is taken from
the cooling towers.
2.3.1 Simulation Basis
The existing operations of the Montour ash sluicing
system were simulated by means of the model shown in Figure 2-3.
As with the cooling tower simulations, this was done in order
to verify the model and establish a basis for comparison. The
model used information about the composition and flows of the
makeup water and the fly and bottom ash as well as the percent
solids in the sludge and pond evaporation, as inputs. From
this information the flows and compositions of all the streams
in Figure 2-3 were calculated. A detailed description of the
ash sluicing model including a brief discussion of the sub-
routines in Figure 2-3 is presented in Appendix E.
1-15
-------
1
M
O>
MAKEUP FOR
FLY ASH SYSTEM
WTRINP
C.T.
LOWOOW
POND
(NONE)
POND OVERFLOW
MAKEUP FOR BOTTOM ASH SYSTEM
ASHiNP
BOTTOM
ASH
3
BOTTOM ASH SLURRY
7
8
16
9
10
7 SLUICE WATER
VAPORIZED
8 SLUDGE
9 FLY ASH SLUICE
VAPORIZATION
10BTM ASH SLUICE
VAPORIZATION
ORDER OF PROCESS CALCULATION: 1, 2. 3, 4. 5, 6, 7, 8, 9, 10*
Figure 2-3. Ash sluicing simulation model.
-------
Several assumptions were made in modeling the ash
sluicing system with this simulation. These include:
1) Ionic reactions taking place in the liquid
phase are rapid and thus at equilibrium.
2) Solid-liquid equilibrium is achieved in
the ash pond, with the exception of CaSOij
which is allowed to remain supersaturated.
3) Ash dissolution is essentially complete
before the slurry reaches the pond.
4) All solids precipitation occurs in reaction
vessels or the pond.
The input data required to simulate the once-through
ash sluicing system at Montour are presented in Table 2-4.
The fly ash, bottom ash, and water flows were obtained from
data supplied by PP&L. Typical slurries of 5% solids were
modeled. The evaporation rate was calculated using average
ambient conditions for wind speed, temperature, and humidity.
The sluice water composition used was blowdown from the cooling
towers at 1.5 cycles. Sample calculations are presented in
Appendix K.
Ash leaching and bench scale recirculating ash sluic-
ing studies were performed to obtain soluble species data. The
results of these studies are presented in Appendix D.
2.3.2 Simulation Results
This section describes the results from the ash
sluicing simulations of existing operations. Two simulations
were performed. The first did not allow C02 transfer in the
ash pond and the second allowed the C02 in the pond to come to
equilibrium with the atmosphere.
Table 2-5 presents the results of these simulations.
The compositions of the pond liquor and fly ash slurry sampled
at the plant are compared to those predicted by the model.
The concentrations of calcium, magnesium, sodium,
chloride, sulfate and nitrate are not affected by the degree of
1-17
-------
TABLE 2-4. MONTOUR EXISTING ASH SLUICING INPUT DATA
FLOWS (per unit)
Fly Ash, kg/min
(Ib/min)
Bottom Ash*, kg/min
(Ib/min)
Fly Ash Sluice Water, i/sec
(GPM)
Bottom Ash Sluice Water, £/sec
(GPM)
Pond Evaporation, £/sec
(GPM)
SLUICE WATER COMPOSITION
Cooling Tower Slowdown (§1.5 cycles, ing/2,
Calcium
Magnesium
Sodium
Chloride
Carbonates (as CO^)
Sulfates (as SO"^)
Nitrates (as NOa)
POND DEPOSITS (wt. % solids)
SOLUBLE FLY ASH SPECIES (wt. 7. solids)
CaSOi*
MgO
MgSO,
Na20
520
(1140)
560
(1230)
175
(2750)
185
(2960)
0.6
(10)
42.5
8.2
12.1
33.0
10.9
8.2
100.3
50.0
1.025
0.008
0.066
0.054
*Includes mill rejects
1-18
-------
TABLE 2-5. MONTOUR EXISTING ASH SLUICING OPERATIONS
Composition, mg/£
Calcium
Magnesium
Sodium
Chloride
Carbonates (as CO 3)
Sulfates (as SOO
Nitrates (as NO^)
pH
Relative Saturations *
CaC03
Mg(OH)2
CaS0lt'2H20
Equilibrium Partial Pressure
of C02, atm.
Fly Ash
Plant Data
142.0
10.4
18.4
38.0
3.8
267.0
13.3
8.9
6.83
2.49 x 10~3
.105
3.18 x 10~5
Slurry
Model
193.0
17.1
12.1
33.2
10.9
486.0
8.2
9.66
8.14
.090
.206
1.05 x 10~6
Plant Data
98.9
10.0
16.2
33.0
24.0
197.0
6.8
8.7
8.08
3.0 x 10~5
0.066
3.74 x 10~"
Pond Liquor
Model
(No C02 Transfer)
115.5
12.6
12.2
33.2
11.3
288.0
8.2
9.4
3.23
7.6 x 10~3
0.102
2.68 x 10"6
Model
(CO 2 Equilibrium)
115.5
12.6
12.2
33.2
16.3
288.0
8.2
7.63
0.112
2.2 x 10~S
0.102
3.3 x lO"1*
*Critical values, above which scale potential exists, are 1.3-1.4 for CaSO^*2H20, about 2.5 for CaCOs,
and about 3.4 for Mg(OH)2 (see Appendix C)
-------
C02 transfer as long as no solids are allowed to form. Both
magnesium and chloride simulated concentrations are very close
to those found at the plant for the ash pond liquor. The
sodium and nitrate values are slightly lower in the simulation
and the calcium is higher. The sulfate concentration is sub-
stantially higher in the simulations than in the plant samples
due to the high sulfur content of the ash which was used in the
leaching studies (see Appendix D). Although this high sulfur
ash is probably not typical, it should make the calculated
values for the relative saturation of CaSOt, higher than would
ordinarily be expected. This will make any recycle/reuse
options obtained from these data conservative relative to CaSOit
scaling.
The pH, carbonate concentration, and the relative
saturation of CaCOs were affected by the degree of C02 transfer.
The pH in the sample was between the value found with no C02
transfer and the value with C02 equilibrium. The carbonate in
the sample was greater than either simulation but it was closer
to the value calculated from the simulation with C02 equilib-
rium. The relative saturation of CaSOi, predicted by the simu-
lations were higher than the sample value due to the higher
sulfate concentration. The relative saturation of CaC03 was
lower in the simulations than the sample value due to the lower
carbonate concentrations. The relative saturation of Mg(OH)2 in
the simulations bound the sample value.
The high relative saturation of CaC03 in the sample
and in the case of no C02 transfer indicates that the ash
sluicing system may be forming CaC03 solids. The erosive
properties of the slurried ash may be keeping the walls of the
slurry line "clean" of CaC03 scale.
The results of these simulations indicate that some
degree of C02 transfer does occur in the pond. They also verify
that this model can be successfully used to simulate the ash
sluicing operations at Montour.
1-20
-------
3.0 TECHNICAL ALTERNATIVES
A modular approach to studying water recycle/reuse
alternatives at Montour was used in that the major plant water
systems were divided into two subsystems to form separate pro-
cess simulations. One subsystem consists of the cooling tower
from one plant, with associated treatment facilities (where
necessary), hold tanks, and condenser. The other subsystem
consists of the ash handling system. The studies for each sub-
system will be discussed separately. The effects of increasing
the cycles of concentration in the cooling towers and of differ-
ent quality makeup water (different magnesium levels) are pre-
sented first. Then the use of cooling tower blowdown in a
recirculating ash sluice system is evaluated. The effects of
carbon dioxide mass transfer between the atmosphere and the
pond liquor are also investigated. Magnesium was chosen as the
makeup parameter to be studied since calcium and sulfate vari-
ations were studied at other plants. Magnesium is an important
species because of the numerous complexes which may form.
3.1 Cooling Tower System
The existing operations simulations indicated that the
cycles of concentration may be increased in the cooling towers
without scaling with respect to calcium sulfate or calcium car-
bonate. Although calcium sulfate scaling must be controlled by
calcium removal, calcium carbonate scale can be controlled by
acid treatment of the circulating water. This section first
presents a description of the simulation bases used, then a
discussion of the results with respect to increased cycles of
concentration and magnesium in the makeup water.
3.1.1 Simulation Basis
The process model used to simulate alternatives for
cooling tower operation is identical to that used for existing
operations (Figure 2-2). Acid treatment for calcium carbonate
scale control was implemented to keep the CaC03 relative satur-
ation between 0.8 and 1.0.
A total of eight simulations were performed for alter-
native cooling tower operations. Six simulations were performed
with the existing makeup water quality and cycles of concentra-
tion of 2.5, 4.0, 8.0, 10.5, 14.0, and 20.0. Two other simula-
tions were performed with magnesium concentrations of one-third
and tripled, at 14.0 cycles of concentration. In order to main-
tain a pH near the sample value, the magnesium was "added" as
1-21
-------
MgCl2. Chloride was chosen as the counter ion because it does
not complex appreciably with other ions.
All of the alternative cooling tower simulations were
performed for summer operation of one tower since these condi-
tions represent the case of maximum blowdown rates. Increased
evaporation rates realized during the summer months necessitate
an increase in blowdown rate over that required during the winter
months to maintain a constant value for cycles of concentration.
The only changes in the input data for the first six
alternative simulations are the values for cycles of concentra-
tion. The adjusted makeup water compositions used in the last
two cooling tower simulations (different magnesium levels) are
shown in Table 3-1, along with the composition used in the first
six simulations.
3.1.2 Effect of Increased Cycles of Concentration
The simulation results from six alternative operating
modes of the cooling towers are presented in Table 3-2. As ex-
pected, at higher cycles of concentration acid addition was re-
quired to control CaCOs scale potential but in no case did the
CaSO n relative saturation become large enough to necessitate
calcium removal.
The two simulations at 2.6 and 4.0 cycles of concen-
tration did not require acid addition. The simulations at 8.0
and 10.5 cycles did indicate a need for acid addition. Much
smaller blowdown rates were attained at the higher cycles. In
order to increase the cycles of concentration to 14.0 and 20.0
the drift was reduced from its design value of 32 £/sec (500
GPM). At these reduced drift rates (see Table 3-2) zero blow-
down was attained without CaSCK scale.
Where acid addition was used, the relative saturation
of CaCOs was kept between 0.8 and 1.0. This is well below the
critical scaling value of 2.5 (see Appendix C). This was done
to insure that upsets in the calcium or carbonate levels would
not cause scaling.
The simulation at 20 cycles of concentration indicates
that the cooling towers at Montour might be able to be operated
at zero discharge. The drift in this simulation was assumed to
be 351 less than the design value reported by PP&L. Since the
1-22
-------
TABLE 3-1. ADJUSTED WATER MAKEUP COMPOSITIONS
FOR INCREASED MAGNESIUM LEVELS*
Carbonates (as CO 3)
Sulfates (as SO^)
Nitrates (as NO^)
Chloride
Calcium
Magnesium
Sodium
pH
1/3 Magnesium
(mg/A)
6,0
67.0
5.5
11.3
28.4
1.83
8.1
8.4
Normal Magnesium
(mg/Jl)
6.0
67.0
5.5
22.0
28.4
5.5
8.1
8.3
Triple Magnesium
(ing/ A)
6.0
67.0
5.5
54.1
28.4
16.5
8.1
8.3
*Magnesium levels were investigated since calcium and sulfate were considered
at other plants.
1-23
-------
TABLE 3-2. SIMULATION RESULTS FOR INCREASED CYCLES OF CONCENTRATION
Cycles of
Concentration
2.6
4.0
8.0
10.5
M
K> 14-0
20.0
Cooling Tower Slowdown
Flow
I/sec
213.
98.9
24.3
9.64
0.
0.
GPM
3380.
1570.
386.
153.
0.
0.
Relative
CaCOa
.146
.447
.953
.880
.898
.987
Saturations**
CaSO.,'2H20
.046
.086
.229
.332
.479
.749
ACB
11
11
12
12
12
12
Index
.39
.94
.38
.40
.45
.57
PH
7.69
7.87
7.92
7.89
7.77
7.84
Sulfuric
kg/day
0.
0.
74.
112.
121.
145.
Acid Rate*
Ib/day
0.
0.
162.
246.
266.
320.
Drift
H/sec
32
32
32
32
30
21
Rate
GPM
500
500
500
500
477
326
*as 100% H2SO,.
**Critical values, above which scale potential exists, are 1.3-1.4 for CaSOi,'2H20 and
about 2.5 for CaC03 (see Appendix C)
-------
air flow rate in the Montour towers remains essentially constant
regardless of the season, the drift should not vary widely.
The ACB Index was developed by the Asbestos Cement
Pipe Manufacturers Association in order to quantify the potential
for corrosive action by the cooling water on cooling tower fill.
The calculations required to compute this index are presented
in Appendix K. A value below 10 indicates highly aggressive
conditions, and a value above 12 indicates a non-aggressive
condition. Figure 3-1 is a plot of the ACB Index as a function
of cycles of concentration as found in the simulations performed
for Montour. Below 4 cycles of concentration there is a steep
slope, whereas above 8 cycles the slope is significantly less.
Above 8 cycles pH control by H^SOu addition was necessary in
order to prevent CaCOs scale formation. The advantage of run-
ning these cooling towers at high cycles of concentration with
acid addition is dramatically displayed in this graph. In all
cases where acid is used the ACB Index indicates a non-corrosive
environment.
In Table 3-3 the relative saturations of many scale
forming species at 20.0 cycles of concentration are reported.
Since all of these species are subsaturated at 20.0 cycles they
should also remain subsaturated under less concentrated situa-
tions. The relative saturation of CaC03 and CaSOi, are well
below their critical scaling values of 2.5 and 1.3-1.4, respec-
tively. The silicate with the largest relative saturation is
MgsSiaOs(OH)^ (serpentine) with a value of 0.259, which implies
that this solid is thermodynamically unstable in this water.
3.1.3 Effect of Magnesium Concentration in the Makeup Water
In addition to the simulations performed at varying
cycles of concentration, two other simulations were performed
at 14 cycles and different magnesium concentrations in the makeup
water. The results of these simulations are presented in Table
3-4.
The results of these simulations point out that as
the concentration of magnesium increases the amount of acid
required to keep the relative saturation of CaCOs between 0.8
and 1.0 decreases. Similarly increases in magnesium causes the
relative saturation of CaSCs to decrease. This is easily ex-
plained in that increasing the magnesium level from 1.83 mg/£
to 16.5 mg/£ causes the activity coefficients of the calcium,
carbonate, and sulfate ions to decrease. Calcium decreased from
1-25
-------
I
NJ
cq
o
1t
10
456
7 8 9 10
(CYCLES)
20 30
100
Figure 3-1.
Asbestos cement pipe manufacturers index as a
function of cycles of concentration.
-------
TABLE 3-3. RELATIVE SATURATION OF SCALE FORMING
SPECIES AT 20 CYCLES OF CONCENTRATION
IN MONTOUR COOLING TOWERS
Solid Species Relative Saturation
Ca(OH)2 6.46 x 10"l °
CaC03 0.987
CaSOt,-2H20 0.749
CaHPO,* 0.019
Ca3(P(K)2 3.8 x 10~3
CaH2SiO<, 1.0 x 10'"
Ca(H3Si04)2 5.0 x 10""
Mg(OH)2 9.76 x 10~5
MgC03 7.04 x 10"5
Mg3Si205(OH)4 0.259
Mg2Si305(OH)6 2.49 x 10~"
Mg(Si02)3(OH)2 0.031
Si02 0.105
1-27
-------
TABLE 3-4. SIMULATION RESULTS WITH DIFFERENT
MAGNESIUM CONCENTRATIONS (14 CYCLES)
Magnesium in
the Makeup
Water (mg/£) 1.83 5.5 16.5
Relative Saturations*
CaC03
CaSOi>-2H20
Mg(OH)2
Activity Coefficients
Ca^
col
so"
Sulfuric Acid
kg /day
Ib/day
pH
0.94
0.52
3.5 x 10"5
0.50
0.49
0.47
137
301
7.89
0.90
0.48
1.1 x 10'1*
0.49
0.48
0.45
121
266
7.80
0.89
0.38
2.0 x IQ~*
0.44
0.43
0.40
119
261
7.77
^Critical values, above which scale potential exists, are 1.3-1.4
for CaS0lt-2H20, about 2.5 for CaC03, and about 3.4 for Mg(OH)2
(see Appendix C).
1-28
-------
0.50 to 0.44, sulfate decreased from 0.47 to 0.40, and carbonate
decreased from 0.49 to 0.43. The increased magnesium levels
cause increased complexing between sulfate, carbonate, and magne-
sium which results in lower activities as well as the lowered
activity coefficients. Thus, the increased presence of magnesium
is seen to inhibit the scaling potential of other species.
In addition to the beneficial effects of increased
magnesium there exists a potential hazard of Mg(OH)2 scale. As
seen in Table 3-4, the relative saturation of Mg(OH)2 does in-
crease with increasing concentrations of magnesium, but remains
very low for the water quality expected at Montour.
3.1.4 Conclusions
From the results presented above, the following con-
clusions can be made concerning the cooling towers at Montour:
1) Sulfuric acid addition is required to control
CaCOs scale when the cooling towers are operated
above 4 cycles of concentration.
2) Operation of the tower with zero blowdown and
design drift will cause the towers to operate
near 14 cycles of concentration.
3) Cooling tower fill corrosion can be minimized
if the towers are operated above 8 cycles of
concentration.
4) Increased magnesium concentrations decrease
the scale potential of the two important scale
forming species, CaCCb and CaSCK .
3.2 Ash Handling System
The existing operations simulations indicated that
some recycle of the ash pond liquor to the ash sluicing opera-
tion could be accomplished without CaSCK scaling. Four alterna-
tives using recirculating ash sluicing systems were studied.
The first three alternatives attained zero discharge and used
Na2C03 softening to control gypsum scale. The fourth used a
blowdown from the ash pond to control the gypsum scale. Differ-
ent makeup water qualities were used in the first three alter-
natives and the fourth alternative used the same makeup water
as Alternative 3.
1-29
-------
3.2.1 Simulation Basis
The process model used to simulate alternatives for
ash sluicing is identical to that used for existing operations
(Figure 2-3). Because of the non-reactivity of the bottom ash
it was exclusively sluiced with pond water. The more reactive
fly ash was sluiced with a mixture of pond water and makeup
water. Evaporation from the pond was calculated to be 1.3 Si/sec
(20 GPM) using the empirical method shown in Appendix K.
The ash flow rates and characteristics were the same
as the values used for existing operations simulations. All
simulations were performed at about 5 wt. 70 solids. In the
cases where softening was required it was assumed that a slip-
stream was taken from the pond recycle line and the calcium con-
centration was reduced to 50 mg/£ in this slipstream to account
for treatment inefficiencies. Actual equilibrium values ranged
from 22 to 29 mg/£.
Table 3-5 presents the input data that was used for
the alternative sluicing operations. Only that data which are
different from the data used for existing operations are tabu-
lated. The makeup water is 8 cycle cooling tower blowdown for
Alternative 1, 20 cycle cooling tower blowdown for Alternative
2, and river water for Alternatives 3 and 4. Alternative 4 has
a larger makeup water requirement because it has a blowdown
stream.
3.2.2 Recirculating Ash Systems
The simulation results for the recirculating ash
sluicing alternatives are shown in Table 3-6. Equilibrium with
respect to C02 between the atmosphere and the ash pond was
assumed. The effects of C02 transfer are discussed in Section
3.3.
The degree of recycle achievable in the ash sluicing
system will depend upon the CaSCK^HaO relative saturation in
the fly ash slurry liquor since gypsum scale is of greater con-
cern than that of CaC03 or Mg(OH)2. Gypsum scale is very diffi-
cult to remove from process vessels and equipment once it is
formed but CaC03 and Mg(OH)2 scale most likely can be dissolved
by acid washing. Both CaC03 and Mg(OH)2 solubilities are pH
dependent.
1-30
-------
TABLE 3-5. RECIRCULATING ASH SLUICING INPUT DATA
Makeup Water
Makeup Flow Ca4"1" Me4"1" Na+ Cl" COT NOT
Water Source I/sec GPM mg/J. mg/fc mg/A mg/fc mg/T mg/i
Alternative 1
8 Cycle
Cooling Tower
Blowdown 18 291 227. 43.9 64.5 176. 30.3 43.9
Alternative 2
20 Cycle
Cooling Tower 18 291 567. 110. 161. 440. 22.7 110.
Blowdown
Alternative 3
River Water 18 291 28.4 5.5 8.1 22. 6.0 5.5
Alternative 4
River Water 44 700 28.4 5.5 8.1 22. 6.0 5.5
SO^ the Fly Ash*
mg/l Sluice System
555. 89
1,430. 89
67.6 89
67.6 73
*89% recycle corresponds to zero discharge from the ash handling system. The remaining 11% of the sluice
water is lost from the system by evaporation and occlusion with the solids in the pond.
-------
TABLE 3-6. RECIRCULATING ASH SLUICING RESULTS
<-C02 EQUILIBRIUM IN THE POND)
Total System
Makeup Water
M
U>
ro
H/sec
Alternative 1 1,000
Alternative 2 950
Alternative 3 985
Alternative 4 1,035
CPU
16,000
15,000
15,600
16,400
Sluice Tank Effluent
Relative Saturations*
CaC03 CaSO.,-2H20 Hg(OH)2
6.69 1.06 0.02
5.06 1.07 0.02
7.52 1.06 0.02
29.13 0.98 0.16
CalciuB
I of Rate
pH recycle I/sec GFM noles/sec
9.07 30 87 1,380 1.056
9.01 36 103 1,624 1.366
9.15 27 77 1,220 0.876
9.76 0000
Slowdown
H/sec GPM
0 0
0 0
0 0
51 810
*Critical values, avove which scale potential exists, are 1.3-1,4 for CaSOt^HzO, about 2.5 for CaCO 3, and about 3.4 for Mg(OH)2 (see Appendix C).
-------
In the first three alternatives, softening was employed
in the recycle stream to reduce the calcium content of the recy-
cle. The rate of calcium removal required was determined by
the relative saturation of CaSO^ in the slurry line. In each
case the relative saturation was near 1.0, well below the criti-
cal scaling level of 1.3-1.4. In the fourth case the rate of
ash pond effluent was determined so that the relative saturation
of CaSO^ was slightly below 1.0.
The calcium was removed from a slipstream taken from
the recycle line. The calcium concentration in the slipstream
was assumed to reach 50 mg/£. Actual equilibrium concentrations
ranged from 22 to 29 mg/£. The percent of the recycle flow that
was treated was determined by the required calcium removal rate.
Sample calculations are presented in greater detail in Appendix K.
In all four cases the relative saturation of CaC03 in
the slurry was above the critical scaling value of 2.5. This
value was actually smallest for the worse quality water because
of the greater calcium removal rate required to control gypsum
scale. GaC03 scale formation in the slurry line may possibly be
reduced by installing a reaction tank prior to the sluice line.
Sizing this tank is critical to the successful operation of this
ash sluicing configuration. Additional data should be taken on
a pilot scale so that the reaction tanks may be accurately sized
before implementing this alternative. Pilot or bench scale tes-
ting to determine the level of acid washing that is sufficient
to prevent the line from plugging should be conducted before
these alternatives are implemented.
3.2.3 Effect of CaSOlt-2E20 Supersaturation in the Pond
Recycle Water
If the pond recycle water in a recirculating ash
sluice system proposed for Montour remains supersaturated with
respect to gypsum, scaling may occur in the fly ash .sluice line.
The degree of supersaturation in the pond recycle water cannot
be accurately quantified but will depend on the degree of tur-
bulence in the pond and on the residence time in the pond. The
greater the degree of mixing in the pond due to thermal gradi-
ents or wind turbulence, the more desupersaturated the liquor
will become. Longer residence times will also encourage gypsum
precipitation.
1-33
-------
However, the lack of CaSO^.ZHgO crystals in the pond
will discourage any precipitation and therefore, limit the de-
gree of desupersaturation. Since ponds are generally not very
well mixed, the pond will most likely remain supersaturated
with respect to gypsum as long as no chemical treatment is used,
and scaling may occur. Pilot or bench scale testing may provide
information to more accurately determine the degree of desuper-
saturation.
In the simulations presented in Section 3.2.2, the
pond water was allowed to remain supersaturated with respect to
CaSOjj. This presented a worse case operation with respect to
the amount of calcium that was required to be removed in order
to inhibit CaS04 scale formation in the slurry line. Thus, if
any precipitation does occur in the pond in actual operation,
the softening load would be less than predicted from the simu-
lations .
3.2.4 Effect of Carbon Dioxide Mass Transfer
Six additional cases were studied to determine the
effects on the operation of the ash sluicing system of carbon
dioxide mass transfer between the process liquor and the atmos-
phere. The results from these additional cases along with the
three base cases previously discussed are shown in Table 3-7.
The three base cases used were similar to the first
three alternatives presented in Table 3-6. In the base case
simulations the pond was assumed to be in equilibrium with the
atmosphere with respect to carbon dioxide but no transfer was
allowed to occur in the slurry tank. For each base case two
more simulations were performed where: (1) No C02 transfer in
the tank or pond was allowed, and (2) C02 equilibrium with the
atmosphere was attained in both the tank and the pond. For pur-
poses of comparison softening was not employed in any of these
simulations, as was done in the previous section.
From Table 3-7 it can be seen that the relative satur-
ation of CaSCK in the slurry line is not strongly influenced
by the degree of C02 transfer. In all three systems the rela-
tive saturation of CaSCK was slightly higher (less than 2%) for
the simulations where no C02 transfer was allowed. For all the
simulations the relative saturation of CaSO^ remained near
three.
1-34
-------
TABLE 3-7. THE EFFECT OF C02 TRANSFER IN RECIRCULATING
ASH SLUICING SYSTEMS AT MONTOUR
Makeup
Water*
A
A
A
B
B
B
C
C
C
CO Equilibrium
in Sluice Tank
No
Yes
No
No
Yes
No
No
Yes
No
CO Equilibrium
in Pond
Ho
Yes
Yes
No
Yes
Yes
No
Yes
Yes
Fly Ash
Slurry
Relative Saturations**
CaC03
6.31
3.20
17.77
4.84
3.02
16.24
2.21
2.76
17.65
CaSO^'OTjO
3.19
3.15
3.15
3.83
3.78
3.79
2.85
2.80
2.80
Mg(OH)2
2.47
1.1 x 10"*
0.02
2.72
1.3 x 10" "•
0.02
2.86
7.5 x 10"s
0.02
PH
10.10
7.90
8.98
10.04
7.90
8.92
10.22
7.88
9.06
Pond
PH
10.14
7.71
7.71
10.06
7.69
7.69
10.24
7 73
7.73
*Makeup Water Sources:
A: 8 cycle cooling tower blowdown
B: 20 cycle cooling tower blowdown
C: River Water
**Critical values, above which scale potential exists, are 1.3--1.4 for CaSON-2H20, about 2.5 for CaC03, and about
3.4 for Mg(OH); (see Appendix C)
-------
The pH of the slurry effluent is strongly affected
by the amount of C02 dissolved in the slurry water. In all
three systems the highest pH's (over 10) occurred when no C02
transfer was allowed. The lowest pH's (under 8) occurred when
C02 equilibrium was attained in both the pond and the slurry
tank.
The relative saturations of Mg(OH)2 and CaC03 are
very pH dependent. As expected, the relative saturation of
Mg(OH)2 varied directly with the pH in the slurry. The lowest
relative saturations of Mg(OH)2 (near 10"1|) occurred when C02
equilibrium in both the tank and the pond brought the pH's
below eight. The largest relative saturations (near 2.5)
occurred when the lack of C02 transfer in both the tank and
the pond allowed the pH to remain high.
The relative saturation of CaC03 is pH dependent
because of the effect pH has on the carbonate-bicarbonate equi-
librium. The variation in the relative saturation of CaC03 is
due to the variation in the concentration of C07, which depends
on both the pH and the total amount of carbonate species in the
water. Because C02 was free to enter the system in some of the
cases a simple relationship between pH and the relative satura-
tion of CaC03 was not seen. In all three systems the largest
relative saturation of CaC03 (above 16) occurred when equilibrium
between the atmosphere and the pond was allowed and no C02 trans-
fer took place in the slurry tank. This large relative satura-
tion occurred because the pond served as a source of C03 and
the pH in the tank did not reach as low a level as it did when
C02 transfer occurred in the tank. Therefore, the reduction in
pH caused by C02 transfer in the tank reduced the relative sat-
uration of CaC03 more than the increase in total amount of car-
bonate species increased it for the cases where C02 equilibrium
occurred in both the tank and the pond.
These results show that the best operating conditions,
with respect to scale formation, would exist when C02 equilibrium
was encouraged in both the tank and the pond. Although C02
transfer has not been quantified in this study, pilot scale
studies to determine the optimum ash sluicing recycle configur-
ation may provide data to allow a more accurate account of the
level of C02 transfer in actual operations.
1-36
-------
3.2.5 Arsenic Discharges
From the ash leaching studies discussed in Appendix D
information about the amount of arsenic in the ash was obtained.
Although Appendix D is concerned purely with calculations about
the input species required for the computer simulations, using
the same method, it was found that arsenic composed about
2.75 x 10 ** weight percent of the ash.
An estimate of the concentration of arsenic in the
pond liquor for the four alternatives discussed was made by per-
forming a mass balance around the ash sluicing system. It was
assumed that all of the arsenic was leached from the ash and
remained in solution. It was further assumed that all of the
arsenic entered the system with the ash and left in the sludge
water and pond discharge. In the first three alternatives the
arsenic concentration rises to 1.3 mg/K,. In the fourth alter-
native the arsenic concentration is about 0.6 mg/& because a
discharge of 25 £/sec (500 GPM) is taken from the pond for each
unit. The calculations required for the fourth alternative are
presented below.
As (in with ash)
As (in discharge) =
H20 (discharged) + H20 (sludge)
1000 mg/sec
(25.5 + 17.8) a/sec
= 0.55 tng/fc
Actual arsenic concentrations may be lower than these
values since equilibrium or rate relationships between the ar-
senic in the solid phase and arsenic in the liquid phase were
not considered. In the closed-loop bench-scale recirculating
studies, arsenic concentrations were found to be between 0.09 and
0.20 ppm, with an average value of 0.14 ppm. These values are
significantly less than the calculated values of 1.3 and 0.6 ppm.
3.2.6 Conclusions
From the results presented above the following conclu-
sions can be made about the ash sluicing system at Montour.
1-37
-------
1) Any recirculating ash sluicing
system at Montour will require
either softening or a blowdown
to control CaSOt, scale in the
slurry line.
2) The amount of softening required
is dependent on the degree to
which CaSO,, precipitates in the
pond.
3) The degree of C02 transfer in
the tank and the pond strongly
affects the relative saturation
of CaC03 and Mg(OH)2 but does
not significantly affect the
scale potential.
3. 3 Summary
From the results of the cooling tower and ash sluice
system simulations discussed in the previous sections, four al-
ternatives for reducing plant discharges are considered techni-
cally feasible. These are:
1) Cooling tower operation at 8
cycles with 24 5,/sec (380 GPM)
blowdown and the design drift
rate of 31 a/sec (500 GPM) and
recirculating ash sluicing
using cooling tower blowdown
as makeup. This alternative
requires acid treatment in the
towers and slipstream softening
for calcium removal from the ash
pond recycle. (Zero discharge)
2) Cooling tower operation at 20
cycles with 20 £/sec (325 GPM)
blowdown and no drift, and re-
circulating ash sluicing, using
cooling tower blowdown as makeup.
This alternative also requires
acid addition in the towers and
softening in the ash system.
(Zero discharge)
1-38
-------
3) Cooling tower operation at 20
cycles with no blowdown and a
drift rate of 20 £/sec (325 GPM)
and recirculating ash sluicing,
using river water as makeup.
Again, acid addition in the
towers and softening in the ash
system are required. (Zero dis-
charge)
4) Cooling tower operation at 20
cycles with no blowdown and a
drift rate of 20 £/sec (325 GPM)
and recirculating ash sluicing
using river water as makeup.
This alternative only requires
acid treatment in the towers.
[Discharge - 51 H/sec (810 GPM)]
All of these alternatives will require the addition
of sulfuric acid in the cooling towers for CaC03 scale control
and reaction tanks prior to the fly ash sluice line to minimize
CaC03 and Mg(OH)2 scale formation in the line. Adjustment of
the pH of the ash pond overflow may be required in Alternative 4,
depending on the amount of carbon dioxide mass transfer occurring
in the pond. (Section 3.2.4) The first three alternatives
employ soda-ash softening of a portion of the ash pond recycle.
This treatment is necessary to prevent CaS04 scale potential in
the sluice line. The fourth alternative controls the gypsum
scale potential without softening by employing a small blowdown
of 51 £/sec (810 GPM) from the pond.
The first two alternatives differ only in the assump-
tion concerning the drift rate in the cooling towers. The oper-
ating conditions are essentially the same. Therefore, these two
are not really different alternatives but represent the extremes
in water quality that might be seen from a cooling tower opera-
ting with a blowdown of about 22 SL/aec (350 GPM). The quality
of this blowdown stream has significant effects on the ash
sluicing system since it serves as makeup to the system.
Alternatives 3 and 4 assume that the drift rate would
not fall below 65% of the design value. Even with this small
drift rate the cooling towers could be operated at zero discharge
using sulfuric acid to prevent CaCOs scale. The ash sluicing
makeup water can be supplied directly from the river or pond
1-39
-------
since there will be no cooling tower blowdown. This higher
quality water allows for less treatment of the ash recycle than
would be needed in Alternatives 1 and 2. Alternative 4 repre-
sents the smallest discharge that could be expected if soften-
ing were not used.
It should be emphasized here that none of the alter-
natives should be implemented before more information is gath-
ered from a bench or pilot scale test program to determine
1) the actual size of reaction tank required in the sluice
system, 2) the quantity and frequency of acid wash water re-
quired to minimize CaCOs and Mg(OH>2 scale formation, and 3) the
true drift rate in the cooling towers.
An economic analysis based on rough cost estimates
for these alternatives is presented in the next section.
1-40
-------
4.0 ECONOMICS
This section provides rough cost estimations for
implementing each of the technically feasible alternatives dis-
cussed in Section 3.0. Both rough capital costs and operating
costs are presented. The assumptions used in calculating these
costs are briefly outlined. It is emphasized that these values
are only rough estimates for comparative purposes.
The capital cost summary for the four technically
feasible alternatives is presented in Table 4-1. All the al-
ternatives involve sluicing the fly and bottom ash at about
5 wt. 70 solids and require identical tanks and agitators.
Since the flows do not differ significantly among alternatives
the pumping and piping capital costs are the same. The capital
expenditures for softening vary directly with the size of the
treated stream.
The costs reported for the fly ash slurry tanks in-
clude two 39,000 gallon, carbon steel, neoprene-lined tanks,
with mixer support structures and baffles. These tanks were
sized to give a fifteen-minute residence time for 2580 GPM of
fly ash slurry from each unit. The costs are for nominal foun-
dations and plumbing. The terrain and soil characteristics may
require special site preparation which will add to tank instal-
lation costs, and the costs for interconnecting plumbing and
piping will also be a function of the particular site. Field
erection rates are based on national average rates but can vary
widely with the specific location and labor pool used.
The costs for the agitators are based on two 10 hp,
electrically-driven, neoprene-coated agitators. The costs for
both the tanks and the agitators were obtained from (GU-075) in
1970 dollars and upgraded to 1976 using a factor of 1.56 (based
on Chemical Engineering Index).
The first three alternatives require approximately
10,000 GPM of the pond water to be recycled to both units.
The fourth alternative requires about 9,000 GPM of the pond
water for recycle. The costs reported for piping include the
cost of 1,400 feet of 24-inch carbon steel pipe with average
fittings, flanges, shop coating, and wrapping. The values were
estimated from (GU-075) in 1970 dollars and upgraded to 1976
using a factor of 1.56.
1-41
-------
TABLE 4-1. CAPITAL COSTS1 FOR WATER RECYCLE/REUSE ALTERNATIVES AT MONTOUR
Fly Ash Slurry Tanks2
Agitators
Pond Overflow Recycle
Pumps
Pond Overflow Recycle
Piping
V Additional Fly Ash
N5
Slurry Pumps
Sodium Carbonate3
Softening
Contingency (20%)
Contractual Fees (3%)
TOTAL
Alternative 1
100,000
14,700
93,400
144,000
42,000
126,500
104,000
15,600
640,200
Alternative 2
100,000
14,700
93,400
144,000
42,000
148,900
108,600
16,300
667,900
Alternative 3
100,000
14,700
93,400
144,000
42,000
111,900
101,200
15,200
622,400
Alternative 4
100,000
14,700
93,400
144,000
42,000
0
78,800
11,800
484,700
'1976 dollars (rough cost estimates for comparative purposes)
2 includes wear liner and agitator supports
3$91.7/GPM
-------
Pump costs were estimated based on cast steel pumps
with electric motor drivers. A labor-to-material ratio of
0.36 was used for installation costs. Engineering was assumed
to be 107o of the combined labor and material cost (GU-075) .
The softening costs were estimated from (NE-107).
A cost of $91.7/GPM of installed capacity was used. The cost
for the four alternatives vary because of the variation in the
slipstream rate required for softening.
The operating cost summary for the four technically
feasible alternatives is presented in Table 4-2. The operating
cost vary directly with the capital costs. Thus, the alterna-
tive requiring the most amount of capital also requires the
largest operational budget.
The power consumption is the same for all but the
fourth alternative. This is because the recycle is about
9,000 GPM in Alternative 4 as opposed to about 10,000 GPM in
the others. The costs are based on a wholesale price of 2£/
kW-hr to the utility.
The cooling tower treatment costs vary directly with
the cycles of concentration of the recirculating flow in the
towers. In Alternative 1, the cooling towers are operated at
8 cycles of concentration compared to the other alternatives
where the cycles of concentration is increased to 20. The
treatment costs only include the cost of sulfuric acid at
$60/ton.
The softening costs are the dominant operating cost
other than capital charges. Alternative 4 costs significantly
less than the others because softening is not employed and
capital charges are smaller. The softening costs include only
the cost of chemicals at $69 per million gallons treated. The
capital charges were estimated as 1570 (MC-136) of the capital
investment shown in Table 4-1 for each alternative based on a
30-year lifetime.
The costs presented in this section are merely rough
estimates. They are presented here in order to compare the
relative cost of each alternative. They do not include any
savings that might occur because of reduced makeup water re-
quirements .
1-43
-------
TABLE 4-2. OPERATING COSTS1
Alternative 1 Alternative 2 Alternative 3
Cooling Tower2
Acid Treatment 2,800 5,600 5,600
Power Consumption3
Agitators
Recycle Pumps 34,200 34,200 34,200
M
1
£ Softening Chemicals" 40,000 47,100 35,400
Capital Charges5 96,000 100f200 93.400
TOTAL 173,000 187,100 168,600
(mils/kW-hr) (.016) (.018) (.016)
Alternative 4
5,600
24,600
0
72,700
162,900
(.010)
!1976 dollars per year based on 80% load factor
2$60/ton for sulfuric acid
32c/kW-hr
4$69/106 gal (chemicals cost only)
5157o per year based on 30-year lifetime
-------
Appendix J. Recycle/Reuse Options at Colstrip (Montana Power Company)
1.0 INTRODUCTION
This appendix describes the analysis of the water
system at Montana Power Co.'s (MFC) Colstrip plant, under EPA
Contract No. 68-03-2339, Water Recycle/Reuse Alternatives in
Coal-Fired Power Plants. A summary of the important results
is presented in this section. Colstrip was chosen along with
four other plants for evaluation of technical and economic
feasibility of various water recycle/reuse options.
Three major topics are discussed in this appendix:
1) Existing Operations Modeling
2) Alternatives Modeling
3) Economics
The major water systems at the two-unit, 700 Mw Col-
strip plant are the cooling tower and combined S02/particulate
scrubbing systems. Colstrip is designed for and is achieving
zero discharge through brine concentration of the cooling tower
blowdown and a disposal pond for the scrubber sludge.
The results of the existing operations simulations
of the cooling towers compare well to the sample data obtained
at the plant with respect to CaC03 and CaSOi»*2H20 scale poten-
tial. The simulations performed indicate that the towers are
presently operating at about the maximum cycles of concentra-
tion obtainable without approaching gypsum scale potential.
The results of the existing operations simulation of
the scrubbing system compare well to the sample results for the
scrubber recycle stream with respect to magnesium, sodium, and
chloride. Calcium and sulfate concentrations were lower in
the sampled stream, most probably due to reduced load and/or
reduced SOa content in the gas from design conditions, resul-
ting in a lower mass removal rate of sulfur from the gas.
Process alternatives at Colstrip were studied on a
modular basis with the cooling system being one module and the
scrubbing system being the other. A total of six cooling tower
simulations were performed to compare treatment alternatives
and determine the effects of calcium and sulfate concentrations
J-l
-------
in the makeup water on slipstream treatment rate. The first
set of simulations comparing treatment alternatives shows that
with the use of slipstream softening, as opposed to the present
pretreatment system, cycles of concentration may be increased
from 13.5 to 20. The towers may only be operated at a maximum
of about 11 cycles of concentration without any softening based
on the makeup water sampled.
The additional simulations showed that calcium had a
greater effect on slipstream treatment rate than sulfate in the
makeup water. Doubling the calcium concentration increased the
treatment rate about 37070 whereas doubling the sulfate concen-
tration only increased the treatment rate about 37%.
No scale potential was found for any phosphate or
silica solids even at 20 cycles of concentration, which is the
highest value considered.
A total of four additional scrubbing simulations were
performed to examine the effects of flue gas ash content, slur-
ry percent solids, and makeup water source on the scrubber oper-
ation at Colstrip. As the ash rate into the scrubbers increases,
and as the percent solids in the circulating liquor decreases,
CaSOn*2H20 scale potential is increased. Therefore, burning a
coal of significantly higher ash content or operating the system
at low solids in the recycle loop could cause scaling problems
because of the small percentage of calcium sulfate seed crystals
present in the slurry.
The use of either untreated river water or cooling
tower blowdown as makeup (excluding demister wash) to the scrub-
bing system does not have a significant impact on the scaling
tendency of the system. The use of cooling tower blowdown in-
creases the total dissolved solids and chloride levels, but these
levels should not be high enough to cause corrosion problems.
Table 1-1 presents a summary of the two combined sys-
tem alternatives for the Colstrip water system as compared to
existing operations and the relative costs of each alternative.
All flows reported in Table 1-1 refer to those produced from
both units. It should be noted here that this analysis was per-
formed to study general water recycle/reuse alternatives. Actual
implementation of any of these alternatives would require a more
extensive investigation of process parameter variability. More
data on makeup water quality, scrubber variations and seasonal
flow variations would be required before a detailed design could
be made.
J-2
-------
TABLE 1-1. SUMMARY OF WATER RECYCLE/REUSE OPTIONS AT COLSTRIP
Existing
Conditions
Alternative
One
Alternative
Two
C-i
i
Cooling Tower Makeup
Source
Cycles of Concentration
in Cooling Towers
Cooling System Treatment
Treatment Rate,
H/sec (GPM)
Cooling Tower Slowdown
Rate, 2,/sec (GPM)
Scrubber Makeup Source
Plant Makeup Rate
«7sec (GPM)
Plant Discharge Rate
Jl/sec (GPM)
Costs: l
Capital, 1976 $
Operating, 1976 $/yr. 2
(mils/kwh)
Softened River Water
13.5
Makeup Softening
423 (6710)
23.6 (376)
Softened River Water,
Brine Concentrator
Distillate
423 (6710)
0.
Softened River Water
13.5
Makeup Softening
397 (6300)
23.6 (376)
Untreated River Water
20
Slip-stream softening
18 (284)
14.6 (230)
Cooling Tower Blowdown, Cooling Tower Slowdown,
Untreated River Water Untreated River Water
423 (6710)
0.
159,000
-237,000
(-.046)
423 (6710)
0.
275,000
-217,000
(-.044)
1These rough cost estimates were made to compare technically feasible options and do not include
a "difficulty to retrofit" factor.
2Includes capital cost amortization at 15% per year.
-------
The first alternative does not involve any changes
in operation of the cooling towers but uses cooling tower blow-
down and untreated river water as scrubber makeup as opposed to
softened river water and brine concentrator distillate as is
presently done. A capital cost of $159,000 is reported for
piping modifications and new pumps. However, a net operating
savings is shown due to a large savings in brine concentrator
operation because of the reduced flow. Only enough cooling
tower blowdown is sent to the brine concentrator to provide
the boiler makeup requirements (only one brine concentrator
needed).
Alternative 2 includes using slipstream treatment in
the cooling tower system in addition to the system changes of
Alternative 1. The towers are operated at 20 cycles of concen-
tration resulting in decreased blowdown. Again only enough
cooling tower blowdown to provide the boiler makeup is sent to
the brine concentrator. A higher capital cost is reported due
to the conversion to slipstream treatment in the cooling system.
The increased capital charges result in a lower operating ex-
pense savings for this alternative. The savings in brine con-
centrator operating costs represents the major savings of both
of these alternatives.
Although the Colstrip plant is achieving zero discharge,
more effective cascading of the water streams in the plant may
be achieved which results in a decrease in operating costs from
the existing level. The capital and operating costs reported
in this appendix do not include any savings which could have
been realized if the Colstrip water system had been designed
for the most effective cascading of aqueous streams. A savings
in capital investment could have been achieved by designing the
cooling towers for slipstream treatment (was not considered re-
liable enough at the time of Colstrip design) and by using only
one 150 GPM capacity brine concentrator as opposed to the two
200 GPM capacity units presently used. The capital savings
associated with purchasing one 150 GPM brine concentrator versus
two 200 GPM units totals about $1.9 million based on $7 750/GPM
(LE-239). '
J-4
-------
2.0 PLANT CHARACTERISTICS
The Montana Power Co. (MPC) Colstrip Plant is located
in Colstrip, Montana, and consists of two 350 Mw coal-fired
units. The coal burned at Colstrip is taken from a mine located
adjacent to the plant and contains from .2 to 170 sulfur and 6.1
to 12.6% ash (average values reported by MPC are .77% and 8.59%,
respectively). The heating value of the coal is approximately
8500 Btu/lb as received. Colstrip is designed to achieve zero
discharge with cooling towers, a recirculating bottom ash sluice
system, and combined particulate and S02 scrubbing with disposal
ponds.
This section of the appendix describes the character-
ization of Colstrip's water system. First, an overall plant
water balance is presented which shows the major in-plant flows
and chemical analyses for the streams which were sampled. Then
a detailed discussion of each of the major water consumers is
presented, including a description of the process model and the
input data used to simulate design conditions at Colstrip. The
computer simulation results are compared to the chemical analyses
of the spot samples. Areas of Colstrip's water system which
show a potential for increased water recycle/reuse are also
identified and discussed.
2.1 Water Balance
A flow schematic for the Colstrip water system is
shown in Figure 2-1. The major water consumers which include
the cooling tower and scrubbing systems are shown in this dia-
gram. This figure presents design flows and results of the
chemical analyses of the samples taken at Colstrip. A more
detailed description of the samples taken and analytical pro-
cedures used is presented in Appendix B. As shown in Figure 2-1,
makeup water for the plant is taken from the Yellowstone River
and stored in a surge pond. The water taken from the pond is
treated with cold lime softening for calcium removal. The lime-
treated water is then used as makeup water to the cooling tower
and scrubbing systems. Softening wastes are sent to the scrub-
ber ponds.
Cooling tower blowdown is piped to two 12.6 I/sec
(200 GPM) capacity brine concentrators. The distillate provides
the demineralizer feed (design rate of about 90 GPM, actual rate
of 100-200 GPM). Excess distillate is combined with lime-soft-
ened water and used as scrubber makeup. The concentrated waste
from the brine concentrators is disposed of in two one-acre lined
ponds.
J-5
-------
C i. ARTfCA TrOfit
o
L/Mf
sof rev/wo
POMO
X
X
X
X
V r TANK
Figure 2-1. Montana Power Company Colstrip Plant water balance.
-------
C-4
01
A V-
t_ .
F —•—A-
H (
i V-
i--
^ KM ret
'TO JCHUMMS
OOHLT*.
ivvtfw nr>«r
SFt:rcic
-------
Streaa Number
Strean Same
flo.: »MetrlC
English
PH
Calciua
Magnesium
Sodiim
Potassium
Chloride
Carbonate (as C03)
Sulfate (aa SOH)
Sulflte (a* SO]}
/ U/l \
Phogphate (aa PO*)
Sllicatea (as BIO,)
Suspended Solids
Dlaaolved Solids
Plane
Makeup
Water
423 t/aec.
6.710 gpm
6.7
57.9
19.5
53.5
4.2
22
17.3
174
1.8
0.002
440
O
Softened
Makeup
Uater
423 H/sec.
6,710 iff
10.3
39.9
10.7
53.1
4.2
17
6.0
IBS
'
1.3
0.0016
360
<3>
Cooling
Tower
25. 2 t/>ec.
400 gpm
6.7
533
193
710
50.3
266
34.8
3,820
0.26
5.0
0.0014
6,000
«>
Bottom
Ash
Sluice
Water
38. 2 I/sec.
605 gpm
10.4
72!
70
295
13.1
79
7.2
2,780
1.4
0.0048
4,200
Scrubber
Recycle
Slurry*
963 I/sec.
15.260 gp»
3.9
504
5,050
458
21.9
129
52.2
19,400
300
31
7.7
29,200
Hash Tray
Recycle
Tank Liquor
3.4
519
2,925
153
11.5
67
25.2
10,600
1,560
22
0.88
16,300
Pond
Recycle
105 I/ sec.
1,665 gpm
5.5
484
1,550
305
13.1
70
9.6
9.000
400
0.01
24
0.0056
13.690
<£>
Effluent
Tank
Slurry"
19.9 t/sec.
315 gpa
6.4
497
2,075
315
15.5
74
31.2
11,600
100
0.028
25
1.36
17,200
<$>
Pond
Liquor
4.8
464
1,600
345
13.1
74
9.0
9,521
21.4
0.01
14,400
<5>
Boiler
Makeup
5.6 t/aec.
08 gp.
«x
Brine
Concentrator
Uater To
Scrubbers
19.7 I/sec.
312 gpm
^X
Line
Treated
Water To
Scrubbers
25.5 i/sec.
404 gpm
10.3
39.9
10.7
53.1
4.2
17
6.0
188
1.3
0.0016
360
<£>
Flue
Gas*
810,000 m'/hr
477.000 acfm
*Flowa reported are for each scrubber module.
**Flows are reported aa design flow under full load operation for both i
except where noted.
Figure 2-1. (Continued)
-------
The portion of Figure 2-1 encircled by the dotted line
represents one scrubber train. There are three identical trains
on each of the two generating units. The scrubbing system makeup
water is added along with lime to the recycle tank in each train.
The dust-laden, S02-rich flue gas enters the scrubber venturi sec-
tion at the top of each train and flows down cocurrently with the
scrubber recycle liquor. The gas then is channeled through a 180°
bend and flows upward through the spray section for S02 removal.
The scrubbing liquor which is sprayed countercurrently to the gas
stream falls into the recycle tank and the clean gas passes
through a mist eliminator section and exits at the top of the
scrubber. The exit gas then passes through a steam reheat section
and an induced draft fan before being vented through the stack.
Mist eliminators are washed by a separate recircula-
ting stream. The wash water is collected by a wash tray and
recycled through a wash tray recycle tank. There is one wash
tray recycle tank for every three modules. A portion of the
wash water is pumped to the wash tray pond for solids settling.
Clear liquor is returned to the spray headers. Lime-softened
makeup water and brine concentrator distillate are added to re-
place water lost through evaporation and occlusion with the solids
A bleed stream is taken from the scrubber recycle tank,
diluted to about 67o solids with pond recycle liquor in the efflu-
ent tank, and pumped to the pond system. At the present time,
scrubber solids are dredged from Pond A and slurried to a dis-
posal pond. Excess pond liquor is recycled to the scrubbers.
There is one effluent tank for every three modules.
Bottom ash at Colstrip is sluiced to the bottom ash
pond in a recirculating system. Clear liquor from the bottom
ash pond clear well is used as sluice water. Makeup to this
system results from plant drainage water which flows into the
bottom ash pond. A blowdown from the bottom ash sluicing sys-
tem results from bottom ash pond overflow into Pond B.
There are no aqueous discharges from the Colstrip
Plant. Water losses occur through cooling tower evaporation
and drift, scrubber evaporation, pond evaporation, solids oc-
clusion, and boiler losses.
The first step in characterizing the chemistry of the
Colstrip water system is to examine the results of the spot
samples taken. The measured species concentrations were input
to the equilibrium program and several parameters were calculated
J-7
-------
which determine the tendency of the liquor sampled to form
chemical scale and to absorb or desorb C02 from the atmosphere
Another parameter calculated checks the internal consistency
of the sample and is a measure of the analytical accuracy.
Table 2-1 presents a summary of the parameters calcu
lated by the equilibrium program for each of the samples taken
at Colstrip. Relative saturations for CaCO 3 , Mg(OH)2, and
CaS04-2H20 are given in the first three columns. These para-
meters indicate the tendency of the stream to form scale. Cri
tical values for relative saturation of each species, above
which scale formation is likely, are 2.5 for CaCO 3 , 3.4 for
Mg(OH)2, and 1.3-1.4 for CaSCK^HaO. (See Appendix C)
Only one stream, the bottom ash sluice water, showed
a tendency to form CaCO 3 scale (relative saturation of 4.55).
The erosive character of an ash slurry stream may prevent exces-
sive buildup of the CaCO 3 scale, since no plugging problems
have been encountered. The bottom ash sluice water also showed
a relative saturation for CaSOt|'2H20 above the critical range
of 1.3-1.4. Gypsum formation has been noted in the bottom of
the boiler where the bottom ash sluice water contacts the hot
ash.
The pond recycle, effluent tank, and pond liquor sam-
ples all showed CaSOil-2H20 relative saturations in the critical
range. Although no scaling has been reported, operation of the
scrubbers in this range may result in some gypsum scale formation
over a long period of time. None of the streams sampled showed
a tendency to form Mg(OH)2 scale (the highest relative satura-
tion calculated was .264 in the bottom ash sluice water).
Equilibrium partial pressures of C02 above the liquor
sampled were calculated by the equilibrium program and show the
tendency of a stream to absorb or desorb C02 when in contact with
the atmosphere. A value less than 3 x 10""* atm. indicates a
tendency to absorb C02 and a value greater indicates a tendency
to desorb COa. The value for the cooling tower blowdown sample
is higher than 3 x lO"1* indicating that complete C02 equilibrium
is not achieved in the cooling towers.
Percent residual electroneutrality is a parameter cal-
culated to determine the internal consistency of each sample with
pH specified. A value of +15% is considered acceptable. A more
detailed explanation of how this parameter is calculated along
with a description of the equilibrium program is presented in
Appendix E (p. E-41).
J-8
-------
TABLE 2-1. PARAMETERS CALCULATED BY EQUILIBRIUM PROGRAM FOR COLSTRIP SAMPLES
Stream Name
Surge Pond
Softened Makeup
Water
Cooling Tower
Slowdown
Bottom Ash Sluice
Water
Scrubber Recycle
Slurry
Wash Tray Recycle
Slurry
Pond Recycle
Effluent Tank
Slurry
Pond Liquor
CaC03
.0026
1.08
.051
4.55
1.87xlO"5
1.2xlO~6
2.2xlO~5
1.5xlO~6
1.5xlO~6
Relative
Mg(OH)2
7.8xlO~9
.027
5.9xlO~7
.264
7. 2xlO~ 1
5.2X10"1
2.7X10'1
2.0xlO~l
3-lxlO"1
Saturations*
CaSOH-2H20
.041
.034
1.11
1.67
0 1.00
1 .82
0 1.38
1 1.31
1 1.30
Equilibrium partial
pressure of C02 , atm
1.91xlO~3
1. 00x10" 7
4.28xlO~3
2.59xlO~8
.511
.240
2.40xlO~3
.0127
2.96xlO~3
% Residual
Elect roneutrality
+23.8
+ 7.3
-10.5
- 7.9
+11.0
+21.2
-10.6
-13.9
-13.7
*Critical values, above which scale potential exists, are 1.3-1.4 for CaSCK • 2HaC>, about 2.5 for CaCOa ,
and about 3.4 for Mg(OH)a (see Appendix C)
-------
2.2 Cooling Tower System
Each of the two 350 Mw units at Colstrip has an
independent cooling system with one cooling tower per unit.
Each tower has seven fans with a combined capacity of about
1.7 x 107 m3/hr (1 x 107 ACFM). Water circulates at a design
rate of 6510 H/sec (103,200 GPM) between the condenser and
cooling tower for each unit. A blowdown stream is removed
from the circulating water after the condenser. Makeup water
is added to replace the water lost in the blowdown and through
evaporation and drift in the tower.
Presently, blowdown is removed from the cooling sys-
tem at a rate which sets the cycles of concentration between 10
and 15, depending on the makeup water quality. Cycles of con-
centration may be defined as the ratio of blowdown species con-
centrations to makeup species concentrations. In terms of flow
rates, cycles of concentration is:
r _ E + B + D
L B + D
where C = cycles of concentration
E = evaporation rate
B = blowdown rate
D = drift rate
This equation shows that as the blowdown decreases, the cycles
of concentration increases, assuming that evaporation and drift
remain constant.
The following sections present the model used to
simulate the Colstrip cooling tower system and the results of
existing operations simulations. A more detailed description
of the tower operating parameters is also presented in the
following sections.
2.2.1 Simulation Basis
Existing operations simulations were performed for
the Colstrip cooling tower system to verify the validity of the
simulation model in predicting scaling tendencies in the tower
and to determine any potential for increased recycle/reuse
J-10
-------
This section first briefly discusses the model followed by a
description of the operating parameters used as inputs to the
model. A detailed description of the process model is included
in Appendix E.
The process simulation flow scheme shown in Figure 2-2
was used to model cooling tower operations at Colstrip. This
is a generalized cooling tower model with capabilities of sim-
ulating sulfuric acid addition and slipstream softening for
calcium removal. Only acid addition was used for existing op-
erations .
Given the inputs of air flow, temperature and compo-
sition, makeup water composition, flow and temperatures of the
circulating water, drift rate, and cycles of concentration, the
model performs iterative calculations around the cooling loop
to determine the blowdown, evaporation and makeup rates and
compositions for all water streams. An acid addition rate is
determined to keep the CaCO3 relative saturation within a spec-
ified range. If slipstream softening is required (determined
by model) the slipstream and chemical addition rates are cal-
culated.
Several assumptions are inherent in performing this
simulation with the subroutines shown in Figure 2-2. These
assumptions are given below:
1) Equilibrium exists between COa and tUO in the
atmosphere and cooling tower exit water.
2) The temperature of the cooled water stream
approaches the wet bulb temperature of ambient
air within a predictable range.
3) The compositions and temperatures of the cooled
water and drift streams are equal.
4) Ionic reactions taking place in the liquid
phase are rapid and thus at equilibrium.
The assumption involving the temperature of the
cooled water stream is a recognized design parameter in cooling
tower evaluation and gives a good approximation. The assump-
tion concerning the temperature and composition of the drift
stream should be very close to actuality as is the assumption
in regard to H20 gas-liquid equilibrium. The assumption with
J-ll
-------
CTGES
(RECYCLE
GUESS)
2
WTRINP
(SULFUFUC
ACID)
HLDTK3
(SUMMER)
I
I-1
i-o
ALKINP
[(LIME OR
SODA ASHI
5
10
CHEMICAL
WASTE
ORDER OF PROCESS CALCULATIONS:
1. 2, 3. 4, 5. 6. 7. (12. 13, 8. 9. 10. 9. 11. 6. 7). 8
6 ^. OUTLET AIR
DRIFT
L-X
I 5
-»•
15
11
CHMTRT
(SOFTENER)
10
•« 13
^m —
14
DIVOR6
(TEE)
9
16
OIVDR5
(TEE)
8
9
n_ 11
CLRHTR
(CONDENSER!
13
AIR
SLOWDOWN
MAKEUP WATER
ACID
SOFTENING CHEMICALS
Figure 2-2. Colstrip cooling tower simulation flow scheme
OUTLET AIR
DRIFT
SLOWDOWN
CHEMICAL WASTE
02-1270-1
-------
regard to CO2 equilibrium is conservative since the partial
pressure of CO 2 in actual cooling towers tends to be greater
than the equilibrium value. The lower equilibrium concentra-
tion of carbonate species, assumed in the model, causes the
pH to be slightly higher in the model than in actual opera-
tion. The higher pH causes the relative saturation of CaCO3
to increase more than the lowered carbonate species concentra-
tion causes it to decrease.
The data used as input to this model is presented in
Table 2-2. Some of this information was obtained directly
from MFC. Other inputs were calculated from MFC data, local
meteorological data, and sample analyses. Air, drift, and
circulating water flow rates were obtained from MFC as were
the approach, condenser AT, and condenser outlet temperatures.
Ambient air wet and dry bulb temperatures were obtained from
local climatological data for Billings, Montana, for 1976.
The water makeup composition was obtained from the spot sam-
ple taken at the Colstrip plant.
2.2.2 Simulation Results
This section describes the results from the simula-
tion of existing cooling tower operations at Colstrip. One
simulation was performed for summer operation and one for win-
ter operation. Table 2-3 presents the results of these two
simulations, along with the plant data concerning actual oper-
ation at the time the spot samples were taken.
The blowdown flow for summer operation compares well
to the value reported by MFC for summertime operation (12.7
vs. 11.8 £/sec). The blowdown pH's for the simulations are
slightly higher than the sample value. This is most probably
due to higher sulfate in the sample through excess acid addi-
tion at the plant. Also, higher sodium and magnesium values
in the sample will allow increased complexing between sulfate
and these cations. The higher cation solution can therefore
tolerate higher sulfate levels while maintaining a relatively
constant gypsum relative saturation. (Simulation value of
1.01 versus sample value of 1.11.)
Comparison of the blowdown compositions shown in
Table 2-3 indicates that the system may not have been at steady
state during sampling. For example, although the calcium con-
centrations agree very well for all cases, the magnesium and
carbonate concentrations are higher in the sample and nitrate
J-13
-------
TABLE 2-2. INPUT DATA FOR COLSTRIP COOLING TOWER SIMULATIONS
Winter
Summer
FLOWS
Air, m3/hr
(ACFM)
Drift, £/sec
(GPM)
Circulating water, Jl/sec
(GPM)
1.7xl07
(l.OxlO7)
1.3
(20)
6,510
(103,200)
1.7xl07
(l.OxlO7)
1.3
(20)
6,510
(103,200)
TEMPERATURES
Ambient Air, °C
Approach, °C
Condenser AT, °C
Wet Bulb, °C
Condenser Outlet, °C
-1.1
(30)
11.9
(21.5)
17.6
(31.7)
-4.4
(24)
25
(77)
20.6
(69)
11.9
(21.5)
17.6
(31.7)
12.8
(55)
42.2
(108)
ADDITIONAL DATA
Relative Humidity, %
Cycles of Concentration
Makeup Water Composition mg/&
Calcium
Magnesium
Sodium
Chloride
Carbonate (as COs")
Sulfate (as S0i»~)
Nitrate (as N03~)
42
13.5
39.9
10.7
40.3
17.0
6.0
188.0
1.4
40
13.5
39.9
10.7
40.3
17.0
6.0
188.0
1.4
J-14
-------
TABLE 2-3. COLSTRIP EXISTING COOLING TOWER OPERATIONS SIMULATION RESULTS
Cn
Simulations (13.5 cycles of concentration)
Cooling tower blowdown
Flow, gpm per tower
(£/sec)
PH
Composition, mg/£
Calcium
Magnesium
Sodium
Chloride
Sulfate (as S0i,~)
Carbonate (as C03 )
Nitrate (as N03~)
Relative Saturations**
CaC03
CaS04-2H20
Partial Pressure C02, atm
Plant data
202.*
(12.7)
6.7
533
193
710
266
3820
34.8
11.2
.051
1.11
4.3 x 10~3
Case 1
Winter Operation
138.
(8.7)
7.2
535.8
143.7
541.6
228.3
2652.
10.1
18.8
.06
1.01
4.5 x 10~"
Case 2
Summer Operation
187.
(11.8)
7.2
533.7
143.1
539.5
227.4
2644.
6.5
18.7
.097
.93
3.8 x Kf"
* Based on 13.5 cycles and makeup rate from MFC.
** Critical values, above which scale potential exists, are 1.3-1.4 for CaSO-»'2H20 and
about 2.5 for CaCOs (see Appendix C)
-------
is lower in the sample. These discrepancies may also be due
to nonhomogeneous sampling and/or analytical errors as well as
non-steady-state operation.
The CaC03 and CaSO.»'2H20 relative saturations agree
well between the sample and simulations. The values for CaC03
(.05-.097) indicate that the towers are operating in a very
safe mode since the critical value for scale formation is about
2.5. This, along with the higher sample sulfate concentration
and lower sample pH, supports the possibility of excess acid
addition in the towers. Normal pH in the cooling system ranges
from 7.8-8.0. The relative saturations for CaS0^2H20 (.93-1.11)
indicate that the towers are presently running at about the max-
imum cycles of concentration with respect to gypsum scale since
the critical range for scale formation is 1.3-1.4. However,
increased cycles may be obtainable if slipstream treatment is
used instead of pretreatment. This allows more effective soft-
ening since the circulating water is more concentrated than
the makeup water.
2.3 Scrubbing System
Each of the two 350 Mw units has three parallel scrub-
bing trains for removal of particulates and S02 from the flue
gas. The basic flow scheme was described and sample analyses
were presented in Section 2.1. This section of the appendix
presents an analysis of the design scrubber operating conditions
based on the sample analyses and operating data for the Colstrip
plant. First, the simulation basis is presented, including a
brief model description and a discussion of the input data used
to simulate design conditions at Colstrip. Then the results of
the simulations are compared to the sample results.
2.3.1 Simulation Basis
A process simulation of the Colstrip scrubbing system
operating at design conditions was performed to characterize the
system and to determine if a potential for water recycle/reuse
exists with the present configuration. This section first
briefly discusses the model, followed by a description of the
operating parameters used as inputs to the model. A detailed
discussion of the process model is included in Appendix E.
The process simulation flow scheme shown in Figure
2-3 was used to model the scrubbing system at Colstrip (see
Figure 2-1 for process flow diagram)'. This model calculates
J-16
-------
2 CLRHTR PMPFAN
~~ 14 15^ 15
/WTRMKP\
V
kFLUE GASV
SUMMER 17 ^
16
8
14
DIVDR3
SCRUBS
10
(SCRUBBER)
4
~~l
RATHLD
1 1
(HOLD TANK)
4 l» r-
" 1
I
DIVDER
9
r
RATHLD
13
(HOLD TANK)
I"
ASPND1
6
(POND)
FLUE
I*[MAK
SETTLED SOLIDS! w*
ADDITIONAL MAKEUP WAI
ORDER OF PROCESS CALCULATIONS:
3
DIVDR3
12
SUMMR1
6
16
^-EVAPORATION
rj A (j .,._.,..^— , , , -i
l»Ao •»• tjYSIBb _£__
EUP 6 5
|ME_Z_». (OVERALL *
5 _ SYSTEM 18 __
' fc" *" BALANCE) *"
STACK GAS
STACK GAS
SETTLED SOLIDS
POND EVAPORATION
1. 2. 3. 4. 5. 6. 7, 8. 9. 10. 5, 6. 7. 8. 9. 16. (10, 11, 12.) 13, 14, 15 *
Figure 2-3. Colstrip scrubbing simulation flow scheme
(see Figure 2-1 for process flow diagram)
02-1265-1
J-17
-------
all stream compositions and flow rates using precipitation rate
kinetics for CaSCK'2H20 and CaS03*%H20, which are the solids
formed in lime/limestone scrubbing systems, and various input
parameters. These parameters characterize the operating condi-
tions for a particular scrubbing system and include flue gas
flow and composition, fly ash rate and composition, makeup water
composition, lime addition rate, tank volumes, scrubber feed
flow rate and percent suspended solids, percent oxidation in the
system, and percent solids in the sludge.
Iterative calculations are performed around the scrub-
bing loop (boxes 10, 11, 12) through the scrubber vessel (SCRUBS)
and the scrubber recycle tank (RATHLD) until relative satura-
tions and stream compositions satisfy the rate equations. Then
calculations are performed for ancillary equipment such as the
reheat and fan requirements, and to determine stream composi-
tions around the effluent tank (box 13). Makeup water require-
ments are calculated by an overall system balance (SYSTB5, box
Several assumptions are inherent in performing this
simulation with the model outlined above. These are enumera-
ted below:
1) The scrubber exit gas is saturated with
respect to water.
2) Equilibrium exists between CO2 in the stack
gas and liquor in the scrubber bottoms.
3) The scrubber bottoms and stack gas tempera-
tures are the adiabatic saturation tempera-
ture of the flue gas.
4) All oxidation was assumed to occur in the
scrubber.
5) All solids precipitation occurs in reaction
vessels (subroutine RATHLD).
6) Ionic reactions taking place in the liquid
phase are rapid and thus at equilibrium.
The data used as input to this model is presented in
Table 2-4. Some of this data was obtained directly from MFC
while some of it was calulated from MFC data or sample analyses
J-18
-------
TABLE 2-4. INPUT DATA FOR COLSTRIP SCRUBBING SIMULATION*
FLUE GAS
Flow, m3/hr 779,000
(ACFM) (458,000)
Temperature, °C 141
(°F) (291)
Composition, mole %
S02 .079
C02 14.1
02 3.47
N2 72.8
H20 9.52
Fly Ash Rate, kg/min 331
(Ib/min) (728)
SYSTEM PARAMETERS
S02 Removal Efficiency, % 74
Oxidation, % 90
Particulate Removal Efficiency, % 99-6
Scrubber Feed Rate, £/sec 963
(GPM) (15,260)
Scrubber Slurry Solids, wt. % 12.
Recycle Tank Volume, m3 380
(ft3) (13,370)
Effluent Tank Volume, m3 43
(ft3) (1,520)
Effluent Tank Solids, wt. % 6.
Sludge, wt. % solids 50
MAKEUP WATER COMPOSITION, mg/£
Calcium 39.9
Magnesium 10.7
Sodium 40.3
Chloride 17.0
Carbonates (as C03 ) 6.0
Sulfates (as SC-O 188.
Nitrate (as N03~) 1.4
* All flows and tank volumes are for each scrubber module.
J-19
-------
The flue gas flow and composition were calculated from the coal
composition and firing rate and the boiler excess air rate.
The flue gas temperature entering the scrubber was obtained
from MFC. Results of these calculations are presented in Ap-
pendix K.
All of the system parameters were supplied by MFC as
design conditions. The makeup water composition was taken from
the sample analyses after adjustment to minimize residual elec-
troneutrality (see Appendix E). The fly ash composition was
determined from the model by specifying the lime addition rate
and adjusting the leachable content of the fly ash to match the
species concentrations measured in the system. Leaching tests
were performed for fly ash from MFC's Corette Plant which fires
the same coal as Colstrip. Results of these studies are shown
in Appendix D.
2.3.2 Simulation Results
This section describes the results from the simula-
tion of design scrubber operations at Colstrip. Before the
simulation was performed, a sample consistency calculation
around the effluent tank was performed since gypsum relative
saturations in the system varied from 1.0 to 1.38. Since all
input and output streams around the effluent tank were sampled,
this consistency check could be made. Table 2-5 shows the
stream dissolved species concentrations and the results of the
consistency calculation shown as a percent deviation defined as:
I species in - £ species out x 100
(£ species ±n + T, species out)/2
A detailed description of the calculation technique is given
in Appendix K.
As the results show in Table 2-5, the deviations for
all species except carbonate and sulfite are acceptable. The
excessive variation in sulfite is probably due to a small amount
of oxidation occurring in the tank. Sulfite represents such a
small portion of the total sulfur that 1-2% oxidation would
adequately explain this inconsistency. The carbonate variation
may be explained by CO 2 transfer between the process liquor and
the atmosphere, and/or analytical errors.
J-20
-------
TABLE 2-5. SAMPLE CONSISTENCY ERRORS AROUND EFFLUENT TANK AT COLSTRIP
10
Element
Calcium
Magnesium
Sodium
Potassium
Chloride
Carbonate
Sulfite
Total Sulfur
Nitrate
Pond recycle
484
1,550
305
13.1
70
9.6
400
9,000
130
Stream Concentrations,
Scrubber blowdown
504
5,050
458
21.9
129
52.2
300
19,400
161
me/8,
Effluent tank
497
2,075
315
15.5
74
31.2
100
11,800
118
% Deviation*
-1.9
+4.4
+5.3
-5.6
+8.3
-58.3
+117.1
-8.5
+13.8
* Defined as E in - E out
3
(E in + E out)/2
100
-------
These calculations uphold the measured concentrations
and therefore the calculated relative saturations. The low
relative saturation (1.0) of CaSOl|-2H20 in the scrubber recycle
liquor indicates that some calcium or sulfate from the fly ash
may be dissolving in the effluent tank causing the relative
saturation to increase across the tank.
The results from the process simulation of design
conditions are presented in Table 2-6. The calculated pH,
suspended solids, relative saturations, and composition for
the scrubber recycle slurry, effluent tank slurry, and pond
recycle liquor are compared to the sample values for these
streams .
The calculated pH of 4.98 for the recycle slurry
compares well to the design value of 5 but under actual opera-
tion on the day of sampling, the pH was 3.9. The effluent tank
pH's compared reasonably well but the pond recycle pH was
higher in the sample than calculated. These pH variations
indicate that the scrubbing system was not running under true
steady-state conditions. This is due to the long residence
time in the pond which makes the system time response to pro-
cess changes quite slow.
The relative saturations in these three streams for
CaSO^-ZHzO were quite different in all cases. However, in one
case, the recycle slurry, the sample value was lower than the
calculated value but in the other two cases the calculated
values were lower. Again, this can be explained by non-steady
state operation and/or analytical errors.
The critical range of values for CaSO^^HaO relative
saturation for scale formation is 1.3-1.4. Gypsum relative
saturations as high as 1.38 were found in the Colstrip scrub-
bing system indicating that operation at design conditions is
very near scaling. Some nucleation may be occurring, but the
erosive nature of the fly ash could be keeping vessel walls
clean, or breaking up gypsum crystals to form very fine seeds
to increase precipitation rates. The fly ash may also be pro-
viding nucleation sites for gypsum since no scale has been re-
ported at Colstrip.
The stream compositions compared well in the recycle
slurry with respect to magnesium, sodium and chloride. Calcium
and sulfate concentrations were lower in the sampled stream,
J-22
-------
TABLE 2-6. COLSTRIP SCRUBBING SIMULATION RESULTS FOR DESIGN CONDITIONS
(-1
i
S3
Stream
pH
Suspended Solids, wt%
Relative Saturations***
CaSCV2H20
CaC03
Composition, mg/&
Calcium
Magnesium
Sodium
Chloride
Total Sulfur (as SO,,")
Sulfite (as S03~)
Carbonate (as C03 )
Nitrate (as N03~)
Scrubber
Sample
3.
7.
1.
1.9x10"
504
5,050
480
129
19,400
300
52.
161
recycle slurry
Calculated
9* 4.98
7** 12.6
00 1.41
5 l.lxlO"3
733
5,285
444
117
24,560
3,560
2 153
9.6
Effluent tank
Sample
4.4
1.36
1.31
1.5xlO~6
497
2,075
330
74
11,800
100
31.2
118
Calculated
4.77
6.0
1.07
3.1xlO~"
542
5,690
478
126
22,250
3,650
147
10.3
Pond
Sample
5.5
neg.
1.38
2.2xlO~5
484
1,550
318
70
9,000
400
9.6
130
recycle
Calculated
4.77
0.0
1.0
3.1x10""
498
6,010
506
133
23,440
3,750
142
10.9
* Design pH is 5
**Design is 12% solids
***Critical values, above which scale potential exists, are 1.3-1.4 for CaSOit*2H20, and
about 2.5 for CaCOs (see Appendix C)
-------
most probably due to reduced load and/or reduced S02 content
in the gas from design conditions, resulting in a lower mass
removal rate of sulfur from the gas.
The large magnesium variations in the other two
streams may be explained by non-steady state, since magnesium
concentrations throughout the system should be approximately
the same at steady state.
The lower sample sulfite values are probably a re-
sult of increased oxidation over the design value of 9070.
Lower SOa in the flue gas generally results in increased oxi-
dation due to a higher oxygen to sulfur ratio.
One possibility for more efficient recycle/reuse
at Colstrip might be to use a combination of cooling tower
blowdown and river water or just river water as makeup to
the scrubbing system without treatment. The effects of makeup
water composition on scaling tendencies in the scrubbing system
will be examined in Section 3.0.
J-24
-------
3.0 TECHNICAL ALTERNATIVES
A modular approach to studying water recycle/reuse
alternatives at Colstrip was used in that the major plant water
system was divided into two subsystems to form separate process
simulations. One subsystem consists of the cooling towers, with
associated treatment facilities (where necessary), hold tanks,
and condensers. The other subsystem consists of the combined
S02 and particulate scrubbers including the disposal ponds.
The studies for each subsystem will first be discussed separ-
ately. The effects of increasing the cycles of concentration
versus treatment alternatives in the cooling towers and of
using poorer quality makeup water (increased calcium and sul-
fate) are presented first. Then the effects of flue gas ash
content, slurry percent solids, and makeup water composition
on the scrubbing system operation are discussed. After inves-
tigating the water subsystems, possible alternatives for more
efficient water recycle/reuse in the overall system are out-
lined.
3.1 Cooling Tower System
The existing operations simulations indicated that
the towers were operating at about the maximum cycles of con-
centration without using slipstream softening. However, with
slipstream treatment for calcium removal, the cycles may be
increased resulting in a reduction in makeup water require-
ments for the cooling system. The amount of treatment re-
quired will depend on the circulating liquor calcium and sul-
fate concentrations. This section first presents a brief
description of the simulation basis, followed by discussions
of the results with respect to treatment alternatives and cal-
cium and sulfate concentrations in the makeup water.
3.1.1 Simulation Basis
The process model used to simulate alternatives for
cooling tower operation is identical to that used for existing
operations. Slipstream softening was necessary to keep the
CaSOi^^HaO relative saturation below the critical range for
sciling of 1.3-1.4. Convergence criterion for this relative
saturation is between 0.8 and 1.2 in the computer model.
A total of six simulations were performed for alter-
native cooling tower operations. One simulation was performed
with the existing makeup water composition at 20 cycles of con-
centration. Another simulation was run to determine the maximum
J-25
-------
cycles using untreated makeup water. Four additional simula-
tions at 20 cycles of concentration were run with variations
in calcium and sulfate concentrations in the makeup water to
determine slipstream treatment rates.
All of the alternative cooling tower simulations were
performed for summer operation since under these conditions a
maximum blowdown rate is achieved. Increased evaporation rates
realized during the summer months necessitate an increase in
blowdown rate over that required during the winter months to
maintain a constant value for cycles of concentration. All
values reported in this section refer to one unit unless spec-
ified otherwise.
The only changes in tne input data for all of the
alternative simulations are the values for cycles of concentra-
tion and makeup water composition,
3.1.2 Cooling Tower Makeup Treatment Alternatives
The results from the simulations with untreated water
and with slipstream treatment are compared to the existing sys-
tem operation with treated makeup water in Table 3-1. Column
A represents operation without any softening of the makeup water
or a slipstream of the circulating water. Column B represents
the existing system with softening of the makeup water, and
Column C represents operation with treatment of a slipstream
for calcium removal.
Without any softening in the cooling tower loop, only
11 cycles of concentration can be obtained as compared to 13.5
cycles with pretreatment, before CaSOit-^HaO relative saturation
approaches a value above 1.0. The makeup water rate for no
treatment is slightly higher because of the lowered cycles of
concentration. The lowered cycles also result in a small de-
crease in acid requirements and an increase in blowdown rate
from 11.8 I/sec (187 GPM) to 15.0 £/sec (238 GPM) per tower.
Approximately 820 kg/day (1800 Ib/day) of CaC03 is removed from
the makeup water per tower with pretreatment, allowing cycles
to rise to 13.5.
When slipstream treatment is employed, cycles of con-
centration can be further increased. Slipstream treatment also
only requires treatment of about 9 £/sec (142 GPM) for 20 cycles
whereas pretreatment requires that about 175 £/sec (2770 GPM) of
J-26
-------
TABLE 3-1.
TREATMENT ALTERNATIVES FOR COLSTRIP
COOLING TOWER OPERATION
Treatment Method*
Cycles of Concentration
Makeup Water Race, 2V sec
(GPM)
Acid Addition Rate, kg/day**
(Ib/day)
Treatment Race, I/ sac
(GPM)
Calcium Removal Rate, kg CaCOj/day
(Ib/day)
Slowdown
Flow, I/sec
(GPM)
pH
Composition, mg/S.
Calcium
Magnesium
Sodium
Chloride
Carbonate (as CO] )
Sulfate (as SOU )
Nitrate (as H03~)
Temperature, °C
(°F)
Relative Saturations***
CaCOj
CaSOu-2H20
A
11.0
178
(2820)
127
(279)
0
(0)
0
(0)
15.0
(238)
7.5
631
117
439
531
22.3
2140
18.5
42.2
(108)
.92
.96
B
13.5
175
(2770)
152
(334)
175
(2770)
820
(1800)
11.8
(187)
7.2
534
143
540
227
6.5
2640
18.7
42.2
(108)
.097
.93
C
20.0
170
(2700)
161
(354)
9.0
(142)
1040
(2290)
7.3
(115)
7.6
587
212
1450
964
25.5
3930
33.6
42.2
(108)
.86
1.02
* Method A - No treatment
Method B • Pretreatment (existing operations)
Method C - Slipstream treatment
** as 100* H2SOu
*** Critical values, above which scale potential exists, are 1.3-1.4 for
and about 2.5 for CaCCh (gee Appendix C)
J-27
-------
water be softened. The reason for this dramatic difference is
that the slipstream of recirculating water is much more concen-
trated than the makeup water and the required calcium removal
can be achieved by treating a smaller stream with a higher cal-
cium level.
About 1040 kg/day (2290 Ib/day) of CaCO3 is removed
from the slipstream of each tower at 20 cycles as compared to
820 kg/day (1800 Ib/day) for pretreatment. Also, operating at
20 cycles of concentration results in a 3870 reduction in blow-
down flow from 11.8 £/sec (187 GPM) with pretreatment (13.5
cycles) to 7.3 a/sec (115 GPM) with slipstream treatment.
The level of slipstream treatment required will de-
pend on the calcium and sulfate concentrations in the makeup
water. Since these parameters may vary with time, the effects
of calcium and sulfate concentrations on slipstream treatment
were determined and are presented in the following sections.
3-1.3 Effects of Calcium Concentration in Makeup Water
at 20 Cycles of Concentration
The results for the three simulations performed to
determine the effects of makeup water calcium concentration on
the magnitude of slipstream treatment are presented in Table
3-2. The calcium concentration in the makeup water was varied
from 39.9 mg/£ to 80 mg/£ in the three runs made. The required
slipstream treatment and calcium removal rates were calculated
and increased as expected with increases in makeup calcium con-
centration. Slipstream rates varied from 2.8 £/sec (45 GPM)
to 13.2 H/sec (210 GPM). Slipstream rate is plotted versus
calcium concentration in the makeup water in Figure 3-1.
It should be noted here that this curve applies only
to the makeup water composition considered. Variations in sul-
fate will definitely affect the slipstream rate and variations
in other species' concentrations may cause significant changes
due to chemical complexing. This curve is valid for the com-
positions considered and is presented to show trends in the
system.
J-28
-------
250-
200 <
5
o.
O
DC
K
UJ
t-
Ui
CL
5
<
in
oc
O)
(L
J
«o
150
100
SO
30
40 50 60 70
CALCIUM CONCENTRATION IN MAKEUP WATER. MG/L
80
—i
90
Figure 3-1
Slipstream rate as a function of makeup calcium
concentration at Colstrip.
-------
TABLE 3-2.
SIMULATION RESULTS FOR CALCIUM VARIATIONS
IN THE MAKEUP WATER AT COLSTRIP
Cycles of Concentration
Calcium in Makeup
Water, mg/£
Slowdown
Composition, mg/&
Calcium
Magnes ium
Sodium
Chloride
Carbonates (as C03)
Sulfates (as S(H)
Nitrates (as N03)
pH
Relative Saturation*
CaCOs
CaSCv2H20
Sulfuric Acid Rate, kg/day**
Ib/day
Slipstream Rate, Si/sec
(GPM)
Calcium Removal Rate
kg CaCOa/day
(Ib CaC03/day)
Low
Calcium
20
39-9
606
212
1015
336
21.6
3910
27.7
7.7
.98
1.13
146
(320)
2.8
(45)
341
(749)
Medium
Calcium
20
57.9
587
212
1450
964
25.5
3930
33.6
7.6
.86
1.02
158
(348)
9.0
(142)
1040
(2290)
High
Calcium
20
80.0
654
212
1890
1740
24.8
3950
33.6
7.6
.96
1.05
174
(384)
13.2
(210)
1725
(3800)
*Critical values, above which scale potential exists, are 1.3-1.4 for
CaS04*2HaO and about 2.5 for CaCOs (see Appendix C) '
**100%
J-30
-------
3.1.4 Effect of Sulfate Concentration in Makeup Water at
20 Cycles of Concentration~~
Since the basis for defining slipstream treatment rates
is the relative saturation of CaSOi+^HzO, the concentration of
the sulfate species in the makeup water will have a significant
effect on these slipstream rates. As the sulfate increases, the
relative saturation of gypsum will increase, necessitating addi-
tional calcium removal to prevent gypsum scale formation.
Two additional simulations were performed to quantify
the differences in required slipstream treatment rates with
changes in the makeup water sulfate concentration. The results
from these two additional simulations along with the case using
untreated makeup are shown in Table 3-3. The three values for
sulfate considered were 125, 188, and 376 mg/£. The case with
188 mg/& represents the existing plant makeup water composition.
Slipstream rates varied from 9.0 £/sec (142 GPM) for the exis-
ting case (Simulation No. 4) to 4.9 &/sec (77 GPM) for the lower
sulfate concentration and 12.3 H/sec (195 GPM) for the higher
sulfate concentration.
The calculated slipstream rates are plotted versus
sulfate concentration in the makeup water in Figure 3-2. Again,
it should be noted that this curve applies only to the makeup
water considered. Variations in calcium have been shown to affect
slipstream rates in the previous section. In addition, variations
in other species' concentrations may cause significant variations
in slipstream rate due to the formation of ionic complexes in the
water. If the magnesium bicarbonate process (TH-192) is used for
slipstream treatment, the magnesium concentration in the system
will increase and treatment requirements may be reduced. The
effects of magnesium on CaSOit«2H20 relative saturation were
presented in Appendix I. The curve in Figure 3-2, however, repre-
sent trends in the system as a function of makeup water sulfate
concentration.
3.1.5 Summary of Cooling Tower Alternatives
The first set of simulations comparing treatment alter-
natives shows that with the use of slipstream softening, as
opposed to the present pretreatment system, cycles of concentra-
tion may be increased in the cooling towers. Increasing the
cycles from 13.5 (existing) to 20 reduces the blowdown rate from
11.8 £/sec (187 GPM) to 7.3 I/sec (115 GPM) and requires a slip-
stream rate of 9.0 H/sec (142 GPM) for each tower.
Additional simulations were performed which showed
the magnitude of effects of calcium and sulfate concentrations
J-31
-------
250T
200
C-l
to
S
0.
o
<
IT
Ul
2
111
IT
ui
DC
t-
05
150
100
50
100
150 200 250 300 350
SULFATE CONCENTRATION IN THE MAKEUP WATER. MG/L AS SO<
400
Figure 3-2. Slipstream rate as a function of makeup sulfate
concentration at Colstrip.
-------
TABLE 3-3.
SIMULATION RESULTS FOR SULFATE VARIATIONS
IN THE MAKEUP WATER AT COLSTRIP
Cycles of Concentration
Sulfate in Makeup
Water, mg/£ as SOit
Slowdown
Composition, mg/£
Calcium
Magnes ium
Sodium
Chloride
Carbonates (as CO 3)
Sulfates (as SOu)
Nitrates (as NO^)
pH
Relative Saturations*
CaC03
CaSO^«2H20
Sulfuric Acid Rate, kg/day**
(Ib/day)
Slipstream Rate, if sec
(GPM)
Calcium Removal Rate
kg CaCOa/day
(Ib CaCOa/day)
Low
Sulfate
20.
125.
751.
212.
633.
964.
23.5
2610.
33.6
7.5
.87
1.09
102.
(225)
4.9
(77)
736.
(1620)
Medium
Sulfate
20.
188.
587.
212.
1450.
964.
25.5
3930.
33.6
7.6
.86
1.02
158.
(348)
9.0
(142)
1040.
(2290)
High
Sulfate
20.
376.
501.
212.
3350.
964.
31.0
7670.
33.6
7.8
.90
1.03
177.
(389)
12.3
(195)
1200.
(2640)
*Critical values, above which scale potential exists, are 1.3-1.4 for
CaSOit'2H20 and about 2.5 for CaCOs (see Appendix C)
**100%
J-33
-------
in the makeup water on slipstream treatment rate. Graphs of
these effects showed that calcium had a greater effect on the
rate than did sulfate since the slope of the treatment rate
versus concentration plot for calcium was steeper than that
for sulfate. Doubling the calcium concentration increased the
treatment rate about 370% from 2.8 £/sec (45 GPM) to 13.2 £/sec
(210 GPM), whereas doubling the sulfate concentration only in-
creased the treatment rate about 37%, from 9.0 £/sec (142 GPM)
to 12.3 Jl/sec (195 GPM).
Operating the cooling towers at higher cycles of con-
centration may cause species besides gypsum and calcium carbon-
ate to become supersaturated and possibly form scale. Table
3-4 shows relative saturations for silica and phosphate solids
in the cooling tower blowdown at 20 cycles. The silica solid
with the largest relative saturation is Si02 with a value of
.12 which is well below saturation. The highest magnesium-silica
solid relative saturation is for Mg(Si02)a(OH)2 (sepiolite) with
a value of 0.20.
TABLE 3-4. RELATIVE SATURATIONS OF SCALE-FORMING SPECIES FOR
20 CYCLES WITH EXISTING MAKEUP WATER AT COLSTRIP*
Species
Ca(OH) 2
CaC03
CaSO^Hj-O
CaHPO,
Ca3(POO2
Mg(OH)2
MgC03
Si02
Mg2Si305(OH)6
Mg3Si205(OH)lt
Mg(Si02)3(OH)2
CaH2SiO.,
Ca(H2SiO,,)2
Relative Saturation
9
5
2
6
6
1
5
.6 x 10 1 °
0.60
1.13
0.015
.4 x 10"*
.5 x 10" *
.7 x 10~5
0.12
.1 x 10~5
0.012
0.20
.8 x 10""
.2 x 10""
day and a slipstream treatment rate for calcium removal of 142
GPM.
J-34
-------
3.2 Ash Handling (Particulate and SO2 Scrubbing) System
The simulation performed for design conditions in the
scrubbing system indicated the relative saturation of gypsum in
the system at steady state is at the upper level of the critical
range for scale formation of 1.3-1.4. Although increased recycle/
reuse cannot be employed in the scrubbing system since the sys-
tem is already at zero discharge, more efficient use of water in
the overall plant water system may be achieved by using a differ-
ent makeup water source in the scrubbers.
This section examines the effects of makeup water as
well as flue gas ash content and slurry percent solids on the
scale-free operation of the Colstrip scrubbing system. First,
a brief description of the simulation basis is given. Then
discussions of the effects of flue gas ash content, slurry per-
cent solids, and makeup water composition on scrubber operation
are presented.
3.2.1 Simulation Basis
The process model used to examine the effects of
operating parameters on scrubber operation is identical to that
used for design operation.
Four additional simulations were performed. One was
run with all parameters identical except the flue gas ash con-
tent was lowered by 3070 to represent operation with coal of a
lesser ash content. Another simulation was performed with the
base case data except the scrubber recycle suspended solids
were reduced to about 7.5% as opposed to the design value of
12%. Finally, two additional simulations were performed with
cooling tower blowdown and untreated water as makeup sources to
the s crubbers.
3.2.2 Effects of Flue Gas Ash Content
The simulation results for reduced flue gas ash content
are compared to the design operation results in Table 3-5. The
fly ash in the flue gas entering the scrubber was lowered from
the design rate of 331 kg/min (728 Ib/min) to 230 kg/min (505
Ib/min). The pH of the recycle slurry was maintained at 5 as
this is the design value. The lower ash content in the flue gas
lowers the fraction of the recirculating solids which is inert
ash, thereby increasing the fraction of the solids which is gyp-
sum. Table 3-5 shows that the portion of the slurry solids which
is CaS(K'2H20 increased from 38.9% to 47.2%.
J-35
-------
TABLE 3-5.
EFFECT OF FLUE GAS ASH CONTENT
ON COLSTRIP SCRUBBER OPERATION
Simulation No.
(Design Conditions)
Flue Gas Ash Flow, kg/min
(Ib/min)
Scrubber Recycle Slurry
PH
Liquor Composition, mg/S
Calcium
Magnesium
Sodium
Chloride
Carbonates (as C03 )
Sulfates (as S0i»~)
Sulfite (as SOs")
Nitrate (as N0s~)
Solid Composition, wt %
CaSOif'2H20
Inert (ash)
CaS03'l/2H20
Relative Saturation*
CaSCK'2H20
331
(728)
4.98
733
5,285
444
117
153
21,000
3,560
9.6
38.9
57.3
3.8
1.41
230
(505)
5.18
715
4,267
448
132
156
17,800
2,135
10.9
47.2
47.8
5.0
1.37
*The critical value, above which scale potential exists, is 1.3-1.4
for CaS04*2H20
The increase in recirculating gypsum solids provides
more precipitation sites, allowing more efficient solids forma-
tion and, therefore, a lower gypsum relative saturation in the
liquor. The simulation of lower ash content in the flue gas
predicted a reduction in gypsum relative saturation from 1.41
to 1.37. This means that operation with less than the design
ash rate will be more conducive to nonscaling operation, and
that operation with a higher ash rate may cause scaling problems,
However, the presence of erosive ash may cause the gypsum crys-
tals to be broken up, providing more precipitation sites and,
therefore, counteract to some extent the increased scale poten-
tial caused by higher ash content in the recirculating solids.
J-36
-------
3.2.3
Effects of Slurry Solids Content
The design value for slurry solids content is 12%. The
slurry solids content will have an effect on scrubber operation
since these recirculating solids provide the precipitation sites
for CaSOi»*2H20 solid formation in the reaction tank. The results
of the simulation performed at about 7.5% solids are compared to
the design operation (12% solids) results in Table 3-6.
TABLE 3-6.
EFFECT OF SLURRY SOLIDS CONTENT
ON COLSTRIP SCRUBBER OPERATION
Flue Gas Ash
Rate, kg/min
(Ib/min)
Simulation No.
1
(Design Conditions)
331.
(728)
2
331.
(728) •
Scrubber Recycle Slurry
Suspended Solids, wt. %
PH
Liquor Composition, mg/£
Calcium
Magnesium
Sodium
Chloride
Carbonate (as CO3)
Sulfate (as SO^)
Sulfite (as S0~3)
Nitrate (as NO^)
Solid Composition, wt. %
CaS(V2H20
Inert (ash)
CaS03'%H20
Relative Saturation*
12.6
4.98
733.
5,285.
444.
117.
153.
21,000.
3,560.
9.6
38.9
57.3
3.8
1.41
7.6
5.09
897.
5,590.
469.
124.
167.
22,300.
3,930.
10.2
38.7
57.9
3.4
1.73
*The critical value, above which scale potential exists, is 1.3-1.4 for
J-37
-------
The solids composition remained essentially constant, but the
gypsum relative saturation rose from 1.41 to 1.73, which is sig-
nificantly above the critical range for scale formation of 1.3-
1.4. This rise in relative saturation is a result of the decrease
in gypsum solids circulating around the scrubbing loop which de-
creases the number of precipitation sites and, therefore, the pre-
cipitation rate. Continuous operation at lower solids content
will most likely cause some scale formation at Colstrip.
3.2.4 Effects of Makeup Water Composition
Two simulations were performed to determine the
effects of makeup water composition on scrubber operation. One
was run with untreated makeup water and one with existing cool-
ing tower blowdown as makeup water. The results from these
two simulations (Nos. 3 & 5) are compared to design conditions
in Table 3-7. The differences between using treated makeup
(design operation) and using untreated makeup (Case 3) are
hardly noticeable. An increase in calcium level in the makeup
water did not have a significant effect on liquor composition,
solid composition, or gypsum relative saturation in the system.
This is due to the fact that the calcium added with the makeup
water to the reaction tank represents only a small portion
(about .077o for Case 3) of the total liquid phase calcium pre-
sent in the tank. In the overall system, the calcium added
through the makeup water represents 0.470 of the total calcium
entering the system for Simulation 3 (untreated makeup water).
The last column in Table 3-7 shows the results from
the simulation with existing cooling tower blowdown as makeup
water. Although the composition of the scrubber recycle liquor
changed, the relative saturation of CaSOi»-2H20 did not change
appreciably. The more concentrated cooling tower blowdown
causes the concentrations in the scrubbing system to be in-
creased. However, in the more concentrated liquor, more chem-
ical complexing is occurring, and the activity coefficients of
the dissolved species are affected such that the relative sat-
uration of gypsum is virtually unchanged. The increase in
chlorides from 117 mg/£ to 1560 mg/Jl should not make the chloride
level high enough to cause corrosion problems (UH-007). Problems
could possibly be encountered in the mist eliminators if cooling
tower blowdown is used. However, a combination of river water
and cooling tower blowdown or just river water could perhaps
be used as mist eliminator wash. Cooling tower blowdown may be
saturated with respect to CaSCK'2H20 if the towers are operated
at high cycles of concentration and therefore could cause gypsum
scaling problems in the demister if used for demister wash
J-38
-------
TABLE 3-7. EFFECTS OF MAKEUP WATER COMPOSITION
ON COLSTRIP SCRUBBER OPERATION
Simulation No.
Makeup Water Source*
Flue Gas Ash Rate, kg /rain
(Ib/min)
Scrubber Recycle Slurry
Suspended Solids, wt %
PH
Liquor Composition, mg/S,
Calcium
Magnesium
Sodium
Chloride
Carbonate (as COa )
Sulfate (as SOO
Sulfite (as S03=)
Nitrate (as NOa")
Solid Composition, wt %
CaSOi,-2H20
Inert (ash)
CaS03-l/2H20
Relative Saturation**
CaSOu^HaO
* Makeup Waters, mg/JZ.
Calcium
Magnesium
Sodium
Chloride
Carbonate (as CO 3")
Sulfate (as S0i,=)
Nitrate (as N03~)
Relative Saturation of
1
(Design Operation)
Treated
Makeup
331
(728)
12.6
4.98
733
5,285
444
117
153
21,000
3,560
9.6
38.9
57.3
3.8
1.41
Treated
Makeup
39.9
10.7
40.3
17.0
6.0
188.
1.4
0.03
3
Untreated
Makeup
331
(728)
12.6
4.97
740
5,284
444
335
154
20,700
3,552
11.7
38.9
57.2
3.9
1.41
Untreated
Makeup
57.9
10.7
40.3
48.7
17.3
188.
1.7
0.04
5
Cooling Tower
Slowdown
331
(728)
12.6
5.04
686
6,190
3,870
1,560
150
29,400
3,700
128
39.9
56.3
3.8
1.42
Cooling Tower
Slowdown (13.5 cycles)
533.7
143.1
539.5
227.4
6.5
2,644.2
18.7
0.93
**The critical value, above which scale potential exists, is 1.3-1.4 for
CaSOi,'2HzO
J-39
-------
3.2.5 Summary of Scrubbing Alternatives
The first two simulations characterized the system
response to flue gas ash content and slurry percent solids.
As the ash rate increases, the fraction of the circulating
solids that is gypsum decreases, and the relative saturation
of CaSO^.ZI^O increases. As the percent solids in the circu-
lating liquor decreases, the amount of circulating gypsum
solids decreases and the relative saturation of CaSO^^H^
increases. Burning a coal of significantly higher ash content
or operating the system at low solids in the recycle loop may
cause scaling problems in the scrubbers at Colstrip.
The last set of simulations showed that the use of
either untreated water or cooling tower blowdown as makeup to
the scrubbing system does not have a significant impact on the
scaling tendency of the system. The use of cooling tower blow-
down will increase the total dissolved solids and chloride
corrosion problems (UH-007). The use of cooling tower blowdown
as mist eliminator wash exclusively could cause scaling prob-
lems but dilution of the cooling tower blowdown with river
water could possibly control scaling in the mist eliminators.
3.3 Combined System Alternatives
From the results of these simulations, cycles of
concentration can be increased through the use of slipstream
softening in the cooling system, and cooling tower blowdown
and raw makeup water may be used in the scrubbing system. Two
technical alternatives are outlined here. The economics associ-
ated with these alternatives are presented in Section 4.0.
The first alternative involves using treated water
for cooling tower makeup as is presently done. A portion of
the cooling tower blowdown could be used as feed to one brine
concentrator as needed for boiler makeup. The remainder of
the cooling tower blowdown can then be used in combination with
untreated river water as makeup to the scrubbers. The makeup
water requirements for this alternative will be the same as
the present amount, but softening requirements will be reduced
and only one brine concentrator is required.
The second alternative involves using raw river water
as cooling tower makeup and operating the towers at 20 cycles
of concentration with slipstream softening. As in Alternative 1,
a portion of the cooling tower blowdown can be used to feed one
J-40
-------
brine concentrator as needed for boiler makeup, and the remain-
der can be used in combination with untreated water as scrubbing
makeup. In this case the overall plant water requirements are
unchanged but only one brine concentrator is used. The flow to
the softener is decreased dramatically but the chemical require-
ments are increased due to an increase in the calcium removal
rate.
Rough economic estimates are presented in the next
section so that these two alternatives may be compared.
J-41
-------
4.0 ECONOMICS
This section provides rough cost estimates for imple-
menting each of the two alternatives discussed in the previous
section. Both rough capital costs and operating costs are pre-
sented. The assumptions and techniques used in calculating
these costs are briefly outlined. It should be emphasized here
that these economics are only rough estimates for comparative
purposes. It should also be noted that the following costs are
concerned with using the existing equipment at Colstrip and
do not reflect any savings which could have been realized if
more effective water recycle/reuse had been used at Colstrip
initially.
A capital cost summary for the two alternatives is
shown in Table 4-1. The system modifications for Alternative 1
include piping for using water before softening as scrubber
makeup and for using cooling tower blowdown as scrubber makeup.
Also included is additional pumps for transporting the cooling
tower blowdown to the scrubbing system. Alternative 2 modi-
fications include all of the changes for Alternative 1 plus
additional pumps and piping to convert to slipstream softening
instead of pretreatment in the cooling tower system.
Four-inch carbon steel pipe with average fittings,
flanges, shop coating, and wrapping was assumed for cooling
tower blowdown and slipstream softening streams. Six-inch
pipe was used for untreated water to the scrubbers. A labor
to material ratio of 0.8 was used to determine installation
costs. Engineering costs (direct and indirect) were assumed
to be 7.270 of the combined labor and material cost (GU-075).
Cast steel pumps with electric motor drivers were
used for cooling tower blowdown and slipstream softening.
A labor material ratio of 0.36 was used for installation costs.
Engineering was assumed to be 10% of the combined labor and
material cost (GU-075).
All pump and piping costs were upgraded from 1970
dollars to 1976 dollars using a factor of 1.56 (based on Chem-
ical Engineering Index).
An operating cost summary for the two alternatives
is presented in Table 4-2. Negative operating costs in this
table represents savings over existing operation. In the first
J-42
-------
TABLE 4-1. CAPITAL COSTS FOR WATER RECYCLE/
REUSE ALTERNATIVES AT COLSTRIP
Item
Alternative One
(1976 dollars)
Alternative Two
(1976 dollars)
Piping for raw water
to scrubbers
Piping for cooling tower
blowdown to scrubbers
Additional pumps and
drivers for cooling
tower blowdown
Piping to convert to
slip stream softening
Additional pumps and
drivers for slip stream
softening
Contingency (20%)
Contractual Fees (3%)
Total
5,000
60,000
64,000
26,000
4,000
159,000
5,000
60,000
64,000
25,000
69,000
45,000
7,000
275,000
J-43
-------
TABLE 4-2. OPERATING COSTS FOR WATER RECYCLE/
REUSE ALTERNATIVES AT COLSTRIF1
Alternative One Alternative Two
Item (1976 dollars/yr) (1976 dollars/yr)
Lime for softening
($35/ ton)
Brine concentrator
operation ($2/1000 gal)2
Capital Charges
(15% per year)3
Total
(mils/kw-hr)
-1,000
-260,000
24,000
-237,000
(-.046)
1,000
-260,000
42,000
-217,000
(-.044)
1 Based on 80% load factor
2LE-239
3MC-136
J-44
-------
alternative a savings is shown for softening lime requirements.
This is a result of the decrease in softener flow since the
scrubber makeup water is not softened in this alternative.
Alternative 2, on the other hand, shows an additional
operating expense of $l,000/yr for softening. This is due to
an increase in the required calcium removal rate in the cooling
tower system since the towers are operating at 20 cycles of
concentration in Alternative 2. The use of slipstream soften-
ing in this alternative allows treating a small, concentrated
stream as opposed to treating a large, dilute stream as is
presently done. If slipstream softening had been designed
into the cooling system initially a capital cost savings for
softening could have been realized due to the much smaller
equipment required. The costs reported here, however, are
based on using the existing softening equipment at Colstrip.
An operational savings is shown for brine concentra-
tor operation for both alternatives in Table 4-2. This esti-
mate is based on an operating cost of $2/1,000 gal (LE-239).
The savings results from reducing the brine concentrator feed
rate from about 25 £/sec (400 GPM) to 5.7 H/sec (90 GPM).
Only one of the two brine concentrators is used in these al-
ternatives. The extra unit will insure that boiler feed water
is always available when the unit is running. If this more
effective cascading of water at Colstrip had been used initial-
ly, only one brine concentrator with 9.5 £/sec (150 GPM) capa-
city would be required as opposed to the two 12.6 ft/sec (200
GPM) units being operated presently. This represents a capital
cost savings that could have been realized of about $1.9 mil-
lion based on $7,750/GPM (LE-239). The economics reported
here do not reflect this savings but only consider what can
be done with the existing equipment.
The last operating expense shown in Table 4-2 is
capital charges. These costs were estimated as 15% (MC-136)
of the capital investment shown in Table 4-1 for each alter-
native based on a 30-year lifetime. The net operating costs
do not vary greatly between alternatives. The major differ-
ence is a result of the increased capital charges for^slip-
stream softening equipment. In both cases, a net savings
was calculated due to the large savings in brine concentrator
operation.
J-45
-------
APPENDIX K
POWER PLANT DATA REDUCTION
1.0 APS FOUR CORNERS STATION
Much of the data required to study the water use
system at Four Corners were supplied directly by Arizona Public
Service (APS). Some of the information was calculated. This
section presents the required calculations .
1. 1 Four Corners Scrubbing System
Calculations were performed to estimate the particu-
late removal efficiency, the S02 oxidation rate, and the flue
gas composition at Four Corners. A study was performed to es-
timate the reactivity of the ash produced at Four Corners.
These calculations and the results of the batch dissolution
studies are presented in this subsection. These results were
used to simulate the scrubbing system at Four Corners.
1.1.1 Four Corners Scrubbing System Particulate
Removal Efficiency
Inlet Loading
Unit 1 & 2 = 332 Ib/min/train x 4 trains = 1328 Ib/min
Unit 3 = 408 Ib/min/train x 2 trains = 816 Ib/min
Outlet Loading
Unit 1 & 2 = 1.08 Ib/min/train x 4 trains =4.32 Ib/min
Unit 3 =1.37 Ib/min/train x 2 trains =2.74 Ib/min
Efficiency (Eff)
Unit 1 & 2 Eff = 1324'32 x 10° = 99-67%
Unit 3 Eff = 816' x 10° = 99'667°
Average Eff = 99.67%
K-l
-------
1.1.2 Four Corners S02 Oxidation Rate
Scrubber 1A Liquor
Total Sulfur =28.6 mmoles/£
SO3 = 0.10 mmoles/£
% Oxidation = 28-^".°-1 x 100 = 99.65%
Zo. o
Scrubber 3A Liquor
Total Sulfur =30.5 mmoles/£
SO3 = 0.75 mmoles/£
% Oxidation = 30-^0"g°-75 x 100 = 97.54%
Average % Oxidation = 98.6%
1.1.3 Four Corners Flue Gas Composition Calculations
Coal Composition
Component
C
H
S
N
0
H20
Ash
Total
wt%
52.62
3.81
0.69
1.19
9.02
11.69
20.98
100.00
Btu/lb = 9300
K-2
-------
Basis: 100 Ib coal; 3% 0? in boiler exit gas
Coal in
Component
C
H
S
N
0
H20
Ibs
52.62
3.81
0.69
1.19
9.02
11.69
Ib moles
4.385
3.81
0.02156
0.085
0.5638
0.6494
Required moles of 02 = 4.385 + %(3.81) + .02156 - %(.5638)
= 5.077 Ib moles
Nitrogen out = ^f (02 in air) + %(.085)
Oxygen out = 02 in air - 5.077
Flue Gas Out (Ib moles) = F = moles C02 + moles H20 + moles S02
+ moles N2 + moles 02
F = 4.385 + 2.554 + 0.02156
~79
-f
(02 in air) + %(.085)
02 in air - 5.077
Moles 02 out = 02 in air - 5.077 = .03F
so, 02 in air = .03F + 5.077
Substituting into above equation for gas flow and
solving:
F = 30.67 Ib moles
K-3
-------
Moles 02 out = (.03)(30.67) = 0.920 Ib moles
Moles N. out
79
IT
.03(30.67) + 5.077
+ %(0.85) = 22.6 Ib moles
These calculations result in the following gas composition
exiting the boiler
Component moles mole%
C02 4.385 14.39
H20 2.554 8.38
S02 0.02156 0.0707
N2 22.60 74.15
02 0.920 3.02
If 1070 leakage is occurring in air preheater there are an addi-
tional 3.07 moles of air in the gas
Moles N2 = .79 (3.07) = 2.4
Moles 02 = .21 (3.07) = 0.64
Gas Composition Entering Scrubber
Component moles mole7,
C02 4.385 13.08
H20 2.554 7.62
S02 0.2156 0.0643
N2 25.00 74.59
02 1.560 4.65
K-4
-------
1.1.4 Batch Dissolution Results with Fly Ash
From Four Corners
Fly Ash Composition
Component wt%
P205 0.35
Si02 56.44
Fe203 3.80
A1203 27.98
Ti02 1.06
CaO 3.04
MgO 1.19
S03 0.55
K20 0.77
Na20 2.32
Unidentified 2.50
Total 100.00
Existing operations: (Ash dissolution from data in ash leaching
results) ; pH = 3.0, 2.0 wt °L solids in slurry.
The calculations below show the results of the leaching
studies. The actual values used in existing operations simula-
tions were adjusted slightly to obtain the scrubber blowdown pH
observed. The values used were 4470 for CaO dissolution and 1-5%
for MgO as compared to the respective calculated values of 41. TL
and 1.1%.
CaO Dissolution
4.52 imnoles Ca dissolved 40 mg Ca .1 liter = 9.05 mg Ca dissolving
liter X mmole Ca X 2 g ash g ash
9.05 x 10"3 g Ca 56 g CaO g ash 10Q = 41.7% CaO
g aSh X 40 g Ca 0.0304 g CaO total dissolving
K-5
-------
RESULTS OF CHEMICAL ANALYSIS FROM LEACHING OF ASH SAMPLES
Calcium
Magnesium
Sodium
Potassium
Sulfate
AT CONSTANT
Free. Ash1
pH 3.0
(mmole/£)6
15.5
.23
.52
.12
1.2
pH FOR FOUR
Free. Ash2
pH 6.0
(mmole/Jl)6
13.4
.10
.48
.06
.31
CORNERS
Free. Ash3
pH 8.5
(mmoleM)6
11.7
.07
.48
.04
.56
POWER STATION
Free. Ash1*
pH 3.0
4.52
.07
.12
.04
.08
Free. Ash5
pH 8.5
(mmole/Jl) 7
3.02
.01
.08
<.02
.01
^intained pH of 3.0 by adding HC1.
Maintained pH of 6.0 by adding HC1.
Maintained pH of 8.5 by adding HC1.
Maintained pH of 3.0 by adding HC1.
5Maintained pH of 8.5 by adding HC1.
6All values represent mmole/£ of soluble species in leachate after 14 grams
of ash were leached in 186 grams of water at constant pH.
7A11 values represent mmole/5. of soluble species in leachate after 4 grams
of ash were leached in 196 grams of water at constant pH.
K-6
-------
MgO Dissolution
0.07 mmoles^Mg dissolved ^ 24.3 mg Mg 0.1 liter = 0.08 mg dissolving
liter mmole Mg x 2 g ash ~ g ash
8. x 10 5 g Mg dissolving 40.3 g MgO g ash
g ash X 24.3 g Mg X 0.0119 g MgO total X 10° ~
1.1 % MgO dissolving
Alternatives: (Ash dissolution from data in ash leaching re-
sults) ; pH = 6.0, 7.0 wt. 70 solids in slurry.
CaO Dissolution
13.4 mmoles Ca dissolved 40 mg Ca .1 liter
liter x mmole Ca 7 g ash
7.66 mg Ca dissolving
g ash
7.66 x 10"3 g Ca 56 g CaO g ash ,QO =
g ash x 40 g Ca x .0304 g CaO total x
35.4% CaO dissolving
MgO Dissolution
10 mmoles Mg dissolved 24.3 mg Mg .1 =
liter mmole Mg 7 g ash
.035 mg Mg dissolving
g ash
3.5 x 10~5 g Mg dissolving 40.3 g MgO _ g ash .„
g ash 24.3 g Mg 0.119 g MgO total
0.5% MgO dissolving
K-7
-------
2.0 GPC PLANT BOWEN
Much of the data required to study the water use
system at Bowen were supplied directly by Georgia Power Company
(GPC). Some of the information was calculated. This section
presents the calculations which were required to perform simu-
lations of the cooling tower and ash sluicing systems.
2.1 Bowen Cooling Towers
Estimates of the ambient air composition and tempera-
ture were calculated for the cooling tower simulations. These
calculations are presented below.
2.1.1 Climatological Data
Wet Bulb Dry Bulb Relative
Month Temp.,°F Temp.,°F Humidity,%
Dec. 74 41 44 75
Jan. 75 42 47 71
Feb. 43 47 73
Mar. 44 50 66
Apr. 52 60 61
May 64 70 74
Jun. 68 74 73
Jul. 71 75 83
Aug. 72 77 84
Sep. 65 69 81
Oct. 58 62 77
Nov. 48 54 69
Winter Average Conditions (Dec., Jan., Feb.)
Wet Bulb Temp. 42°F
Dry Bulb Temp. 46°F
Relative Humidity 73%
Summer Average Conditions (Jun., Jul., Aug.)
Wet Bulb Temp. 70°F
Dry Bulb Temp. 75°F
Relative Humidity 80%
K-8
-------
Air Flow (Example Calculation) - Winter
Cooling Tower 3 or 4 :
Design flow = 100 x 106 Ib/hr (wet bulb
temp. = 25 °F, 75% relative humidity)
Water content = 0.0024 Ib H20/lb bone dry air (BDA)
i no v i n 6
Ib BDA = = 9-976 x 107 lb/hr
For wet bulb temp. = 42 °F, 7370 relative humidity
Water content = 0.0047 Ib H20/lb BDA
Total Flow = 1.0047 x 9.976 x 10 7 = 100.2 x 106 lb/hr
100.2 x 106 Ib hr Ib-mole 359 scf 506°R
ALm hr x 60 min x 29 Ib x Ib-mole 492°R
= 2.12 x 10 7 ACFM
Air Composition (Example Calculation) - Winter
Basis: 1 Ib BDA
N2 .0272 Ib moles
02 .0072
C02 1.03 x 10"5
H20 2.61 x 10~"
.034671 Ib moles
mole- II 0 -0047 lb H2° - ° 61 T 10'"
moles II20 lg ib/ib_moie -•ei 'c 1U
-, . „ 1 lb BDA -,n moles Nj
nnr»l nc N- — - ... -TT 7M a >
mole fraction
0.7845
0.2077
0.0003
0.0075
1.0000
lb moles H20/lb BDA
29 Ib/lb-mole total moles
0.0272 lb moles N2/lb BDA
n = 1 lb BDA 71 moles 02
U2 29 Ib/lb-mole x ' total moles
0.0072 lb moles 02/lb BDA
K-9
-------
i rn - 1 lb BDA ~ -, «- 4 moles C02 =
moles CO 2 - 29 Ib/lb-mole x J total moles
1.03 x 10" 5 lb moles C02/lb BDA
2. 2 Bowen Ash System (Units 3 or 4)
Estimates of the ash and water flow rates were calcu-
lated for the ash sluicing simulations. Also required were the
result of the ash leaching studies which measured the reactivity
of the fly ash. These calculations are presented below.
2.2.1 Fly Ash
Precipitator Inlet - 3.0 gr/scf
Precipitator Outlet - 0.025 gr/scf
Flue Gas - 3.0 x 106 ACFM @ 300 °F, latm
Flv A^h - (3.0-O.Q25)gr 3.0 x 10s acf 60 min 492 R
Fly Ash -- ^ - x ffi x h x
hr 76Q R
= 49>520 lb/hr
T t- i AT-,- 890,000 kW-hr 9000 Btu lb coal . 11 lb coal
lotai Asn hr x kw_hr x 11)500 Btu x lb coal
= 76,620 lb/hr
Bottom Ash = 76,620 - 49,520 = 27,100 lb/hr
2.2.2 Total Ash from all Units
Fly Ash = 205,620 lb/hr
Bottom Ash = 68,140 lb/hr
Total = 273,760 lb/hr
Sluice water required for 10% slurry of all ash:
GPM = 273.760 lb ash hr gal .90 lb H20 _ 4950
Er x 60 min x 8.3 lb H20 x .10 lb ash ~ GPM
K-10
-------
2.2.3 Recirculating Ash System Flows
Slowdown from cooling towers (§15.0 cycles = 1507 GPM
Water required for fly ash slurry = 3716 GPM
Water required for bottom ash slurry = 1230 GPM
Fly ash makeup = 1491 GPM
Bottom ash makeup = 0 GPM
Fly ash recycle = 3716 - 1491 = 2225 GPM
Bottom ash recycle = 1230 GPM
2.2.4 Pond Evaporation
Reference: PA-121
Enthalpy of evaporation = (73 + 7.3W) (eg - e&)
where W = average wind speed, mph
where e = saturation vapor pressure of H20, mmHg
where es = existing vapor pressure of HaO, mmHg
cl
For Atlanta from January '75 to November '75, average
wind speed = 8.4 mph
e = 1.102 mmHg
s
e = 0.944 mmHg
cl
Hevap = [73 + (7.3 x 8.4)] [1.102 - .944] = .01765 ^^ day
Pond area = 250 acres = 1.089 x 107 ft2
K-ll
-------
day hr
x
., _. .01765 Ib H20 v -i nQQ v in ft. „
Evaporation = £t _ day x 1.089 x 10 ft x 60 min
x 8 Sa]^ = 18.4 GPM = 69.4 kg/min
2.2.5 Ash Dissolution
Leaching studies results are shown below.
RESULTS OF CHEMICAL ANALYSIS FROM LEACHING OF ASH SAMPLES
Calcium
Magnesium
Sodium
Potassium
Sulfate
AT CONSTANT pH
Prec. Ash1
pH 6.0
(mmole/£) *
24.2
.23
3.0
.90
10.0
FOR PLANT BOWEN
Prec. Ash2
pH 8.5
(mmole/5,) **
14.2
.06
1.1
.72
9.22
Prec. Ash3
pH 10.4
(mmole/5,) 4
10.5
.01
1.3
.56
7.71
Maintained pH of 6.0 by adding HC1.
Maintained pH of 8.5 by adding HC1.
Maintained pH of 10.4 by adding HC1.
''All values represent mmole/5, of soluble species in leachate
after 14 grams of ash were leached in 186 grams of water at
constant pH.
K-12
-------
Using pH 10.4, TL slurry
Assuming all sulfate is from
- 7.71 mmole CaSO^ , R,. , 136.1 mg g
it - -- ; - x . 185 kg x - 5 - "• x , nnn - x
& 6 mmole 1000 mg
I4g1ash x 100x-9 = 1-25 wt% CaSO,, dissolving before
pond (90% of ash) .
- (10.5-7.71) mmole CaO .186kg 56 Itng
mmole lOIDOmg
1
14g ash
x 100 = 0.21% at pH 10.4
ir n - (14.2-9.2) mmole CaO .186 kg 56.1 mg g
/oCa° ~ - 1 - x x mmole x 1000 mg
... 1 , x 100 = 0.37% at pH 8.5
14g ash
Avg =0.3 wt%
This average was used since CaO dissolution is more pH de-
pendent than CaSOi, and is rapid.
K-13
-------
3.0 PSC COMANCHE PLANT
Much of the data required to study the water use
system at Comanche were supplied directly by Public Service
of Colorado (PSC) . Some of the information was calculated.
This section presents the calculations which were required to
simulate cooling tower and ash sluicing operations.
3. 1 Comanche Cooling Towers
Estimates of the ambient air and composition were
calculated for the cooling tower simulations. These calcula
tions are presented below.
3.1.1 Inlet Air Composition
Average Winter Conditions = 32 °F, 51% relative humidity
Inlet Air = 0.0036
Bone Dry Air (BDA) = 78.98% N2, 20.99% 02 , 0.03% C02
mole fraction H20 = _ 0.0036 Ib H20 _ 29 Ib air
1 Ib BDA + 0.0036 Ib H20 x mole air
mole H20
x 18 Ib H20
= 0.005779 Ib moles H20/total Ib moles moist
air
mole fraction N2 = 78.98 moles N2 100 moles BDA
100 moles BDA x 100.6 moles moist air
= 0.7851 Ib moles N2/lb mole moist air
mole fraction 02 = 20. 99 moles Q2 100 moles BDA
100 moles BDA x 100.6 moles moist air
= 0.2087 Ib moles 02/lb mole moist air
K-14
-------
mole fraction CO = 0.03 mnlp.s rna 100 moles BDA
2 100 moles BDA 100.6 moles moist air
= 2.98 x 10'"* Ib moles C02/lb mole moist air
3.2 Comanche Ash System
Estimates of the ash flow rates, the pond evaporation
and the ash reactivity were calculated for the ash sluicing
simulations. These calculations are presented below.
3.2.1 Ash Flows
Coal Characterization (From Comanche Data)
Heating Value, Btu/lb Percent Ash
7887 8.65
8131 8.09
8286 5.45
8223 7.42
9021 6.67
Avg. 8310 Btu/lb 7.26% ash
Fly Ash = 78%, Bottom Ash = 22?0 of total ash
Heat Lost in Cooling Water:
143 200 gal 1 Btu 25.93°F 8.33 Ib 60 min
'min * X "Ib^F x X "g5I x hr
= 1.856 x 109 Btu/hr removed in
condenser with a AT of 25.93°F
K-15
-------
Heat Input:
Assuming the plant is about 40% efficient
Coal Rate = 1.856 x 109 Btu/hr Ib coal _ o 799 1o5
— • ,* • -f- X >VA'-I"/I" T\ *- — J • / Z. i, A. J_U
0.6 8310 Btu coal/hr
= 6.20 x 103 Ib
coal/min
Assuming precipitator efficiency of 95%:
Fly Ash = 6.20 x 103 Ib coal 7.26 Ib ash 0.78 x 0.95
min x 100 Ib coal x
= 333.7 Ib fly ash
min
Bottom Ash = 6.20 x 103 Ib coal 7.26 Ib ash 0.22 = 99 Ib
min x 100 Ib coal x min
bottom
3.2.2 Pond Evaporation ash
He (Enthalpy of evap.) = (73 + 7.3W) (e - e ) from (PA-121)
S 3.
where W = average wind speed, mph
eg = saturation pressure or H20, mmHg
ea = existing partial pressure of H20, mmHg
Average Summer Conditions: 74°F, 51% humidity, wind speed= 7 mph
e = 15.46 mmHg
S
eo = 10.95 mmHg
Si
K-16
-------
He = (73 + 7.3(7)) (15.46 - 10.95) Btu
He = 559.69 Btu
ft" day
H20 loss = 559.69 Btu x Ib
ft* day1060 Btu
ft" day
x 5410 ft2
x 453.6 g x day x hr
Ib 24 hr 3600 sec
H20 loss = 151.1 g/sec
3.2.3
Ash Dissolution
Using pH 8.5, 7% slurry
From the leaching study, the amounts of dissolved species are
Ca
Mg
Na
K
SO 3
mmoles/kg H20
45.34
1.77
1.07
0.15
6.99
The actual salts present are not known, but for convenience in
using these numbers in computer simulations, the sulfate was
combined with calcium and the remaining cations were expressed
as oxides:
mmoles/kg H20
CaSCK 6.99
CaO 38.35
MgO 1.77
Na20 0.61 (K combined with Na)
For a 7% slurry: 7 g ash x 1000 g
93 g H20 leg
= 75.27 g ash
H20
K-17
-------
Converting moles to grams:
mg/kg H20
CaSCK 951
CaO 2151
MgO 71.3
Na20 37.8
On an ash basis, this yields:
mg/g ash
CaSCU 12.63
CaO 28.58
MgO 0.947
Na20 0.503
Based on analogy with the leaching studies of the Bowen Plant
and Four Corners Plant, for a pH of 11.0 the dissolution will
be less than at 8.5. An assumption of a 20% reduction in dis-
solution will give conservative results (i.e., slightly more
dissolution than will probably occur). This gives:
wt 70 in ash, dissolved
CaS04 1.01
CaO 2.287
MgO 0.077
Na20 0.041
Inerts 96.585
K-18
-------
4.0 PP&L MONTOUR
Much of the data required to study the water use
system at Montour were supplied directly by Pennsylvania Power
and Light (PP&L). Some of the information was calculated.
This section presents the calculations which were required to
simulate the cooling tower and ash sluicing systems.
4.1 Montour Cooling Towers
For the cooling tower simulations estimates of the
ambient air composition and temperature, and the heat load on
the towers were calculated. In addition to these calculations
a sample calculation estimating the ACS index is presented.
4.1.1 Montour Climatological Data
Weather data were obtained from PP&L in order to
simulate typical weather conditions at the Montour plant.
These data were used to calculate ambient air conditions as
well as air flow rates through the tower. Averages for
December 1975 and August 1976 are reported in Table 4-1.
TABLE 4-1. AVERAGES FOR DECEMBER 1975 AND AUGUST 1976
Month
December 1975
August 1976
Wet Bulb
Temp., °F
28.5
64.4
Dry Bulb
Temp., °F
31.8
69.1
Relative
Humidity %
68.6
79.1
PP&L also supplied actual operating data from the
condensers for December 1975 and August 1976. These data were
used to obtain estimates for the cooling water temperatures,
the range, and the approach. Average values were used and sep-
arate calculations were performed for each tower. The air flow
rate was estimated using the method described in "Managing Waste
Heat with the Water Cooling Tower" (DI-057). This method uses
the concept of a rating factor, which takes into account the
wet bulb, the range, and the approach. These results are pre-
sented in Table 4-2.
K-19
-------
TABLE 4-2. COOLING TOWER OPERATING CONDITIONS
Air in Dry Bulb, (°F)
Air in Wet Bulb, (3F)
Water in Temp. , (°F)
Air Out Temp. , (°F)
Water Out Temp. , (°F)
Range (°F)
Approach (°F)
Air Flow (Ib/hr)
Aug. '76
//I
70.2
64.4
111.9
103
84.2
28
20
90 x 10s
Aug. '76
#2
70.2
64.4
109.6
100
81.3
28
17
100 x 10s
Dec. '75
#1
31.8
28.5
94.3
85
61.8
32.5
33
105 x 10 5
Dec. '75
n
31.8
28.5
99.7
91
60.7
39
32
95 x IQ'"
It should be noted that the air flow rates remained
essentially the same in December as in August. In fact, there
was as much variation between towers in a given month as between
months. For this reason the average of these four values was
used in the simulations.
The composition of the input air was calculated on
the basis of the relative humidity. It was assumed the ratio
of N2:02:C02 remained constant and that the change in the mole
fraction of the water changed the mole fraction of the others.
An example calculation using August 1976 data follows:
Basis: 1 Ib BDA mole fraction
N2 .0272 moles .7747
02 .0072 .2050
C02 1.03 x 10"s .0003
H20 7.22 x 10"4 .0200
Total .0351 moles 1.0000
K-20
-------
Moles of snecips = (mole fraction of BDA)
Moles or species (molecular weight of air)
N:
.79
29.0
Ib
Ib mo 1 e
= .0272 moles/lb BDA
02 =
.21
Tb
29.0 Ib mole
= .0072 moles/lb BDA
C02 =
.0003
29.0 Ib mole
= 1.03 x 10"5 moles/lb BDA
Using the psychrometric chart we find under these conditions that
there is:
Ib BDA
Therefore:
moles
•°13 ' = 7.22 x 10~" Ib moles/lb BDA
18.016
4.1.2
Montour Keat Load
No direct information was sent by PP&L on the expec-
ted heat load on the cooling towers. Information was available
on the inlet and outlet condenser temperatures as shown in
Table 4-2. These data were input into the computer model and
the heat load was calculated using an energy balance. As an
independent check on these calculations, the cooling require-
ments were calculated from other data sent by PP&L. The results
from this second method are presented in Appendix I. The follow-
ing outlines the assumptions and calculations made to obtain
these results.
K-21
-------
Assume: Full load electric production
Electricity = 750 MW
= 1.79 x 108 cal/sec
(2.56 x 109 Btu/hr)
Assume:
Design evaporation takes care of
95% of full load cooling
Evaporation = 430 I/sec
(6800 GPM).
AHvap = 586 ca1/^111
(1054 Btu/lb)
Cooling load = Evap. x AH,T__ x
V tip
- 2.64 x 108 cal/sec
(3.77 x 109 Btu/hr)
Data from 3/13/75: Plant operating at 752 MW
Flue gas temp. = 140°C
(285°F)
Flow rate = 3.96 x 106 m3/hr
(2.332 x 106 acfm)
Ambient temp. = 21°C
(70°F)
Heat capacity - .26 cal/gm
(.26 Btu/lb)
Density of air = 8.5 x lO"1* gm/cm3
(0.53 lbm/ft3)
K-22
-------
Stack losses = (V)(p)(AT)(Cp)
= .29 x 108 cal/sec
(.42 x 109 Btu/hr)
Heat input = 4.817 x 108 cal/sec
(6.879 x 109 Btu/hr)
These data can be checked for consistency by taking
the difference between the total heat in and the total heat out
The difference can easily be attributed to other losses if it
is reasonably small.
Other losses = Input - (Electricity + Cooling
Load + Stack Losses)
= 9.0 x 106 cal/sec
(1.2 x 108 Btu/hr)
This value is much less than any of the other heat
losses and this confirms that the assumptions made in these
calculations are reasonable.
4.1.3 Montour ACB Index
The ACB Index is used to determine the corrosive
effects of cooling water in a cooling tower with the specific
fill used at PP&L. This index, developed by the Asbestos Cement
Pipe Manufacturers Association, is calculated from the following
expression:
ACB = pH + Log (Calcium X Alkalinity)
with calcium and alkalinity expressed
as mg/£ of calcium carbonate
As an example, the calculations performed for cooling
water at 4 cycles of concentration are presented below. The
results for all the simulations performed are presented in
Appendix I.
K-23
-------
pH = 7.87
Ca = 113 mg/fc (as Ca1 r)
= 283 mg/A (as CaC03)
Alkalinity = 24.8 mg/£ (as C03=)
=41.3 mg/fc (as CaC03)
ACB = 7.87 + Log [ (283) • (41 . 3) ]
= 11.94
4.2. Montour Ash System
For the sluicing runs estimates of the fly ash, bot-
tom ash, pyrite, and water flows were made. Due to the amount
of data made available, some of these flow rates were calculated
by more than one method. Also required were calculations of
the pond evaporation rate, the ash reactivity, and the calcium
removal rate. All of these calculations are presented in this
subsection.
4.2.1 Montour Fly Ash
Fly ash flow rates for one plant were estimated by
the two methods shown below.
Method A
Assume: 9970 efficiency from the electrostatic
precipitator
Typical dust flow was reported as 685
Ib/hr.
Fly Ash = l_E - Dust
- 685 Ib/hr
— — ~~ ~
67815 Ib/hr
30825 kg/hr
....
Ib/hr
K-24
-------
Method B
From data collected from 5/11/76 to 5/13/76
Fly Ash Sluice = 2730 GPM
Weight 70 Solids = 4.79
Fly Ash - (wt. of water)
- (13.6x10' Ib/hr)
= 68400 Ib/hr
= 31100 kg/hr
The second value was used in the simulations, because
it was larger and represented a worse case. The first value
served as a very good check on the reliability of the second
calculation.
4.2.2 Montour Bottom Ash and Mill Rejects
Data were available on the amount of bottom ash and
mill rejects that were sluiced from 5-11-76 to 5-13-76. Sluicing
of these solids is not done on a continuous basis, but is run
intermittently. From these data the solid flow rates were cal-
culated (normalized to a continuous basis) .
Bottom Ash Sluice = 1680 GPM = 8.37 x 105 Ib/hr (using
density of 8.3 Ib/gal)
Weight % Solids =4.46
Bottom Ash = (8.37 x 10 5 Ib/hr)
= 39100 Ib/hr
- 17760 kg/hr
Mill Reject Sluice = 1550 GPM = 7.72 x 10s Ib/hr
Weight % Solids =0.54
K-25
-------
Mill Rejects = (7.72 x 105 Ib/hr) (10o°:5Q 55)
= 4200 Ib/hr
= 1900 kg/hr
For purposes of the simulation the bottom ash and
mill rejects sluicing operations were combined. Since the
mill rejects are actually sluiced at one-tenth the percent
solids used for the bottom ash the total mass flow of the
mill rejects was increased by a factor of ten. This has the
effect of requiring the same amount of water to sluice the
mill rejects as is actually needed and allowing the mill re-
ject sluicing operation to be combined with the bottom ash
sluicing for purposes of simulation.
Bottom Ash + Mill Rejects = 81,100 Ib/hr
36,760 kg/hr
The Montour station has reported that their cooling
towers operate on the average near 2 cycles. The blowdown from
the cooling towers operating at 2 cycles is less than that
which was required to sluice all the ash at 570 solids. In or-
der to have the ash sluiced at 570 solids with the blowdown from
the cooling towers operating at 2 cycles, the amount of bottom
ash was reduced from 36,760 kg/hr to 33,500 kg/hr. This has
the effect of increasing the ratio of reactive fly ash to non-
reactive bottom ash. This is a conservative assumption when
applied to the recycle alternatives presented in this report,
in that it increases the concentration of the ionic species in
the ash pond, and the scaling potential.
4.2.3 Montour Pond Evaporation
The evaporation rate was estimated using a technique
presented in PA-121. Weather conditions were those found in
August 1976.
E = [(73 + 7.3 W) (e - e )]/AH
s a vap
w = wind speed, mph
e = saturation vapor pressure of H20, in. Hg
S
ea = existing vapor pressure of H20, in. Hg
AH = enthalpy of evaporation, Btu/lb
K-26
-------
E = evaporation flux, lb/ft2 - day
E = [{73 + 7.3(5.3)}](.712 - .618)/1050
= .011 lb/ft2 - day
The area of the ash pond has been estimated to be
2.4 x 106 ft2.
Evaporation rate = (2.4 x 106) (.0100)
= 2.4 x 101* Ib/day
=20.1 gpm
= 1.26 £/sec
An evaporation rate of .63 i/sec (10 gpm) was used
in the simulations, since half the pond was used by each unit.
4.2.4 Montour Calcium Removal Rate
Some of the ash sluicing results presented in Appen-
dix I employ sodium carbonate softening in order to reduce cal
cium levels and prevent CaSC\ scale formation in the sluice
line. A slipstream is taken from the recycle line where the
calcium concentration is lowered to 50 mg/£. The equations
used to determine the calcium removal rate and the size of
the slipstream are presented below.
r
- C
- CI
(MW)(H2,0)r
Ca' = Ca" x + Ca(l-x) (3)
CR = x (Ca - Ca")(H20)r/(MW) (4)
Ca = (CI - CR)(MW)/(H20)0 (5)
K-27
-------
where:
Ca = calcium in the pond with softening (mg/£)
Ca' = calcium in the recycle after softening (mg/£)
Ca" = calcium in the slipstream after softening (mg/i)
C = calcium in the slurry before softening (mole/sec)
C' = calcium in the slurry after softening (mole/sec)
CR = calcium removal rate (moles/sec)
CI = calcium input from the ash and the make-up
water (moles/sec)
RS = relative saturation of CaSCs in the
slurry stream before softening
RS" = relative saturation of CaSCU in the
slurry stream after softening
(H20)r = water in the recycle stream (&/sec)
(H20) = water in the sludge (£/sec)
s
MW = atomic weight of Ca (mg/mole)
x = fraction of the recycle used for
slipstream treatment
! j Equation (1) assumes that the activity coefficients
of Ca and SCU remain constant and gives an estimate of the
desired calcium level in the slurry stream in order to elimin-
ate CaSCK scale. Equation (2) calculates the required calcium
level in the recycle to reach C' in the slurry stream. Equa-
tion (3) determines the minimum slipstream that must be treated
to obtain a concentration of Ca' in the recycle. Equation (4)
calculates the calcium removal rate for a given sized slipstream.
Equation (5) is an overall mass balance. In the following,
these equations are used to determine the softening required
for Alternative 1 using 8 cycle cooling tower blowdown as makeup
water.
c' - (57157) 6-749 a>
C' = 2.037 moles/sec
K-28
-------
= 2.037 - .7533
(4.0LxlOH)(143J~ (2)
Ca' = 389 mg/JZ,
The last three equations were combined to form Equation (6)
and solved for the calcium removal rate (CR) , the fraction of
the recycle that must be softened (x) , and the concentration
of calcium in the pond (Ca).
Ca = (CI-MW)+(Ca-*(H20)r)
(H20) (1 + (H20 /(H20q)
o L b
= (.753-4.01x10") + (389-143)
(143) (1 +' (143/16.8) )
Ca = 537 mg/S,
CR = (Ca - Ca')(H20)r/MW (3 + 4)
CR = (537 - 389)(143)/4.01x10"
CR = 0.528 moles/sec
x
x = .304
4.2.5 Montour Ash Dissolution
In order to simulate the ash sluicing system at Montour
it was necessary to obtain information on the different soluble
species in the ash. In order to make an estimate of the amount
of calcium, magnesium, sodium, and sulfate that is leached from
this ash, batch dissolution studies were carried out.
K-29
-------
4.2.5.1 Procedure
Four experiments, were performed using the Montour ash,
two at pH 6 and two at pH 8.1. Under each pH condition the
weight percent solids was varied. In each case, 200 grams of
deionized water were used with either 10 or 20 grams of ash.
Under these conditions, weight percent solids of 4.3% and 9.1%
were attained. In order to maintain pH 6, HC1 was added and the
amount was recorded as a function of time. No acid was necessary
to attain pH 8.1, but the pH was recorded as a function of time.
Figure 4-1 is a plot of HC1 added versus time for both
the 4.8 percent and the 9.1 percent slurries. From this graph
it can be seen that more than half the alkalinity, 53 percent and
64 percent, respectively, is leached from the ash in the first
fifteen minutes. Figure 4-2 is a plot of pH versus time for both
slurries when the pH was allowed to float (pH = 8.1). Within
the first fifteen minutes almost all of the changes in pH had
occurred. This evidence suggests that most of the alkalinity
in the ash, in a five to ten percent slurry, should be leached
out in the first fifteen minutes.
4.2.5.2 Results
The liquors from the four slurries were analyzed for
calcium, magnesium, sulfate, and sodium. These analyses gave
the concentration of each species which was then used to obtain
the reactive amount of each on a weight percent basis. Table 4-3
lists the results obtained under each operating condition. The
values are reported as weight percent of the dry ash.
TABLE 4-3. ASH LEACHING RESULTS
5% Slurry 10% Slurry
Calcium (as Ca)
Magnesium (as Mg)
Sodium + Potassium (as Na)
Sulfate (as SOO
pH = 6
(wt.%)
0.329
0.019
0.041
0.768
pH = 8.1
(wt.%)
0.287
0.017
0.041
0.797
pH = 6
(wt . %)
0.317
0.018
0.040
0.759
pH = 8.1
(wt.%)
0.273
0.016
0.038
0.778
Average
(wt.%)
0.302
0.018
0.040
0.776
K-30
-------
i
oo
04 8 12 16 20 24 28 32 36 40 44 48 52 56 60 64 68 72 76 80 84 88 92 96
Figure 4-1. Fly ash leaching studies at pH 6.0.
02-1165-1
-------
I
OJ
ro
PH
9.
8.
7.0
6.0
5.0
4.0
FLY ASH LEACHING STUDIES AT PH 8.1
4.8% SOLIDS
0 4 8 ' 12 16 20 24 28 32 36 40 44 48 52 56 60 64 68 72 76 80 84 88 92 96
TIME (WIN)
O2-1166-1
Figure 4-2. Fly ash leaching studies at pH 8.1.
-------
This study was performed under idealized conditions
substantially different from those which the ash would see in'
actual operation. These results should represent the worst case
since deionized water was used. The higher ionic strength of the
slurry water in an actual plant would probably leach less from
the ash. ^In order to simulate actual conditions, a bench-scale
ash sluicing experiment has been performed using more representa-
tive water. The results of this experiment are presented in
Appendix D.
The data in Table 4-3 are only estimates of the reacti-
vity of the ash used at Montour. The average values in column
five were used to generate the inputs required for the ash sluic-
ing simulations. The required calculations are presented below.
Convert from wt. 70 to mmoles/gm:
( 302)
Ca = 100 (4!01 x 10-2) = 0.0753 mmoles/gm
MS = 100 (2'.438x 10-2) = °-0074 mmoles/gm
Na = IQO (2.30 x 10-2) = °-0174 ™moles/gm
so" - 100 A'.Mi 10-*) • °-0803
Combine anions and cations to account for all species and convert
to wt. %:
CaSO,, = 0.0753 mmoles/gm = 1.025 wt. %
MgSOit = 0.0055 mmoles/gm = 0.066 wt. %
MgO = 0.0019 mmoles/gm = 0.008 wt. %
NaO = 0.0174 moles/gm = 0.054 wt. %
The rest of the ash was assumed to be inert and equal to 98.847
wt. %.
K-33
-------
5.0 MFC COLSTRIP
Much of the data required to study the water use
system at Colstrip were supplied directly by Montana Power
Company (MFC). Some of the information was calculated. This
section presents the methods used to calculate the data re-
quired to simulate the cooling tower and S02~particulate scrub-
bing systems.
5.1 Colstrip Cooling Towers
Estimates of the ambient air composition and temper-
ature were calculated to simulate the cooling towers at Colstrip
These calculations are presented in this subsection.
5.1.1 Colstrip Climatological Data
Weather data was obtained for Billings, Montana, for
January-October 1976, These data were used to calculate ambi-
ent air conditions. Averages for summer and winter operation
are shown, in Table 5-1.
TABLE 5-1. AVERAGE CLXMATQLOGICAL DATA FOR BILLINGS, MONTANA
Wet Bulb Dry Bulb Relative
Month Temp . , F . Temp . , F_. Humidity, %
Summer 55 69 40
(June, July, Aug.)
Winter 24 30 42
(January, February)
The composition of the input air was calculated on
the basis of the relative humidity. It was assumed that the
ratio of N2 to 02 to C02 remained constant and the change in
the mole fraction of the water changed the mole fraction of
the others. An example calculation for winter operation follows
K-34
-------
Basis: 1 Ib. Bone Dry Air (BDA)
Moles Mole %
N2 .0272 78.8
02 .0072 20.9
C02 1.03 x 10"5 .03
H20 7.77 x IP"5 .23
Total .0345 100.
Moles of st>ecies = moj-e fraction of BDA
Moies or species molecular weight of air
-79 , \= .0272 moles/lb BDA
.( -»lb V.
\29 0 H .-^ /
\*~ y • w 1 ^ mnl *s/
02 = I -^ lb H -0072 moles/lb BDA
Ib mole
= f -QQQ^ \= 1<
\29 0 m • /
\^y-u Ib mole /
C02 = f ife }= 1.03 x 10-5 moles/lb BDA
lb mole
Using the psychrometric chart for a wet bulb temperature of 24°F
and a dry bulb temperature of 30°F, the water content is:
.0014 lb H20/lb BDA or .0023 lb moles H20/lb mole BDA
Therefore:
moles H20 = \ 2°^ = 7.77 x 10'5 lb moles H20/lb BDA
lo.UID
5.2 Colstrip Scrubbing System
Estimates of the flue gas composition and the ash
reactivity were calculated to simulate the scrubbing system
at Colstrip. The methods used to calculate these data and
perform the sample consistency check are presented in this
subsection.
K-35
-------
5.2.1 Colstrip Flue Gas Composition
The average coal analysis reported by MFC was used to
determine the flue gas composition entering the scrubbers. The
particulate loading supplied by MFC of 725 Ib/min ash (design
value) was used with the calculated flue gas composition as in-
put to the model. The coal used and the resulting flue gas com-
positions are shown in Table 5-2.
TABLE 5-2. COAL AND FLUE GAS COMPOSITIONS
Coal
(wt. fraction)
C
H
0
N
S
Moisture
Ash
.579
.038
.119
.0086
.0085
.15
.097
Flue
Gas
(mole 70)
C02
H20
02
N2
SO 2
Flow, acfm
(both units)
14.
9.
3.
72.
2.
1
5
5
8
08
75 x 106
No chloride content of either the coal or flue gas was
reported, but the scrubbing simulations predicted chloride levels
well, indicating that very small amounts, if any, of chloride
enter the scrubbing system by the flue gas.
A firing rate of 400 tons/hr (both units) with 21%
excess air fired was used to determine the gas flow. The flue
gas temperature of 291°F was supplied by MFC.
K-36
-------
5.2.2
Colstrip Effluent Tank Sample Consistency
x
Scrubber Slowdown
7.770 solids
0% Solids
Pond Return
Effluent
Tank
1.36% solids
Slurry to Pond
Let x = flow of scrubber blowdown,
y = flow of pond return, and
z = flow of slurry to pond.
By material balance
x + y = z
By solids balance (assuming negligible solid formation and/or
dissolution):
.077x = .0136z
or
.077
z =
.0136
x = 5.662x
y = z - x = 4.662x
K-37
-------
Now a species balance can be written as follows with
Ci. being the concentration of species i in stream j:
Cix(x) + C.y(y) = Ciz(z)
substituting for z and y,
Cix(x) + Ciy(4.622x) = Ciz(5.662x)
dividing by x f
C.x + 4.662Ciy = 5.662Ciz
Since flows have been eliminated, the sample consistency
may be checked as follows:
ZIn = Cix + 4.662Ciy
EOut = 5.662C.
_L 2
Sin - ZOut
% Error = (TTn + S0ut772
For each species, the concentrations in the three streams are
known, and the 70 error may be calculated. The results of this
calculation are presented in Appendix J.
K-38
-------
5.2.3 Colstrip Ash Dissolution
Three experiments were performed to characterize the
ash from Colstrip coal. The actual samples were taken from
MFC s J.E. Corette plant in Billings, Montana. This plant burns
the same coal as Colstrip. The reason this sample was used was
that a dry ash sample at Colstrip was not obtainable.
•The three experiments involved slurrying 20 grams of
ash in 180 m£ of deionized water with the pH held constant by
HC1 addition. pH values of 4.0, 6.0, and 8.0 were used to
characterize the alkalinity leached from the ash as a function
of pH. After all the alkalinity was leached (no more acid add-
ition required) a sample of the leachate was analyzed for cal-
cium, magnesium, sulfate, and sodium. No appreciable sodium
was found in the Corette ash.
Table 5-3 presents the results of the three experi-
ments performed for the MFC ash. The values for concentration
in the final leachate were used to calculate the leachable spe-
cies as a fraction of the dry ash.
TABLE 5-3. ASH LEACHING RESULTS
Species
CaSO,,
CaO
MgO
Na20
pH=4
(wt%)
0.78
6.8
1.1
« .*.
pH=6
(wt%)
0.85
4.8
.35
"
pH=8
(wt%)
0.81
4.1
.10
"
These results show that calcium dissolution from the
ash is strongly dependent on pH, varying from 6.8% at pH4 to
4.1% at pH8. Magnesium dissolution is considerably smaller but
is still very pH dependent. The amount of sulfate leached from
the ash (assumed as CaSOn) does not vary significantly with pH
but remains at about .8%.
K-39
-------
Since this ash did not actually come from the Colstrip
plant, these results were not used in the simulations. Instead,
the ash composition was calculated based on the overall alkali**
riity required and the design lime addition rate. These calcula-
ted values are shown in Table 5-4 along with the fly ash analysis
provided by MFC.
TABLE 5-4. COMPARISON OF CALCULATED, SAMPLE,
AND MPC FLY ASH REACTIVITY
Species
CaSO,
CaO
MgO
Na20
Sample
(pH=4.0)
(wt%)
0.78
6.8
1.1
~ —
Calculated
(wt%)
--
17.9
1.3
0.035
MPC
Data
(wt%)
21.9
4.95
0.31
The calculated values are much closer to the MPC data
than the sample values. The sample values possibly vary due to
the ash being taken from another plant. The scrubber environ-
ment may also cause increased leaching of the ash species due
to the acidic species sorbed. If the sample value for calcium
is used for the ash, then the lime addition rate to the system
would be 4000 kg/hr (8800 Ib/hr) as opposed to the design rate
of 760 kg/hr (1670 Ib/hr).
K-40
-------
APPENDIX L
ASH CHARACTERIZATION FOR
FOUR CORNERS, BOWEN, AND COMANCHE
FLY ASHES
ABSTRACT
A study under EPA Contract No. 68-02-1319 was made to
perform a preliminary characterization of ash dissolution and
C02 mass transfer occurring in fly ash sluicing operations of
coal-fired steam-electric generating stations. This task was
performed in support for EPA Contract No. 68-03-2339, Water
Recycle/Reuse Alternatives in Coal-Fired Steam-Electric Power
Plants.The three plants studied in the Water Recycle/Reuse
programs are also the sites of the ashes used in this study.
These plants are: (1) Public Service of Colorado, Comanche Plant;
(2) Arizona Public Service, Four Corners Plant; and (3) Georgia
Power Company, Plant Bowen.
Three major types of tests were performed: (1) carbon
dioxide sorption, (2) bench-scale closed-loop sluicing, and
(3) fly ash leaching and batch dissolution. The carbon dioxide
sorption tests were made so that an estimate of the amount of
C02 absorbed from the atmosphere in a settling pond of an ash
sluicing system could be made. Experiments with both agitated
and stagnant systems were made at pH values of 9.0 and 11.0.
The bench-scale closed-loop sluicing tests were made
to characterize the dissolution of fly ashes as a function of
liquor composition. The variables considered include makeup
water composition, sluice tank residence time, carbonate sorption
in the settling pond, sluice-line residence time and ash compo-
sition.
The ash leaching tests were performed to determine
the total alkalinity available in each ash and the rate at
which it dissolves as a function of liquor pH. Batch dissolu-
tion experiments were made to test the effect of liquor compo-
sition on the ash dissolution characteristics. The results are
qualitatively compared to the results of the closed-loop sluicing
tests.
L-l
-------
1.0 INTRODUCTION
This project was performed under EPA Contract No.
68-02-1319 to gather information concerning the chemical charac-
teristics of ash sluicing systems in support for EPA Contract
No. 68-03-2339, Water Recycle/Reuse Alternatives in Coal-Fired
Steam-Electric Power Plants"! In this section, background infor-
mation for the project is first presented followed by a summary
of the program.
1.1 Background
To evaluate the ash sluicing systems studied in the
water recycle/reuse project, dissolution characteristics of the
ash and the degree of carbon dioxide mass transfer between the
atmosphere and process liquor must be known.
This task was performed to provide data concerning
COz mass transfer between process liquors and the atmosphere,
and the dissolution characteristics of the ashes from the plants
under study in the water recycle/reuse program. This informa-
tion is vital in predicting scaling tendencies throughout the
ash sluicing systems for CaC03 , Mg(OH)2, and CaS0lt«2H20. The
following section will present a description of the tests per-
formed and a brief summary of the results.
1.2 Summary
Three major types of experiments were performed using
the ashes from: 1) Public Service of Colorado, Comanche Station;
2) Arizona Public Service, Four Corners Station; and 3) Georgia
Power, Plant Bowen. The three types of tests are: 1) carbon
dioxide sorption, 2) closed-loop sluicing, and 3) leaching and
batch dissolution.
1.2.1 Carbon Dioxide Sorption
The C02 sorption experiments were performed using
liquors with pH values of 9.0 and 11.0 in both stagnant and
agitated vessels. Carbonate concentrations were measured at
various depths in the stagnant runs and at one location for the
agitated runs as a function of time.
L-2
-------
Carbonate concentrations in the pH 11 stagnant runs
were rougly three times as great as in the pH 9 stagnant runs,
indicating a significant enhancement of C02 sorption at higher
pH values. Higher sorption rates were also obtained by instal-
ling an instrument fan to blow air across the surface of the
container to prevent the formation of a C02-poor layer of air
at the air-water interface.
Rapid aqueous dispersion of C02 was observed in the
stagnant experiments as evidenced by the uniformity of carbonate
concentrations at various depth levels. The results of the C02
sorption tests are discussed in greater detail in Section 2.0.
1.2.2 Closed-Loop Sluicing
Bench-scale closed-loop sluicing tests were performed
to characterize the dissolution of fly ashes as a function of
liquor composition. The bench-scale model included a mix tank
where ash and makeup water were added, a sluice line, and a
settling pond.
The closed-loop tests for ashes from all three plants
showed high calcium carbonate relative saturations, indicative
of a potential scaling problem. All three sets of runs indicated
low magnesium concentrations although supersaturation of Mg(OH)2
was noted in the Comanche and Bowen tests due to high hydroxide
concentrations.
Both CaO and CaSOi* were major species dissolving from
the ashes from Comanche and Bowen. Gypsum (CaSCK •2H20) super-
saturation was observed in both cases, probably resulting from
rapid CaO dissolution to provide high calcium levels and disso-
lution of calcium sulfate (in a form other than gypsum) to pro-
vide both calcium and sulfate ions. Both gypsum and magnesium
hydroxide were below saturation levels for all of the runs using
Four Corners fly ash.
Another observation was that lower pH values in the
system (obtained by bubbling C02 into the pond liquor) tend to
enhance CaO dissolution more than CaSO., dissolution. Detailed
discussions of the results and graphs prepared to show the dis-
solution characteristics of each ash are presented in Section
3.0.
L-3
-------
1.2.3 Leaching and Batch Dissolution
Leaching tests were performed to determine the avail-
able alkalinity and the concentrations of chemical species dis-
solved from the ashes at various pH values. Batch dissolution
experiments were conducted to ascertain the effects of liquor
composition on ash dissolution by means of a simple laboratory
test. The pH of the liquor in the leaching studies was main-
tained by adding a measured amount of acid, whereas in the batch
dissolution tests, the pH was not controlled.
The leaching tests indicated that the major source of
alkalinity in all three cases is CaO. Calcium and sulfate were
the major species leached from the Comanche and Bowen ashes.
Sulfate was leached from the Four Corners ash but at a much
smaller level than the other ashes. Also, as in the closed-loop
sluicing tests, lower pH values enhanced the dissolution of cal-
cium more than sulfate.
The batch dissolution characterizations confirmed the
results of the leaching tests in that calcium is the major spe-
cies dissolving. The calcium dissolution produces high pH values
(10-12) and calcium carbonate supersaturation. For the Comanche
ash, calcium sulfate precipitation was evidenced by a decrease
in sulfate concentration, presumably by gypsum formation, al-
though gypsum relative saturations were less than one.
For the Bowen ash, calcium and sulfate were the major
dissolving species although lower concentrations were encountered
than in the closed-loop sluicing experiments. Liquor pH values
as compared to those measured in the closed-loop runs were simi-
lar for the Comanche ash, slightly lower for the Bowen ash, and
slightly higher for the Four Corners ash. A more detailed dis-
cussion of the leaching and batch dissolution results is presented
in Section 4.0.
L-4
-------
2.0 CARBON DIOXIDE SORPTION TESTS
The purpose of this subtask was to determine the
degree of carbon dioxide mass transfer from the atmosphere into
ah aqueous medium. This information is necessary to determine
the effect of carbon dioxide transfer upon the chemical equili-
brium of the sluice water of an ash sluicing facility. Carbon
dioxide mass transfer rates are necessary in the determination
of the pH of the settling pond of an ash sluicing facility.
2.1 Technical Approach
Carbon dioxide sorption characteristics of aqueous
mediums at pH levels of 9 and 11 were determined. The effects
of depth and agitation of the system were also examined. Both
agitated and stagnant systems were used.
The tests were performed in 60.8£ (16 gal) linear
polyethylene containers. Rubber septums were mounted in the
walls of the test containers so that samples could be taken
directly by a syringe. The samples were then injected into a
nondispersive infrared C02 analyzer. The sampling ports of the
stagnant system were positioned 5.1, 20.3, 40.6, and 61.0 cm
(2, 8, 16, and 24 inches) below the surface of the test liquor.
A peristaltic pump was used to mix the test medium of the agitated
system by transferring liquor from the top to the bottom of the
test container at a rate of 250 ml/min. Due to the consistency
of the test medium, the agitated system necessitated only one
sampling port which was 15.2 cm (6 inches) below the surface.
Figure 2-1 is a depiction of the test containers indicating the
positions of the sample ports for both the stagnant and agitated
systems.
2.2 Experimental
2.2.1 pH = 11, Nonbuffered Test Medium
The initial experiment was performed to measure the
carbon dioxide sorption rate into an aqueous medium of deionized
water with the pH adjusted to 11 by sodium hydroxide. The test
solutions were adjusted to this pH at the start of the run, and
the pH of the agitated system was periodically readjusted to this
value. The pH of the stagnant system was not readjusted during
the experiment so as not to disturb the system. Samples were
drawn 17, 48, and 89 hours into the run, and the results appear
in Table 2-1.
L-5
-------
5.1 cm
20.3 cm
40.6 cm
61.0 cm
I
15.2 cm
Stagnant
Agitated
Figure 2-1. Test containers.
-------
TABLE 2-1. C02 SORPTION - NONBUFFERED pH = 11
Time
Initial
17 Hours
48 Hours
89 Hours
STAGNANT SYSTEM
5.1 cm below Surface <.5 mg/1
<.001 mg/l/cm2
20.3 cm below Surface <.5 mg/1
<.001 mg/l/cm2
40.6 cm below Surface <.5 mg/1
<.001 mg/l/cm
61.0 cm below Surface < *.5 mg/1
<.001 mg/l/cm2
3.7 mg/1
.004 mg/l/cm2
.0002 mg/l/cm2/hr
2.8 mg/1
.003 mg/l/cm2
.0002 mg/l/cm2/hr
2.4 mg/1
.003 mg/l/cm2
.0002 mg/l/cm2/hr
2.6 rag/1
.003 mg/l/cm2
.0002 mg/l/cm2/hr
12.6 mg/1
.014 mg/l/cm2
.0003 mg/l/cm2/hr
10.4 mg/1
.011 mg/l/cm2
.0002 mg/l/cm2/hr
11.6 mg/1
.013 mg/l/cm2
.0003 mg/l/cm2/hr
11.1 mg/1
.012 mg/l/cm2
.0002 mg/l/cm2/hr
20.1 mg/1
.022 mg/l/cm2
.0002 mg/l/cm2/hr
19.8 mg/1
.022 mg/l/cm2
.0002 mg/l/cm2/hr
19.7 mg/1
.022 mg/l/cm2
.0002 mg/l/cm2/hr
19.3 mg/1
.021 mg/l/cm2
.0002 mg/l/cm2/hr
AGITATED SYSTEM
<.5 mg/1 5.3 mg/1 15.6 mg/1 37.0 mg/1
<.001 mg/l/cm2 .006 mg/l/cm2 .017 mg/l/cm2/hr .041 mg/l/cm2
.0004 mg/l/cm2/hr .0004 mg/l/cm2/hr .0005 mg/l/cm2/hr
Volume of the test container was 60.83 liters.
Surface area of the test container was 907.92 cm2.
-------
2.2.2 pH = 11, Buffered Test Medium
This experiment was also performed at pH 11, but the
pH was maintained by buffering the test medium with KHaPCK ,
NaXDH, and HCl. Also small instrumentation fans were positioned
to blow air across the surface of the test liquor. The purpose
of these fans was to prevent the creation of a diffusion layer
in the air just above the liquor. If the air above the test
medium is stagnant, the air in contact with the liquor will have
a reduced CCh concentration due to the sorption of COa by the
water. A diffusion layer will then be formed in the air above
the test medium whereby this low CO2 concentration air will be
replenished by the air of higher COa concentration just above
it. Table 2-2 contains the data collected in this experiment.
Samples were taken 24, 48, and 70 hours into the run.
2.2.3 pH = 9, Buffered Test Medium
A pH of 9 was maintained in this run by using sodium
tetraborate decahydrate as a buffer. The instrumentation fans
were again positioned to prevent the depletion of CO2 above the
test liquor. Samples were withdrawn 20, 44, and 50 hours into
the run. The data collected in this experiment appears in
Table 2-3.
2.3 Data
Tables 2-1, 2-2, and 2-3 contain the data collected
during the experiments. The concentration of carbon dioxide
appears first, then the concentration as a function of surface
area, and last, the sorption rate of carbon dioxide expressed
as a concentration flux.
2.4 Conclusions
The results of the nonbuffered pH 11 run indicated
that the rate of carbon dioxide sorption is higher in an agitated
system than in a stagnant system. This would be anticipated
due to more water of lower concentration with respect to COa
being made available at the surface for reaction. The carbon
dioxide concentrations at all the sampling points of the stagnant
system were similar with only a small increase in CO2 concentra-
tions close to the surface of the test solution. A more concen-
trated layer of CO2 was expected near the surface. However
dissolved carbon dioxide diffuses rapidly in aqueous media.
Evidence of fast diffusion of dissolved carbon dioxide was
apparent in each experimental run from the uniformity of concen-
trations in the stagnant systems.
L-8
-------
TABLE 2-2. C02 SORPTION - BUFFERED pH = 11
Time
Initial
24 Hours
48 Hours
70 Hours
STAGNANT SYSTEM
5,1 cm below Surface <.5 mg/1
<.001 mg/l/cm2
20.3 cm below Surface
40.6 cm below Surface
<.5 mg/1
<.001 mg/l/cm2
<-5 mg/1
<.001 mg/l/cm
61.0 cm below Surface <.5 mg/1
<.001 mg/l/cm2
6.7 mg/1
.007 mg/l/cm2
.0003 mg/l/cm2/hr
6.7 mg/1
.007 mg/l/cm2
.0003 mg/l/cm2/hr
7.0 mg/1
.008 mg/l/cm2
.0003 mg/l/cm2/hr
7.5 mg/1
.008 mg/l/cm
.0003 mg/l/cm2/hr
15.8 mg/1
.017 mg/l/cm2
.0004 mg/l/cm2/hr
15.1 mg/1
.017 mg/l/cm2
.0004 mg/l/cm2/hr
15.1 mg/1
.017 mg/l/cm2
.0004 mg/l/cm2/hr
15.4 mg/1
.017 mg/l/cm2
.0004 mg/l/cm2/hr
24.5 mg/1
.027 mg/l/cm2
.0004 mg/l/cm /hr
24.0 mg/1
.026 mg/l/cm2
.0004 mg/l/cm2/hr
24.5 mg/1
.027 mg/l/cm2
.0004 mg/l/cm2/hr
23.6 mg/1
.026 mg/l/cm2
.0004 mg/l/cm2/hr
AGITATED SYSTEM
< .5 mg/1
< .001 mg/l/cm2
7.1 mg/1
.0008 mg/l/cm2
.0003 mg/l/cm2/hr
14.2 mg/1
.016 mg/l/cm2
.0003 mg/l/cm2/hr
23.3 mg/1
.026 mg/l/cm2
.0004 mg/l/cm2/hr
Volume of the test container was 60.83 liters.
n
Surface area of the test container was 907.92 cm .
Instrument fan positioned to blow air across water surface.
-------
TABLE 2-3. C02 SORPTION - BUFFERED pH = 9
Time
Initial
20 Hours
44 Hours
50 Hours
STAGNANT SYSTEM
5.1 cm below Surface <.5 mg/1
<.001 mg/l/cm2
20.3 cm below Surface <.5 mg/1
<.001 mg/l/cm2
40.6 cm below Surface <.5 mg/1
<.001 mg/l/cm
61.0 cm below Surface <.5 mg/1
<.001 mg/l/cm2
2.1 mg/1
.002 mg/l/cm2
.0001 mg/l/cm2/hr
2.6 mg/1
.003 mg/l/cm2
.0001 mg/l/cm2/hr
1.9 mg/1
.002 mg/l/cm2
.0001 mg/l/cm2/hr
1.8 mg/1
.002 mg/l/cm2
.0001 mg/l/cm2/hr
4.1 mg/1
.005 mg/l/cm2
.0001 mg/l/cm2/hr
3.0 mg/1
.003 mg/l/cm2
.0001 mg/l/cm2/hr
4.1 mg/1
.005 mg/l/cm2
.0001 mg/l/cm2/hr
4.8 mg/1
.005 mg/l/cm2
.0001 mg/l/cm2/hr
5.6 mg/1
.006 mg/l/cm2
.0001 mg/l/cm2/hr
4.8 mg/1
.005 mg/l/cm2
.0001 mg/l/cm2/hr
5.0 mg/1
.006 mg/l/cm2
.0001 mg/l/cm2/hr
5.6 mg/1
.006 mg/l/cm2
.0001 mg/l/cm2/hr
AGITATED SYSTEM
<.5 mg/1 1.0 mg/1 2.2 mg/1 2.8 mg/1
<.001 mg/l/cm2 .001 mg/l/cm2 .002 mg/l/cm2 .003 mg/l/cm2
.0001 mg/l/cm2/hr <.0001 mg/l/cm2/hr .0001 mg/l/cm2/hr
Volume of the test container was 60.83 liters.
Surface area of the test container was 907.92 cm2.
Instrument fan positioned to blow air across water surface.
-------
The experiments with a buffered pH 11 and an instru-
mentation fan blowing air across the test liquor surface resulted
in a higher rate of carbon dioxide sorption. After 48 hours,
CCh concentrations at 5.1 cm from the surface were 12.6 mg/J, for
the case without the fan and 15.8 mg/£ for the run with the fan.
The results at the other depth levels also confirm the increased
sorption rate with the fan. The constant replenishment of C02
at the surface of the test medium by the fan accounts for the
higher C02 sorption rate. The C02 sorption rate of the agitated
system under these conditions was lower than the C02 sorption of
both the agitated system of the previous experiment (nonbuffered
pH 11) and the stagnant system under the same condition (buffered
pH 11).
At pH 9, the rate of C02 sorption is quite low. At
normal atmospheric carbon dioxide concentrations, carbon dioxide
will equilibrate with an aqueous medium at a pH of approximately
8.3. Therefore, as the pH of an aqueous medium approaches 8.3,
the rate of carbon dioxide sorption becomes slower due to the
decreased driving force. The results of the pH 9 experiment,
as compared to the other pH 11 experiments, indicate this to be
the case. After approximately 50 hours, the C02 concentration
at 5.1 cm below the surface for the pH 9 system was 5.6 mg/£ as
compared to 15.8 mg/Jl for the buffered pH 11 system. Other depth
levels confirm the decreased sorption rate for the stagnant pH
9 system. The agitated pH 9 system showed a C02 concentration
of only 2.8 mg/£ as compared to 14.2 mg/£ for the buffered pH 11
system after approximately 50 hours.
L-ll
-------
3.0 BENCH-SCALE SLUICING TESTS
Recent emphasis on water recycle/reuse in the electric
power industry has induced power plants to investigate the
feasibility of recycling water which has been used to sluice
coal ash. This system is known as a closed-loop ash sluicing
facility. The engineering involved in designing such a facility
necessitates the prediction of scaling potentials of CaCOs,
Mg(OH)2 and CaS04-2H20 so that the system can be designed to
control possible scaling problems. To predict scaling poten-
tials for these species, the dissolution characteristics of the
coal ash must be known. Therefore, it is important to investi-
gate the ash dissolution characteristics which will be involved
in such an ash handling facility. A bench-scale, closed-loop
ash sluicing facility was built to study the dissolution charac-
teristics of the ash in a system of this type. Measurements
were made to determine the chemical composition of the water at
various locations in the system. The values obtained will aid
in the prediction of scaling potentials for CaCOs, Mg(OH)2, and
CaSOit-2H20 in closed-loop ash sluicing facilities.
3.1 Technical Approach
A depiction of the laboratory scale ash sluicing
facility which was built to simulate a closed-loop ash handling
system is shown in Figure 3-1. Water from the settling pond
was pumped to the mixing tank, a 6-liter (1.6 gal) Plexiglass
cylinder where the coal ash is mixed with the sluice water. The
slurry formed was allowed to flow by gravity from the mixing
tank to the settling pond. An actual ash sluicing facility has
a pipeline to the settling pond through which the ash slurry is
pumped. The bench-scale model has such a pipeline simulation
but this phase of the bench-scale model could not be used con-
tinuously because of flow stoppage due to plugging. The method
of gravity flow from the mixing tank to the settling pond was
adopted because dissolution occurs quickly in the mixing tank.
Therefore, the majority of dissolution occurs in the mixing tank
with only a minor fraction occurring in other portions of the
system.
Batch dissolution studies indicate that the major
portion of the dissolution of the ash occurs within 3 minutes
and the mixing tank has a residence time of over 6 minutes. The
sluice line of the bench-scale model from the mixing tank to the
settling pond, simulating an actual sluice pipeline, was used
during sampling routines to determine the effect upon chemical
composition of the liquor caused by sluicing the ash slurry.
Makeup water was fed into the mixing tank to simulate the re-
plenishment of water lost from the system due to evaporation,
L-12
-------
ASH STORAGE HOPPER
tr1
i
t-1
CO
MAKEUP
WATER
WORM SCREW FEEDER
SETTLING POND
OVERFLOW
Q - SAMPLE POINTS
SLOWDOWN
Figure 3-1. Bench-scale simulation model of ash pond facilities
02- 140 1-1
-------
leakage, or seepage. The chemical composition of this liquor
varied among runs to simulate the composition of actual makeup
water streams of the power plants studied. All of the liquid
flows and/or slurries of the bench-scale model were controlled
by peristaltic pumps. The fly ash was fed into the mixing tank
by a Model SCR-20 precision volumetric screw feeder manufactured
by Vibra Screw, Inc. The settling pond was constructed of fiber-
glass and had a capacity of 454& (120 gal).
The liquor chemical compositions of the system must be
determined at steady-state for values which can be effectively
used in a computer model of the closed-loop ash sluicing facility.
For the system to be at steady-state, the chemical composition
of the liquor entering the mixing tank from the settling pond
must be constant, and the chemical composition of the liquid
flowing from the mixing tank to the settling pond must be con-
stant. The equation
y 4. = (y - y- )e~t^T + y.
•'out wo -'in' -'in
describes the system surrounding the settling pond assuming no
reaction occurs, where yout is the concentration of the pond
overflow, yo is the initial pond overflow concentration, yin is
the inlet concentration to the pond, t is the number of hours
the experiment has run, and T is the residence time of the
settling pond. This equation may be used to make a rough esti-
mate for the time necessary for the system to reach steady-state.
The flow rates were controlled to produce a T value of
10 hours. Chemical composition values should be approximately
at steady-state at 30 hours, or e-30/10, at which time
(Ye - yin)e~t'T reduces to a negligible value such that essen-
tially y0ut = Yin- Tne experiments were conducted for five res-
idence times of the settling pond to more realistically approach
steady-state. More than three residence times will probably be
necessary to achieve steady-state since the incoming stream of
the settling pond is not constant as it is affected by the out-
going stream of the settling pond and precipitation and/or disso-
lution in the system.
Only one parameter was changed between runs using fly
ash from any one of the plants. No more parameters were changed
between runs so that correlations among runs could be made with
confidence.
L-14
-------
The samples collected were immediately measured for
pH, temperature, and EMF using calcium and divalent cation-
specific electrodes. Ionic activities of calcium and magnesium
were calculated from graphs obtained from values gathered by
measuring EMF values of calcium standards with the calcium and
divalent cation-specific ion electrodes. Calcium and magnesium
ionic activities along with the relative saturation of the spe-
cies under investigation were also calculated by the chemical
equilibrium computer program. Direct analytical methods were
used to measure calcium, magnesium, sodium, total sulfate, and
chloride concentrations of the samples which were collected.
These samples were filtered, acidified, and diluted when collec-
ted. The carbonate concentrations of the various streams were
measured by nondispersive infrared analysis. The carbonate sam-
ples collected were preserved using a NH^OH-EDTA buffer system.
The tests were performed with fly ash collected from
three coal-fired electric generating plants:
1) Comanche Steam Electric Station of
Public Service Company of Colorado,
2) Four Corners Power Plant of Arizona
Public Service, and
3) Plant Bowen of Georgia Power Company.
3.2 Experimental
3.2.1 Comanche Steam-Electric Station
Five experiments were performed using fly ash from
the Comanche Steam-Electric Station. The initial experiment
was performed using a makeup water approximating a probable
makeup water stream at the Comanche plant, cooling tower blow-
down.
The sulfate concentration of the makeup water was
doubled for the second run. All other parameters remained the
same.
L-15
-------
The pH levels obtained during the first two runs were
above the anticipated pH level of an ash sluicing system. An
experimental run at a different pH value was desirable so that
correlations could be drawn upon chemical compositions at varying
pH values. Therefore, for the third run, a fan was used to blow
air across the surface of the settling pond to try to increase 1
the rate of carbon dioxide sorption and lower the pH of the sys-
tem. The makeup water used during this run was identical to the
makeup water of the first run.
The parameter change in the fourth experiment was to
reduce the residence time of the mixing tank from 6.4 minutes
to 3.2 minutes. The purpose of this adjustment was to determine
if a decrease of the reaction time in the mixing tank would alter
the chemical composition of the system.
A significantly lower pH value was obtained during
the fifth run by bubbling an air stream spiked with carbon di-
oxide through the settling pond. The bubbler system was sub-
merged only 2.5-7.6 cm (1-3 in) below the surface of the liquor
of the settling pond to prevent agitation of the system which
would probably cause additional ash dissolution in the settling
pond. The carbon dioxide flow into the system was 1.0 £/min.
3.2.2 Plant Bowen
The first experiment using Plant Bowen fly ash was
performed at the conditions stated in Section 3.1, Technical
Approach, and a makeup composition the same as the makeup water
used in Comanche Steam-Electric Station Runs 1, 3, 4, and 5.
The makeup water composition was not changed so that dissolution
characteristics could be correlated among ashes from different
plants, if deemed necessary.
The second experiment was run using the same makeup
water composition as in the second Comanche fly ash run, which
was twice the sulfate concentration of the makeup water of the
first run. For the third Bowen run, water with a composition
similar to that of the Bowen cooling tower blowdown was used
as makeup.
The carbon dioxide bubbling system was used to reduce
the pH of the settling pond for the fourth experiment using
Plant Bowen coal ash. Carbon dioxide was bubbled through the
settling pond at a rate of 0.5 £/min.
L-16
-------
3.2.3 Four Corners Power Plant
The first run using coal ash from the Four Corners
Power Plant was performed using the same makeup water composition
that was used for the first runs of the other plants. The make-
up water composition simulated the chemical composition of the
cooling tower blowdown stream of Comanche Steam-Electric Station.
Other conditions of the experiments using Four Corners Power
Plant coal ash were outlined in Section 3.1, Technical Approach.
The makeup water composition of the second experiment
simulated the ash pond effluent of the Four Corners Power Plant.
This stream would be a probable makeup water source for a sluic-
ing facility at this plant. No other conditions were varied.
The third run had a makeup water composition which was
the same as that of the first run using Four Corners Power Plant
coal ash. The carbon dioxide bubbler was installed and carbon
dioxide was bubbled through the surface of the settling pond at
a rate of 0.5 £/min.
3.3 Results
The experimental data gathered from the ash dissolution
characterizations at steady-state appears in Appendix LA of this
report. Operating conditions of each experiment also appear in
these tables. Tables B-l, B-2, and B-3 in Appendix LB of this
report contain the calcium concentrations of the sampled streams
of each experiment. These values were used to determine if each
run was at steady-state. Also included in Appendix LB are the
parameters used to correlate the data from the three sets of runs.
3.4 Conclusions
3.4.1 Comanche Steam-Electric Station
Two plots were drawn to describe the dissolution of
calcium in the mix tank of the bench-scale closed-loop ash sluic-
ing facility. Figure 3-2 shows a graph describing the calcium
dissolution as a function of the activity of the calcium ion.
Figure 3-3 depicts the dissolution of calcium as a function of the
activity product of calcium sulfate. It is important to note that
towards the upper end of the graph (higher dissolution rates of
calcium) the curve should begin to flatten out as the solubility
product constant of calcium sulfate dihydrate (KSp = 2.4 x 10~
at 25°C) is approached due to the precipitation of gypsum.
L-17
-------
.25 .50 .75 1.00 1.25 1.50 1.75
Calcium Dissolution Rate in Mix Tank (MMole/Min)
Figure 3-2. Comanche steam-electric station calcium dissolution
rate in mix tank versus aCa in mix tank.
L-18
-------
10"
Sfl 1 n—
(J 1U
O5
0
i*
10-
.25 .50 .75 1.00 1.25 1.50 1.75
Calcium Dissolution Rate in Mix Tank (MMole/Min)
Figure 3-3. Comanche steam-electric station calcium dissolution
rate in mix tank versus a0 acri in mix tank.
Ca S0i|
*Point inconsistent due to sulfate concentration not being at
steady-state.
L-19
-------
Sulfate dissolution was described basically in the
same manner. Figure 3-4 is a plot of sulfate dissolution versus
the activity of sulfate. Figure 3-5 shows the sulfate dissolu-
tion as a function of the activity product of calcium sulfate.
The lesser slope of the graph of Figure 3-5 as compared to the
graph of the calcium dissolution versus the activity product of
calcium sulfate (Figure 3-3) indicates that sulfate arrives in
the system primarily through the dissolution of calcium^sulfate,
whereas calcium arrives through the dissolution of calcium oxide
as well as calcium sulfate.
Calcium carbonate precipitation in the mix tank as a
function of the activity product of calcium carbonate in the
mix tank is shown in Figure 3-6. The rate of precipitation of
calcium carbonate was calculated as the rate of change of car-
bonate across the mix tank. This curve occurs in the range of
relative saturation of 5-25 for calcium carbonate. The high
relative saturation values are not primarily due to the carbon-
ate concentration which is low, but to the rapid dissolution of
calcium oxide and calcium sulfate and the resulting high calcium
concentration which produces activity products of calcium car-
bonate well above the solubility product constant of calcium
carbonate. The precipitation rate of calcium carbonate is not
sufficient to reduce the concentration of calcium in the system
such that the relative saturation of calcium carbonate would be
close to one. The point associated with the first run is high
and inconsistent with the others because the carbonate concen-
tration of the settling pond had not reached steady-state.
Analysis of the tap water used to fill the pond provided a
carbonate concentration of 0.38 mmoles/&. This value is con-
siderably higher than the carbonate concentrations of the ex-
periments with the exceptions of the first run and the fifth
run which had the carbon dioxide bubbler installed. The carbon-
ate concentration of the tap water approximates the final carbon-
ate concentration of the settling pond of the first run indicating
that calcium carbonate precipitation had not diminished the
carbonate concentration to steady-state conditions.
Although the mixing tank is supersaturated with respect
to magnesium hydroxide, this is not due to the presence of high
concentrations of magnesium but to the high concentrations of
hydroxide. Therefore, magnesium hydroxide will precipitate even
with values of magnesium which are too low to calculate accurate
dissolution rates. However, the amount of magnesium hydroxide
which would_precipitate is insignificant. The fifth run had a
lower pH which increased the magnesium dissolution. The streams
were still supersaturated with respect to Mg(OH)2 but the amount
of magnesium hydroxide which would precipitate still would not
be substantial.
L-20
-------
© 2
10"
c
(TJ
f->
X
•H
z
C
10-
O
i*
10-
©5
-.75 -.50 -.25 0.0 .25 .50 .75 1.00
Sulfate Dissolution Rate in Mix Tank (MMole/Min)
Figure 3-4. Comanche steam-electric station sulfate dissolution
rate in mix tank versus agQ in mix tank.
*Point inconsistent due to sulfate concentration not being at
steady-state.
L-21
-------
10'
,-5-
O
10-
1*
10"
-.75 -.50 -.25 0.0 .25 .50 .75 1.00
Sulfate Dissolution Rate in Mix Tank (MMoLe/Min)
Figure 3-5. Comanche steam-electric station sulfate dissolution
rate in mix tank versus ar acn in mix tank.
La oUi^
*Point inconsistent due to sulfate concentration not being at
steady-state.
L-22
-------
10"
10"
c
G L*
•50 1.00 1.50 2.00 2.50 3.00 3.50
CaCOs Precipitation Rate in Mix Tank (MMole/Min)
Figure 3-6.
Comanche steam-electric station CaCO 3 precipitation
rate in mix tank versus &raarQ in mix
*Point inconsistent due to carbonate not being at steady-state.
L-23
-------
The fourth run using Comanche Steam-Electric Station
coal ash had a reduced residence time in the mix tank, from 6.4
minutes to 3.2 minutes. The sulfate dissolution in the mix tank
during the fourth run was significantly higher than the sulfate
dissolution in the first and third runs although all three runs
had similar parameters. The fourth run sulfate dissolution was
higher than the sulfate dissolution of the second run which had
a higher sulfate concentration in the makeup water. Specifically,
the first and third runs had negative dissolution rates, about
-0.6 mmole/min. This was due primarily to calcium sulfate dihy-
drate precipitation. The sulfate dissolution of the fourth run
was a positive rate of 0.88 mmole/min. The increase in dissolu-
tion rate of the fourth run could be misleading in that the in-
crease was probably not an increase in the dissolution rate of
sulfate but a decrease in the precipitation rate of gypsum due
to the shorter residence time in the mix tank.
The dissolution rate of calcium in the mix tank of
the fourth run was greater than the dissolution rate of calcium
in the first and third runs but the difference was not as signi-
ficant as the increase in the dissolution rate of sulfate. This
is due to the fact that there are two sources of calcium avail-
able for dissolution, CaO and CaSOi,. The two sources allow cal-
cium to enter the system at a faster rate than sulfate. This
accounts for the positive calcium dissolution rates of the first
three runs even though the dissolution rates of sulfate were
negative. The fact that all the runs had higher calcium disso-
lution rates than sulfate dissolution rates supports the fact
that CaO and CaSOij are both readily available sources of calcium.
The fifth run also had higher calcium and sulfate dis-
solution rates than the first three runs. The increase in disso-
lution was probably caused by the lower pH of the system. Noting
that the calcium dissolution rate approximates that of the fourth
run, the sulfate dissolution of the fifth run is lower than the
sulfate dissolution of the fourth run. This could be due to ad-
ditional gypsum precipitation because of the longer residence
time of the mix tank in the fifth run. However, if this were
the cause, the calcium dissolution rate would not approximate
that of the fourth run due to gypsum precipitation. The reason
that there is a greater difference between the calcium dissolution
rate and the sulfate dissolution of the fifth run as compared to
the fourth run, is that lower pH values enhance the potential to
dissolve calcium oxide more than calcium sulfate.
L-24
-------
There are indications that the first run using Comanche
Steam-Electric Station coal ash was not at steady-state with re-
spect to sulfate even though sampled at three residence times of
the settling pond as described in Section 3.1. The first run
had fresh water in the settling pond whereas the water of the
settling pond was not renewed between runs using the same coal
ash. The fresh water has a lower pH, and, as stated in the
previous paragraph, there is a greater potential to dissolve
CaO as opposed to CaS04 at lower pH values. Initial higher
concentrations of calcium would tend to impede calcium sulfate
dissolution, requiring more time to reach steady-state.
Table B-4 of Appendix LB of this report contains the
information which was used to derive the graphs which describe
the dissolution of Comanche Steam-Electric Station coal ash.
3.4.2 Plant Bowen
Four graphs were drawn to describe the dissolution and/
or precipitation rates occurring in the mix tank of the four runs
using Plant Bowen coal ash. Figure 3-7 is a plot of the calcium
dissolution in the mix tank versus the activity of the calcium
in the mix tank, and Figure 3-8 is a plot of the sulfate disso-
lution in the mix tank versus the activity of sulfate in the mix
tank. Figure 3-9 is a plot of the sulfate dissolution of the mix
tank, which approximates the calcium sulfate dissolution since
calcium sulfate is the major source of sulfate from the ash, ver-
sus the activity product of calcium sulfate. The entire curve is
above the solubility product constant of calcium sulfate dihy-
drate (Ksp = 2.4 x 10~5 at 25°C). Calcium sulfate dihydrate
would not dissolve at these conditions yet the curve indicates an
increase in dissolution above the solubility product constant of
calcium sulfate dihydrate. An explanation is that the calcium
sulfate of the ash is not in the form of calcium sulfate dihy-
drate. There are several other forms of calcium sulfate which
are considerably more soluble than calcium sulfate dihydrate.
Two are a-CaSCs and a-CaSCH -%H20. These two forms have solubili-
ties of 0.63 to 0.8 grams per 100 milliliters of solution, where-
as calcium sulfate dihydrate has a solubility of 0.2 to 0.22
grams per 100 milliliters of solution.1 These values and Figure
3-9 indicate that the calcium sulfate is present in one of the
more soluble forms which dissolves in the mix tank. Solids
analysis by X-ray diffraction was not attempted since the sulfate
solids represent only a small fraction of the total solids pre-
sent.
1GM-061: Gmelin, Gmelin Handbuch de anorg. Chemie, 8. Auflage,
Calcium, Tiel B. Lieferung 3, (1961).
L-25
-------
10
6 -
a
u
n
0 1.00 2.00 3.00 4.00 5.00 6.00 7.00
Calcium Dissolution Rate in Mix Tank (MMole/Min)
Figure 3-7. Plant Bowen calcium dissolution rate in
mix tank versus ac in mix tank.
L-26
-------
10
2
c
o
C/5
CO
©
©
Q i*
0 .50 1.00 1.50 2.00 2.50 3.00 3,i()
Sulfate Dissolution Rate in Mix Tank (MMole/Min)
Figure 3-3. Plant Bowen sulfate dissolution rate in
mix tank versus aon in mix tank.
*Point inconsistent due to sulfate concentration not being at
steady-state.
L-27
-------
10
6-
4-
ac
c
o
iH
X
2-
0 .50 1.00 1.50 2.00 2.50 3.00 3.50
Sulfate Dissolution Rate in Mix Tank (MMole/Min)
Figure 3-9. Plant Bowen sulfate dissolution rate in
mix tank versus aCaaso in mix tank.
*Point inconsistent due to sulfate concentration not being at
steady-state.
L-28
-------
Figure 3-10 is a plot of the calcium carbonate preci-
pitation rate in the mix tank versus the activity product of
calcium carbonate. As in the Comanche Steam-Electric Station
ash dissolution characterization experiments, the curve appears
in the range of relative saturation values for calcium carbonate
of 10-30. The precipitation of calcium carbonate is insufficient
to reduce the ionic activity of calcium to a level where the
solubility product of calcium carbonate will approach the solu-
bility product constant of calcium carbonate.
The majority of the streams were supersaturated with
magnesium hydroxide in all of the experiments with Plant Bowen
coal ash. The activity product of magnesium hydroxide was above
the solubility product constant of magnesium hydroxide due to
high hydroxide concentrations even with low magnesium concentra-
tions. Therefore, even though many streams were supersaturated
with respect to magnesium hydroxide, there is no potential scaling
problem because the magnesium concentrations are too low to allow
substantial magnesium hydroxide precipitation.
Table B-5 of Appendix LB of this report contains the
information used to derive the graphs which describe the dissolu-
tion of Plant Bowen coal ash.
3.4.3 Four Corners Power Plant
Three experiments were performed with coal ash from
the Four Corners Power Plant. However, dissolution rates were
too low to plot any data over a sufficient range. If this data
were plotted over an insufficient range, erroneous graphs could
possibly be drawn. Therefore, since the dissolution of the ash
could not be illustrated with certainty, graphs were not used
to depict the dissolution of fly ash from Four Corners Power
Plant. The only species that could possibly present a scaling
problem is calcium carbonate. All the streams in all of the
experiments were supersaturated with respect to calcium carbonate.
This is due primarily to the extremely low solubility of calcium
carbonate. The concentrations of calcium were relatively low as
compared to calcium concentrations from other plants. Carbonate
concentrations were somewhat higher due apparently to carbonate
available from the fly ash. Still the low concentrations and
meager dissolution rates of calcium prevent the creation of a
scaling problem. All of the streams of all of the runs were sub-
saturated with respect to calcium sulfate dihydrate and magnesium
hydroxide.
L-29
-------
10
c
10"
-.25 0 .25 .50 .75 1.00 1.25 1.50
CaCOi Precipitacion Rate in Mix Tank (MMole/Min)
Figure 3-10. Plant Bowen CaCO3 precipitation rate in
mix tank versus aa in mix tank.
*Point inconsistent due to carbonate concentration not being at
steady-state.
L-30
-------
4.0 BATCH DISSOLUTION AND LEACHING TESTS
. In conjunction with the bench-scale studies, batch
dissolution characterizations and leaching studies of the coal
ash were conducted. The leaching studies were performed to de-
termine the amount of alkalinity available from each of the coal
ashes tested and the concentrations of several chemical species
dissolved from the coal ashes at various pH conditions. The
batch Dissolution characterization experiments were conducted
to gain information concerning liquor composition effects in
simple laboratory tests which may correlate to the information
gathered from the ash dissolution characterization studies so
that information desired for computer modeling could be derived
by this simpler method. Liquor pH values were adjusted periodi-
cally in the leaching experiments but no pH control was exerted
for the batch dissolution experiments.
Realistically, these tests will not be able to provide
the total information necessary for a computer model of a closed-
loop ash sluicing facility. But a favorable correlation of the
batch dissolution characterizations could reduce the number of
runs of the bench-scale closed-loop ash sluicing facility neces-
sary to model an ash sluicing facility. Even though the results
of the batch dissolution characterizations and the ash dissolu-
tion characterizations will be compared, values obtained by the
batch dissolution characterizations were not used in the computer
models. The results of the experiments did not cover a large
enough variation in operating and ash parameters to justify
correlation by a computer model.
The purpose of this subtask was not to ascertain if
favorable correlations exist such that the information gathered
through the batch dissolution characterizations could justifi-
ably be used to computer-model a closed-loop ash sluicing faci-
lity. The batch dissolution characterizations were conducted to
gather the information available by performing experiments of
this type and to have a data bank available so that correlations
could be attempted between the two sources of information.
4.1 Technical Approach
4.1.1 Ash Leaching
To determine the available alkalinity and the rate of
dissolution of that alkalinity from the coal ashes, the coal
ashes were subjected to liquors of varying pH. The pH of the
L-31
-------
liquor was held constant by the addition of acid. The equiva-
lents of acid added were recorded as was the time of the addi-
tion. In this manner, the available alkalinity of the respec-
tive fly ashes, present predominantly as CaO and MgO, can be
determined as a function of pH and time. Analyses for calcium,
magnesium, sodium, potassium and total sulfate were performed
on resulting liquors which were collected after all available
alkalinity had been leached.
The ash leaching studies as well as the batch disso-
lution characterizations were performed with fly ash collected
from the following three plants:
1) Comanche Steam-Electric Station of
Public Service Company of Colorado,
2) Plant Bowen of Georgia Power Company, and
3) Four Corners Power Plant of Arizona Public
Service.
4.1.2 Batch Dissolution
The batch dissolutions were performed by varying the
chemical composition of the test liquors to which the coal ashes
were subjected. pH, EMF readings of calcium and divalent cation
specific ion electrodes, and time were monitored continuously
using a chart recorder. These values are presented in Appendix
LC of this report. The EMF values obtained were used to calcu-
late the ionic activities of calcium and magnesium. EMF read-
ings were taken of standard solutions and used to derive stan-
dard curves. The EMF readings recorded during the tests were
then used to calculate ionic activities from these graphs. Each
final sample was analyzed for calcium, magnesium, sodium, total
sulfate and chloride. Samples were also taken at 10 and 20
minutes into the experiment, and analyzed for calcium and mag-
nesium. The temperatures of the solutions were also recorded to
determine if the reactions involved were endothermic or exother-
mic to such a degree as to effect the temperature of the system.
With the information gathered, the dissolution properties of
calcium and magnesium of the coal ashes can be interpreted as a
function of chemical composition and pH.
L-32
-------
4.2 Experimental
4.2.1 Ash Leaching
The leaching experiments were performed in 400-ml
beakers. A magnetic stirring bar and a mechanical stirrer were
used to mix the slurry. The experiments which called for a 7%
slurry were a combination of 186 grams of water and 14 grams of
the coal ash. A 27» slurry was formed by combining 196 grams of
water and 4 grams of the coal ash. The electrodes were positioned
in the test liquor, the coal ash added and, as quickly as possible
thereafter, the pH was adjusted by addition of hydrochloric acid
of known concentration. The test reactor was covered to prevent
the effects on pH caused by the sorption of carbon dioxide. The
desired pH was maintained by the repeated addition of acid as
necessary. The amount of acid added and the time of addition
were recorded with each pH adjustment. The experiment was con-
ducted until all available alkalinity had been leached. Fly ash
from the Four Corners Power Station was examined at pH levels
of 3.0, 6.0, and 8.5 with a 1°L slurry of ash and water, and at
pH levels 3.0 and 8.5 with a 2% slurry. pH values of 6.0, 8.5
and 10.4 were used to describe the Plant Bowen fly ash using a
7% slurry. Fly ash collected from Comanche Steam-Electric
Station was examined using a 770 ash slurry and pH values of
6.0 and 8.5.
4.2.2 Batch-Dissolution
Five chemical compositions were used as the aqueous
media in the experiments for the batch dissolution characteriza-
tions. In this report, the chemical compositions are identified
by run number as follows:
Run No. 1 - Deionized water
Run No^ 2 - A 1:1 mixture of deionized water
and the resultant liquor of Run
No. 1. (Therefore, the initial
concentrations of the second run
were one-half of the final concen-
trations of the first run.)
Run No. 3-25 ppm carbonate (0.417 mmole/£)
500 ppm sulfate (5.21 mmole/£)
259 ppm sodium (11.3 mmole/&)
L-33
-------
Run No. 4 - 25 ppm carbonate (0.417 mmole/£)
2500 ppm sulfate (26.0 mmole/£)
1220 ppm sodium (52.8 mtnole/ £)
Run No. 5 - 100 ppm carbonate (1.67 mmole/£)
500 ppm sulfate (5.21 mmole/£)
316 ppm sodium (13.8 mmole/jj.)
Each coal ash was examined using each of the test liquors listed
above monitoring pH and EMF values for calcium and divalent cation
specific ion electrodes using a chart recorder. Separate runs
were necessary for each species so that the information could be
continuously recorded using the strip chart recorder. The 10-
minute sample was taken from the calcium specific ion electrode
run, the 20-minute sample from the divalent cation specific ion
electrode; and the 30-minute sample was taken at the end of the
pH run. The samples were taken in this manner to interface the
three runs necessary. For each of the three ashes tested, a 1°L
mixture of the ash and the test liquor was used. During the di-
valent cation experiments, the temperature of slurry was measured
and recorded. Calcium standards were analyzed after each run
involving a specific ion electrode so that the EMF values ob-
tained could be used to draw graphs from which ionic activities
for calcium and magnesium could be determined.
4.3 Results
Tables 4-1, 4-2, and 4-3 contain the results of the
ash leaching studies. Table 4-4 presents the amount of leach-
able species at pH 6 as a percent of the ash. The results of
the batch dissolution studies are shown in Table 4-5.
4.4 Conclusions
4.4.1 Ash Leaching
4.4.1.1 Comanche Steam-Electric Station
The leaching studies of coal ash from Comanche Steam-
Electric Plant show that calcium and sulfate are the major species
leached. Magnesium and sodium are less significant constituents
(see Table 4-1). The sulfate to calcium ratio of the leaching
results is 0.172 for pH 6.0 and 0.154 for pH 8.5. These values
indicate that the calcium dissolving into the system arrives
predominantly as CaO, the remaining calcium arriving as
L-34
-------
TABLE 4-1.
RESULTS OF CHEMICAL ANALYSIS FROM LEACHING OF ASH
SAMPLES AT CONSTANT pH FOR COMANCHE STEAM-ELECTRIC
STATION
Free. Ash1
pH 6.0
(mmole/2.) 3
Free. Ash1
pH 8.5
(mmole/£)3
Calcium
Magnesium
Sodium
Potassium
Sulfate
53.9
3.5
1.3
--
9.28
42
1
1
6
.2
.6
.0
.11
.50
Maintained pH of 6.0 by adding HCl.
Maintained pH of 8.5 by adding HCl.
3A11 values represent mmole/& of soluble species in leachate
after 14 grams of ash were leached in 186 grams of deionized
at constant pH.
TABLE 4-2.
RESULTS OF CHEMICAL ANALYSIS FROM
LEACHING OF ASH SAMPLES AT CONSTANT
pH FOR PLANT BOWEN
Free. Ash1
pH 6.0
(mmole/£)'t
Prec. Ash 2
pH 8.5
(mmole/ £,) **
Prec. Ash 3
pH 10.4
(mmole/ #,) **
Calcium
Magnesium
Sodium
Potassium
Sulfate
24.
.
3.
.
10.
2
23
0
90
0
14.
.
1.
.
9.
2
06
1
72
22
10
1
7
.5
.01
.3
.56
.71
Maintained pH of 6.0 by adding HCl.
Maintained pH of 8.5 by adding HCl.
Maintained pH of 10.4 by adding HCl.
''All values represent mmole/A of soluble species in leachate
after 14 grams of ash were leached in 186 grams of deionized
water at constant pH.
L-35
-------
TABLE 4-3 RESULTS OF CHEMICAL ANALYSIS FROM LEACHING
OF ASH SAMPLES AT CONSTANT pH FOR FOUR CORNERS
POWER STATION
Calcium
Magnesium
Sodium
Potassium
Sulfate
Free. Ash1
pH 3.0
(mmole/£)6
15.5
.23
.52
.12
1.2
Free. Ash2
pH 6.0
(mmole/£)s
13.4
.10
.48
.06
.31
Free. Ash3
pH 8.5
(mmole/£)6
11.7
.07
.48
.04
.56
Free. Ash4
pH 3.0
(mmole/JO 7
4.52
.07
.12
.04
.08
Free. Ash5
pH 8.5
(mmole/£)7
3.02
.01
.08
<.02
.01
Maintained pH of 3.0 by adding HC1.
Maintained pH of 6.0 by adding HC1.
Maintained pH of 8.5 by adding HC1.
Maintained pH of 3.0 by adding HC1.
Maintained pH of 8.5 by adding HC1.
All values represent mmole/£ of soluble species in leachate after 14 grams of
ash were leached in 186 grams of deionized water at constant pH.
All values represent mmole/£ of soluble species in leachate after 4 grams of
ash were leached in 196 grams of deionized water at constant pH.
TABLE 4-4. LEACHABLE SPECIES FROM ASH SAMPLES AT pH 61
Species2
CaO
MgO
Na20
K20
CaSOi,
Comanche
(wt. % of ash)
3.
0.
0.
-
1.
3
2
4
-
7
Bowen
(wt. % of ash)
1.
0.
0.
0.
1.
0
01
10
06
8
Four Corners
(wt . % of ash)
1.
0.
0.
0.
0.
0
006
04
004
06
'Experiments performed with a 1% slurry with pH adjusted period-
ically to 6.0.
A11 sulfate is assumed to enter the system as
2
L-36
-------
TABLE 4-5. ASH DISSOLUTION CHARACTERIZATIONS - BATCH DISSOLUTIONS
r1
LO
Calcium
Magnesium
Sodium
Sulfate
Chloride
Carbonate
pH
£» Or-
T,°F
aCa
*Kg
aCa>
3Mg'
Calcium
Magnesium
Sodium
Sulfate
Chloride
Carbonate
pH
T. F
aCa
aKg
aCa'
a*g'
Calcium
Magnesium
Sodium
Sulfate
Chloride
Carbonate
pH
T, °F
aCa
*Mg
aCa'
•Kg'
Explanation of
C - denotes
F - denotes
B - denotes
C-1C30 min)
6. 18
0.002
0.78
0.666
0.259
0.14
11.9
74.5
1.67 x 10" '
3.3 x 10""
3.20 x 10"'
4.15 x 10"'
B-1OO min)
10.3
0.010
1.19
9.38
0.586
0.16
10.8
76.5
1.97 x 10"'
7.8 x 10""
3.95 x 10" 3
3.65 x 10"'
F-l(30 min)
4.43
0.008
0.59
0.972
0.045
0.21
11.7
74.2
1.10 x 10" 3
9.5 x 10""
2.43 x 10"3
2.22 x 10"'
Satrple Headings
Comanche Steam-Electric
Four Corners Power Plant
Plant Bowen
C-2(30 min)
6.73
0.004
1.24
0.649
0.397
0.08
12.0
75.3
1.60 x 10"'
8.0 x 10"*
3.32 x 10"3
6.73 x 10"7
B-2(30 min)
15.1
0.014
1.79
14.1
0.952
0.26
10.8
77.8
3.46 x 10"'
3.2 x 10""
5.05 x 10"3
4.51 x 10"6
F-2(30 min)
4.71
0.005
0.80
1.61
0.090
0.16
11.6
75.8
1.05 x 10" !
8.0 x 10""
2.51 x 10" 3
1.47 x 10"6
Run #1 -
Station Run *2 "
Run #3 -
C-3(30 min)
2.12
0.002
11.5
0.627
0.138
0.20
12.0
75.5
7.0 x 10""
7.0 x 10""
1.04 x 10" 3
3.40 x 10"'
B-3(30 min)
10.6
0.010
11.4
14.8
0.327
0.28
10.8
72.7
1.35 x 10"3
1.29 x 10"3
3.41 x 10" 3
3.17 x 10"6
F-3(30 rain^
3.98
0,004
10.8
6.00
0.115
0.11
11.7
74.2
9.0 x 10""
3.5 x 10""
1.61 x 10" 3
9.47 x 10"'
C-4<30 min)
0.96
0.002
56.5
18.1
0.138
0.24
12.0
74.5
4.2 x 10"*
9.5 x 10~*
2.48 x 10""
2.62 x 10"'
B-4(30 min)
10.6
0.012
52.9
35.4
0.358
0.32
10.7
73.3
3.80 x 10" '
2.36 x 10" 3
2.79 x 10"6
F-4(30 min)
5.15
0.003
53.8
26.3
0.110
0.15
11.8
74.9
8.4 x 10""
9.9 x 10"*
1.22 x 10" 3
4.61 x 10"'
7% Slurry D.I. Water 25
77, Slurry 1:1 D.I
Liquor
77, Slurry 25 ppm
-l»t C-T 0 c ___
. Water - 100
from Run #1 500
COj-500 ppm SO- 2500
C-5(30 min)
4.26
0.002
13.7
4.36
0.144
0.16
11.9
75.0
4.0 x 10""
1.68 x 10"'
1.76 x 10"'
3.70 x 10"'
B-5(30 min)
9.62
0.008
13. 5
14. 9
0.400
0.27
10.8
74.0
1.28 x 10" '
1.S2 x 10"3
3.06 x 10"'
2.50 x 10"6
F-4(30 min)
3. 20
0.010
13.4
5 . 42
0.118
0.28
11.7
76.0
6.1 x 10""
6.7 x 10"'
1.28 x 10"3
2.24 x 10"6
ppm CDs = 0.417 rar.ole/^ CO j
pprn CO 5 = 1.67 mnole/i CO,
ppn SO* - 5.21 mraole/t SO.
ppm SO,, - 26.0 mmole/l S0»
• H ' • :j
-------
Since the magnesium and sodium concentrations are not signifi-
cant, the sulfate concentration is the result mainly of CaSCs
dissolution as shown in Table 4-4.
The plot of meq acid/g fly ash versus time (Figure 4-1)
obtained from the data gathered during the leaching studies of
Comanche Steam-Electric Station ash show that at pH 6.0, 1.60 meq
OH~/g ash or 22.4 milliequivalents of hydroxide were leached from
the ash. Using the idea stated above, and calculating the amount
of alkalinity which will be made available from the species ana-
lyzed, 18.1 milliequivalents of hydroxide were leached. This
value is 10.6% below the figure obtained from the graph. Figure
4-1 indicates that 1.10 meq OH~/g fly ash or 15.4 milliequiva-
lents of hydroxide were leached at pH 8.5. Calculating the
amount of alkalinity from the analyses of the leachate, 14.1
milliequivalents of alkalinity were leached. The alkalinity
calculated from the analyses is 4.470 less than the value obtained
from the graph. These values indicate that additional alkalinity
in forms other than the oxides of calcium, magnesium, sodium, or
potassium is present in the ash of Comanche Steam-Electric Sta-
tion. Another possibility is that sulfate could be more avail-
able to the system in some form other than calcium sulfate or as
a salt of one of the species listed above.
4.4.1.2 Plant Bowen
Calcium and sulfate are the major constituents leached
from Plant Bowen coal ash. Magnesium, sodium, and potassium are
much less concentrated in the ash. There is a significant dif-
ference in the sulfate to calcium ratio among the different pH
values at which the ash was leached. The sulfate to calcium
ratio at pH 6.0 is 0.413, but this ratio increased to 0.649 at
pH 8.5, and to 0.734 at pH 10.4. It follows from these values
that at a low pH value there is a tendency for more calcium in
the form of CaO to be leached than CaSCs which is the other major
form of calcium found in the fly ash. However, at higher pH
values, less alkalinity was leached and CaSCs was the major fora
by which calcium was dissolved into the system. The sulfate be-
ing dissolved into the system was a result chiefly of CaSO^ dis-
solution since magnesium, sodium and potassium are present in
much smaller quantities than calcium (see Table 4-4) .
Plant Bowen ash has available alkalinity of 0.43 meq
OH /g ash as derived from the plot of meq acid/g fly ash versus
time. At pH 6.0, 6.03 milliequivalents of hydroxide were leached
during the experiment; 6.1 milliequivalents of hydroxide were
calculated to have been leached, arriving in the form of the
L-38
-------
r1
OJ
4
a
.so
7% SLURRY pH=6.0
7% SLURRY pH = 8.5
.10
20
40 TIME (MINUTES) 6°
Figure 4-1. Comanche steam-electric station
meq acid/g fly ash versus time.
-------
oxides of calcium, magnesium, sodium and potassium using the
analyses of the leachate. At pH 8.5, 0.14 meq OH'/g fly ash
and 1.95 milliequivalents of hydroxide leached were the values
obtained from Figure 4-2; 2.2 milliequivalents of hydroxide
were leached as calculated from the analyses of the leachate.
No plot of meq acid/g fly ash versus time was made for the
leaching study of Plant Bowen ash at pH 10.4 due to the low
amount of alkalinity leached. The agreement between the values
above indicates that the sulfate of the system is derived from
CaSCK and that CaO is the major source of alkalinity found in
Plant Bowen coal ash.
4.4.1.3 Four Corners Power Plant
Calcium is the single major constituent leached from
coal ash of the Four Corners Power Plant. The sulfate leached
is significant but well below the amount from the ashes of the
two other plants as shown in Table 4-4. The amounts of magnesium,
sodium and potassium leached were insignificant. Leaching a 77o
slurry of the ash, the amounts of calcium and sulfate leached
were more pronounced than that of a 2% slurry. However, the sul-
fate to calcium ratio from the 770 slurry was 0.077 at pH 3.0,
0.023 at pH 6.0, and 0.049 at pH 8.5. The sulfate to calcium
ratios from the 27, slurries were insignificant. Leaching of the
Four Corners Power Plant coal ash indicates that the available
alkalinity is in the form of CaO, which is also the major form
of calcium present in the ash.
The values of milliequivalents of hydroxide leached
as CaO derived from the plots of meq acid/g fly ash versus time
(Figures 4-3 and 4-4) agree well with the values obtained by
calculating the hydroxide leached using the concentrations ob-
tained by the analyses of the leachate for a 770 slurry. The
corresponding values are as follows:
pH 3.0 pH 6.0 pH 8.5
Calculated from analyses
of leachate 5.52 5.00 4.28
Derived from meq acid/g
fly ash versus time plots 5.85 5.32 4.25
L-40
-------
1.80
7% SLURRY pH = 8.0
7% SLURRY PH«=8.5
i.oo
<
<
i
.80
40 60
TIME (MINUTES*
80
100
Figure 4-2. Plant Bowen meq acid/g fly ash versus time.
-------
1.SO
I
-P-
1.00
g
o
o
111
3
.50
7% SLURRY pH-3.0
7% SLURRY pH - 6.0
7% SLURRY pH =8.5
40
TIME IMINUTES)
60
100
Figure 4-3.
Four Corners power station
meq acid/g fly ash versus time
-------
1.50
2% SLURRY PH=3.0
2% SLURRY pH =8.5
I
4^
LO
o
o
i
.so
TIME (MINUTE3I
Figure 4-4. Four Corners power station
meq acid/g fly ash versus time.
-------
These values show the agreement between the analyses
of the leachate and Figure 4-3. There is little doubt that CaO
is the major compound leached from Four Corners Power Plant^as
described by the analyses. The leaching performed with a 2%
slurry also agreed well between the two methods of determining
the alkalinity leached. At pH 3.0, 0.54 meq OlT/g fly ash and
2.16 milliequivalents of hydroxide leached were derived from
Figure 4-4; 1.77 milliequivalents of hydroxide were calculated
from the analyses to have been leached. For pH 8.5, 1.18 milli-
equivalents of hydroxide was calculated to have been leached
using the analyses values and 0.29 meq OH~/g fly ash or 1.16
milliequivalents of hydroxide was derived from Figure 4-4.
4.4.2 Batch Dissolutions
The following section is a semiqualitative discussion
of the results of the batch dissolution characterizations. These
results are also compared to the closed-loop ash sluicing experi-
mental results. It should be stated here that the ionic activi-
ties derived from the EMF values of the specific ion electrodes
are significantly different from the ionic activities calculated
by the chemical equilibrium computer program. The matrix to
which the specific ion electrodes were subjected adversely
affected the reliability of these electrodes. The liquor compo-
sitions were continuously changing, thereby changing the elec-
trode calibration. Continuous calibration is not possible since
the liquor composition changes are unknown when the experiment
is being run. Results from the batch dissolution experiments
were presented in Table 4-5.
4.4.2.1 Cotnanche Steam-Electric Station
The first batch dissolution characterization run with
coal ash from Comanche Steam-Electric Station shows an increase
in the concentration of calcium, the predominant species present,
throughout the run. There was a final concentration of calcium
of 6.18 mmole/£. No other species had higher than 1 mmole/£
concentration.
The calcium concentration increased somewhat, from
3.09 mmolar to 6.73 mmolar during the second run. This increase
is only 50% of the calcium increase of the first run. The chlor-
ide concentration doubled but no other species indicated a signi-
ficant increase in concentration.
L-44
-------
The third run, which was spiked with 0.417 mmole/£
carbonate and 5.21 mmole/£ sulfate, indicates calcium sulfate
precipitation. The calcium concentration decreased by more than
50% from the initial quick dissolution of more than 2.5 mmole/£
and the sulfate concentration decreased by 4.5 mmole/£. The
carbonate concentration also decreased by one-half indicating
calcium carbonate precipitation.
The fourth run had an initial sulfate concentration of
26.0 mmole/£. This concentration was reduced to 18.1 mmole/£
while the calcium concentration was less than 1 mmole/£ after 30
minutes. At 10 minutes, the calcium concentration was 4 mmole/£,
due probably to the quick dissolution of calcium oxide. The
carbonate concentration was reduced to slightly more than 0.2
mmole/£ which is approximately the final concentration of car-
bonate of each of the runs.
The fifth run also had a decrease in the calcium con-
centration, but a change of less than one mmole/£. The sulfate
concentration also decreased by less than one mmole/£. The solu-
tion had a carbonate concentration initially of 1.67 mmole/£,
and, finally, as in the previous runs, 0.2 mmole/£•
The information gathered from the last three runs indi-
cates that calcium carbonate was precipitated. Calcium carbo-
nate was supersaturated in the final liquors of each run. The
reduction in the carbonate concentration which correlates to the
calcium carbonate precipitation would not explain the amount of
the decrease of the calcium concentration after the initial
dissolution of calcium. The sulfate concentration had also
decreased significantly in these runs. However, computer chemi-
cal equilibrium relative saturation values do not indicate cal-
cium sulfate supersaturation which is necessary for precipitation
of calcium sulfate. Also, the resulting liquor is below a rela-
tive saturation of one with respect to calcium hydroxide.
One possible explanation for this observation is that
with a sluice liquor high in sulfate, the initial quick dissolu-
tion of CaO causes gypsum nucleation and precipitation to near
saturation. Then, as calcium carbonate precipitates, the calcium
concentration is lowered resulting in a subsaturated gypsum
solution at the end of the run.
Magnesium hydroxide was supersaturated in the liquors
of all the runs due to the high pH values. The concentrations
of the species analyzed were well below the concentration values
L-45
-------
obtained from the ash dissolution characterizations. However,
the pH values were very similar.
4.4.2.2 Plant Bowen
All five of the batch dissolution characterization
experiments for Bowen ash exhibited almost the same characteris-
tics. In each instance, approximately 10 mmole/£ of calcium,
10 mmole/Jl sulfate, 1 mmole/& sodium, and less than 1 mmole/&
chloride were dissolved into the liquor. Magnesium dissolution
was quite low. The carbonate concentrations of the last three
runs were reduced by calcium carbonate precipitation. Calcium
carbonate was supersaturated in each of the liquors of the exper-
iments as calculated by the chemical equilibrium program (rela-
tive saturations ranged from 13-26).
As compared to the ash dissolution characterizations,
the batch dissolution characterizations exhibited lower calcium
and sulfate concentrations. The pH of each of the runs using
fly ash from Plant Bowen was lower than the pH observed during
the ash dissolution characterizations of Plant Bowen fly ash.
4.4.2.3 Four Corners Power Plant
All five runs with fly ash from the Four Corners Power
Plant exhibited an increase in the calcium concentrations with
an upper limit of slightly more than 5 mmoles/£. The sulfate,
sodium, and chloride concentrations increased, although not sub-
stantially. Calcium carbonate precipitation probably occurred
during the final three runs, being indicated by a decrease in
carbonate concentration and supersaturation of calcium carbonate.
Even though magnesium concentrations were very low,
the liquors of each run were supersaturated with magnesium hy-
droxide due to the high hydroxide concentrations.
The calcium concentrations of the batch dissolution
characterizations were in the same range as the calcium concen-
trations of the ash dissolution characterizations. Other species
analyzed from the batch dissolution characterizations were not as
concentrated as in the ash dissolution characterizations. The pH
values observed during the batch dissolution characterizations
were more than one pH unit higher than the pH values observed
during the ash dissolution characterizations.
L-46
-------
APPENDIX LA
TEST PARAMETERS AND CHEMICAL ANALYSES
FOR CLOSED-LOOP SLUICING TESTS
L-47
-------
TABLE A-l. COMANCHE STEAM-ELECTRIC STATION RUN NO. 1
Settling Pond
31 Hrs. 26 Hrs.
'Calcium 4.30 4.92
'Magnesium .002 .002
'Sodium 6.36 4.45
1 Sulfate .49 .75
'Chloride .675 .559
'Carbonate .43 2.44
pH 11.9 11.7
T°C 21 19.5
a 7.9 xlO""* 1.52xlO~3
aM 1.4 xlO"3 2. 04x10" 3
aCau 2. 22x10" 3 1.99xlO~3
ay u 4.72xlO~7 4.89xlO~7
Feed Rates to Mixing Tank:
Sluice water 710 ml/min.
Makeup water 40 ml/min.
Coal ash 56 g/min.
Makeup Water Composition:
Calcium 5.1 nnnole/£
Magnesium 2.7 mmole/£
Sodium 6.3 tnmole/£
Sulfate 10.1 mmole/£
Carbonate .04 mmole/£
Chloride 1.4 mmole/£
Nitrate .25 mmole/£
pH 8.9
Mixing
31 Hrs.
5.52
.002
7.22
.20
.805
.29
11.9
23
1.1 xlO 3
1.1 xlO~3
2.83xlO~3
4.24xlO~7
Tank2
26 Hrs.
5.85
.004
6.04
.24
.656
.42
11.1
21
1. 28x10" 3
1. 75x10" 3
3. 34x10" 3
1.88xlO~6
Sluice
31 Hrs. 3
4.96
.002
7.12
.17
.705
.18
11.8
9.3 xlO~"
1.1 xlO~3
2.71xlO~3
5. 25x10" 7
Line
26 Hrs.3
5.21
.002
5.72
.08
.625
.40
11.5
1.8 xlO"3
8.0 xlO~"
2.95xlO"3
7. 18x10" 7
Concentration in mmole/£.
Residence time of mixing tank is 6.4 minutes.
Residence time of sluice line is 10 minutes.
Ionic activities calculated by chemical equilibrium computer program.
L-48
-------
TABLE A-2. COMANCHE STEAM-ELECTRIC STATION RUN NO. 2
Settling Pond
48 Hrs. 43 Mrs.
1 Calcium
1 Magnesium
i ^ i j
Sodium
1 _
^ulfate
Chloride
1 Carbonate
pH
T°C
aCa
^g
o
aCa5
Sfo5
6.32
.016
18.8
7.84 ,
1.12
.31
11.9
23 _„
2.50x10
3.0 8x10" 3
2.25xlO~ 3
2.86xlO~6
6.95
.017
18.2
7.82
1.01
.70
12.0
23 -s
2.20x10
3.1 xlO~3
2.34xlO~3
2.60xlO~6
Mixing Tank2
48 Hrs. 43 Hrs.
7.18
.004
20.5
7.78
1.09
.18
12.15
24
6.7 xlO *
2.66xlO~3
2. 38xlO~ 3
4. 6 8x10" 7
7.93
.009
19.4
8.61
1.03
.47
11.9
23.5 ,
4.75x10
2.83x!0"3
2.72xlO~3
1.54xlO~6
Sluice Line
48 Hrs.3 48 Hrs." 43 Hrs.
7.39
.016
20.5
8.58
1.11
.15
12.0
——
9.3 xlO""
2.15xlO~3
2. 54x10" 3
2.49xlO~6
7.07
.028
21.0
8.69
1.15
.27
12.0
9.3 xlO~"
1.92xlO~3
2.40x10" 3
4. 32x10" 6
7.32
.007
20.6
8.81
1.06
11.6
1.42xlO~
2.15xlO~
2.62xlO~
1.69x10"
3
14
3
3
6
Feed Rates to Mixing Tank:
Sluice water
Makeup water
Coal ash
710 ml/rain
40 tnl/min
56 g/min
Makeup Water Composition:
Calcium
Magnesium
Sodium
Potassium
Sulfate
Carbonate
Chloride
Nitrate
pH
5.1 mmole/£
2.7 mmole/£
24.9 mmole/£
1.7 mmole/£
20.2 mmole/H
.04 mmole/i
1.4 mmole/H
.3 mmole/S.
9.2
Concentration in mmole/£.
2Residence time of mixing tank is 6.4 minutes.
'Residence time of sluice line is 10 minutes.
''Residence time of sluice line is 5 minutes.
5Ionic activities calculated by chemical equilibrium computer program.
L-49
-------
TABLE A-3. COMANCHE STEAM-ELECTRIC STATION RUN NO. 31
Settling Pond
49 Hrs. 44 Hrs.
2Calcium 7.68 8.68
2Magnesium .005 .008
2Sodium 18.4 13.8
2Sulfate 6.78 6.51
2Chloride 1.24 1.18
2Carbonate .26 .25
pH 12.2 12.0
T°C 21 21
aca — 7.9 xlO-*
^ 9.1 xlO"3
aCae 2. 69x10" 3 3. 27x10" 3
a^j e 6. 63x10" 7 1.44xlO"6
Feed Rates to Mixing Tank:
Sluice water 710 ml/rain
Makeup water 40 ml/min
Coal ash 56 g/min
Makeup Water Composition:
Calcium 5.1 mmole/£
Magnesium 2.7 mmole/S,
Sodium 6.3 mmole/£
Sulfate 10.1 mmole/Jl
Carbonate .04 ramole/£
Chloride 1.4 mmole/£
Nitrate .25 mmole/£
pH 8.9
Mixing
49 Hrs.
8.49
.006
14.4
5.98
1.25
.12
12.2
23
3.01xlO~3
7. 08x10" 7
Tank3
44 Hrs.
8.46
.004
13.0
6.66
1.23
.14
12.1
23
9.6 xlO"*
3.2 xlO"3
3. 04x10" 3
5. 53x10" 7
49 Hrs.
8.07
.006
15.2
6.67
1.22
.21
12.2
24.5
2.71x10"
6.32xlO~
Sluice Line
" 49 Hrs.5
7.39
.006
15.0
7.24
1.24
.11
12.2
24.5
2
2
3 2. 46x10" 3 2
7 6. 31x10" 7 7
44 Hrs.1*
8.41
.006
13.4
6.59
1.20
.15
12.2
23
. 37x10" 3
.1 xlO"3
.9 2x10" 3
. 04x10" 7
*Fan positioned to blow across surface of settling pond.
Concentration in mmole/£.
Residence time of mixing tank is 6.4 minutes.
^Residence time of sluice line is 10 minutes.
Residence time of sluice line is 5 minutes.
Ionic activities calculated by chemical equilibrium computer program.
L-50
-------
TABLE A-4. COMANCHE STEAM-ELECTRIC STATION RUN NO. 4
Settling Pond
50 Hrs. 45 Hrs.
'Calcium
Magnesium
1 Sodium
'Sulfate
'Chloride
'Carbonate
PH
T°C
3Ca
^g
aCa5
^S5
8.98
.004
14.1
6.26
1.18
.058
11.95
22
2. 20x10" 3
1. 40x10" 3
3.46xlO~3
7. 38x10" 7
9.04
.007
9.63
5.50
1.28
.122
12.0
22
1. 27x10" 3
2. 56x10" 3
3. 55x10" 3
1.22xlO~6
Mixing
50 Hrs.
10.4
.006
10.3
7.63
1.20
.063
11.9
23
2. 55x10" 3
1. 66x10" 3
3. 86x10" 3
1. 10x10" 6
Tank2
45 Hrs.
10.7
.004
10.1
6.94
1.24
.072
11.9
23
1.73x10 3
2. 44x10" 3
4. 06x10" 3
7. 41x10" 7
Sluice Line
50 Hrs. 3 50 Hrs." 45 Hrs. 3
9.97
.012
17.8
6.93
1.24
.080
11.95
23
1.22x10 3
2. 67x10" 3
3. 68x10" 3
2. 06x10" 6
9.76
.008
16.0
6.80
1.25
.170
11.95
23
2.04x10 3
2.17x10" 3
3. 61x10" 3
1.37xlO~6
9.68
.013
8.50
6.40
1.20
.122
11.95
22.5
7.25x10"'*
2. 84x10" 3
3.72xlO~3
2.33xlO~6
Feed Rates to Mixing Tank:
Sluice water 710 ml/min
Makeup water 40 ml/min
Coal ash 56 g/tain
Makeup Water Composition:
Calcium
Magnesium
Sodium
Sulfate
Carbonate
Chloride
Nitrate
PH
5.1 mmole/£
2.7 mmole/£
6.3 mmole/Jl
10.1 mmole/£
.04 mmole/Jl
1.4 mmole/Jl
.25 mmole/X,
8.9
Concentration in mmole/JL
2Residence time of mixing tank is 3.2 minutes.
3Residence time of sluice line is 10 minutes.
""Residence time of sluice, line is 5 minutes.
5Ionic activities calculated by chemical equilibrium computer program.
L-51
-------
TABLE A-5. COMANCHE STEAM-ELECTRIC STATION RUN NO. 5
Settling Pond
50 Hrs. 45 Hrs.
'Calcium 11.2 11.0
'Magnesium .450 .849
'Sodium 16.6 20.0
'Sulfate 12.6 11.4
'Chloride 1.20 1.21
'Carbonate 3.93 6.87
pH 7.1 7.45
T°C 24 24
a 7.9 xlO"" 1.05x10 3
aM 2.4 xlO"3 2.75xlO~3
a,, 5 3. 79x10" 3 3. 74x10" 3
Ca
a.. 5 1.64x10"" 3.12x10 "
^Ig
Feed Rates to Mixing Tank:
Sluice water 710 ml/min
Makeup water 40 ml/min
Coal ash 56 g/min
Makeup Water Composition:
Calcium 5.1 mmole/X.
Magnesium 2.7 mmole/A
Sodium 6.3 mmole/Jl
Sulfate 10.1 mmole/H
Carbonate .04 mnole/&
Chloride 1.4 mmole/X,
Nitrate .25 mmole/Jl
pH 8.9
Mixing
50 Hrs.
12.6
.055
10.2
13.4
1.26
.150
11.2
24
2. 30x10" 3
1. 52x10" 3
4. 22x10" 3
8.70xlO~5
Tank2
45 Hrs.
9.71
.594
16.4
12.3
1.18
.392
10.8
24
2.52x10 3
9.8 xlO~"
3. 46x10" 3
1. 94x10" "
50 Hrs.
11.9
.188
16.5
14.0
1.26
.210
9.8
25
1.12x10"
2.16x10"
3.94x10"
6.58x10"
Sluice Line
3 50 Hrs."
12.0
.103
19.4
13.6
1.24
.143
10.7
25
3 2. 90x10" 3 5
3 6.2 xlO~" 2
3 3. 99x10" 3 3
5 3.39xlO~5 5
45 Hrs.3
10.2
.164
14.9
12.1
1.20
.263
10.7
24
.7 xlO "
. 63x10" 3
. 51x10" 3
. 56xlO~ 5
'Concentration in mmole/JZ..
2Residence time of mixing tank is 3.2 minutes.
'Residence time of sluice line is 10 minutes.
''Residence time of sluice line is 5 minutes.
5Ionic activities calculated by chemical equilibrium computer program.
L-52
-------
TABLE A-6. PLANT BOWEN RUN NO. 1
Settling Pond
48 Hrs. 43 Hrs.
1 Calcium
Magnes ium
1 Sodium
'Sulfate
'Chloride
Carbonate
PH
T°C
aCa
aM
aCa*
SL u
16.4
.014
8.32
10.3
1.26
.281
11.9
23
2. 87x10" 3
2.9 xlO~3
5. 59x10" 3
2. 44x10" 6
16.0
.008
12.6
9.95
1.32
.187
11.85
23
9.9 xlO""*
5.2 xlO~3
5. 54x10" 3
1.49xlO~6
Mixing
48 Hrs.
22.8
.010
4.66
12.7
1.25
.292
12.1
23.5
3. 13x10" 3
5.4 xlO~3
7. 00x10" 3
1.25xlO~6
Tatik2
43 Hrs.
22.4
.005
5.92
12.6
1.23
.240
12.05
23.5
7.9 xio"1*
7.7 xlO~3
6.99xlO~3
6.73xlO~7
Sluice Line
48 Hrs.3 43 Hrs.3
22.9
"• .014
8.66
12.8
1.28
.155
12.1
25
4. 07x10" 3
5.2 xlO~3
6. 89x10" 3
1.59xlO~6
23.0
.006
6.68
12.5
1.29
.213
12.0
24
4. 16x10" 3
3.6 xlO~3
7. 24x10" 3
8.44xlO~7
Feed Rates to Mixing Tank:
Sluice water
Makeup water
Coal ash
710 ml/min
40 tnl/min
56 g/min
Makeup Water Composition:
Calcium
Magnesium
Sodium
Sulfate
Carbonate
Chloride
Nitrate
PH
5.1 mmole/JZ.
2.7 mmole/2
6.3 mmole/£
10.1 mmole/£
.04 mmole/£
1.4 mmole/5,
.25 mmole/£
8.9
'Concentration in mmole/ft.
2Residence time of mixing tank is 6.4 minutes.
^Residence time of sluice line is 10 minutes.
''Ionic activities calculated by chemical equilibrium computer program.
L-53
-------
TABLE A-7. PLANT BOWEN RUN NO. 2
Settling Pond
50 Hrs. 45 Hrs.
'Calcium
'Magnesium
1 Sodium
1 Sulfate
'Chloride
'Carbonate
PH
T°C
aCa
aM
aCas
SrfoS
20.7
.020
15.6
23.3
1.28
.169
11.3
20.5
2. 02xlO~ 3
2. 13xlO~ 3
5. 80x10" 3
S.llxlO"6
19.6
.012
12.3
22.0
1.31
.109
11.3
19.5
2.44xlO~3
1. 96x10" 3
5. 64x10" 3
3.18xlO~6
Mixing Tank2
50 Hrs. 45 Hrs.
23.0
.102
15.2
26.2
3.66
—
10.6
22
1. 17x10" 3
3. 3 3x10" 3
6.22xlO~3
2.87xlO~5
22.0
.072
14.1
24.7
1.33
.125
11.1
21
1. 38x10" 3
3. 02x10" 3
6. 08x10" 3
1.92xlO~5
Sluice Line
50 Hrs.3 50 Hrs." 45 Hrs.3
22.5
.090
14.5
26.2
1.28
.119
10.85
22.5
1. 17x10" 3
3. 33x10" 3
6. 06x10" 3
2.44xlO~5
21.8
.062
15.9
26.4
1.29
.112
11.0
22.5
3.25xlO~3
9.0 xlO"1*
5. 82x10" 3
1.62xlO~s
21.4
.074
14.5
25.2
1.24
.147
10.75
21.5
3.6 xlO~3
1.2 xlO~3
5. 86x10" 3
2.08xlO~5
Feed Rates to Mixing Tank:
Sluice water
Makeup water
Coal ash
710 ml/min
40 ml/min
56 g/min
Makeup Water Composition:
Calcium
Magnesium
Sodium
Potassium
Sulfate
Carbonate
Chloride
Nitrate
PH
5.1
2.7
24.9
1.7
20.2
.04
1.4
.3
9.2
mmole/S,
mmole/£
mmole/&
mmole/S.
mmole/X,
mmole/2.
mmole/£
mmole/X.
'Concentration in mmole/X,.
2Residence time of mixing tank is 6.4 minutes.
Residence time of sluice line is 10 minutes.
""Residence time of sluice line is 5 minutes.
5Ionic activities calculated by chemical equilibrium computer program.
L-54
-------
TABLE A-8. PLANT BOWEN RUN NO. 3
Settling Pond
48 Hrs. 43 Hrs.
1 Calcium
Magnesium
1 Sodium
'Sulfate
1 Chloride
1 Carbonate
pH
T°C
aCa
"Kg
*u,.5
18.4
.015
11.4
21.0
.992
.160
11.3
24.5
3.6 xlO 3
1.2 xlO"3
5. 25x10" 3
3. 65x10" 6
18.3
.012
13.4
20.6
.983
.091
11.3
24
2.62xlO~3
3.1 xlO~3
5. 28x10" 3
2.98xlO~6
Mixing
48 Hrs.
20.0
.048
14.2
21.9
.958
.110
11.1
25
1.9 xlO~3
2.9 xlO~3
5. 68x10" 3
1.25xlO~s
Tank2
43 Hrs.
19.9
.048
17.6
22.1
.977
.092
11.1
25.5
2. 87x10" 3
2.8 xlO~3
5.61xlO~3
1. 24x10" 5
Sluice Line
48 Hrs. 3 48 Hrs." 43 Hrs. 3
20.0
.044
13.6
22.0
.976
.075
11.05
26.5
3.9 xlO~3
2.0 xlO~"
5. 65x10" 3
1. 14xlO~ 5
20.0
.035
14.3
22.1
.949
.147
11.2
26.5
4.5 xlO~3
3.0 xlO~"
5. 59x10" 3
8.48xlO~6
20.6
.040
18.2
22.2
.980
.119
11.2
25
1.43x10
4.3 xlO~
5.80xlO~
9.99xlO~
3
3
3
6
lg
Feed Rates to Mixing Tank:
Sluice Water 710 ml/min
Makeup water 40 ml/min
Coal ash 56 g/min
Makeup Water Composition:
Calcium
Magnesium
Sodium
Sulfate
Carbonate
Chloride
PH
.75 mmole/2.
.35 mmole/Z
1.0 mmole/£
.10 mmole/£
1.71 mmole/X,
.24 mmole/Jl
7.8
'Concentration in mmole/£.
2Residence time of mixing tank is 6.4 minutes.
Residence time of sluice line is 10 minutes.
''Residence time of sluice line is 5 minutes.
slonic activities calculated by chemical equilibrium computer program.
L-55
-------
TABLE A-9. PLANT BOWEN RUN NO. 4
Settling Pond6
47 Hrs. 42 Hrs.
'Calcium
'Magnesium
'Sodium
'Sulfate
'Chloride
'Carbonate
PH
T°C
aCa
aw-
Mg
a^.s
18.5
.494
4.54
18.5
1.33
2.19
8.4
23
3. 13x10" "
4.5 x!0~3
5. 65x10" 3
17.4
.450
4.36
17.3
1.37
2.32
8.4
23
1.83x10""*
4.0 xlO~3
5.43xlO~3
Mixing Tank2
47 Hrs. 42 Hrs.
21.2
.334
5.12
20.8
1.38
.333
10.6
23
7.9 xlO 3
6. 25x10" 3
19.8
.475
4.72
19.7
1.34
.222
10.0
23
4.7 xlO~3
1.0 xlO~3
6. 00x10" 3
Sluice Line
47 Hrs.3 47 Hrs." 42 Hrs.3
21.7
.054
4.98
20.8
1.34
.210
11.0
24
3.20xlO~3
2.4 xlO"3
6.38xlO~3
21.6
.392
5.01
21.3
1.34
.439
10.2
24.5
2 . 90x10" 3
2.4 xlO"3
6. 28xlO~ 3
20.1
.368
4.78
19.8
1.38
.205
10.3
24
1. 37x10" 3
3.9 xlO"3
6. 05x10" 3
1.62x10'" 1.51x10'*
Feed Rates to Mixing Tank:
Sluice water 710 ml/min
Makeup water 40 ml/min
Coal ash 56 g/min
Makeup Water Composition:
1.01x10"" 1.52x10 *
1.12xlO~" 1.20xlO~" 1.16x10'"
Calcium
Magnesium
Sodium
Sulfate
Carbonate
Chloride
Nitrate
PH
5.1 mmole/2,
2.7 mmole/d
6.3 mmole/fc
10.1 mmole/&
.04 mmole/S,
1.4 mmole/d
.25 mmole/&
8.9
'Concentration in mmole/X,.
2Residence time of mixing tank is 3.2 minutes.
Residence time of sluice line is 10 minutes.
''Residence time of sluice line is 5 minutes.
5Ionic activities calculated by chemical equilibrium computer program.
6Carbon dioxide bubbled through settling pond at the rate of 0.5 H/min.
L-56
-------
TABLE A-10. FOUR CORNERS POWER PLANT RUN NO. 1
Settling Pond
46 Hrs. 41 Hrs.
rCalcium 2.51 1.93
Magnesium .207 .232
'Sodium 8.05 12.8
'Sulfate 2.40 2.36
'Chloride 1.47 1.42
'Carbonate .506 .369
PH 10.55 10.4
T°C 22 22
aCa 2.9 xlO~" 5.35x10 "
a^ 5.0 xlO~* 3.2 xlO~*
aCas 1.26xlO~3 9.83xlO~"
a^j 5 9.72xlO~s 1.14xlO~"
Feed Rates to Mixing Tank:
Sluice water 710 ml/min
Makeup water 40 ml/min
Coal ash 56 g/min
Makeup Water Composition:
Calcium 5.1 mmole/Jl
Magnesium 2.7 mmole/2,
Sodium 6.3 mmole/£
Sulfate 10.1 mmole/£
Carbonate .04 tmnole/fe
Chloride 1.4 mmole/£
Nitrate .25 mmole/£
pH 8.9
Mixing
46 Hrs.
2.64
.259
7.62
2.81
1.44
.372
10.5
23
7.0 xlO~5
7.8 xlO~"
1.32xlO~3
1.23xlO~"
Tank2
41 Hrs.
2.26
.259
8.14
2.74
1.48
.420
10.4
23
4.26xlO~"
6.3 xlO""
1.13xlO~3
1.24xlO~'1
46 Hrs.
2.60
.223
13.6
2.94
1.46
.246
10.55
24
1.9 xlO~
7.5 xlO~
1.28xlO~
1.05xlO~
Sluice Line
3 46 Hrs."
2.60
.187
2.93
2.93
1.40
.375
10.6
24
"4.7 xHf*
" 4.7 xlO~"
3 1.30xlO~3
11 8.68xlO~5
41 Hrs. 3
2.12
.207
2.17
2.17
1.42
.410
10.45
24
6.23xlO~"
3.2 xlO~"
1. 12x10" 3
1.02xlO~"
1 Concentration in mmole/Jl.
2Residence time of mixing tank is 6.4 minutes.
Residence time of sluice line is 10 minutes.
''Residence time of sluice line is 5 minutes.
slonic activities calculated by chemical equilibrium computer program.
L-57
-------
TABLE A-11. FOUR CORNERS POWER PLANT RUN NO. 2
Settling Pond
46 Hrs. 41 Hrs.
'Calcium
'Magnesium
'Sodium
'Sulfate
'Chloride
'Carbonate
PH
T*C
*MB
*Ca*
*M«S
5.06
.170
10.4
6.72
2.19
.458
10.6
19
1. 15x10" 3
8.7 xlO~*
2. 06x10" 3
6. 80x10" s
5.19
.174
11.1
6.40
2.07
.655
19.6
19
1.74x10 3
4.6 xlO~"
2. 10x10" 3
6. 84x10" s
Mixing Tank2
46 Hrs. 41 Hrs.
6.16
.189
12.6
7.24
2.32
.379
10.6
21
7.9 xlO"11
1.4 xlO~3
2. 46x10" 3
7. 44x10" 5
5.90
.196
10.8
7.18
2.24
.633
10.6
20
1. 24x10" 3
3.6 xlO~*
2. 32x10" 3
7. 5 2x10" 5
Sluice Line
46 Hrs.3 46 Hrs." 41 Hrs.3
5.78
..150
14 Hi
7.58
2.28
.342
10.55
22.5
1. 24x10" 3
1. 13x10" 3
2. 27x10" 3
5.82xlO~5
5.62
.216
10.8
7.54
2.34
.439
10.4
22.5
1.56x10" 3
6.3 x!0~"
2. 21x10" 3
8. 45x10" 5
5.58
.172
12.0
7.19
2.24
.423
10.6
21.5
1. 33x10" 3
8.7 xlO~*
2. 22x10" 3
6. 70x10" 5
Feed Rates to Mixing Tank:
Sluice water 710 ml/min
Makeup water 40 ml/min
Coal ash 56 g/min
Makeup Vater Composition:
Calcium
Magnesium
Sodium
Potassium
Sulfate
Carbonate
Chloride
•itrate
PH
16
1.8
13
.3
21.3
.5
5
.3
9.1
mmole/£
mmole/S.
mmole/fc
mmole/fc
mmole/£
mmole/£
•mole/ft
•mole/l
1Concentration in mmole/l.
2Kesldence time of mixing tank is 6.4 minutes.
'Residence time of sluice line is 10 minutes.
"Residence time of sluice line is 5 minutes.
1Ionic activities calculated by chemical equilibrium computer program.
L-58
-------
TABLE A-12. FOUR CORNERS POWER PLANT RUN NO. 3
Settling Pond6
50 Hrs. 45 Hrs.
Calcium 3.40 3.88
Magnesium .739 .736
'Sodium 4.48 4.92
'Sulfate 4.00 3.80
'Chloride 1.38 1.40
Carbonate 2.92 2.99
PH 8.2 8.2
T°C 23 23
aCa 6.5 xlO~3 1.05xlO~ 3
a^jg 9.7 xlO~*
aCas 1.61xlO~3 1. 85x10" 3
£L. 5 3.65x10"" 3.66xlO~"
5
Feed Rates to Mixing Tank:
Sluice water 710 ml/min
Makeup water 40 ml/min
Coal ash 56 g/min
Makeup Water Composition:
Calcium 5.1 mmole/Jl
Magnesium 2.7 mmole/Jl
Sodium 6.3 mmole/Jl
Sulfate 10.1 mmole/Jl
Carbonate .04 mmole/Jl
Chloride 1.4 mmole/J,
Nitrate .25 mmole/Jl
pH 8.9
Mixing
50 Hrs.
3.34
.744
4.93
4.20
1.30
2.61
8.7
23
2.80xlO~3
1.7 xlO~3
1.54xlO~3
3.52xlO~1'
Tank2
45 Hrs.
3.27
.750
4.76
4.10
1.30
2.60
8.7
22
9.7 x!0~"
9.0 xlO~"
1.52xlO~3
3. 58x10" "
Sluice Line
50 Hrs. 3 50 Hrs. "
3.50 3.08
.740 .704
4.98 5.03
4.30 4.40
1.37 1.36
2.32 2.03
8.7 8.9
23 23
3.43xlO~3 7.2 xlO~3
1. 61xlO~ 3 1.40xlO~3
3.51x10"" 3. 28xlO~ 3
45 Hrs. 3
3.43
.734
5.12
4.10
1.38
2.41
8.7
23
1.21x10 3
6.6 xlO~"
1.59xlO~3
3.51xlO~"
'Concentration in nunole/d.
2Residence time of mixing tank is 3.2 minutes.
3Residence time of sluice line is 10 minutes.
''Residence time.-.of sluice line is 5 minutes.
5Ionic activities calculated by chemical equilibrium computer program.
6Carbon dioxide bubbled through settling pond at the rate of 0.5 fc/min.
L-59
-------
APPENDIX LB
CORRELATION PARAMETERS FOR
CLOSED-LOOP SLUICING TESTS
L-60
-------
, «_ . Appendix LB contains the information which was used to
determine it the ash dissolution characterization experiments
were sampled at steady-state. The information which was used to
derive the graphs which illustrate the dissolution occuring in
the mix tank of the bench-scale model of a closed-loop ash
sluicing facility also appears in Appendix LB.
Tables B-l, B-2, and B-3 co.ntain calcium concentra-
tions of samples which were taken throughout the ash dissolution
characterization experiments. The calcium concentrations con-
tained in these tables were used to determine if the final sam-
ples were collected after the system had achieved steady-state.
Tables B-4 and B-5 contain the information which was
used to depict the dissolution of fly ash from Comanche Steam-
Electric Station and Plant Bowen. The tables contain the disso-
lution rates of various species and related ionic activities
and activity products which were used to characterize the disso-
lution of the fly ash of the respective plants. Although the
same information was provided for Four Corners Power Plant ash
dissolution (Table B-6), the data did not provide a sufficient
range over which the data could be plotted. Therefore, possibly
erroneous graphs could be drawn. For this reason, illustration
of the dissolution of Four Corners Power Plant fly ash was con-
sidered inappropriate.
L-61
-------
TABLE B-l. CALCIUM CONCENTRATIONS OF SAMPLED STREAMS DURING
BENCH-SCALE EXPERIMENTS FOR COMANCHE1
Run No. 1
Run No. 2
Run No. 3
Run No. 4
Run No. 5
Time
(Hours)
5
10
15
21
26
31
8
14
18
23
39
43
48
6
12
19
26
35
44
49
17
26
33
40
45
50
17
24
40
45
50
Settling
Pond
5.25
5.32
5.41
4.98
4.92
4.30
4.62
4.19
4.39
5.99
5.86
6.95
6.32
5.09
5.81
6.68
6.70
7.75
8.68
7.68
5.36
5.65
7.40
8.26
9.04
8.98
8.87
7.74
9.49
11.0
11.2
Mixing
Tank
6.96
6.84
6.39
5.57
5.85
5.52
4.32
6.24
6.69
7.98
8.03
7.93
7.18
6.70
6.86
8.64
8.38
8.83
8.46
8.49
7.16
6.23
9.20
9.90
10.7
10.4
7.43
10.9
9.73
9.71
12.6
Sluice
Line
6.77
5.44
5.21
4.54
5.21
4.96
5.59
5.71
6.49
7.73
7.36
7.32
7.39
5.58
6.07
8.17
7.75
8.87
8.41
8.07
6.43
8.19
8.14
9.16
9.68
9.97
7.97
8 19
10.2
11.9
Concentrations in iranole/£,
L-62
-------
TABLE B-2.
CALCIUM CONCENTRATIONS OF SAMPLED STREAMS DURING
ASH DISSOLUTION CHARACTERIZATION OF PLANT BOWEN
COAL ASH1
Run No. 1
Run No. 2
Run No. 3
Run No. 4
Time
(Hours)
15
24
30
39
43
48
10
18
26
40
45
50
12
20
28
36
43
48
15
22
38
42
47
Settling
Pond
6.29
10.0
12.7
13.9
16.0
16.4
15.5
_ _
16.7
18.5
19.6
20.7
17.2
16.9
18.3
18.0
18.3
18.4
9.35
11.5
17.9
17.4
18.5
Mixing
Tank
11.7
15.7
19.3
16.6
22.4
22.8
16.3
17.9
19.1
20.7
22.0
23.0
17.8
19.6
19.4
19.3
19.9
20.0
12.5
14.5
16.4
19.8
21.2
Sluice
Line
10.5
16.4
18.2
16.4
23.0
22.9
16.4
17.5
18.6
20.4
21.4
22.5
18.9
19.6
20.0
19.3
20.6
20.0
12.7
14.9
17.8
20.1
21.7
Concentrations in mmole/Jl.
L-63
-------
TABLE B-3
CALCIUM CONCENTRATIONS OF SAMPLED STREAMS DURING
ASH DISSOLUTION CHARACTERIZATION OF FOUR CORNERS
POWER PLANT COAL ASH1
Run No. 1
Run No. 2
Run No. 3
Time
(Hours)
15
24
31
36
41
46
13
27
36
41
46
5
30
38
45
50
Settling
Pond
1.14
1.28
1.58
1.66
1.93
2.51
2.66
3.10
4.71
5.19
5.06
1.72
3.40
3.42
3.88
3.40
Mixing
Tank
1.21
1.57
1.73
2.00
2.26
2.64
3.28
3.95
4.69
5.90
6.16
2.25
3.08
3.27
3.27
3.34
Sluice
Line
1.41
1.58
1.85
2.22
2.12
2.60
3.08
4.03
4.60
5.58
5.78
2.25
2.93
3.34
3.43
3.50
Concentrations in mmole/£.
L-64
-------
TABLE B-4. COMANCHE STEAM-ELECTRIG STATION DISSOLUTION AND PRECIPITATION RATES,
IONIC ACTIVITIES AND ACTIVITY PRODUCTS IN MIX TANK
Calcium Sulfate CaC03
Dissolution Dissolution Precipitation
Mmole/min Mmole/min Mmole/min
*Ca
co
r1
i
Ln
Run
Run
Run
Run
Run
No.
No.
No.
No.
No.
1
2
3
4
5
.88
.69
.71
1.22
1.29
-.60
-.54
-.73
.87
.70
.089
.087
.097
-.004
2.68
2.83
2.38
3.01
3.86
4.22
x 10 3
x 10~3
x 10~3
x 10~3
x 10~3
8.51
2.78
2.12
2.59
4.22
x 10
x 10"
x 10"
x 10"
x 10"
5
3
3
3
3
2.41
6.61
6.37
9.98
1.78
x 10~7
x 10"6
x 10~6
x 10~6
x 10~5
1.23
5.78
4.62
2.76
6.41
x 10 7
— g
X JL \J
x 10~8
x 10~8
x 10~8
-------
TABLE B-5. PLANT BOWEN DISSOLUTION AND PRECIPITATION RATES,
IONIC ACTIVITIES AND ACTIVITY PRODUCTS IN MIX TANK
Calcium
Dissolution
Mmole/min
f
i
Run No.
Run No.
Run No.
Run No.
1
2
3
4
5.25
2.35
1.91
2.56
Sulfate
Dissolution
Mmole/min
1.
2.
1.
2.
81
30
51
06
CaC03
Precipitation a~ a_,_
.. - / . La bUij
Mmole/mxn
-.018
-.015
.100
1.31
7.00
6.22
5.68
6.25
x 10 3
x 10~3
x 10~3
x 10~3
3.23
6.47
5.72
5.47
x 10~3
x 10"3
x 10~3
x 10~3
aCa
2.26
4.02
3.24
3.42
*so.
x 10~5
x 10~5
x 10~5
x 10~5
aCa
1.51
6.70
4.93
1.56
aco3
x 10~7
x 10~8
x 10" 8
x 10~7
-------
TABLE B-6. FOUR CORNERS POWER PLANT DISSOLUTION AND PRECIPITATION RATES,
IONIC ACTIVITIES AND ACTIVITY PRODUCTS IN MIX TANK
Calcium
Dissolution
Mmole/min
tp Run No. 1 0.0
ON
^ Run No. 2 .39
Run No. 3 -.10
Sulfate
Dissolution
Mmole /min
.01
-.19
-.09
CaC03
Precipitation a a a a
,, «• / , v*3. Ov/L. \sEL J\Jli
Mmole/min
.082 1.32 x 10~3 1.42 x 10~ 3 1.88 x 10~6
.061 2.46 x 10~3 2.87 x 10~ 3 7.06 x 10~6
.117 1.54 x 10~3 1.98 x 10" 3 3.04 x 10~6
aCa 3C03
9.16 x 10~8
1.25 x 10~7
6.94 x 10~8
-------
APPENDIX LC
pH AND EMF VALUES OF CALCIUM AND
DIVALENT CATION SPECIFIC ELECTRODES FOR
BATCH DISSOLUTION TESTS
L-68
-------
Appendix LC contains the charts obtained by instrument-
ally monitoring pH and EMF values of calcium and divalent cation
specific ion electrodes from the experiments conducted during
the batch dissolution characterizations. The tests were performed
upon coal ash from three coal-fired electric generating plants.
1) Comanche Steam-Electric Plant of
Public Service Company of Colorado,
2) Four Corners Power Plant of Arizona
Public Services, and
3) Plant Bowen of Georgia Power Company.
Experiments were conducted upon a 7% slurry of fly ash from
each of the plants and liquors of varying chemical compositions.
Each fly ash was subjected to five various chemical
compositions which are identified by run number as follows:
Run No. 1 - Deionized water
Run No. 2 - A 1:1 mixture of the final liquor
of the first run and deionized water
Run No. 3 - 25 ppm carbonate (0.417 mmole/£)
500 ppm sulfate (5.21 mmole/£)
259 ppm sodium (11.3 mmole/&)
Run No. 4 - 25 ppm carbonate (0.417 mmole/2.)
2500 ppm sulfate (26.0 mmole/2,)
1220 ppm sodium (52.8 mmole/£)
Run No. 5 - 100 ppm carbonate (1.67 mmole/O
500 ppm sulfate (5.21 mmole/£)
316 ppm sodium (13.8 mmole/Ji)
L-69
-------
f
o
-240-
-250
-260 .
-270 •
-280 .
10
20
Time (Minutes)
COMANCHE STEAM-ELECTRIC STATION
DIVALENT CATION SPECIFIC ION ELECTRODE - RUN NO. 1
-------
-230-
-240-
b,-250
§
-260-
10
20
Time (Minutes)
COMANCHE STEAM-ELECTRIC STATION
DIVALENT CATION SPECIFIC ION ELECTRODE - RUN NO. 2
-------
I
-J
to
-230 -I
-240
-250 -I
-260
-270 -I
-280 -I
10
20
30
Time (Minutes)
COMANCHE STEAM-ELECTRIC STATION
DIVALENT CATION SPECIFIC ION ELECTRODE - RUN NO. 3
-------
-230 -
-240 -
r<
oj
-250
§
-260
-270
-280
20
30
Time (Minutes)
COMANCHE STEAM-ELECTRIC STATION
DIVALENT CATION SPECIFIC ION ELECTRODE - RUN NO. 4
-------
-230 •
I
•vj
-P-
-240 .
-250
-260
-270
-280 -
10
20
30
Time (Minutes)
COMANCHE STEAM-ELECTRIC STATION
DIVALENT CATION SPECIFIC ION ELECTRODE - RUN NO. 5
-------
t-1
Ui
-10
-20 -
-30
-in
10
—i—
20
Time (Minutes)
T~
30
COMANCHE STEAM-ELECTRIC STATION
CALCIUM SPECIFIC ION ELECTRODE - RUN NO. 1
-------
-10
-20
t-1
I
-vl
-30 •
10
20
~30~
Time (Minutes)
COMANCHE STEAM-ELECTRIC STATION
CALCIUM SPECIFIC ION ELECTRODE - RUN NO. 2
-------
-30
-40
t*
-vl
-50 '
-60 -
-70
-80 -
-90 -
10
20
—T
30
Time (Minutes)
COMANCHE STEAM-ELECTRIC STATION
CALCIUM SPECIFIC ION ELECTRODE - RUN NO. 3
-------
-30 H
-40 J
-50
t-1
I
*-J
oo
-60 H
-70 1
-80
10
20
30
Time (Minutes)
COMANCHE STEAM-ELECTRIC STATION
CALCIUM SPECIFIC ION ELECTRODE - RUN NO. 4
-------
f
vo
-10
-20
-30 -
-40 -
-50
-60
-70-1
-80
-90
-100
10
20
Time (Minutes)
COMANCHE STEAM-ELECTRIC STATION
CALCIUM SPECIFIC ION ELECTRODE - RUN NO. 5
-------
f
I
00
o
13
12
11
10
9
8
7
6 J
10
20
30
Time (Minutes)
COMANCHE STEAM-ELECTRIC STATION
pH - RUN NO. 1
-------
12 A
tl
IT1
I
00
10 A
10
20
30
Time (Minutes)
COMANCHE STEAM-ELECTRIC STATION
pH - RUN NO. 2
-------
13
12
11
oo
S3
10
10
20
30
Time (Minutes)
COMANCHE STEAM-ELECTRIC STATION
pH - RUN NO. 3
-------
13 ^
12
11
I
00
10
10
20
Time (Minutes)
30
COMANCHE STEAM-ELECTRIC STATION
pH - RUN NO. 4
-------
12
i
CO
10
in . 20
Time (Minutes)
30
COMANCHE STEAM-ELECTRIC STATION
pH - RUN NO. 5
-------
-240
-250
i
00
-260
-270-
-280-
10 20
Time (Minutes)
FOUR CORNERS POWER PLANT
DIVALENT CATION SPECIFIC ION ELECTRODE - RUN NO. I
-------
tr*
i
oo
-240
-250
-260
10
20
30
Time (Minutes)
FOUR CORNERS POWER PLANT
DIVALENT CATION SPECIFIC ION ELECTRODE - RUN NO. 2
-------
-240
-250
IT1
I
00
-260
-270
-280
-290
10 20
Time (Minutes)
30
FOUR CORNERS POWER PLANT
DIVALENT CATION SPECIFIC ION ELECTRODE - RUN NO. 3
-------
00
00
-2/40 -
-250 .
-260
-270
10
20
30
Time (Minutes)
FOUR CORNERS POWER PLANT
DIVALENT CATION SPECIFIC ION ELECTRODE - RUN NO. 4
-------
f
I
00
-240-
-250 -
-260
-270
-280
-290
10
20
Time (Minutes)
FOUR CORNERS POWER PLANT
DIVALENT CATION SPECIFIC ION ELECTRODE - RUN NO. 5
-------
-30
-40 -
f
o
-50 -
10
20
30
Time (Minutes)
FOUR CORNERS POWER PLANT
CALCIUM SPECIFIC ION ELECTRODE - RUN NO. 1
-------
-20 -
-30-
vO
-40 -
10
20
30
Time (Minutes)
FOUR CORNERS POWER PLANT
CALCIUM SPECIFIC ION ELECTRODE - RUN NO. 2
-------
-40
-50 -
b.
£
-60
I
v£>
N>
-70 -
-80
—r-
10
-i—
20
Time (Minutes)
—T~
30
FOUR CORNERS POWER PLANT
CALCIUM SPECIFIC ION ELECTRODE - RUN NO. 3
-------
-40 -
-50 -
-70 -
10
20
30
Time (Minutes)
FOUR CORNERS POWER PLANT
CALCIUM SPECIFIC ION ELECTRODE - RUN NO. 4
-------
VO
-50
-60 -
T.
UJ
-70 -
-80
-90
—i—
10
—i—
20
30
Time (Minutes)
FOUR CORNERS POWER PLANT
CALCIUM SPECIFIC ION ELECTRODE - RUN NO. 5
-------
tr1
I
*£>
Ul
13
12
11
10
10 20
Time (Minutes)
30
FOUR CORNERS POWER PLANT
pH - RUN NO. 1
-------
11-
12
IT1
I
VO
11
10
10
20
Time (Minutes)
FOUR CORNERS POWER PLANT
pH - RUN NO. 2
30
-------
12
U
10
10
20
Time (Minutes)
FOUR CORNERS POWER PLANT
pH - RUN NO. 3
30
-------
11
sr1
i
v£>
OO
11
10
JO 20
Time (Minutes)
FOUR CORNERS POWER PLANT
pH - RUN NO. 4
-------
U 1
10
10 20
Time (Minutes)
30
FOUR COMERS POWER PLANT
pH - RUN NO. 5
-------
r1
i
t—'
o
o
-230
-240 -
s -250
-260
-270
-280
10 20
Time (Minutes)
30
PLANT BOWEN
DIVALENT CATION SPECIFIC ION ELECTRODE - RUN NO. 1
-------
-230
-240
ir1
i-1
o
-270 -I
-280
-290
10
20
Time (Minutes)
PLANT BOWEN
DIVALENT CATION SPECIFIC ION ELECTRODE - RUN NO. 2
-------
o
ro
-230
-240 -
-2riO
-260
-270 -
20
30
Time (Minutes)
PLANT BOWEN
DIVALENT CATION SPECIFIC ION ELECTRODE - RUN NO. 3
-------
o
LO
-211) -|
£ -230
ui
-21,0
JO
20
30
Timp (Minutes)
PLANT BOWEN
DIVALENT CATION SPECIFIC ION ELECTRODE - RUN NO. 4
-------
i
M
O
-230
-250 -
-260
10 20
Time (Minutes)
30
PLANT BOWEN
DIVALENT CATION SPECIFIC ION ELECTRODE - RUN NO. 5
-------
sr>
i
t->
o
Ln
-10
-2(1
-30
-40
-50
10 20
Time (Minutes)
PLANT BOWEN
CALCIUM SPECIFIC ION ELECTRODE - RUN NO. 1
-------
-30 -
in
—i—
20
Time (Mi nutcs)
—r-
30
PLANT BOWEN
CALCIUM SPECIFIC ION ELECTRODE - RUN NO. 2
-------
-10
-30
-50
r1
i
£ -60
-80
-1 00
10 20
Time (Minutes)
PLANT BOWEN
CALCIUM SPECIFIC ION ELECTRODE - RUN NO. 3
-------
t-1
I
M
O
OO
-40
-50
-70
-80
10 20
Time (Minutes)
PLANT BOWEN
CALCIUM SPECIFIC ION ELECTRODE - RUN NO. 4
-------
o
VO
-20 -
-40 -
-60
-80
-100 -
10 20
rime (M i nutOR)
—T~
30
PLANT BOWEN
CALCIUM SPECIFIC ION ELECTRODE - RUN NO. 5
-------
11
in
9
8
7
10
20
Timo (Minutes)
PLANT BOWEN
pH - RUN NO. 1
30
-------
10 20
Time (Minutes)
PLANT BOWEN
pH - RUN NO. 2
30
-------
12 •
11
10
in 20
Time (Minutes)
PLANT BOWEN
pH - RUN NO. 3
30
-------
n
10
10 20
Time (Mi nu t PS)
PLANT BOWEN
pH - RUN NO. 4
-------
11
JO
10 20
T into (M Lnu Les )
30
PLANT BOWEN
pH - RUN NO. 5
-------
REFERENCES
BA-465 Babcock and Wilcox Company, Steam, its generation and
use, 37th ed., N.Y., 1955.
CH-335 Chu, Tien-Yung J., Peter A. Krenkel, and Richard J.
Ruane, "Reuse of ash sluicing water in coal-fired
power plants", in Proceedings of the Third National
Conference on Complete Water Reuse: Symbiosis as a
Means of Abatement for Multi-Media Pollution, Cin-
cinnati, June 1976, Lawrence K. Cecil, ed., N.Y.,
AIChE, 1976.
DE-165 Derrick, A. E., "Cooling pond proves to be the econo-
mic choice at Four Corners", Power Eng. 1963 (Nov.),
46.
DI-057 Dickey, Joe Ben, Jr. and Robert E. Gates, Managing
waste heat with the water cooling tower. 2nd ed.
Mission, Kansas, Marley Co., 1973.
DI-170 Dionex Corporation, Analytical ion chromatography,
Models 10 and 14, operation and maintenance manual.
Palo Alto, CA, Jan. 1976.
EL-094 Electrical World Directory of Electric Utilities,
74-75, 93rd Edition. N.Y., McGraw-Hill, 1974.
FE-001 Feitknecht, W. and P. Schindler, "Solubility con-
stants of metal oxides, metal hydroxides, and metal
hydroxide salts in aqueous solution", Pure Applied
Chem. 6, 130-160, 197-99 (1963).
L-115
-------
FE-102 Federal Power Commission, Steam-electric plant air
and water quality control data for the year ended
December 31, 1972. Atlanta, Georgia, 1973.
FU-001 Fuoss, R. M., "Ionic association, pt. 3, the equili-
brium between ion pairs and free ions", J. ACS 80,
5059-61 (1968).
GH-001 Ghosh, S. R. , S. C. Ghosh and D. Roy, "Nitrogen Oxides
Absorption. Design of Packed Towers", Technology 2
(3), 149-59 (1965); C. A. 65: 3358d.
GJ-001 Gjaldbaek, J. K., "Untersuchungen uber die loslichkeit
des magnesium-ydroxyds", II. Die loslichkeitsprodukte
und die dissoziationskonstante der magnesium hydroxyde",
Z. Anorg. Chem. 44, 259-88 (1925).
GU-075 Guthrie, K. M., Process plant estimating evaluation
and control. Solana Beach, CA, Craftsman Book Co.,
1974.
JA-105 Jaulmes, P. and S. Brun, "Solubilisation et precipi-
tation des sels peu solubles d'acides moyens ou
faibles", Trav. Soc. Pharm. Mont. 25(2). 98 (1965).
KL-052 Klein, David H. , Melvin D. Smith, and James A. Driy,
"Homogeneous nucleation of magnesium hydroxide",
Talanta 14, 937-40 (1967).
KL-053 Klein, David H. and James A. Driy, "Heterogeneous
and homogeneous nucleation of strontium sulphate",
Talanta 13, 289-95 (1966).
L-116
-------
KL-054 Klein, David H. and Louis Gordon, "Nucleation in
analytical chemistry, II. Nucleation and preci-
pitation of silver chloride from homogeneous
solution", Talanta 13, 177-86 (1959).
LE-201 Leung, Paul and Raymond E. Moore, "Water consumption
study for Navajo Plant", in Proceedings of the Ameri-
can Society of Civil Engineers, Power Division, 97
(P04), 8564 (1971).
LE-239 Lepper, F. R. , Jr., Private communication, Resources
Conservation Co., 3 June 1976.
LI-001 Linden, J. Van der, J. Soc. Chem. Ind. 36, 96 (1917).
LI-068 Linke, William F., comp., Seidell's solubilities.
Inorganic and metal-organic compounds, 4th ed.,
2 vols., Princeton, NJ, D. Van Nostrand, 1958, 1965.
MC-136 McGlamery, G. G., et al., Detailed cost estimates
for advanced effluent desulfurization processes.
Interagency Agreement EPA IAG-134 (D), Pt. A.
Research Triangle Park, NC, Control Systems Lab.,
NERC, 1974.
NA-205 National Coal Association, Steam-electric plant
factors, 24th ed. , WAshington, D.C., 1974.
NE-107 Nelson, Guy R. , Water recycle/reuse possibilities:
power plant boiler and cooling systems, final report.
EPA 660/2-74-089. Corvallis, Oregon, National Envi-
ronmental Research Center, Thermal Pollution Branch,
September, 1974.
L-117
-------
OT-001 Ottmers, D. M., Jr., Simulation of NAPCA venturi
system. Technical Note 200-002-4, Austin, Texas,
Radian Corporation, November 1969.
PA-121 Patterson, W. C., J. L. Leporati, and M. J. Scarpa,
"The capacity of cooling ponds to dissipate heat",
Proc. Amer. Power Conf. 33, 446 (1971).
PA-227 Paimer-Hostik & Assoc., Private communication,
Houston, Texas, 2 August 1976.
PE-161 PEDCo-Environmental Specialists, Inc., Summary
Report-Flue Gas Desulfurization Systems - April
1975. EPA Contract No. 68-02-1321, Task No. 6,
Cincinnati, Ohio, 1975.
PE-R-277 Perry, John H., Chemical engineers handbook, 5th
edition, New York, McGraw-Hill, 1973.
RE-211 Resources Conservation Co., Feasibility study for
Arizona Public Service Company Four Corners power
station. Renton, Washington, April 13, 1976.
RO-266 Rogers, A. N. and L. Awerbuch, "Reuse or disposal
of salts from power and desalination plant wastes",
in Proceedings of the 3rd National Conference on
Complete Water Reuse: Symbiosis as a Means of
Abatement for Multi-Media Pollution, Cincinnati,
June 1976, Lawrence K. Cecil, ed., N.Y., AIChE, 1976
TH-192 Thompson C. G. and G. A. Mooney, Recovery of lime
and magnesium in potable water treatment. EPA 600/
2-76-285, Montgomery, Alabama, Black, Crow, and
Eisdness, Inc., December 1976.
L-118
-------
UH-007 Uhlig, Herbert H., Corrosion and corrosion control,
an introduction to corrosion science and engineering
N.Y., Wiley, 1963.
L-119
-------
TECHNICAL REPORT DATA
(Please read Inantctions on the reverse before completing)
1- REPORT NO.
EPA-600/7-78-055b
3. RECIPIENT'S ACCE3SI Of* NO.
•). TITLE AND SUBTITLE
Water Re cycle/Re use Alternatives in
Coal-fired Steam-electric Power Plants: Volume II.
Appendixes
5. REPORT DATE
March 1978
6. PERFORMING ORGANIZATION CODE
7. AUTHORIS)
8. PERFORMING ORGANIZATION REPORT NO.
James G. Noblett and Peter G. Christman
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Radian Corporation
P.O. Box 9948
Austin, Texas 78766
10. PROGRAM ELEMENT NO.
EHE624
11. CONTRACT/GRANT NO.
68-03-2339
12. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Industrial Environmental Research Laboratory
Research Triangle Park, NC 27711
13. TYPE OF FIEPORT AND PERIOD COVERED
13. TYPE OF FIEPORT AND
Final; 6/75-2/78
14. SPONSORING AGENCY CODE
EPA/600/13
is. SUPPLEMENTARY NOTES TERL-RTP project officer is Frederick A. Roberts, EPA/ERC,
200 S. 35th St. , Corvallis, OR 97330 (503/420-4715).
16. ABSTRACT,
The report gives results of an investigation of water recycle/treatment/
reuse alternatives in coal-fired power plants. Five power plants from representative
U.S. regions were studied. The major water systems encountered were cooling, ash
sluicing, and SO2/particulate scrubbers. Results were used to provide general im-
plementation plans for the various options identified. Computer models were used
to identify the degree of re circulation achievable in each water system without for-
ming scale. The effects of makeup water quality and various operating parameters
were determined for each water system. Several alternatives for minimizing water
requirements and discharges were studied for each plant, and rough cost estimates
were made for comparison. An implementation plan is presented for each water sys-
tem and is divided into phases, including system characterization, alternative eval-
uation, pilot studies, and full-scale implementation. This volume presents detailed
studies for each plant, plant selection methodology, laboratory ash sluicing
studies, kinetics for CaCOS and Mg(OH)2 precipitation, and model descriptions.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. COSATl Field/Group
Pollution
Water Treatment
Filtration
Circulation
Combustion
Cooling Water
Scrubbers
Sulfur Oxides
Electric Power Plants Dust
Coal Mathematical Model
Pollution Control
Stationary Sources
Water Recycle/Reuse
Ash Sluicing
Par ticu late
13B
07D
10B
21B
13A
07A
07B
11G
12A
3. DISTRIBUTION STATEMENT
Unlimited
19. SECURITY CLASS (Tins Report}
Unclassified
21. NO. OF PAGES
576
20. SECURITY CLASS (This pagej
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
L-120
------- |