p O A U.S Environmental Protection Agency Industrial Environmental Research
,„»• I f^ Off -e of Research and Development Laboratory
Laboratory
Research Triangle Park, North Carolina 27711
EPA-600/7-78-055a
March 1978
           WATER RECYCLE/REUSE
           ALTERNATIVES IN COAL-FIRED
           STEAM-ELECTRIC POWER PLANTS
           Volume I. Plant Studies and
           General Implementation Plans
           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.
                           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-600/7-78-055a
                                       March 1978
      WATER RECYCLE/REUSE
  ALTERNATIVES IN COAL-FIRED
STEAM-ELECTRIC POWER PLANTS:
      Volume I.  Plant Studies and
    General Implementation Plans
                      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

-------
11

-------
                           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 SC-2/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 S02/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 CaCCh and
Mg(OH)2, and a description of the models used in this study.
                               iii

-------
IV

-------
                           VOLUME I
                           CONTENTS

                                                            PAGE
          ABSTRACT	  iii
          CONTENTS, VOLUME I 	   iv
          CONTENTS, VOLUME II 	viii
          FIGURES, VOLUME I 	xxii
          TABLES, VOLUME I 	xxiii
          ACKNOWLEDGEMENT 	   xv
PLANT STUDIES
1.0       INTRODUCTION 	    1
          1.1  Program Objectives 	    1
          1.2  Summary	    1
          1.3  Conclusions and Recommendations 	    4
2.0       COOLING SYSTEMS 	    7
          2.1  Process Description 	    8
          2.2  Process Variables 	   13
               2.2.1  Cycles of Concentration	   13
               2.2.2  Relative Saturation 	   14
          2.3  Treatment Options 	   15
               2.3.1  Acid Treatment	   16
               2.3.2  Calcium Removal 	   18
          2.4  Recycle/Reuse Alternatives in Cooling
               Towers 	   26
3.0       ASH SLUICING SYSTEMS 	   30
          3.1  Process Description 	   30
          3.2  Process Variables 	   33
               3.2.1  Ash Reactivity	   34
                               v

-------
                           VOLUME I
                           CONTENTS
                          (Continued)

                                                             PAGE
                3.2.2  Makeup Water Quality 	    36
                3.2.3  Carbon Dioxide  Transfer 	    40
                3.2.4  Degree of Recycle	    44
           3.3  Recycle/Reuse Alternatives in Ash Sluicing
                Systems 	    44
4.0        SO2/PARTICIPATE SCRUBBING SYSTEMS 	    47
           4.1  Process Description	•	    47
                4.1.1  Gas Cleaning	•---    49
                4.1.2  Solids Precipitation	•	    49
                4.1.3  Solid/Liquid Separation 	    50
           4.2  Process Variables 	    51
                4.2.1  S02 Removal Rate	    52
                4.2.2  Ash Removal Rate	•	    54
                4.2.3  Ash Pond Recycle Rate	    58
                4.2.4  Slurry Solids Concentration 	    58
                4.2.5  Reaction Tank Size	    61
                4.2.6  Liquid-to-Gas Ratio 	    63
                4.2.7  Makeup Water Quality 	    66
           4.3  Recycle/Reuse Alternatives in SOa/
                Particulate Scrubbing  Systems 	    68
5.0        COMBINED SYSTEMS 	    70
           5.1  APS Four Corners 	    70
           5.2  PSC Comanche	    73
           5.3  GPC Bowen	    76
           5.4  PP&L Montour	    79
           5.5  MPC Colstrip 	    81
           5.6  Recycle/Reuse Alternatives in Combined
                Systems 	    83

-------
                            VOLUME I
                            CONTENTS
                           (Continued)
                                                            PAGE
GENERALIZED IMPLEMENTATION PLANS
1.0       INTRODUCTION	  87
2.0       COOLING TOWERS 	  89
          2.1  Phase I:   Cooling Tower  Characterization 	  92
               2.1.1  Identification of Process Variables -  92
               2.1.2  Sampling Program	  96
          2.2  Phase II:  Evaluation of Operating
               Alternatives 	  97
               2.2.1  Evaluation Criteria 	  98
                      2.2.1.1  Scale Potential 	  98
                      2.2.1.2  Treatment Alternatives 	 100
               2.2.2  Model Application 	 103
          2.3  Phase III:   Cooling Tower Modifications 	 105
               2.3.1  Equipment 	 105
               2.3.2  Implementation	 106
3.0       ASH SLUICING	 110
          3.1  Phase I:   Ash System Characterization 	 112
               3.1.1  Identification of Process Variables - 112
               3.1.2  Sampling	 114
               3.1.3  Laboratory Studies 	 114
          3.2  Phase II:  Evaluation of Operating
               Alternatives 	 117
               3.2.1  Evaluation Criteria	 118
                      3.2.1.1  Scale Potential 	 119
                      3.2.1.2  Treatment Alternatives 	 119
               3.2.2  Model Application 	 121
          3.3  Phase III:   Ash Sluicing Pilot Studies 	 122
          3.4  Phase IV:  Full-Scale Implementation 	 126
               3.4.1  Equipment 	 126
               3.4.2  Implementation	 127
                                vii

-------
                           VOLUME I
                           CONTENTS
                          (Continued)
                                                            PAGE
4.0       S02/ PARTICIPATE SCRUBBING	 128
          4.1  Phase I:  System Characterization 	 130
               4.1.1  Identification of Process Variables - 130
                      4.1.1.1  Scrubber and Demister Data - 131
                      4.1.1.2  Reaction Tank Data	135
                      4.1.1.3  Solids Concentration Data -- 137
               4.1.2  Sampling Plan	 137
                      4.1.2.1  Analytical Measurements 	 139
                      4.1.2.2  Process Measurements 	 142
          4.2  Phase II:  Alternative Evaluation 	 145
               4.2.1  Evaluation Criteria 	 145
                      4.2.1.1  Scale Potential 	 146
                      4.2.1.2  Dissolved Solids
                               Concentration 	 148
               4.2.2  Model Application 	 148
          4.3  Phase III:  Pilot-Scale Studies 	 150
               4.3.1  Converting to Closed-Loop Operation - 150
               4.3.2  Changing Makeup Water Source 	 152
          4.4  Phase IV:  Full-Scale Operations 	 153
5.0       SUMMARY	 154
REFERENCES  			 156
                                vzn

-------
                           VOLUME II
                            CONTENTS

                                                        PAGE
         CONTENTS, VOLUME I 	    iv
         CONTENTS, VOLUME II 	  viii
         FIGURES	  xxii
         TABLES 	xxiii
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
                              IX

-------
                           VOLUME II
                            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  CaC03	  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
                               x

-------
                           VOLUME II
                            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 3		  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
                              XI

-------
                           VOLUME II
                            CONTENTS
                          (Continued)
                                                        PAGE
       5.3  Discussion of Results 	   C-30
       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_]_
2.0    MODEL DESCRIPTIONS 	   E-2
       2.1  Ash Sluicing (Bowen, Montour,
            Comanche) 	   E-3
                              xix

-------
                           VOLUME II
                            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
                              xiii

-------
                           VOLUME II
                            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
                             xiv

-------
3.0
                           VOLUME II

                            CONTENTS

                          (Continued)
                                                        PAGE
2.3


Ash Handling Systems 	
2.3.1 Simulation Basis 	
2.3.2 Simulation Results 	
TECHNICAL ALTERNATIVES 	 • 	 	
3.1







3.2












3.3
Cooling Tower System 	
3.1.1 Simulation Basis 	
3.1.2 Effect of Increased Cycles
of Concentration 	
3.1.3 Effect of Calcium Concentra-
tion in the Makeup Water 	
3.1.4 Summary of Cooling Tower
Alternatives 	
Ash Handling Systems 	
3.2.1 Simulation Basis 	
3.2.2 Once-Throueh Ash Sluicing
System 	
3.2.3 Recirculating Ash Sluicing
System 	
3.2.4 Effects of Carbon Dioxide
Mass Transfer 	
3.2.5 Effect of CaSO^'ZI^O
Supersaturation in the Pond
Recycle Water 	
3.2.6 Summary of Ash Sluicing
Operations 	
Conclusions 	
ECONOMICS 	
G-16
G-17
G-20
G-22
G-22
G-22

G-23

G-27

G-33
G-33
G-34

	 G-34

	 G-37

	 G-40


	 G-43

	 G-44
	 G-44
	 G-47
4.0
APPENDIX H - RECYCLE/REUSE OPTIONS AT COMANCHE
             (PUBLIC SERVICE OF COLORADO)
                              xv

-------
                           VOLUME II
                            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	  X_]_
 2.0'   PLANT CHARACTERISTICS 	  X_5
       2.1  Water Balance	  j_5
       2.2  Cooling Tower System	  I_g
                             xvi

-------
                           VOLUME II
                            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 CaS
-------
                           VOLUME II
                            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
            3.2.1  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
                            xvi 11

-------
                           VOLUME II
                            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
            2^2.4  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-15
            3.2.2  Pond Evaporation	  K-16
            3.2.3  Ash Dissolution	  K-17
                             xix

-------
                           VOLUME II
                            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
                               xx

-------
                          VOLUME II
                           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
       2.1  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
                             xxi

-------
                          VOLUME II
                           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
                             xxi i

-------
                           VOLUME I
                           FIGURES

                                                             PAGE
PLANT STUDIES
Figure 2-1.  General cooling tower system flow scheme. 	  9
Figure 2-2.  Calculated acid requirements as a function of
             cycles of concentration. 	 19
Figure 2-3.  Gypsum relative saturation variation with
             cycles of concentration. 	 22
Figure 2-4.  Solution deviations from ideality with cycles
             of concentration. 	 23
Figure 2-5.  Slipstream rate as a function of makeup calcium
             concentration at Colstrip. 	 27
Figure 2-6.  Slipstream rate as a function of makeup sulfate
             concentration at Colstrip. 	 28
Figure 3-1.  Typical once-through ash sluicing flow scheme. - 32
Figure 3-2.  Recirculating ash sluicing flow scheme. 	 32
Figure 3-3.  Reactive calcium in fly ashes as a function
             of pH. 	 37
Figure 3-4.  Reactive sulfate in fly ashes as a function of
             pH. 	 38
Figure 4-1.  Typical scrubbing system flow scheme. 	 48
GENERALIZED IMPLEMENTATION PLANS
Figure 2-1.  General cooling tower system flow scheme. 	 90
Figure 2-2.  CaCOs scale potential as a function of cycles of
             concentration in cooling towers without treat-
             ment  (example) . 	 101
Figure 2-3.  Cooling tower blowdown pH as a function of
             cycles of concentration in towers without
             treatment (example). 	 102
Figure 3-1.  Recirculating ash sluicing flow scheme. 	 111
Figure 3-2.  Bench-scale model of ash sluicing facilities. - 116
Figure 3-3.  General pilot scale ash sluicing facility. 	 123
Figure 4-1.  Typical scrubbing system flow scheme. 	 132
Figure 4-2.  Sample points for scrubbing system charac-
             terization.  	 140
                                 XX1L1

-------
                           VOLUME I

                            TABLES
PLANT
Table

Table

Table

Table


Table


Table


Table



Table


Table



Table


Table


Table


Table


Table


Table


Table


Table


Table
                                                       PAGE

STUDIES
1-1.   Selected Plants for Water Recycle/Reuse Study --   2

2-1.   Cooling Tower System Design Parameters 	  10

2-2.   Normalized Cooling Tower Design Parameters 	  12

2-3.   Effects of Slipstream Softening on Treatment
      Rate at Colstrip 	  25
3-1.
3-2
3-3
3-4.
3-5
3-6
4-1.
4-2
4-3
4-4.
4-5
4-6,
4-7,
4-8
Ash Reactivity Determined from Leaching
Studies 	  35

Effects of Makeup Water Quality on Recirculating
Ash Sluice System Scaling Potential 	•--  39

Effects of Makeup Water Quality on Treatment
Required for Closed-Loop Ash Sluicing of
Montour Ash	  41

Effects of CO2 Transfer in Pond on Fly Ash
Slurry Scaling Tendency (Bench-Scale Results) --  42

Effects of CO2 Transfer in Pond on Fly Ash
Slurry Scaling Tendency (Computer Model
Results) 	  43

Effects of Degree of Recycle in an Ash Sluicing
System Using Montour Ash	  45

Effects of SO2 Removal Rate on Four Corners
Scrubber Makeup Requirements 	  53

Effects of Ash Removal Rate on Scrubber Makeup
Requirements 	

Effect of Ash Removal Rate on Colstrip Scrubbing
System Scaling Tendency 	

Effect of Pond Recycle on Four Corners Scrubber
Makeup Requirements 	
Effect of Ash Pond Recycle on Four Corners
Scrubbing System Scaling Tendency 	
                                                        56
                                                        57
Effect of Slurry Solids Concentration on Colstrip
Scrubber Scale Potential 	"

Effect of Reaction Tank Volume on Four Corners
Scrubbing System Scaling Potential 	

Effects of Liquid-to-Gas Ratio on Four Corners
Scrubbing Operation 	
                                                        59
                                                        60
                                                        62
                                                        64
                                                               65
                                XXIV

-------
                           VOLUME I
                            TABLES
                          (Continued)

                                                             PAGE
Table 4-9.  Effects of Makeup Water Quality on Colstrip
            Scrubbing System Scaling Potential 	   67
Table 5-1.  Summary of Recycle/Reuse Options at Four
            Corners 	   72
Table 5-2.  Summary of Water Recycle/Reuse Options at
            Comanche	   74
Table 5-3.  Summary of Technically Feasible Options at
            Bowen	   77
Table 5-4.  Summary of Technically Feasible Options at
            Montour	   80
Table 5-5.  Summary of Water Recycle/Reuse Options at
            Colstrip 	   82
GENERALIZED IMPLEMENTATION PLANS
Table 2-1.  General Data Sheet for Cooling Towers 	   93
Table 4-1.  Scrubber and Demister Data	  133
Table 4-2.  Reaction Tank Data	  136
Table 4-3.  Solids Concentration Data	  138
Table 4-4.  Scrubbing System Characterization Measurements
            Sample Points and Analyses 	  141
Table 4-5.  Process Data Requirements 	  143
                              xxv

-------
                         ACKNOWLEDGEMENT
          This study was conducted by Radian Corporation, Austin,
Texas, for EPA under Contract No. 68-03-2339.  Mr. Fred A. Roberts
of the Corvallis Environmental Research Laboratory was the EPA
Project Officer.  The Radian program staff included Dr. Frank G.
Mesich as Program Manager and Dr. Delbert M. Ottmers as Technical
Project Director for the first year.  Mr. James G. Noblett served
as Technical Project Director for the completion of the project.


          The final report was prepared by Mr. James G. Noblett
and Mr. Peter G- Christman with Dr.  Delbert M. Ottmers providing
review.  Drs.  Ron E. Pyle and Frank B. Meserole coordinated the
precipitation kinetics work.  Dr. Frank B. Meserole also coordi-
nated  the sampling and analyses efforts, and the ash characteri-
zation studies.  Mr. Michael Fuchs and Mr. Michael Ellsworth were
primary contributors to the ash characterization studies.


          Mr.  Wayne A. Gathman was the primary contributor to the
studies at Comanche and Mr. 0. W. Hargrove was a primary contri-
butor  to the Four Corners studies.  The remaining plant studies
were  conducted by Mr. James G. Noblett and Mr. Peter G. Christman.


          Acknowledgement is also given to the personnel from
Georgia Power Co., Arizona Public Service Co., Public Services
of Colorado Co., Montana Power Co.,  Pennsylvania Power and Light
Co.,  and Southern Services, Inc. whose cooperation greatly facil-
itated this work.
                              xxvi

-------
                         PLANT  STUDIES
1.0       INTRODUCTION
          Achievement of the national goal of "zero-discharge" of
pollutants to the environment will necessitate that industrial
water users consider possibilities for recycle or reuse of waste-
water streams.  Electric power plants may presently incorporate
some recycle/reuse of wastewater for expediency or short-term
economy, but limited availability and rising costs of water,
together with water treatment requirements has placed an in-
creased importance on recycle/reuse.


1.1       Program Objectives

          The overall objectives of this project are to provide
general technical and economic evaluations of water recycle/reuse
and treatment options for coal-fired power plants through the
use of computer process models.


          The primary objective is to make a reliable technical
assessment of various recycle/reuse alternatives for five region-
ally representative power plants chosen for study.  This includes
determining the system sensitivity to operating parameters which
are subject to variations.


          A second objective is to prepare rough cost estimates
for the purpose of complementing the technical assessment in
determining the most attractive options for each plant situation
encountered.  These rough cost estimates are for comparative
purposes and are based on nominal sizing of process and treatment
equipment.  Rough cost estimates are included for capital and
operating costs.


1.2       Summary

          The first step in evaluating water recycle/reuse op-
tions for power plants was to select five typical plant situa-
tions.  The four main criteria used in selecting the five plants
for study are location, availability, site characteristics, and
project timing.  Table 1-1 presents the five plants selected for
study and a general description of the water systems found at
each plant.  A detailed discussion of the plants condsidered and
the criteria used for selection is presented in Appendix A.  A
summary of the plant water system alternatives, including rough
economics for comparison purposes, for the five plants studied
is presented in Section 5.0 (beginning on p. 70)..

-------
                TABLE 1-1.   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
Mont our
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
1 Handling2
WSB
WSF
WSB
WSB
WSF
WSF
WSB
WSB
WSF
Part.
Control3
ESP>
Venturi
ESP
ESP
ESP
Venturi
S02
Control'1
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
''DC  = under construction

-------
          Once the five plants were chosen,  spot  samples of the
major water streams in each plant were taken to aid in charac-
terizing each plant's water system.  These samples were taken
for only one day's operation and represent operating conditions
on that day only.  Additional sample analyses to more fully char-
acterize a plant's water system would be required before any of
the water recycle/reuse alternatives in this study would be im-
plemented.  The results from the samples taken are presented in
Appendix B.  Makeup waters to the plants varied from about 60 mg/£
total dissolved solids to about 450 mg/S, total dissolved solids.


          In order to evaluate CaCO3 and Mg(OH)2  scale potential
in new process configurations for power plant water systems, ex-
perimental work was conducted to determine precipitation rates
as a function of relative saturation for these species.  The re-
sults of these experiments (see Appendix C)  show  that at a rela-
tive saturation of about 2.5 for CaCO3 and about  3.4 for Mg(OH)2,
there is a sharp increase in precipitation rates due to nuclea-
tion.  This indicates that at relative saturations above 2.5 for
CaCO3 and 3.4 for Mg(OH)2, scale formation is likely.


          In addition to the experimental studies concerning
CaCO3 and Mg(OH)2, laboratory studies were conducted to deter-
mine the solubility characteristics of fly ash collected from
each of the five plants.  These data were collected to enable
evaluation of recirculating ash sluice systems.  The results of
these studies for the Montour and Colstrip ashes are presented
in detail in Appendix D.  The results for the remaining three
plants are presented in detail in the final  report for EPA Con-
tract No. 68-02-1319, Ash Characterization Studies, which was
performed in support of this program.This  report is included
as Appendix L.


          Computer process models were used  to investigate vari-
ous recycle/reuse alternatives at the five selected power plants.
The models used in this study are discussed  in detail in Appen-
dix E.  Detailed discussions of the studies  conducted at each
plant are presented in Appendices F-J.  It should be noted here
that these analyses were performed to study  general water recycle/
reuse alternatives.  Actual implementation of any of the alter-
natives would require a more extensive investigation of process
parameter variability.  Additional data concerning makeup water
quality, fuel composition and seasonal flow  and load variations
would be required before a detailed design could be made.  The
types of major water systems encountered at  these plants include
cooling towers, bottom and fly ash sluicing, and particulate/S02
scrubbing.
                                -3-

-------
          In Sections 2.0-4.0, the results of the plant studies
are summarized into discussions of water recycle/reuse options
in cooling towers, fly ash sluicing, and particulate/S02 scrub-
bing systems.  Section 5.0 discusses combined system alterna-
tives for each of the five plants studied and presents the rough
economic estimates for comparison of alternatives at each plant.
It should be emphasized here that the rough cost estimates were
made for comparative purposes only and may not be representative
of the actual cost to be expected for implementing particular
alternatives.  Retrofit problems were not considered in the cost
estimates since evaluation of retrofit would necessitate forming
a detailed design of each alternative.


1.3       Conclusions and Recommendations

          The major tool used in this study was a computer simu-
lation package capable of modeling the major water consumers
(cooling tower, ash sluicing, and scrubbing systems) at coal-
fired electric power plants.  The models used were verified using
the results of spot samples taken at five different plants.   The
verified models were used to evaluate various water recycle/reuse
opportunities at each of these five power plants.  The results of
the computer simulations showed that improvements over existing
operations could be made which would reduce water use and dis-
charge .


          In cooling tower cases where CaC03 or CaSCK'ZHzO are
the limiting scale-forming species, recirculation of the cooling
water may be increased so that the entire blowdown can be used as
makeup to another major water consumer(s) in the plant.  Sulfuric
acid addition and/or lime softening may be required to achieve
this degree of recirculation depending on the plant makeup water
quality.  Kinetic studies such as those presented in Appendix C
for CaC03 and Mg(OH)z are recommended for silica-based scale-for-
ming species so that the maximum safe degree of recirculation in
tower systems where these solids are limiting may be more ade-
quately  defined.  Other factors which may limit the cycles of
concentration in cooling towers are the chloride level and the
suspended solids level.  High chloride concentrations may cause
corrosion problems.  High suspended solids concentrations (from
makeup water or scrubbed from the air) can cause erosion and
plugging in the condenser.


          Most fly ash sluicing operations are once-through with
the ash  pond overflow discharged.  This study revealed that recir-
culating fly ash sluice systems may be used but treatment may be
                               -4-

-------
necessary to control gypsum scale potential.  The amount of leach-
able species in the ash and the makeup water quality are the major
parameters which determine the level of treatment required.  The
computer simulations performed for recirculating ash sluicing
systems were based on ash reactivity determined from beaker leach-
ing studies, and represent "worst case" operation.


          Bench-scale experiments showed that ash reactivity
varies considerably with liquor pH and composition.  Pilot or
additional bench-scale studies are recommended to more accur^
ately determine the reactivity of a particular ash before im-
plementing a recirculating fly ash sluice system.


          Carbon dioxide transfer to the pond liquor from the
atmosphere increases CaCO 3 and decreases Mg(.OH) 2 scale poten-
tial in recirculating ash sluice systems.  Installation of re-
action vessels prior to the slurry line may allow controlled
precipitation of these solids to prevent line plugging.  Pilot
or bench-scale studies are recommended to better identify CaC03
and/or Mg(OH)2 scale potential for a particular ash before a re-
circulating system is built.


          Since lime-based S02/particulate scrubbing systems
may be operated in a closed-loop fashion, there is little op-
portunity for recycle/reuse within the scrubbing system.  How-
ever, cascaded water may be used as scrubber makeup water (ex-
cluding demister wash) in the final step of the cascaded sys-
tem, resulting in zero discharge.  Possible sources for the cas-
caded water are cooling tower Slowdown and ash pond overflow.
Combining cascaded water with fresh makeup water and using this
mixture as demister wash may be feasible depending on the re-
spective water compositions and the amount of S02 removed in
the demister.  A careful process analysis and pilot studies are
suggested for determining the quality of water required for de-
mister wash in a particular scrubbing system.


          The results of this study point out that computer models
are effective in identifying water recycle/reuse options in
cooling tower, ash sluicing, and S02/particulate scrubbing sys-
tems.   However, the use of these predictions for designing actual
modifications is limited by the model assumptions and the input
data.  The major assumption made in the cooling tower models is
that complete C02 equilibrium between the circulating water and
the atmosphere is achieved in the tower.  If equilibrium is only
approached, higher total carbonate concentrations than predicted
will be observed in the system.  This will cause the observed pH
                               -5-

-------
to be lower than predicted.  This lower pH will make the scaling
potential of CaC03 lower since the pH has a greater effect on
CaCCh scale potential than the total carbonate concentration.
The cooling tower models' results should therefore represent^
slightly conservative mode of operation with respect to scaling,
but this should have a small effect on acid requirements,  so the
acid rates predicted should be realistic.


          The ash sluicing model predictions are perhaps overly
conservative since the ash reactivity was assumed to be constant
and equal to the amount leached by deionized water.  This repre-
sents a "worst case" operation since bench-scale ash sluicing
experiments showed that more concentrated liquors tended to
leach less from the ash.   The actual degree of C02 transfer oc-
curring in ash ponds also has not been adequately defined.  Com-
puter simulations were performed to bound the level of C02 trans-
fer by assuming equilibrium with the atmosphere in one case and
no transfer in another.  The actual amount of transfer will be
between these two predictions.


          The scrubbing models used in this study are based on
precipitation kinetics data and material and energy balances.  A
key assumption in the scrubbing model is the amount of SOa sorbed
that is oxidized.  Required reaction tank volumes will vary con-
siderably with oxidation since in most cases they are sized based
on CaSOit*2H20 relative saturation.  The cases studied all assumed
greater than 9070 oxidation and therefore represent a worst case
situation.  If less oxidation occurs, the predicted hold tank vol-
ume and CaSOit-2H20 relative saturation will be overly conservative
with respect to gypsum scale formation.


          The input data to all of these models were based on de-
sign operating parameters where available.  Oxidation in the
scrubbing systems was assumed to remain at the level determined
from the spot samples.  Makeup water compositions were also based
on spot samples.  These spot samples represent one day's operation
and do not reflect variations in the water quality.  Although
spot samples are sufficient for preliminary judgements, additional
data concerning process parameter variability would be required to
safely predict conditions under alternative operating conditions.
                               -6-

-------
2.0       COOLING SYSTEMS

          Approximately 40 percent of the energy produced in the
boiler of a coal-fired steam-electric generating plant leaves
the plant in the form of electricity.  The remaining 60 percent
leaves the plant as waste heat with the stack gases and the cool-
ing water.  The energy lost with the stack gases is about 10% of
the energy rejected by the condensers.  Therefore, the largest
single end use for the energy consumed at a power plant is in
low temperature rejection of warm water from the condensers.


          The ultimate disposal of this waste heat can be handled
in two basic ways.  The least expensive way to supply cooling
water to a power plant is via a once-through cooling system.  In
such a system cooling water is taken from a stream, river, lake
or ocean and is rejected, at a higher temperature, at some other
point away from the intake.  This type of cooling is considered
less environmentally acceptable than other methods because of
the biological effects associated with warming natural bodies of
water.  Recently power plants have been built with recirculating
cooling systems which reuse the cooling water and dispose of the
waste heat to the atmosphere.


          Recirculating systems include a broad range of cooling
towers and cooling ponds.  Wet cooling towers are in more general
use because they require less land than do cooling ponds, and are
much less expensive than dry cooling towers.  Wet systems gener-
ally reject heat from the cooling water to the atmosphere by sen-
sible heat exchange (~20%) and evaporation (~80%).  Dry systems
depend entirely on sensible heat transfer to reject their heat.
All recirculating cooling systems are more expensive to build and
operate than are once-through systems.


          Four of the coal-fired steam-electric generating plants
studied employ wet cooling towers to dispose of their waste heat.
The four plants were Georgia Power Company's Plant Bowen, Colorado
Public Service's Comanche SES, Montana Power Company's Colstrip
SES, and Pennsylvania Power and Light's Montour SES.  These cool-
ing systems were studied and alternative operauing modes were
suggested which would decrease the water consumption required
under normal operating conditions.


          In this section the results of these studies will be
presented, including a general description of wet cooling towers,
as well as the four specific systems studied.  This is followed
by a description of the important factors which affect
                                -7-

-------
cooling tower operation and limit the recycle options.   Finally,
the proposed alternative operating modes which resulted from this
study are presented.


2.1       Process Description

          Wet cooling towers serve the function of contacting a
recirculating water stream with the atmosphere.  To obtain the
maximum amount of cooling, these towers bring air and water into
intimate contact on the wetted surface of the fill in the tower.
Theoretically, the water can reach the ambient wet bulb tempera-
ture in the tower, although in actual practice the approach to
the wet bulb is closer to 3-8°C.  The design of large cooling
towers, such as those used in power plants,  is therefore highly-
dependent on the ambient conditions which will be expected during
the operational lifetime of the tower.


          Wet cooling towers are net water consumers.   Losses
occur from the tower in three major ways:  evaporation, drift,
and blowdown.  The sum of these three streams is equal to the
makeup requirements for the tower.  Evaporation accounts for
the majority of the water lost from the tower.  Drift occurs be-
cause fine droplets of water are carried off as a mist due to
the intimate contact of the air and water.  Most cooling towers
have mist eliminators designed to minimize drift.   The blowdown
is a purge stream which is used to control the level of dissolved
solids which will build up in the recirculating water.   Figure
2-1 shows the general flow scheme for a cooling tower system
including the tower and condenser.


          The evaporation and drift rates are set by the design
of the tower, the ambient conditions, and the cooling load.  The
blowdown rate is maintained at a level sufficient to prevent
scale formation in the condenser but in many cooling towers the
size of the blowdown is much larger than required to prevent
scale formation.  By better defining the limits of scale forma-
tion in recirculating cooling systems, the blowdown can be reduced
and thus the net water consumption and discharge from these sys-
tems can be reduced.  The major effort expended in the cooling
tower systems was directed toward reduction of the blowdown re-
quired for safe operation.


          The four cooling systems studied varied in location
size, and design.  Table 2-1 presents information about the de-
sign conditions of the different cooling systems.  The data are
presented for one cooling tower under design summer conditions.
                              -8-

-------
                             EVAPORATION
                                      DRIFT
MAKEUP
                           COOLING
                            TOWER
                          CONDENSER
-*- SLOWDOWN
   Figure  2-1.   General cooling tower system flow scheme
                               -9-

-------
                                   TABLE  2-1.    COOLING  TOWER  SYSTEM  DESIGN PARAMETERS
          Type
                                      Georgia Power  Co.*
                                          Plant  Bowen
                         Hyperbolic  natural draft
                                                    Colorado Public Service
                                                           Comanche
                                                                   Induced draft
                                                       Montana  Power Co.
                                                           Co 1st rip
                        Penn  Bower k l.ii'.hi.
                              Mont our
                                                                                             Induced draft        Hyperbolic natural  diali
          Number of Towers
          Electric Generating

          Capacity per  Tower, MW
                                    700

                                    880
                                                                        350
                                                                                                  350
                                                                                                                            730
          Circulating Water  Rate

          per Tower,  I/sec  (GPM)
                             16.000 (260,000)

                             19.000 (310,000)
                              9.100 (145.000)
6,500 (100,000)
                                                                                                                     15,800 (250,000)
 I
M
O
Temperature Change

Across Condenser.  °C  (°F)



Air 1'low Kate  per

Tower, M'/hr (ACFM)
        14 (26)

        16 (28)



2.5 x 107  (1.45 x 10')

3.5 x 10'  (2.07 x 10')
                                                                      14  (26)
                                                                                      18  (32)
                              16 (28)
                                                               2.7  x  10'  (1 6 x 10')     1.7  x  107  (1.0 x 107)      3.5 x 107 (2.06 x 10')
          Cooling Tower Drift

          Kate,  ft/sec (GPM)
                                 3.3  (52)

                                 3.4  (62)
                                 9.0  (142)
  1.3 (20)
                                                                                    31.b  (500)
         Cooling Water Approach,
         °c no
                                  U  (19)

                                  10  (18)
                                   8  (15)
   12 (22)
  10 (19)
         Evaporation Rate per

         Tower,  H/sec (GPM)
                                340  (5,500)

                                400  (6,500)
                                190  (3.000)
 160 (2,600)
390 (6,200)
         *l'lant Howen has  cooling towers of two different sines.  The firsr line refers to Unila  1  &  2 and the second line

          refers to Units  3  & 4.

-------
Two different sizes of towers, are used at Plant Bowen and the
data for both sets of towers are presented.


          Comanche and Colstrip have induced draft cooling towers
This is the most common type of cooling tower found in the United
States.  Induced draft towers have large fans mounted in the top
of the towers which move the air which contacts the cooling water,
The cooling towers at Bowen and Montour are hyperbolic natural
draft towers.  The natural draft towers do not use fans to move
the air through the tower, but instead take advantage of the
buoyancy effects that occur when the air is heated to a tempera-
ture higher than the ambient dry bulb temperature.  Natural draft
towers therefore require high relative humidities to operate ef-
fectively, such as the conditions found in the eastern part of
the United States.
          The design parameters for the temperature change across
the condenser, and the cooling water approach do not vary much
among the different tower designs.  The remaining parameters do
not compare readily because of the large differences in the power
generating capacities of the different plants that the towers
serve.  Table 2-2 presents parameters for the different plants
on a normalized basis.  The parameters presented in this table
show that they do not vary widely from plant to plant.


          The cooling towers at Colorado Public Service's Com-
anche SES are the most conservatively designed.  The towers at
Comanche are designed to discharge 360 kcal/sec per MW of elec-
tric generating capacity, the largest value for the four plants.
Also, Comanche has the largest air and water flow rates per MW,
with the design air flow rate more than twice as large as the
normalized air flow for towers 1 and 2 at Bowen.  The design
drift rates reported are values that the cooling tower manufac-
turer guarantees the drift will not exceed.  Drift rates are
very hard to measure and estimates vary significantly.  The
values reported in Table 2-2 range from 0.02% to 0.20% of the
recirculating water.  The evaporation rates from the different
towers are in close agreement ranging from 2.1%, to 2.6% of the
total recirculating water flow.


          These cooling systems have been chosen as a sample of
the cooling towers presently in use by the-electric utility in-
dustry in the United States.  In the next section the important
variables that affect the reuse options available to these cool-
ing systems will be identified.
                               -11-

-------
                     TABLE  2-2.    NORMALIZED  COOLING TOWER DESIGN PARAMETERS
 Circulating Water Rate,

 (I/sec por MU (Gl'M/MW)
                                 Georgia Power Co.*
                                     Plant Bowen
                        Colorado Public Service
                               Comanche
      23  (365)

      21  (333)
      26  (112)
                           Montana Power Co.
                               Colstrip
                     Penn. Power &
                           Montour
                                                       19 (300)
                                                                                                              21  (333)
Air Flow Rate,

m'/lir per MW (ACt'M/MW)
3.6 x  10"  (2.1 x 10")

3.9 x  10"  (2.3 x 10'')
7.7 x  13" (/i. 5 x 10")     4.9 x 10" (2.9 x 10")   4.7 x  10" (2.8 x HI")
Cool ing Loud,

kcal/sec per MW (Btu/min-MW)
   320  (7 6 x 10")

   340  (8.1 x lO")
   360  (8.6 x 10")
330 (7  9  x  10")
340 (8. 1  x  10")
Drift Rate,  % of

recirculat ing water
      0.02

      0.02
                               0.10
                                                        0.02
                                                      0. 2U
Evaporation  Rate,

"/„ of recirculating water
      2. 1

      2.1
      2.1
                               2.6
                                                      2.5
•'•'Plant  Bowen has towers of two di f ferent designs.  The first line refers  to Units 1 & 2 and the  second line refers to  Units 3 in 4.

-------
2.2       Process Variables

          The major problem associated with the increased recycle
of cooling water in wet cooling tower systems is the danger of
scale formation in the condenser.  Scale can form on the tubes
of the condenser, reduce the efficiency of the condenser, increase
the back pressure in the turbine, and therefore reduce the effi-
ciency of the plant.  The potential for scale formation is a
function of the temperature of the cooling water as well as the
water quality.  The composition of the recirculating water is a
function of the makeup water quality and the cycles of concen-
tration.  In this section the general effects of cycles of con-
centration on water quality and scale potential will be discussed.
These effects will be quantified in Section 2.3.


2.2.1     Cycles of Concentration

          The cycles of concentration is defined as the ratio of
the recirculating species concentration to the makeup species
concentration and can be expressed in terms of flow rates by
Equation 2.1:


                     r - E + B + D _   M                 /
                     C
                           B + D     B + D


where


          C = cycles of concentration

          E = evaporation rate

          B = blowdown rate

          D = drift rate

          M = makeup rate


This expression  is a direct result of a mass balance of the
conserved dissolved species around the cooling tower.


          Water  enters the system as makeup and leaves by evap-
oration  blowdown, and drift.   The dissolved species enter the
system in the makeup and leave in the blowdown and drift.  The
concentrating effect of the wet cooling tower occurs because of
the large amount  of water which leaves the system via evaporation


                               -13-

-------
          The concept of cycles of concentration can be used to
calculate the concentrations of individual dissolved solids in
the recirculating water from the makeup water concentrations.
As the cycles of concentration increase, the recirculating water
species concentrations increase as does the potential for scale
formation in the condenser.  The blowdown stream is used to con-
trol  the cycles of concentration.  As the blowdown decreases,
the cycles of concentration increases as can be seen from Equa-
tion  2.1.  The major thrust of this study was to minimize the
blowdown requirements from the cooling tower without increasing
the cycles of concentration beyond the point of safe operation.


2.2.2     Relative Saturation

          In this study the potential for scale formation was
quantified using the concept of relative saturation.   Relative
saturation is a measure of the degree of the saturation of a
solution with respect to precipitation of a particular solid
and is related to the thermodynamic stability of that solid.
As an example, the relative saturation of CaC03(s)  can be ex-
pressed with the following equation:
                   R c
                   R-S'CaC03
where
          R.S.£ £Q  = relative saturation of CaC0
                    = activity of free calcium ion in solution


              aCQ=  = activity of free carbonate ion in solution


             KCaCO  = solubility product constant for calcium
                  3   carbonate


          The relationship between the total calcium and carbo-
nate present and the free ion activity is explained in Appendix
E.  Some of the calcium and carbonate will be complexed with
other ions.  Thus, the free ions represent only part of the
total species present.
                              -14-

-------
          When the relative saturation for any species is less
than one, the species is subsaturated and no solid formation
occurs.  When the relative saturation is greater than one, the
species is supersaturated and scale may form.  In the presence
of seed crystals, the relative saturation of a species may ex-
ceed 1.0 without scale formation, but solids precipitation will
occur in a controlled fashion by seed crystal growth.  However,
as relative saturation increases above 1.0, a critical value is
reached where precipitation occurs through crystal growth and
nucleation.  When nucleation occurs, the precipitation rate in-
creases rapidly resulting in conditions favorable to scale for-
mation.
          In this study CaSCK-2H20, CaCOa, Mg(OH)2, and silicate
scale potentials were calculated.  Kinetic studies have been per-
formed which determined that the critical relative saturation of
CaSCH'ZHaO is 1.3-1.4.  Kinetic studies reported in Appendix C
determined that the critical relative saturations for CaC03 and
Mg(OH)2 are about 2.5 and 3.4, respectively.  The kinetics of
silicate scale was not studied in this program and literature
results are insufficient to determine critical relative satura-
tions.


          From the form of Equation 2.2 it can be seen that the
relative saturation of the different scale-forming species should
increase as the concentration of the dissolved solids increase.
In this study, operating conditions for cooling towers were de-
signed to maintain the relative saturation of all species less
than or equal to one for the options presented.  This allowed
for a margin of safety equal to the difference between the cri-
tical relative saturation and one.
2.3       Treatment Options

          The blowdown from a cooling tower system is used to
control the level of the dissolved solids in the recirculating
water.  Blowdown alone can be used to control the scaling po-
tential of the recirculating water as is presently done at Bowen
and Montour.  Other methods also exist which will reduce the
scale potential of particular solids.  These methods include
acid treatment for pH control, lime softening to reduce the
level of calcium, and the use of a wide range of scale inhibi-
tors.  In this study, the use of inhibitors was not considered
due to the proprietary nature of many inhibitors and the com-
plex chemistry involved.  A discussion of the treatment options
which can be used to decrease the discharge from cooling systems
is presented along with the impact of new operating modes in this
section.


                               -15-

-------
2.3.1     Acid Treatment

          Sulfuric acid is often used in cooling towers to con-
trol the pH of the recirculating water.  CaC03, Mg(OH)2, and
some of the silicate scales are more soluble in lower pH solu-
tions.  In this study, sulfuric acid was used to keep these
scales from forming at increased cycles of concentration.  The
relative saturation of Mg(OH)2 was well below one for all cases
where acid addition was used to control CaCO 3 scale potential
so the acid addition rate was calculated based on CaC03 relative
saturation in these studies.


          The relative saturation of CaC03 is directly dependent
on the activities of Ca++ and CO! as can be seen in Equation 2.2,
                   R.S.r rn  = -^	^               (2.2)
                       ua^u 3     ^CaCO 3


The concentration and therefore the activity of the carbonate ion
is dependent on the total carbonate species in solution (H2C03,
HCOI, CaHC03+, CaC03, MgHC03+, MgC03, NaHCO3,  NaCOI,  and COD and
the equilibrium distribution of these ions.  A major equilibrium
relationship is the one between the carbonate and bicarbonate
ions shown in Equation 2.3.
                  HC(r'(a,) *   (aq) + C0°(aq)
From Equation 2.3 it can be seen that an increase in pH (decreas-
ing the H+ concentration) has the effect of shifting the reac-
tion to the right and increasing the concentration of the carbo-
nate ion.


          Since C02 transfer can occur between the air and the
water in the cooling tower, the total amount of carbonate spe-
cies is a function of the equilibrium shown in Equation 2.4.


                  C02(g) +H20(1) JH2C03(aq)           (2.4)



As the pH increases, this equilibrium tends to shift to the right,
increasing the total amount of carbonate species in solution
                               -16-

-------
          The above equations point out  the  strong effect pH has
on the relative saturation of CaC03.  They also show that the
relationship between pH and the relative  saturation of CaCO 3 is
not simple.  It is important to account  for  these factors when
determining the amount of acid that must  be  added to control the
relative saturation of CaCO 3.


          In this study a computer model  of  the cooling tower
systems (Appendix E) was used to determine the amount of acid
treatment required to keep the relative  saturation of CaCO3 be-
low one.  This model takes into account  the  equilibrium rela-
tionships among the dissolved species as  well as the gas/liquid
equilibrium described in Equation 2.4.   Given the inputs of air
flow, temperature and composition, makeup water composition,
flow and temperatures of the circulating  water, drift rate,  and
cycles of concentration, the model performs  iterative calcula-
tions around the cooling loop to determine the blowdown, evap-
oration and makeup rates, and compositions for all water streams.
An acid addition rate is determined to keep  the CaC03 relative
saturation within a specified range (less than 1.0).  If slip-
stream softening is required (determined  by  model) the slipstream
and chemical addition rates are calculated.  Several assumptions
are inherent in performing cooling tower  simulations and are lis-
ted below:
          1)  Equilibrium exists between C02 and H20 in the
              atmosphere and the 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)  Softening calculations are based on a calcium
              reduction to 50 mg/& to account for treatment
              inefficiencies.


          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
C02 equilibrium is conservative since the partial pressure of
C02 in actual cooling towers tends to be greater than the equi-
librium value.  The lower equilibrium concentration of carbonate
                               -17-

-------
 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 de-
 crease .


          Figure 2-2 presents the calculated acid addition
 rate required to control CaCO3 scale as a function of cycles
 of concentration for the cooling systems at Montour and Bowen,
 two systems that presently operate without acid addition.
 This graph shows that the acid required to control CaC03 rises
 with the cycles of concentration in a non-linear fashion. _It
 also shows that the change in acid requirements as a function
 of cycles (the slope) varies with makeup water composition.
 Both plants presently operate their cooling towers at less than
 3 cycles of concentration and relative saturations of CaCO 3 much
 less than one.  Performing simulations of these systems has
 shown  that the makeup water requirements as well as the blowdown
 can be  reduced using only sulfuric acid to inhibit CaC03 scale
 formation.


          The makeup water quality determines which species are
 most likely to scale.  The water used for makeup at Comanche is
 high in silica.  The relative saturation of SiOa in actual plant
 operation with acid addition is 1.36, and that of CaC03 is less
 than 10"3.  When considering alternatives for the Comanche
 cooling towers, the relative saturation of Si02 is much more im-
 portant than CaC03.  Since scaling in the condensers does not
 presently occur at Comanche, the critical relative saturation
 for Si02(c) formation is most likely above 1.36.  Kinetic stu-
 dies,  such as those presented in Appendix C for CaC03 and Mg(OH)2,
 for silica and other silicate solids should be performed to de-
 termine the critical relative saturation necessary for homogene-
 ous nucleation.  When the critical values for the silica and
 other  silicate scale is determined, estimates of the acid re-
 quirements for the Comanche system can be more accurately esti-
 mated .
2.3.2     Calcium Removal

          There are some scale forming species whose solubility
is not affected by the pH of the solution.  For these solids
acid addition is not an effective inhibitor.  In the cooling
tower systems studied, the only scale forming species which could
not be controlled with acid was CaSOn^HzO  (gypsum).  Equation
2.5 presents the equilibrium relationship affecting
solid formation.
                              -18-

-------
VO
I
                     
-------
             CaSO,.2H20(s) J Ca^(aq) +  SO^  + 2H20   (2.5)




The relative saturation of CaS04'2H20 can  be  calculated using
Equation 2.6.


                   a    •   — .  2
  •n q               Ca    S0i»	H20
  ^•CaSO^H.O      K^co . OTI „
                                                          (2.6)
                                   JSbCaS0lt-2H20



where



          R'S'CaSO -2H 0 = relative saturation of  CaSOt,«2H20
                  u'  2    (gypsum)


             K« ^^ -2H 0 = s°l-ul:>ility constant for calcium
                      2    sulfate dihydrate


                   aCa++ = activity of calcium ion in  solution


                   a£Q=  = activity of carbonate ion in  solution

                         = activity of water in solution


                         = molality of calcium ion in  solution

                   mco=  = molality of carbonate ion in  solution

                    nv, n = molality of water in solution
                     n 2 \*

                   YCa-H- - activity coefficient of the calcium
                           ion in solution


                   YSO=  = activity coefficient of the sulfate
                      4    ion in solution


                    YH20 = activity coefficient of water

                             -20-

-------
If ideal solution is assumed  (yr ++ =  Yon=  = Yu n = 1)
                                  -H-
The calcium level and the sulfate level should be functions of
the cycles of concentration.  Figure 2-3 shows how the relative
saturation of CaSC\«2H20 varies with the cycles of concentration
at the three plants where softening was not used.   The lines for
Montour and Comanche are nearly parallel whereas the slope of
the line for Bowen is much  steeper.  At Bowen the sulfate added
with acid treatment is significant relative to the sulfate added
With the makeup water, whereas it is small at Montour and Coman-
che.  In cases where the sulfate ion concentration due to acid
addition is small, Equation 2.7 can further be simplified into
Equation 2.8.


                   •"•Pacn  =    ^  •^•"r'ocn            (2.8)
                      v^CLOvJli.             wCLOvJli


where

                   C = cycles of concentration

                     = the  relative saturation of CaSQu'lHzQ
                       in the makeup water  (one cycle of
                       concentration)
               CaSC\
           R.S.p so  = the relative saturation of
                   "*   in the recirculating water


If the solutions studied were ideal, all of the data plotted in
Figure 2-4 would be on the line which represents Equation 2.8.
As can be seen from the figure, the actual relative saturation
for a given cycles of concentration is substantially less than
that predicted by Equation 2.8.  This is due to the non-ideali-
ties which exist in real systems which cause the activity coef-
ficients of calcium and sulfate to be less than one.  Thus,
the predictions of scale potential arrived at using ideal solu-
tion assumptions can be overly conservative.


          Colstrip is the only plant that needs treatment to
inhibit gypsum scale in their cooling towers under existing
operating conditions.   The makeup water to Colstrip has the_
largest amount of sulfate in solution (about 170 mg/£ as SOO
                              -21-

-------
     o
     01
     u_
     O
     
-------
       100

       90


       30


       70


       60



       SO




       40
       15
   Ul
   oc

   o

   £
   CM


   8
   
-------
in addition to operating the towers at the highest cycles of con-
centration (10-15).   Colstrip presently keeps the relative satu-
ration of CaSCU.2H20 below 1.10 by lime softening of the makeup
water.


          Softening is a technique used to reduce the calcium
concentration in an aqueous stream.  In lime softening, CaO(s)
is added to the liquid stream to increase the pH of the
solution and precipitate CaCO 3/s\ .   From Equation 2.6 it can
be seen that a reduction in the calcium level reduces the rela-
tive  saturation of CaSCU-2H20.  Furthermore, softening reduces
the relative saturation of CaC03, and thus the acid requirements
to prevent CaCO3 scale.  Lime-soda ash softening is similar but
non-carbonate hardness and silica removal are also accomplished.


          Softening is the least expensive treatment that can
be used to control CaSOi»'2H20 scale.  Other methods that could
be used, such as brine concentration and reverse osmosis, have
much  higher operating and capital costs.  If the gypsum scale
potential is not controlled with treatment, the cooling tower
blowdown may be excessive.  This may require expensive tail-end
treatment of the ultimate effluent stream to achieve zero dis-
charge .


          Softening is a more expensive treatment method when
compared to acid treatment because it requires more expensive
equipment and greater maintenance.   For this reason, the size
of the softened stream can have a major impact on the cost of
treatment.  A smaller stream to be treated requires lower cap-
ital  investment for the flocculator and other softening equip-
ment.  Higher concentrations of calcium in the stream to be
treated require smaller flow rates through the softener to re-
move  the same amount of calcium.  The stream with the highest
concentration of calcium in most cooling towers is the recir-
culating water.  Often a small slipstream of the recirculating
water can be softened and remove as much or more calcium than
by pretreating a much larger makeup stream.


          Simulations were performed to determine the amount of
calcium that must be removed to keep CaSOi^HaO subsaturated.
In order to account for deviations from equilibrium it was as-
sumed that a softener could effectively reduce the calcium con-
centration to 50 mg/Jt.  The results of two simulations perfor-
med to compare slipstream treatment and pretreatment for the
Colstrip cooling system are presented in Table 2-3.  In the case
of makeup softening, a treatment rate of 175 £/sec was required
                               -24-

-------
        TABLE 2-3.  EFFECTS OF SLIPSTREAM SOFTENING ON

                    TREATMENT RATE AT COLSTRIP

Cycles of Concentration
Makeup Water Rate, I /sec
(GPM)
Treatment Rate, a/sec
(GPM)
Calcium Removal Rate, kg CaCOs/day
(Ib/day)
Slowdown Composition, mg/£
Calcium
Magnesium
Sodium
Chloride
Carbonate (as COT)
Sulfate (as SO^)
Nitrate (as NO 3)
pH
CaSO^HaO Relative Saturation*
Makeup
Softening
13.5
175.
(2770)
175.
(2770)
820.
(1800)

534.
143.
540.
227.
6.5
2640.
18.7
7.2
.93
Slipstream
Softening
20.0
170.
(2700)
9.0
(142)
1040.
(2290)

587.
212.
1450.
964.
25.5
3930.
33.6
7.6
1.02
*The critical value,  above which scale potential exists is
 1.3-1.4 for CaSO^-2H20.
                              -25-

-------
to achieve 13.5 cycles of concentration.   With slipstream^soft-
ening, however, a treatment rate of only 9 £/sec was sufficient
to allow operation at 20 cycles of concentration.  The calcium
removal rate increased from 820 kg CaC03/day to 1040 kg/day to
allow for the increase in cycles of concentration, but the treat-
ment rate decreased due to the large increase in calcium concen-
tration in the stream being softened.


          An increase in required calcium removal necessitates
a larger slipstream.  From Equation 2.6,  increases in the con-
centration of either calcium or sulfate raise the relative sat-
uration of CaSO^-ZHzO and thus increase the calcium removal rate
required for scale-free operation.  Figures 2-5 and 2-6 show^how
the required slipstream rate increases as a function of calcium
and sulfate concentration in the makeup water when the towers
are operated at 20 cycles of concentration (based on sampled
Col strip makeup water) .


          It should be noted here that these curves apply only
to the makeup water composition considered.  Variations in other
species' concentrations may cause significant changes due to
chemical complexing.  This curve is valid for the compositions
considered and is presented to show trends in the system.


2.4       Recycle/Reuse Alternatives in Cooling Towers

          In any cooling system employing cooling towers the
cooling water is recirculated.  The degree of recirculation is
measured by the cycles of concentration.   Increasing the cycles
of concentration in a cooling tower will reduce the makeup and
the blowdown rates but will increase the potential for scale
formation.  Treatment methods can be employed which will reduce
the scale potential of cooling towers  which operate at high cy-
cles of concentration.


          Two of the plants studied, Bowen and Montour, presently
operate their towers at less than three cycles of concentration.
Operating in this mode the towers do not need any treatment but
do require very large blowdown streams.  Simulations of these
systems have shown that both plants could increase the cycles  of
concentration without the risk of CaCOj scale if sulfuric acid
were used to control the pH of the recirculating water.  These
results show that the blowdown can be reduced by almost an order
of magnitude at increased cycles of concentration using only pH
control to inhibit scale formation.
                               -26-

-------
     250-r
     200
5
£L
O
OC
Ut
S



Ul
IT
ui
IT

CO

Q.
     150
     100
     50
       30
40         50         60         70


    CALCIUM CONCENTRATION IN MAKEUP WATER. MG/L
                                                        80
                                                                  90
  Figure  2-5
  Slipstream rate as a  function  of makeup calcium
  concentration at Colstrip.

-------
                       250
                       200
                    2
                    Q.
                    (9
                    Z
                    UJ
                    5
                       160
00
                       100
                    m
                    a.
                    It
                    m
                       SO
                         100       150       200        250       300        350


                                 SULFATE CONCENTRATION IN THE MAKEUP WATER. MG/L AS SO|
400
                    Figure 2-6.   Slipstream rate as  a  function of  makeup sulfate
                                  concentration at Colstrip.

-------
          The cooling towers at Colstrip presently employ pH
control and are supplied with softened makeup water.  The towers
are operated at 10-15 cycles of concentration without scale for-
mation.  Simulations of this system have shown that even higher
cycles of concentration could be attained if a slipstream from
the recirculating water were softened rather than the makeup.
This can be accomplished at reduced capital costs because the
size of the treated stream is smaller if the slipstream is sof-
tened.
          The cycles of concentration that can be obtained at
Comanche's cooling towers is limited by the potential for silica
scale.  Comanche's towers presently operate supersaturated with
respect to Si02 but no significant scale has been noticed in the
condensers.  Without conclusive kinetic data with respect to
Si02 formation, simulations of the cooling towers are not suffi-
cient to determine what cycles of concentration Comanche can
operate at safely.


          In order to minimize makeup water requirements and
blowdown rates, cooling towers should be operated at the high-
est cycles of concentration which will not produce scale or
corrosion problems in the condensers.  The relative saturation
of many scale forming species can be reduced by acid addition
for pH control.  Softening of either the makeup or a slipstream
will reduce the relative saturation of CaS04-2H20 which cannot
be controlled with acid treatment.  Utilization of these treat-
ment methods may allow many existing cooling towers to increase
their cycles of concentration safely.


          Corrosion problems may be encountered in cooling
towers operating at high cycles of concentration if the chlo-
ride levels are excessive.  High chloride levels in the tower
makeup may therefore limit the degree of recycle.  Removal of
chlorides is very expensive, requiring sophisticated treatment
such as reverse osmosis or brine concentration.
                              -29-

-------
3.0       ASH SLUICING SYSTEMS

          In addition to cooling towers, a major water consumer
encountered at the power plants studied in this program is ash^
sluicing.  Of the five plants considered, two employed wet sluic-
ing for fly ash disposal and all five plants used wet sluicing
for bottom ash disposal.  Georgia Power Company's Bowen Plant and
Pennsylvania Power and Light Company's Montour Plant both use
cooling tower blowdown to sluice fly ash on a once-through basis
for disposal in ash ponds.  The ash pond overflow is discharged
in both cases.  All of the plants sluice bottom ash on a once-
through basis except Four Corners and Colstrip, which have recir-
culating bottom ash disposal systems.


          This section first presents a typical ash sluicing
flow scheme with discussions of the particular systems studied.
The process description is followed by discussions of the vari-
ous operating parameters and their effects on the design and
operation of ash sluicing systems.  The final portion of this
section  discusses how recycle/reuse options in ash sluicing sys-
tems may be incorporated in overall plant water systems to min-
imize water requirements and discharges.


3.1       Process Description

          In coal-fired boilers, two types of ash residue are
created  by  combustion of the coal.  Fly ash is that portion of
the ash  which is carried out of the boiler with the combustion
gases and bottom ash is the ash remaining in the boiler which
collects at the bottom of the boiler.  The relative amounts of
fly ash  and bottom ash produced depend on the type of furnace
in which the coal is fired.  Burning pulverized coal in a dry-
ash furnace will generally result in about 8070 of the ash being
entrained with the flue gas as fly ash.  Cyclone furnaces, how-
ever, retain  70-80% of the coal ash as bottom ash leaving only
20-30% as fly ash  (BA-465).  The fly ash made up 60-80% of the
total ash at the five plants studied, as would be expected from
dry ash  furnaces.


          Bottom ash is typically removed from the boiler by
periodic washing and subsequent sluicing of the ash to a pond
for disposal.  Fly ash must be continuously removed from the
flue gases  to prevent the discharge of  large amounts of partic-
ulate matter into the atmosphere through the stack.  Available
methods  for removing the  fly ash from the flue gas include elec-
trostatic precipitators, mechanical collectors, fabric filters,
and wet  scrubbers.  Fly ash collected by wet scrubbers is disposed
                              -30-

-------
of in a slurry form.  Arizona Public Service's Four Corners
Plant (Units 1-3) and Montana Power Co.'s Colstrip Plant employ
wet scrubbing for fly ash collection.  The Four Corners fly ash
scrubbing system is open loop whereas the Colstrip S02/particulate
scrubbing system is closed loop.  These systems are discussed in
Section 4.0 and Appendices F and J.  The remaining three plants
(Comanche, Bowen, and Montour) and Units 4 and 5 of the Four
Corners Plant collect the fly ash with electrostatic precipi-
tators.


          Once the fly ash has been collected, it must be trans-
ported to a suitable disposal site.  This may be accomplished by
slurrying the ash with water and pumping it to a pond or by
trucking the ash to the disposal site in a dry form.  Fly ash
collected by the precipitators at Comanche is trucked away in
a dry state.  At Bowen and Montour, the collected fly ash is
slurried to ash ponds (at about 5% solids at Montour and 7%
solids at Bowen).  Figure 3-1 represents the type of fly ash and
bottom ash sluicing operations at Montour and Bowen.  The ash is
sluiced to the disposal pond where it settles to 40-50% solids.
The excess water is discharged as ash pond overflow.  Detailed
discussions of the Bowen and Montour water systems are presented
in Appendices G and I, respectively.


          In areas where water is scarce or regulations prohibit
discharging ash pond overflow, a portion or all of the excess
pond water may be recirculated to sluice ash.  Figure 3-2 repre-
sents a recirculating ash sluicing system.  In the case where all
of the excess pond water is recycled, the makeup requirements are
determined by the pond evaporation rate and the sludge solids
concentration.  Recirculating ash sluice systems with blowdown
streams are used at both Colstrip and Four Corners for bottom ash
disposal.  The blowdown at Colstrip is the overflow from the
bottom ash pond to the scrubber ponds.  This type of ash handling
is not typically used for fly ash disposal since the amount of
leachable species (mostly calcium, magnesium, sodium, and sulfate)
in fly ash is generally much greater than bottom ash.


          A recirculating system is much more susceptible to
scaling since dissolved solids in the makeup are concentrated
in addition to the species leached from the ash.  This study
concentrates on the feasibility of using recirculating sluicing
systems for fly ash disposal.  The effects of operating param-
eters and fly ash reactivity on the scaling tendency in the
system are quantified in the following sections.
                               -31-

-------
                   ASH
                                        EVAPORATION
SLUICE 	 fc
WATER
1

SOLID/LIQUID
MIXING


/
POND
*• OVERFLOW

                                    SLUDGE
Figure 3-1.   Typical once-through  ash sluicing flow scheme
                   ASH
                                        EVAPORATION
     MAKEUP
     SLUICE
     WATER
*- SLOWDOWN
                            POND RECYCLE
   Figure 3-2.   Recirculating ash  sluicing flow  scheme.
                             -32-

-------
3.2       Process Variables

          Two methodologies were used to investigate the effects
of operating parameters on the scale potential in recirculating
ash sluicing systems.  The parameters investigated include ash re-
activity,  makeup water quality, C02 transfer with the atmosphere,
and degree of recycle.


          First, parametric studies were performed using the ash
sluicing computer model.  This model calculates the composition
of the ash slurry pond, and recycle liquors and determines the
scale potential in the system.  All flow rates and the amount of
soluble species in the ash are the major inputs to the model.
First, an overall material balance determines the composition of
the pond liquor assuming solid-liquid equilibrium for all species
except gypsum.  The remaining stream compositions are then calcu-
lated based on the amount of leachable species in the ash.  No
precipitation is allowed in the slurry liquor in order to deter-
mine the potential for scale formation by CaSOtt-2H20, CaC03, and/
or Mg(OH)2 solids.  Since most of the calculations are either
material balances or equilibrium predictions, the validity of the
model in predicting scale potential depends on the accuracy of
the ash reactivity input to the model.  The ash reactivity used
as input to the model in this study was obtained from leaching
studies using deionized water.  Using the ash reactivity deter-
mined from the beaker studies is conservative from a standpoint
of scale-free operation because fly ash tends to be more reac-
tive in deionized water than in water with a higher level of dis-
solved solids.  In actual operation the water used to slurry the
ash has a high enough concentration of dissolved solids to depress
the reactivity of the ash.


          There are several assumptions which are inherent in
performing simulations with the ash sluicing simulation.  These
include:


          1)  Solid-liquid equilibrium is achieved
              in the ash pond, with the exception
              of CaSOi»-2H20 which is allowed to re-
              main supersaturated.

          2)  Ash dissolution is essentially complete
              before the slurry reaches the pond, and
              supersaturation of all species is allowed
              in the slurry line.

          3)  All solids precipitation occurs in
              reaction vessels or the pond.
                               -33-

-------
          The long residence time in an ash pond is sufficient
for most of the species to reach solid-liquid equilibrium.  Since
CaSCs*2H20 supersaturation has been observed in scrubber ponds,
the model did not allow gypsum to precipitate in the pond.  The
model is also designed to handle varying degrees of C02 transfer
in the pond.   In many cases two simulations were performed, one
with no C02 transfer and one with C02 equilibrium achieved be-
tween the pond liquor and the atmosphere.   These two cases serve
as a boundary on the actual amount of C02  transfer occurring.
In actual ponds the degree of C02 transfer is generally between
equilibrium and no transfer.  This computer model is discussed
in greater detail in Appendix E.


          The second methodology used to study ash sluicing in-
volved bench-scale operations.  These operations were performed
with ash samples collected at the plants and makeup water similar
to that measured at these plants.  The bench-scale model used a
mix tank to combine the sluice water and the ash which was trans-
ported to a settling tank which simulated the pond.  The pond
liquor was then recirculated to the mix tank where it was added
to fresh ash and makeup water.


          The results of the leaching studies for Montour and
Colstrip ash are presented in detail in Appendix K.  The bench-
scale studies for Montour and Colstrip are discussed in detail
in Appendix D.  The results of the bench-scale and leaching
studies for Bowen, Comanche, and Four Corners ash are presented
in the final report for EPA Contract No. 68-02-1319, Ash Char-
acterization Studies, which was performed in support of this
program and isincluded as Appendix L.


3.2.1     Ash Reactivity

          Leaching studies were performed to determine the maxi-
mum amount of soluble species in the coal ashes from each of the
five power plants studied.  In these studies, a small amount of
ash was mixed with deionized water in a beaker and the pH was
periodically adjusted with HC1 to maintain a constant value.
When no further alkalinity was leached (pH remained constant
without acid addition) the liquor was analyzed for calcium,
magnesium, sodium, and sulfate.  The leachable amount of each
species as a fraction of the ash was then calculated.


          Table 3-1 presents a summary of the results of the
leaching studies performed for the five ashes.  These results
show that calcium and sulfate are the major species leached from
                              -34-

-------
                 TABLE 3-1.  ASH REACTIVITY DETERMINED FROM LEACHING STUDIES*
i
u>
Species ,
wt. %
PH
Ca
Mg
Na
so.
Montour Bowen Four Corners
6.0 8.1 6.0 8.5 10.4 3.0 6.0 8.5
.32 .28 1.3 .76 .56 .83 .71 .62
/\O rt O • ••• ^» — ^ «•« _•« - . —
.04 .04 .12 .10 .07 .02 .02 .02
.76 .79 1.3 1.2 1.0 .15 .04 .07
Coraanche
6.0
2.9
.11
.04
1.2
8.5
2.2
.05
.03
.83
Co Is trip
4.0 6.0 8.0
5.1 3.7 3.2
.66 .21 .06
—
.55 .60 .57
       *Reported as wt. % of ash.

-------
each of the ashes.   Magnesium is present in substantial quantity
in the Colstrip ash only.   The amount of calcium leached decreases
with increasing pH for all cases but the sulfate remains rela-^
tively constant except for the Bowen and Comanche ashes where it
also decreases with increasing pH.   The leachable calcium and
sulfate as a function of pH for each of the ashes are plotted in
Figures 3-3 and 3-4.


          The ash reactivity will have a major impact on the
feasibility of recirculating ash sluicing systems with respect
to gypsum scale formation since significant quantities of both
calcium and sulfate may be leached from the ash.


3.2.2     Makeup Water Quality

          The composition of the makeup water used in a recir-
culating ash sluice system will influence the composition of the
sluice water and therefore the scaling tendency of the system.
Both the bench-scale experiments and the computer model results
showed that poorer quality makeup water (greater amounts of cal-
cium and sulfate) increased the gypsum scale potential in the
fly ash slurry liquor.  Table 3-2 presents results from bench-
scale experiments with the Bowen ash and from computer model
calculations for the Montour ash.  The first two columns repre-
sent two closed-loop bench-scale experiments where only the
makeup water quality was varied.  The first experiment involved
poor quality makeup water and a gypsum relative saturation of
1.68, well above the critical value for scale formation, was
observed.  When a much better quality makeup water was used
(lower calcium and sulfate concentrations) the gypsum relative
saturation encountered was 1.35, which is significantly lower.


          The last two columns in Table 3-2 show the results
from two computer simulations of a closed-loop ash sluicing
system using the Montour ash.  The first simulation involved 8-
cycle cooling tower blowdown as makeup water whereas the second
simulation used river water as makeup.  Again, the gypsum rela-
tive saturation decreased (2.8 versus 3.2) when makeup water
with less calcium and sulfate was used.


          For systems which have gypsum relative saturations
greater than 1.3-1.4 (the critical range for scale formation)
one treatment option is to use soda ash softening of a portion
of the pond recycle liquor.  The magnitude of the treatment will
depend primarily on the ash reactivity but the makeup water
quality will also affect the amount of softening required
                               -36-

-------
    4-
  ra
  U
»
O CO



> I
< U.
    2-
     r
                                   -0	
                                                      Ny COLSTRIP


                                                       COMANCHE


                                                      ^ BOWEN


                                                      Q FOUR  CORNERS


                                                      © MONTOUR
                                                       -0
                                         PH
        Figure 3-3.   Reactive calcium in fly ashes  as a function of pH.

-------
LO
OO
                   1.0
               o
                w
                 ? -5
                       A BOWEN
                       <•> COMANCHE
                       © MONTOUR
                       V COLSTRIP
                       0 FOUR CORNERS
                    n
                               V
                                                                              0
                                                           pH
                     Figure  3-4.  Reactive  sulfate in fly ashes  as a function of pH.

-------
           TABLE  3-2.
EFFECTS OF MAKEUP WATER QUALITY  ON
RECIRCULATING ASH SLUICE SYSTEM  SCALING
POTENTIAL
                               Bench Scale Experiments  Computer Model Results
                                   with Bowen Ash          with Montour Ash*
MAKEUP WATER, mg/£
Calcium
Magnesium
Sodium
Chloride
as
Carbonate (as CO 3)
Sulfate (as S0~)
Nitrate (as N0~)

205.
66.
570.
50.
2.4
1940.
19.

30.
8.5
23.
8.5
100.
10.0
—

227.
44.
65.
176.
30.
555.
44.

28.
5.5
8.
22.
6.
6.8
5.5
 FLY ASH  SLURRY, mg/A
   Calcium
   Magnesium
   Sodium
   Chloride
   Carbonate  (as CO 3)
   Sulfate  (as SOO
   Nitrate  (as
920.
2.4
350.
130.
6.0
2500.
__
800.
1.2
330.
34.
6.6
2100.
__
1690.
120.
67.
182.
4.
4350.
46.
1500.
77.
8.4
23.
1.4
3840.
5.7
           Relative Saturation**  1.68
                         1.35
 3.2
                                                                         2.8
                                 10.6
                        11.1
10.1
                                                                        10.2
 *No solids precipitation was allowed;  calculations were made based on leaching
  results which represent a worst case  situation  (i.e., maximum leachable  spe-
  cies).  Bench-scale experiments gave  gypsum relative saturations of 1.1-1.2
  for the Montour ash with a reaction tank  to allow solids to form.
**Critical value, above which scale potential  exists,  is 1.3-1.4 for CaSOit'2H20.
                                    -39-

-------
Table 3-3 presents the results of three computer simulations of
closed-loop sluicing operations at Montour using slipstream treat-
ment.  As the water quality becomes poorer (cycles in cooling
tower increase) the amount of pond recycle water which must be
treated increases from 2TL using river water as makeup to 30/0 and
36% for 8- and 20-cycle cooling tower blowdown, respectively.


3.2.3     Carbon Dioxide Transfer

          In a recirculating ash sluice system, there is poten-
tial for C02 transfer between the atmosphere and the pond water.
The degree of C02 transfer will affect the pH of the pond return
water and thus the fly ash slurry composition.  Since both CaCOs
and Mg(OH)2 relative saturations are pH dependent (tend to pre-
cipitate at higher pH),  the degree of C02 transfer which occurs
will have a significant effect on the scaling tendency in the
system with respect to these two species.


          Table 3-4 presents the results from two of the bench-
scale experiments where the only variable between runs was the
transfer of C02 in the pond.  For the run with no C02 transfer,
no CaC03 precipitation was noted in the reaction tank even though
the relative saturation was 7.8, which is above the critical value
of 2.5  (see Appendix C).  This may be due to the low carbonate
levels  in the system  (2-5 mg/Jl) .  However, when C02 was bubbled
through the pond liquor, the relative saturation increased to
31.8 and about 2.7 mmole/min of CaC03 precipitated in the reac-
tion tank.  The gypsum relative saturation decreased slightly
(from 0.5 to 0.4) due most likely to the precipitation of calcium
as calcium carbonate.  In every case where the C02 bubbler was
used in the bench-scale experiments, CaC03 precipitation was
noted in the mix tank (ranging from 0.1 to 2.7 mmole/min), indi-
cating  that CO2 transfer between the pond and the atmosphere
could cause CaC03 scaling problems with recirculating ash sluice
systems.


          Table 3-5 presents the results from two of the computer
model calculations where C02 transfer effects were studied.  As
with the bench-scale  experiments, the transfer of C02 in the
pond increased the CaC03 scaling potential in the system.  A sig-
nificant increase in  the amount of CaC03 solids precipitated was
noted along with an increase in relative saturation.  The calcu-
lated amount of CaC03 solids formed increased from 0.03 gmole/sec
for the case with no  C02 transfer to 1.85 gmole/sec for the case
with C02 equilibrium  in the pond.  The relative saturation of
CaC03 in the fly ash  slurry increased from 29.3 to 89.1.
                             -40-

-------
          TABLE 3-3.   EFFECTS OF MAKEUP WATER QUALITY ON
                        TREATMENT REQUIRED FOR CLOSED-LOOP
                        ASH SLUICING  OF MONTOUR ASH*
Makeup Source
MAKEUP WATER COMPOSITION, mg/£
Calcium
Magnesium
Sodium
Chloride
Carbonate (as CO 3)
Sulfate (as SOO
Nitrate (as NOl)
SLIPSTREAM TREATMENT RATE, 5,/sec
% OF POND RECYCLE TREATED
CALCIUM REMOVAL RATE, gmole/sec
FLY ASH SLURRY LIQUOR, mg/A
Calcium
Magnesium
Sodium
Chloride
Carbonate (as COl)
Sulfate (as SOO
Nitrate (as NO^)
PH
RELATIVE SATURATIONS**
CaCOa
CaSO^ZHaO
River 8-Cycle Cooling
Water Tower Slowdown

28.
5.5
8.0
22.
6.0
68.
5.5
38.5
27.
0.88

500.
77.
1150.
22.8
1.4
3840.
5.7
10.3

0.97
1.06

227.
44.
65.
176.
30.
555.
44.
43.7
30.
1.06

500.
120.
1420.
180.
4.2
4350.
45.4
10.1

2.4
1.07
20-Cycle Cooling
Tower Slowdown

567.
110.
161.
440.
23.
1430.
110.
50.8
36.
1.36

510.
135.
1940.
455.
3.3
5060.
114.
10.3

1.7
1.08
 *Based on ash reactivity determined from beaker, leaching studies (worst  case):
 no C02 transfer allowed in  the pond.
&&
 Critical values, above which scale  potential exists, are 1.3-1.4 for
 CaSO^'2H20 and about 2.5 for CaCOs  (see Appendix  C)
                                    -41-

-------
  TABLE  3-4.   EFFECTS  OF C02 TRANSFER IN POND ON  FLY ASH SLURRY
               SCALING  TENDENCY (BENCH-SCALE  RESULTS)


                                               (Colstrip Ash)
                                  No CO2 Transfer       C02 Bubbled in Pond
POND LIQUOR, mg/2,
Calcium
Magnesium
Sodium
Chloride
Carbonate (as CO 3)
Sulfate (as SOl)
Nitrate (as NO 3)
pH
Relative Saturations
CaCOs
Mg(OH)2
CaS04«2H20
FLY ASH SLURRY, mg/S,
Calcium
Magnesium
Sodium
Chloride
Carbonate (as CO 3)
Sulfate (as SO^)
Nitrate (as NOl)
PH
Relative Saturations **

1240.
0.*
27.6
39.1
2.4
595.
14.3
12.6

3.7

0.5

1000.
0.*
27.6
39.1
4.8
691.
14.3
12.6


481.
0.*
52.9
31.2
654.
566.
12.4
7.6

12.6
—
0.4

441.
0.*
25.3
29.1
21.0
614.
14.3
11.7

  Mg(OH)2
 CaC03 PRECIPITATION RATE
  ACROSS MIX TANK,  mmole/min
7.8

0.5


0.0
31.8

 0.4


 2.7
 *Magnesium levels were not detectable due to the high pH's in the system
  (Mg(OH)a precipitation removed virtually all of the liquid phase magnesium
  from the system or prevented the dissolution of magnesium from the ash).
**Critical values, above which  scale potential exists,  are 1.3-1.4  for
  CaSOn'2H20, about 2.5 for CaCOs, and about 3.4 for Mg(OH)2  (see Appendix C)
                                    -42-

-------
 TABLE 3-5.
EFFECTS OF CO2  TRANSFER IN  POND  ON FLY  ASH SLURRY
SCALING TENDENCY (COMPUTER  MODEL RESULTS)
CaC03  SOLIDS FORMED IN
   SYSTEM, gmole/sec
                                                (Bowen Ash)
                                  No  CO2  Transfer
                            .032
                                            CO2 Equilibrium
                                           with Air in Pond
POND LIQUOR, mg/2,
Calcium
Magnesium
Sodium
Chloride
Carbonate (as COa)
Sulfate (as SO^)
Nitrate (as NO? )
pH
Relative Saturations
CaCOa
Mg(OH)2
CaSOit'21120
FLY ASH SLURRY, mg/£
Calcium
Magnes ium
Sodium
Chloride
Carbonate (as COJ)
Sulfate (as S0=)
Nitrate (as NO?)
PH
Relative Saturations**
CaC03
Mg(OH)2***
CaSO^-ZHaO

1170.
.01
187.
33.
0.7
1290.
67.
12.5

1.0
1.0
1.0

1410.
11.
187.
33.
9.0
1850.
67.
11.7

29.3
1460.
1.3

570.
28.
188.
33.
33.
1740.
66.
7.9

1.0
9.3 x 10~s
1.0

1050.
28.
187.
33.
29.
2130.
67.
11.3

89-1
1200.
1.3
1.85
*Equilibrium partial pressure of CO 2 in  pond specified to be equal to  that
 in the atmosphere (3.3 x 10"1* atm) .

**Critical values, above which scale potential exists, are 1.3-1.4 for
  CaSO^'2H20, about 2.5 for CaC03 ,  and about 3.4 for Mg(OH)2 (see Appendix C)

***Solid precipitation was not allowed.  Kinetic studies (Appendix C)  indicate
   that at these relative saturations, the magnesium will precipitate, result-
   ing in very low magnesium concentrations in the liquid phase.
                                    -43-

-------
          These  cases represent the two extremes of what may
 actually happen  in a plant situation.  The samples taken at the
 plants  studied in this program indicate that some C02  transfer
 occurs  in the pond but complete equilibrium is not always  _?
 achieved.  At Bowen the partial pressure of C02 was 2  x 10    atm
 and  at  Montour the partial pressure was slightly above the equili-
 brium value  of 3.3 x 10~4 atm.  The degree of C02 transfer at a
 specific location will have a significant effect on the CaCOa
 relative saturation in the pond recycle and fly ash slurry
 liquors.


 3.2.4    Degree of Recycle

          The previous discussions have considered completely
 closed-loop  ash  sluicing operations and the effects of various
 operating parameters on these systems.  However, in some cases
 completely closed-loop operation may not be desired and treatment
 steps might  be eliminated.  For example, if ash pond overflow is
.used as makeup to a scrubbing system, the ash sluicing network
 would be only partially closed-loop since a blowdown stream to a
 scrubber is  used.  Table 3-6 presents two computer simulation
 cases for a  sluicing system using Montour ash.  In the first
 case (column one) closed-loop operation is employed so that all
 of the  pond  overflow liquor is recycled to the system.  A  slip-
 stream  is taken  from the pond recycle to remove calcium and con-
 trol gypsum  scale potential in the fly ash slurry.  A  slipstream
 treatment rate of about 77 5,/sec was required.


          In the second case, gypsum scale potential is control-
 led by  taking a  blowdown stream from the system to keep the sul-
 fate level  low enough  in the circulating liquor to prevent gypsum
 supersaturation. A blowdown rate of approximately 51  &/sec was
 required  to  maintain a gypsum relative saturation near 1.0 in
 the fly ash  slurry  liquor.  This corresponds to about  1870  of  the
 pond recycle liquor.


 3.3      Recycle/Reuse Alternatives  in Ash Sluicing Systems

          Of the ash  sluicing systems encountered, recirculation
 of the  ash  pond  water  was  only practiced with bottom ash which
 is in general much  less reactive.  Both of the plants  which
 sluice  fly  ash  do  so  on a  once-through basis.  These studies
 revealed  that  a  recirculating fly ash system may be employed
 but treatment may be necessary to prevent gypsum scale formation.
 The amount  of  leachable  species  in the ash and makeup  water
 quality are  the  major  parameters which determine the  level of
 treatment necessary.
                                -44-

-------
          TABLE 3-6.
EFFECTS OF  DEGREE OF RECYCLE  IN AN
ASH  SLUICING SYSTEM USING MONTOUR ASH*

Fly Ash Rate, kg/hr
(Ib/hr)
Pond Recycle Rate to Fly Ash Sluice, A/ sec
(GPM)
Sluice Water Makeup Rate, 2,/sec
(GPM)
Pond Overflow Rate, &/sec
(GPM)
Slipstream Treatment Rate,** A/sec
(GPM)
Fly Ash Slurry, mg/£
Calcium
Magnesium
Sodium
Chloride
Carbonate (as CDs)
Sulfate (as SO^)
Nitrate (as NO 3)
pH
„ - . ***
Relative Saturations
CaCOs
CaSO^-ZHaO
Closed-Loop
Operation
62,200
(137,000)
284
(4,500)
37.8
(600)
0.0
(0.0)
77.0
(1,220)

500
77
1,150
22.8
1.4
3,840.
5.7
10.3

0.97
1.06
Slowdown Taken
from System
62,200
(137,000)
284
(4,500)
44.2
(1,400)
51.1
(800)
0.0
(0.0)

624
37
8.2
22.4
2.2
1,600.
5.6
10.3

3.3
1.0
  *Based on ash reactivity from ash leaching  experiments (worst  case);  no C02
   transfer in the pond.
 **Based on treatment  to 50 mg/& Ca to account for inefficiencies.
***Critical values, above which scale potential  exists, are 1.3-1.4  for
              and about 2.5 for CaC03 (see Appendix C)
                                    -45-

-------
          Determining the quantity of teachable species in
different ashes is a difficult task in that the amount leached
depends on liquor pH and composition.  Bench-scale experiments
revealed that solutions containing more dissolved species leach
less of the ash than solutions with lower dissolved solids con-
centration.  The leaching studies showed the variability in ash
compositions and the pH dependency of the soluble species in
fly ash.  The results of the leaching studies represent the
maximum levels of soluble species in each ash studied and were
therefore used to represent "worst case" operation.   Pilot or
additional bench-scale studies are recommended to determine more
accurately the solubility characteristics of a particular ash
before a recirculating fly ash sluice system is implemented.


          Carbon dioxide mass transfer between the pond liquor
and the atmosphere will affect the CaC03 and Mg(OH)2 scale poten-
tial in a fly ash sluice system.   Increased C02 sorption in the
pond will increase CaCOs scale potential and decrease Mg(OH)2
scale potential.  Since these scales are pH dependent (higher pH
means more scale potential) problems may be encountered with
very alkaline ashes.  However, these scales are typically softer
than gypsum scale and may be eroded by the ash.  Installing
reaction vessels prior to the fly ash slurry line may allow
precipitation of these solids in  a controlled fashion so that
scale formation in the slurry line may be minimized.  Again,
pilot or bench-scale studies are  recommended to more accurately
quantify CaCOs and/or Mg(OH)2 scale potential for a particular
ash before a recirculating system is implemented.
                               -46-

-------
4.0       SO2/PARTICULATE SCRUBBING SYSTEMS

          The third type of major water consumer encountered at
the plants studied in this program is combined particulate and
SO2 scrubbing.   Of the five plants investigated, two had scrubbing
systems.   The Four Corners Plant of Arizona Public Service has ven-
turi particulate scrubbers on three of the five generating units
although some SO2 removal is also obtained.  The Colstrip Plant of
Montana Power Co.  has combined SO2 and particulate scrubbers on
both of the generating units.  Venturi collectors for particulate
removal are followed by spray chambers for SO2 removal.


          This section of the report first presents a typical
scrubbing process flow scheme including discussions of differences
between the typical system and actual systems studied.  The pro-
cess description is followed by discussions of the various oper-
ating parameters and their effects on the design and operation of
scrubbing systems.  The final portion of this section discusses
the potential for scrubbing in various recycle/reuse alternatives
at power plants.  Since most scrubbing systems are designed for
zero discharge, they may possibly be used as receptors for the
final water effluent in a cascaded water system to achieve zero
discharge or reduced blowdown for an entire plant water network.


4.1       Process Description

          A typical scrubbing system may be divided into three
maj or operations:


          1)  gas cleaning,

          2)  solids precipitation, and

          3)  solids concentration.


The combination of these operations is shown in the simplified
scrubbing system in Figure 4-1.  Gas cleaning is accomplished^in
the scrubber vessel, solids precipitation occurs in the reaction
tank, and solids concentration is achieved in the solid-liquid
separator which may be a clarifier, filter, pond, or any combina-
tion of the three.  This section presents a discussion of each of
these major operations observed in scrubbing systems and how  the
two specific systems studied deviate from the typical flow scheme
in Figure 4-1.   More detailed discussions of the scrubbing systems
at Four Corners and Colstrip may be found in Appendices F and J,
respectively.
                               -47-

-------
MAKEUP
 WATER
   FLUE
   GAS
 ALKALI
                 STACK
                  GAS
DEMISTER
             SCRUBBER
              REACTION
               TANK
                                   SOLID/LIQUID
                                   SEPARATION
                                      WASTE
    Figure 4-1.  Typical scrubbing  system flow scheme
                              -48-

-------
4.1.1     Gas Cleaning

          In both particulate and combined SO 2/particulate scrub-
bing systems, the boiler flue gas is contacted with a recircula-
ted slurry in the scrubber vessel where the particulates (fly
ash) and/or S02 are absorbed.  Venturi scrubbers are used for
particulate removal at both Four Corners and Colstrip.  However,
at Colstrip where the scrubbers are designed for combined par-
ticulate and SO2 removal, the venturi sections are followed by
spray sections where most of the SO2 is absorbed.


          Spray towers provide more liquid-gas contact area and
longer residence times than Venturis and are therefore more effi-
cient at removing SO2 from the flue gas.  At Four Corners only
about 30% of the SO2 in the flue gas is removed in the Venturis,
but at Colstrip 74% of the scrubber inlet SO 2 is removed by the
combination of venturi and spray sections.


          To prevent excess carryover of the scrubbing liquor
with the clean gas, mist eliminators are used in the scrubbers.
The mist eliminators are generally washed with fresh makeup water
which falls into the scrubber after being sprayed over the demis-
ters.  At Colstrip, a wash tray collects some of the wash water
so that it may be reused along with the makeup water as demister
wash.


4.1.2     Solids Precipitation

          The second major operation in scrubbing systems is
solids precipitation.  In a lime or limestone based closed-loop
scrubbing system, the S02 which is absorbed must be removed from
the system by precipitation of calcium sulfate and calcium sul-
fite.  The required rate of precipitation is determined by the
rate of absorption of S02 from the flue gas.  At both Four Cor-
ners and Colstrip, greater than 90% of the sulfite formed in the
liquid phase by S02 sorption is oxidized to sulfate, making the
precipitation of gypsum (CaSO,, «2H20) the controlling rate of
solids formation.


          It should be noted that revision of system operation
could alter the sulfite oxidation rate significantly.  Systems
with lower than approximately 15% oxidation can be designed for
the so-called "subsaturated gypsum mode" whereby the calcium sul-
fate coprecipitates with calcium sulfite.  The results presented
here assume high (>90%) sulfite oxidations.
                               -49-

-------
          The precipitation of the absorbed sulfur occurs in  a
reaction tank where adequate time is allowed for crystal^growth
of recirculated calcium sulfate and calcium sulfite particles.
If the residence time of the reaction tank is too small, then the
concentrations of sulfate and sulfite will increase until nuclea-
tion occurs, resulting in scale formation.  The scale will most
likely form in the scrubber where sulfur concentrations are
highest.


          At Four Corners, where scaling has been noted, the re-
action time for solids precipitation is very small (about one
minute), whereas at Colstrip the reaction tank residence time is
about eight minutes.   No scaling has been reported for the Col-
strip scrubbing system.  The effects of reaction tank volume on
system operation will be quantified in Section 4.2 which discus-
ses process variables.


          The sorbed fly ash solids and precipitated sulfur sol-
ids must be removed from the recirculating slurry at a rate suf-
ficient to control the solids concentration in the circulating
slurry at the desired level.  At Colstrip, the blowdown is taken
at a rate sufficient to keep the circulating solids concentra-
tion at about 12%.  At Four Corners, the circulating solids con-
centration is about 27o.  The low value of 2% is used at Four Cor-
ners to minimize erosion problems.  Higher solids concentrations
are more abrasive but require smaller reaction tanks due to the
increased number of precipitation sites.  The blowdown from the
recirculating slurry is pumped to the solid/liquid separation
portion of the system.


4.1.3     Solid/Liquid Separation

          In order to minimize water requirements for a scrubbing
system, solid/liquid separation is employed to recover a portion
of the water used to slurry the waste solids (ash, CaSCU^HzO,
CaS03'%H20).  Higher percent solids also allow for easier disposal
of the sludge as landfill.  This separated water can be recycled
to the scrubbing system to eliminate any aqueous discharge and
reduce makeup water requirements.  The final sludge water content
and the evaporation occurring in the scrubber determine the makeup
water requirements for the system.

          Typical solid/liquid separation techniques include clar-
ification, filtration, and ponding or combinations of these tech-
niques.  At Four Corners the 270 solids slurry blowdown is clari-
fied to about 107o solids (typical operation) and pumped to a pond.
The clarifier overflow is returned to the scrubbing system.  In
                             -50-

-------
the ash pond, the sludge settles to about  50% water and the ex-
cess water (not evaporated or occluded with  sludge) is discharged
rather than recirculated.


          In the Colstrip scrubbing system,  the  12% solids stream
is diluted to about 6% solids with pond water and pumped to the
scrubber ponds where the solids  (ash and CaSO^»2H20, predominant-
ly) settle to about 50% water.  The remaining water which is not
evaporated is recycled to the scrubbing system.


4.2       Process Variables

          The effects of important process variables concerning
scale control were investigated as well as the effects on makeup
water requirements for the two scrubbing systems studied using
the computer models discussed in Appendix  E.  The parameters
affecting the system water makeup requirements are S02 removal
.rate, ash removal rate, and pond recycle rate.   Other parameters
considered include liquid-to-gas ratio, circulating slurry solids
concentration, reaction tank volume, and makeup  water quality.
The following sections contain quantitative  evaluations of the
effects of these parameters on the operation of  particulate and
S02 scrubbing systems.  Detailed analyses  of the Four Corners and
Colstrip scrubbing operations are presented  in Appendices F and
J, respectively.


          The scrubbing models used in this  study calculate all
stream compositions and flow rates in the  system using precipi-
tation rate kinetics for CaSCU^HaO and CaS03*%H20, which are
the solids formed in lime/limestone scrubbing systems, and vari-
ous input parameters.  These parameters 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 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 through the scrubber vessel and  the scrubber recycle
tank (if applicable) until relative saturations  and stream com-
positions satisfy the rate equations.  The calculations are per-
formed for ancillary equipment such as the reheat and fan require-
ments, and to determine additional stream  compositions.  Makeup
water requirements are calculated by an overall  system balance.
                               -51-

-------
          Several assumptions are inherent in performing  scrub-
ber simulations with the model outlined above.  These are enum-
berated below:

          1)  The scrubber exit 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 tem-
              peratures are the adiabatic saturation
              temperature of the flue gas.

          4)  All oxidation was assumed to occur in
              the scrubber.

          5)  All solids precipitation occurs in reac-
              tion vessels.


4.2.1     SO2 Removal Rate

          The effect of SO2 removal rate on scrubber makeup water
requirements may be determined by examining the results of two of
the Four Corners process simulations.  In the high solids opera-
tions case where the circulating solids concentration is 9%, the
clarifier underflow is 30% and the final slurry discharged is ~LTL,
only 307> 862 removal is specified.  50% SOz removal was specified
for Alternative Two.  This alternative mode of operation is with
10% solids in the circulating liquor and 3070 solids in the final
slurry discharged.  Table 4-1 presents the characteristics of the
solid waste for each of these cases as well as the operating
conditions and makeup water requirements.


          Direct comparison of the makeup water requirements for
these two cases as a function of SC>2 removal cannot be made since
the waste suspended solids concentrations are different.  Table
4-1 shows an adjusted makeup water rate based on the solids flow
in the waste and a solids concentration of 30% in the waste for
both cases.  The difference in adjusted makeup water requirements
is only 2.1  5,/sec or about 370 of the total makeup for a change in
SO2 removal from 30% to 50%.


          This effect is small due in part to the ash comprising
more than 90% of the solids in the waste.   Although the amount of
sulfate solids changed considerably, the overall impact is small,
since the sulfate solids represent less than 10% of the total


                              -52-

-------
    TABLE 4-1.   EFFECTS  OF S02 REMOVAL  RATE  ON FOUR CORNERS
                  SCRUBBER MAKEUP REQUIREMENTS
                                        High Solids
                                        Operations
                    Alternative Two
S02 in Flue  Gas, ppm

Ash in Flue  Gas, g/sec
                (Ib/min)

SO2 Removal,  %

Particulate  Removal, %

Final Slurry to Ash  Pond

  Wt. % Solids

  Solids Composition, Wt.  %

    CaSOi»-2H20

    Inert (Ash)

  Solids Flow,  g/sec
               (Ib/min)

Makeup Water Requirements, i/sec
                           (GPM)

  Evaporation

  Occluded with Solids

  Total

Makeup Water Rate  Adjusted to 30%
Solids in Waste, £/sec  (GPM)

  Evaporation

  Occluded with Solids

  Total
     640.

  16,200.
   (2140)

      30.

      99.7
      17.3
 30.0 (475)

 79.0 (1250)

109.0 (1735)
 30.0 (475)

 38.6 (612)

 68.6 (1087)
    640.

 16,200.
  (2140)

     50.

     99.7
     30.
3.9
96.1
16,600.
(2200)
7.9
92.1
17,500.
(2300)
30.0 (475)

40.7 (645)

70.7 (1120)
30.0 (475)

40.7 (645)

70.7 (1120)
                                   -53-

-------
solids.  The total amount of solids changed only by about 5/0.  As
the amount of sulfur removed in relationship to the ash removed
increases, the effects of S02 removal on makeup water requirements
will become more pronounced.


          The effect of changes in S02 removal rate on makeup
water rate will be dampened by the fact that water evaporation
in the scrubber frequently represents a major portion of the
makeup water requirement.  As the evaporation becomes a larger
part of the makeup water, changes in S02 removal rate will have
a decreasing effect on the overall scrubber makeup rate.  At
Four Corners, for the high solids operation case (17% solids in
final waste), the evaporation is about 30% of the total makeup.
However, at Colstrip the evaporation is about 80% of the total
makeup because the final sludge solids concentration is about
50%, and only a small amount of water is lost with the solids.


4.2.2     Ash Removal Rate

          The effect of the ash removal rate on makeup water re-
quirements for combined SOa and particulate scrubbing is more
pronounced than that of the S02 removal rate in the situations
studied.  The scrubbing situations encountered were for low sul-
fur applications.  In the cases studied, most of the solid waste
was ash as opposed to calcium sulfate or calcium sulfite.


          As the ash removal rate is decreased, by burning a coal
of lower ash content or by decreasing load for example,  the makeup
water requirements for the scrubbing system are decreased since
less water is lost from the system by occlusion with the solids.
By an overall mass balance around the scrubbing system,  the makeup
water rate may be calculated by Equation 4.1:


                            - Xn)

                                ~(1 " Xs)                 ^
where
          M = makeup water requirement,  £/sec

          E = evaporation (scrubber + pond),  £/sec

          A = ash removal rate,  g/sec

          p, = density of water,  g/Jl
           J_i

          XD= weight fraction of ash which is soluble
                              -54-

-------
          XA = wei§ht fraction of solids in waste which is ash

          Xg = weight fraction solids in waste stream


          The makeup water requirement is the sum of two compo-
nents.   One is the evaporation in the scrubber and the other is
the water occluded with the solids in the waste stream.  The sec-
ond term in Equation 4.1 represents the water lost with the solid
waste and is directly proportional to the ash removal rate, A.
Table 4-2 presents data from two scrubbing simulations for the
Four Corners Plant and two for the Colstrip Plant which illustrate
the effects of flue gas ash content on scrubbing system makeup
water requirements.  The two cases for Four Corners represent a
reduction in flue gas ash content of 60%, from 16,200 g/sec to
6,480 g/sec.  The water lost through solids occlusion was reduced
from 17.5 £/sec to 8.0 £/sec resulting in an overall makeup water
reduction of about 20% from 50.7 a/sec to 41.2 £/sec.


          Lowering the ash removal rate did not produce a propor-
tional reduction in water lost with the solids since the fraction
of ash in the solids also decreased.  From Equation 4.1, lowering
the fraction of ash in the solids, X., will increase the makeup
water requirements.  A directly proportional decrease of occluded
water would have been to 7.0 5,/sec insteam of the 8.0 £/sec calcu-
lated.   This difference is due to X. changing as well as A, the
ash removal rate, in Equation 4.1.


          The two simulated cases for the Colstrip scrubbing
system produced similar results in that a 30% reduction in ash
removal rate caused only a 14% reduction in occluded water.
Again,  this is due to the change in the solids composition.  The
weight fraction of ash in the solids decreased from 0.568 to 0.474
for these two cases.  Overall makeup water requirements were re-
duced by only about 2%, from 58.2 H/sec to 57.1 5,/sec, since most
of the makeup water requirements are needed to replace evaporated
water rather than occluded water.


          The ash removal rate will also affect the scaling ten-
dencies in the system.  A decrease in ash removal rate will in-
crease the amount of calcium sulfite and calcium sulfate solids
being recirculated.  This provides more crystal sites for preci-
pitation and therefore decreases the tendency of the system to
form scale.  For example, comparison of the reaction tank liquors
for the two Colstrip simulations shown in Table 4-3 shows a de-
crease in calcium sulfate relative saturation from 1.41 to 1.37
when the amount of ash removed is decreased from 5500 g/sec to
3800 g/sec.
                              -55-

-------
                TABLE 4-2.   EFFECTS  OF ASH REMOVAL RATE  ON SCRUBBER MAKEUP REQUIREMENTS
Ln

I

Ash in Flue Gas, g/sec
(Ib/min)
Particulate Removal, %
Total Fraction of Ash Dissolving
SOa in Flue Gas, ppm
SO 2 Removal, %
Solid Waste
Wt. % solids
Solid Composition, wt. %
CaS03'%H20
CaS(V2H20
Inert (Ash)
Solids Flow, g/sec
(Ib/min)
Makeup Water Requirement, A/sec (GPM)
Evaporation
Solids Occlusion
Total
Four
Alternative*
Three
16,200.
(2140)
99.7
.011
640.
50.

50.

8.5
91.5
17,500.
(2300)

33.2 (525)
17.5 (278)
50.7 (803)
Corners
Alternative**
Four
6480.
(860)
99.7
.011
640.
50.

50.

18.8
81.2
8000.
(1060)

33.2 (525)
8.0 (127)
41.2 (652)
Colstrip
Design
Conditions
5500.
(730)
99.6
.193
790.
74.

50.

4.1
39.0
56.8
7740.
(1020)

50.5 (800)
7.7 (122)
58.2 (922)
Low Ash
Conditions
3800.
(500)
99.6
.193
790.
74.

50.

5.2
47.4
47.4
6560.
(870)

50.5 (800)
6.6 (105)
57.1 (905)
            *Repiped scrubber system with adequate reaction tank volume to prevent scale and recycle of ash pond
            overflow.

           **Same as Alternative Three except  ash content of flue gas lowered by 60%.

-------
TABLE 4-3.
                  EFFECT OF ASH REMOVAL RATE ON COLSTRIP
                  SCRUBBING SYSTEM SCALING TENDENCY
                                        Design
                                      Conditions
                                                  Low Ash
                                                Conditions
Ash in Flue Gas ,
            /sec
            Ib/min)
Scrubber Recycle  Slurry
  pH
  Liquor Composition,  mg/£
    Calcium
    Magnesium
    Sodium
    Chloride
    Carbonates  (as  COT)
    Sulfates  (as  SO"^)
    Sulfites  (as  SOl)
    Nitrates  (as  NO 3)
CaSOi^HaO Relative Saturation*
Solid Composition,  wt.  70
  CaSO^«2H20
  Inert (Ash)
  CaS03'%H20
5,500.
 (730)
                                      5.0
                                    733.
                                  5,285.
                                    444.
                                    117.
                                    153.
                                 21,000.
                                  3,560.
                                      9.6
                                      1.41
                                     38.9
                                     57.3
                                      3.8
3,800.
 (500)
                    5.2
                  715.
                4,267.
                  448.
                  132.
                  156.
               17,800.
                2,135.
                   10.9
                    1.37
                   47.2
                   47.8
                    5.0
*The critical value, above which  scale  potential  exists,  is
 1.3-1.4 for CaSO^'2H20.
                               -57-

-------
4.2.3     Ash Pond Recycle Rate

          Although most throwaway scrubbing  systems  are designed
for closed-loop operation (no aqueous  discharges)  some systems
may operate with a blowdown from time  to  time.   This section des-
cribes the impact on the scrubbing system water  balance of oper-
ation with or without recycle of ash pond overflow.   The Four
Corners scrubbing system simulations chosen  to quantify the ef-
fects of ash pond recycle are summarized  in  Table  4-4.   Detailed
operating data and simulation results may be found in Appendix F.


          The first column of Table 4-4 (Alternative Two)  rep-
resents operation of the system with a solid waste suspended
solids concentration of 30%.   No pond water  is recycled to the
system in this case.  Alternative Three represents identical
conditions except the ash pond water is recycled to  the reaction
tank after the waste solids have settled  to  50%,  water.   The sol-
ids composition was changed slightly due  to  changes  in  the liquid
composition in the system, resulting from the recycle of the pond
water.  The major influence on the makeup water  rate,  however,
was the change in solids concentration in the final  waste  from
307o in Alternative Two to 50% for Alternative Three.   The  occlu-
ded water decreased from 40.7 I/sec to 17.5  £/sec  to reduce the
overall makeup requirements from 70.7 5,/sec  to 50.7  £/sec.   The
evaporation component of the makeup water requirements  increased
slightly in Alternative Three due to the  evaporation occurring
in the ash pond.


          Recycle of the ash pond liquor,  in  addition to decreas-
ing makeup water  requirements, raises the  dissolved solids  level
in the scrubbing  system.   The  effect on the system  scaling  ten-
dency is illustrated by the two  simulation cases  presented  in
Table 4-5.  The calcium sulfate  relative saturation of the  scrub-
ber liquor remains unchanged at  1.07 for both cases although the
TDS increases from 4,000 to 5,200 mg-/A.  Therefore, with adequate
reaction time,  recycle  of the  ash pond  overflow to  the reaction
vessel will not cause scaling  problems.  The  total  dissolved
solids concentration increases,  however, from about 4,000 mg/&  to
about 5,200 mg/£.


4.2.4     Slurry  Solids Concentration

          The scrubber  makeup  requirements are not  affected by the
recirculating slurry solids content but a  study of  water recycle/
reuse opportunities in  scrubbing systems necessitates an under-
standing of scaling and the parameters  which  may  be used to control
scaling potential.   The recirculating solids  concentration  will
have a significant impact on the scale  potential  in the system
                               -58-

-------
      TABLE 4-4
EFFECT OF POND RECYCLE ON FOUR CORNERS

SCRUBBER MAKEUP REQUIREMENTS

Ash in Flue Gas, g/sec
(Ib/mln)
Particulate Removal, %
Total Fraction of Ash Dissolving
SOa in Flue Gas , ppm
SO 2 Removal, %
Solid Waste
Wt. % Solids
Solid Composition, Wt. %
CaSCK'ZHaO
Inert (Ash)
Solid Flow, g/sec
(Ib/min)
Pond Recycle Rate, 5,/sec
(GPM)
Makeup Water Requirement , £ /sec
(GPM)
Evaporation
Solids Occlusion
Total
Alternative Two*
16,200.
(2140)
99.7
.011
640.
50.

30.

7.9
92.1
17,500.
(2300)
0.0
(0)

30.0
(475)
40.7
(645)
70.7
(1120)
Alternative Three**
16,200.
(2140)
99.7
.011
640.
50.

50.

8.5
91.5
17,500.
(2300)
20.0
(320)

33.2
(525)
17.5
(278)
50.7
(803)
 *Repiped scrubbing system with adequate reaction time to prevent scale,  no
  ash pond recycle.
**Same as Alternative Two except all ash pond overflow is recycled.
                                  -59-

-------
      TABLE 4-5.   EFFECT OF ASH POND RECYCLE ON  FOUR CORNERS
                    SCRUBBING SYSTEM  SCALING  TENDENCY

                                     Alternative  Two*      Alternative Three**


 Pond Recycle Rate, &/sec                     °-°                   2°'°
                   (GPM)                     (0)                   (320)

 Solid Waste Solids Concentration,
   wt. %                                     30.                    50.
Scrubber Recycle Slurry
PH
Liquor Composition, mg/&
Calcium
Magnesium
Sodium
Chloride
Carbonates (as CO 3)
Sulfate (as SOO
Sulfite (as SOl)
Nitrate (as NO^)
TDS
CaSOit'ZHaO Relative Saturation***
Solids Composition, wt. %
CaSO^'ZHzO
Inert (Ash)

6.9

670.
84.
450.
235.
116.
2400.
33.
16.
4000.
1.07

7.9
92.1

7.0

640.
145.
790.
390.
105.
3100.
35.
27.
5200.
1.07

8.1
91.9
  *Repiped  system with adequate reaction time to prevent  scale and no pond
   recycle.

 **Same as  Alternative Two except  ash pond overflow is recycled.

***The critical value, above which scale potential exists,  is 1.3-1.4 for
   CaSO<*'2H20.
                                   -60-

-------
because the solids provide precipitation sites for CaS03-%H20 and
CaSCK-ZHzO.   A decrease in the number of sites slows precipitation
rates and therefore maintains more calcium and sulfate in solution,
raising the calcium sulfate relative saturation.  This effect is
illustrated by the two simulations of the Colstrip scrubbing sys-
tem shown in Table 4-6.  As the slurry solids concentration was
reduced from 12.670 to 7.6%, the calcium sulfate relative satura-
tion increased to 1.73, which is considerably above the critical
range for scale formation of 1.3-1.4.


          The recirculating slurry solids concentration is there-
fore an important parameter to control as far as scale prevention
is concerned.  A tradeoff does exist, though, in that slurries
with higher solids concentrations are more abrasive.


4.2.5     Reaction Tank Size

          In a lime or limestone based S02 scrubbing system, the
control of CaSO^•2H20 and CaS03*%H20 scale potential is an extre-
mely important consideration.  If scaling conditions are realized
for significant amounts of time in any part of the system, chemi-
cal scale will probably be deposited on equipment which will even-
tually necessitate system shutdown for cleaning.  Pure phase cal-
cium sulfite or calcium sulfate kinetics can be described by the
expression:


                     R = KafCV  (R.S. - 1)               (4.2)


where

          R = precipitation rate

          K = temperature dependent constant

          a = crystal interfacial area per mass of
              precipitating solid

          f = weight fraction of the precipitating
              species in the solid phase

          C = total solids concentration in the slurry

          V = reaction tank volume

       R.S.  = relative saturation of the precipitating
              species.
                              -61-

-------
 TABLE 4-6.  EFFECT OF SLURRY SOLIDS CONCENTRATION ON COLSTRIP

             SCRUBBER SCALE POTENTIAL
Scrubber Recycle Slurry
Suspended Solids, wt. %
pH
Liquor Composition, mg/&
Calcium
Magnesium
Sodium
Chloride
Carbonates (as GOT)
Sulfate (as SO")
Sulfite (as S07)
Nitrate (as NOl)
CaSO^-ZHzO Relative Saturation*
Solid Composition, wt. %
CaSOi*'2H20
Inert (Ash)
CaS03-%H20
Design
Conditions
12.6
4.98

733.
5,285.
444.
117.
153.
21,000.
3,560.
9.6
1.41

38.9
57.3
3.8
Low
Solids
7.6
5.09

897.
5,590.
469.
124.
167.
22,300.
3,930.
10.2
1.73

38.7
57.9
3.4
'•'The critical value,  above which scale potential exists  is
 1.3-1.4 for CaS04*2H20.
                              -62-

-------
The system precipitation rates are  set by  the  S02 removal rate
and oxidation rate.  As can be seen  from Equation 4.2, increases
in the weight fraction of gypsum  in  the solid  phasei the total
solids concentration in the circulating slurry, or the reaction
tank volume should decrease the gypsum relative saturation in the
reaction tank for a specific precipitation rate.  Table 4-7 illus-
trates the effect of reaction tank volume  on relative saturation
for the Four Corners scrubbing system.


          The only difference between Cases One and Two of Alter-
native Three is the reaction tank volume.   This alternative in-
volved recycle of the ash pond overflow in a repiped system with
adequate reaction tank volume for scale control.  As the tank size
was decreased from 37,500 m3 to 21,200 m3,  the gypsum relative
saturation in the tank increased  from 1.07 to  1.13.  The decrease
in reaction time kept more calcium  and sulfate in the liquid phase
and less in the solid phase, resulting in  the  increase in relative
saturation.  The change in solid  phase composition was very small
and did not have a significant impact on the results.  The change
in relative saturation is only about 5% different from what
would be expected by direct ratio using Equation 4.2 (neglecting
changes in solid composition).


          These results are based on the assumption that 98.67<> of
the sorbed S02 is oxidized.  Four Corners  scrubber tests performed
by Arizona Public Service since this study was made indicate that
the amount of sorbed S02 that is  oxidized  is a function of the pH
of the scrubbing liquor.  Lime addition in their venturi scrubbers
at Four Corners raised the pH and reduced  the  oxidation to a level
that permits operation in the "subsaturated gypsum mode".  They
found that the reaction tank volume  required with reduced oxida-
tion is significantly less than that predicted using the 98.6%
oxidation assumption in the simulations.


4.2.6     Liquid-to-Gas Ratio

          The effects of liquid-to-gas ratio  (L/G)on the revised
Four Corners scrubbing system operation are shown in Table 4-8.
The simulation results shown in the  first  column  (Alternative Two,
Case One) represent operation at  the present L/G  at Four Corners
(4.7 £/m3 & STP) but with adequate  reaction time  for scale con-
trol.  The calculated pH for the  scrubber  bottoms stream is 2.9.
This low value could cause corrosion problems_if the system oper-
ated in this manner for extended  periods of time.
                              -63-

-------
     TABLE  4-7.
EFFECT OF REACTION TANK VOLUME  ON FOUR CORNERS
SCRUBBING SYSTEM SCALING POTENTIAL
                                      Alternative Three
                                          Case One*
                                         Alternative Three
                                            Case Two**
 Combined Reaction Tank Volume***, m
                                (ft3)

 Scrubber Recycle Slurry

   Suspended Solids,  wt.  %

   pH

   Liquor Composition,  mg/5.

     Calcium
     Magnesium
     Sodium
     Chloride
     Carbonate (as C0"i)
     Sulfate (as SOlj)
     Sulfite (as SOI)
     Nitrate (as NOl)

   CaSOi^-ZHzO Relative  Saturation****

   Solid Composition, wt. %
                        37,500.
                       1.32 x 106
                           10.0

                            7-0
                          640.
                          145.
                          790.
                          390.
                          104.
                         3,110.
                           35.
                           27.

                            1.07
21,200.
7.5 x 101



    10.0

     7.0
   670.
   140.
   760.
   360.
    93.
 3,120.
    40.
    25.

     1.13

Flue
S02
S02
S02
CaSOit^HaO
Inert (Ash)
Gas Ash Flow, g/sec
(Ib/min)
in Flue Gas, ppm
Removal Rate, %
Oxidation, %
8.0
92.0
16,200.
(2140)
640.
50.
98.6
7.9
92.1
16,200.
(2140)
640.
50.
98.6
   *Repiped Scrubbing System, more than adequate reaction tank volume for scale
    control.
  **Same as Case 1 but reduced reaction tank volume.
 ***Combined volume of all  six proposed reaction tank vessels (one per scrubbing
    train).
****The critical value,  above which scale potential exists, is 1 3-1 4 for
                                    -64-

-------
           TABLE 4-8.   EFFECTS OF LIQUID-TO-GAS  RATIO ON
                         FOUR CORNERS SCRUBBING  OPERATION

Liquid- to-Gas Ratio, £/m3 @ STP
(gal/ 1000 ACF)
Scrubber Effluent Slurry
Suspended Solids, wt. %
PH
Liquor Composition, mg/5,
Calcium
Magnesium
Sodium
Chloride
Carbonate (as C0"i)
Sulfate (as SO=O
Sulfite (as S0=)
Nitrate (as NO!)
CaSOif«2H20 Relative Saturation ***
Solid Composition, wt. %
CaSO^*2H20
Inert (Ash)
Flue Gas Ash Flow, g/sec
(Ib/min)
S02 in Flue Gas, ppm
S02 Removal, %
S02 Oxidation, %
Alternative Two
Case One*
4.7
18.7

10.5
2.9

700.
85.
450.
240.
124.
2,750.
37.
16.
1.16

7.4
92.6
16,200.
(2140)
640.
50.
98.6
Alternative Two
Case Two**
10.0
39.8

10.2
3.9

690.
85.
450.
240.
120.
2,580.
35.
16.
1.14

7.6
92.4
16,200.
(2140)
640.
50.
98.6
  *Alternative Two involves  repiping scrubbing system, adding reaction tank
   volume to control scale.   Ash pond overflow is not recycled.
 **Same  as Case One except L/G  increased.
***The critical value, above which scale potential exists, is 1.3-1.4 for
  CaSCV2H20.
                                   -65-

-------
          The second column in Table 4-8 presents the results
from a simulation with a higher L/G of 10.0 £/m3 @ STP.  The
calculated pH of the scrubber effluent stream was 3.9 as opposed
to the value of 2.9 for the previous case.  This value is still
not ideal from the standpoint of corrosion control but is some-
what better than the 2.9 pH scrubber liquor.  The reason for the
rise in pH from Case One to Case Two is that less SOa is absorbed
per volume of scrubbing liquor so that the change in liquid phase
sulfur concentration across the scrubber is less.  This results
in less of a pH drop across the scrubber.   In both cases the
scrubber feed pH was 6.9.


          A small change in gypsum relative saturation in the
scrubber effluent occurred with the increase in L/G (dropped
from 1.16 to 1.14 when L/G increased from 4.7 £/m3 @ STP to
10.0 2,/m3 @ STP).  This again is due to the smaller sorption
rate of S02 per volume of scrubber liquor and therefore a smal-
ler change in relative saturation across the scrubber.


4.2.7     Makeup Water Quality

          The effects of makeup water quality on scrubbing system
operation and design were determined by using the Colstrip scrub-
bing system data and varying the makeup water composition input
to the simulation model.  Table 4-9 presents the results from
three simulations where the makeup water composition was varied.
The first case represents the design conditions with makeup water
being softened river water.  The other two cases represent using
untreated river water and cooling tower blowdown as makeup water.


          The results from these simulations indicate that the
makeup water has very little effect on the recirculating slurry
scaling tendency.  The relative saturation of gypsum did not
change appreciably between cases.  The liquor composition did
change considerably.  The total dissolved solids concentrations
ranged from 31,000 mg/£ for the base case to 46,000 mg/£ for the
case using cooling tower blowdown as makeup.


          The chloride content of the scrubbing liquor increased
from 120 mg/5, for the base case to 1,560 mg/£ for the case where
cooling tower blowdown is used.  The corrosive properties of this
liquor will depend on the materials of construction for the scrub-
ber.  The feasibility of use of cooling tower blowdown depends
therefore in part on the chloride content of the initial makeup
water to the plant.  Higher chloride levels may cause corrosion
problems in some scrubbing systems.
                             -66-

-------
      TABLE 4-9.   EFFECTS OF MAKEUP WATER QUALITY ON COLSTRIP
                   SCRUBBING SYSTEM SCALING POTENTIAL


Makeup Water Source
Makeup Water Composition, mg/S,
Calcium
Magnesium
Sodium
Chloride
Carbonate (as CO 3)
Sulfate (as SO^)
Nitrate (as NO^)
Scrubber Recycle Slurry
Suspended Solids, wt . %
pH
Liquor Composition, mg/&
Calcium
Magnesium
Sodium
Chloride
Carbonate (as C03)
Sulfate (as SO^)
Sulfite (as SOj)
Nitrate (as NOl)
IDS
CaSOit-2H20 Relative Saturation**
Solid Composition, wt. %
CaSCVZHzO
Inert (Ash)
CaS03-%H20
Case A
Softened
River Water

39.9
10.7
40.3
17.0
6.0
188.
1.4

12.6
5.0

730.
5,290.
440.
120.
150.
21,000.
3,560.
10.0
31,000.
1.41

38.9
57.3
3.8
Case B
Raw
River Water

57.9
10.7
40.3
48.7
17.3
188.
1.7

12.6
5.0

740.
5,280.
440.
330.
150.
20,700.
3,550.
12.0
31,000.
1.41

38.9
57.2
3.9
Case C
Cooling Tower*
Slowdown

534.
143.
540. -
227.
6.5
2,640.
18.7

12.6
5.0

690.
6,190.
3,870.
1,560.
150.
29,400.
3,700.
130.
46,000.
1.42

39.9
56.3
3.8
 *Use of this water as demister wash may not be feasible depending on  the
  amount of S02  sorbed in the demister.
**The critical value, above which scale potential exists,  is 1.3-1.4 for
                                   -67-

-------
          Since scrubbing systems generally use makeup water as
demister wash, the use of cooling tower blowdown may be limited to
other makeup requirements such as pump seal water.  The feasibil-
ity of using cooling tower blowdown as demister wash will depend
on the amount of S02 sorbed in the demisters.   Since cooling tower
blowdown may be nearly saturated with respect to CaSOit-2H20 when
maximum cycles of concentration is achieved in the tower system,
only a small amount of SO2 sorbed in the demisters may cause
scaling problems.  Colstrip pilot studies indicated that cooling
tower blowdown could not be used as demister wash in that system.


          One possible solution to this problem is to dilute the
cooling tower blowdown with a subsaturated stream such as river
water so that when S02 is absorbed,  the critical relative satura-
tion of CaSOi»'2H20 will not be exceeded in the demisters.  Pilot
studies to determine the required dilution under various operating
conditions are needed before a system change to an existing scrub-
bine process could be made.


4.3       Recycle/Reuse Alternatives in S02/Particulate Scrubbing
          Systems

          Makeup water is required in S02/particulate scrubbing
systems to replace water lost through evaporation and through
occlusion with the waste solids.   Evaporative losses include evap-
oration in the scrubber, from process vessels, and in the pond
system if the pond overflow is recycled to the scrubbers.  These
losses are fixed by the flue gas flow, temperature,  and water con-
tent and by the exposed area of process vessels and  the pond system.


          The amount of water lost through solids occlusion depends
on the amount of solids produced and the weight fraction solids of
the waste.  Scrubbing systems may be designed to be closed loop,
i.e., no aqueous discharges.  The Four Corners study (Appendix F)
showed that recycle of the ash pond overflow liquor had no appre-
ciable effect on the recirculating slurry scaling potential.  In
this case, where a pond is used for disposal,  the final sludge
composition will be approximately 50% solids (1 kg of water lost
per kg of solids).


          The amount of solids to be disposed of will depend on
the ash removal rate and the S02  removal rate. The  relative amount
of ash and S02 will determine which removal rate has the most ef-
fect on the amount of solids produced.  At Four Corners, where
about 30 kg of ash are removed per kg of S02,  the S02 removal rate
has very little effect on the scrubber makeup requirements.
                               -68-

-------
          Since scrubbers can be operated in a closed-loop fash-
ion,  there is little opportunity for recycle/reuse but they can
be used as a sink for water streams which might normally be dis-
charged from a power plant such as cooling tower blowdown.  The
use of various qualities of makeup water did not affect the
scaling potential of the scrubber recycle liquor in the Colstrip
scrubbing simulations (Appendix J).  This indicates that the qual-
ity of makeup water not used as demister wash  (for instance, pump
seal water) does not have a major impact on scrubber scaling ten-
dency.  The composition of this water will, however, affect the
chloride content of the recirculating liquor.  Corrosion problems
could possibly occur if the chloride content of the makeup water
is too high.


          The use of water which is saturated  or near-saturated
with respect to gypsum (such as cooling tower  blowdown) as demis-
ter wash is not recommended since the addition of small amounts
of S02 in the demister may cause scale formation.  Combining
cooling tower blowdown with fresh makeup water and using this mix
as demister wash may be feasible in some situations.  A careful
process analysis and pilot studies are suggested for determining
the quality of water required for demister wash in a particular
scrubbing system.
                               -69-

-------
5.0       COMBINED SYSTEMS

          The object of this study was to evaluate methods to
minimize the aqueous discharges and makeup requirements_for
coal-fired steam-electric generating stations.  In Sections 2.0,
3.0, and 4.0, the operating conditions of the major water con-
sumers at power plants were discussed.  In these sections,
methods to reduce the water consumption of the individual pro-
cesses were presented.  It is important to use these methods in
an efficient combination to produce an overall system which will
reduce the discharge requirements of the entire plant at a
reasonable cost.  This section discusses alternative operating
conditions and the associated costs for the water systems at
the Arizona Public Service Four Corners Station, the Public Ser-
vice of Colorado Comanche Station, the Georgia Power Co.  Plant
Bowen, the Pennsylvania Power and Light Montour Station,  and
the Montana Power Co. Colstrip Plant.


          Detailed discussions of the  water systems at these
plants as well as the recycle/reuse alternatives studied for
each are presented in Appendices F through J.  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 inves-
tigation of process variability.  More water quality data would
be required along with additional studies to fully characterize
the ash.
5. 1       APS Four Corners

          The major water consumer other than the cooling pond
(Morgan Lake) at Four Corners is the particulate/S02 scrubbing
system.  Cooling is accomplished by recirculating water from
Morgan Lake.  Bottom ash is sluiced in a recirculating fashion
with Morgan Lake providing the sluice water.   All of the recy-
cle/reuse alternatives for Four Corners concern themselves ex-
clusively with the scrubbing system since scaling problems have
been encountered and an effective water management scheme neces-
sitates scale free operation of the scrubbing system.  Since
Morgan Lake is already part of two recirculating systems (bottom
ash, cooling) no additional recycle/reuse alternatives were in-
vestigated.  To do so would involve considering the effects on
heat dissipation from the pond and is beyond the scope of this
project.
                              -70-

-------
          Four alternatives were investigated for the parti-
culate scrubbing system at Four Corners.  Table 5-1 presents
a summary of these four alternatives  compared to existing op-
erations .


          The results of the first alternative simulation indi-
cates that the present system tankage  capacity is not sufficient
to allow ample gypsum precipitation to prevent scaling.  These
results are based on the assumption that 98.6% of the sorbed S02
is oxidized to sulfate in the scrubbers.  In cases where the
oxidation is less than 98.6%, less tankage  capacity will be re-
quired to prevent scale.  Since gypsum precipitation is the con-
trolling factor at these levels of oxidation, lower oxidation
rates will lower the amount of gypsum to be precipitated and
therefore require a smaller reaction  tank.  In all four alter-
natives oxidation was assumed to remain at  the level measured
at the plant.  Studies conducted at Four Corners after this
work was completed showed that addition of  lime to the venturi
recirculation tank lowered the sulfite oxidation to a level
where the scrubbers operate in the subsaturated mode with re-
spect to gypsum.  The hold tank volume required to prevent
scale in this mode is much smaller than the volumes predicted
in this study where high oxidation was assumed.


          In the second alternative,  a tank capacity of 37,500
cubic meters (1.33 x 106 cubic feet) was simulated.  Gypsum
relative saturations were reduced to  levels below the critical
level required for the on-set 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/1000 scf) 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


          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 Jt/sec (1122 GPM) for Alternative  2 to  about 50.8 £/sec
(807 GPM).  Also, a simulation with ash pond overflow recycle
using a reaction tank volume of 21,200 nr  (7.5 x 105 ft )
showed that a more reasonable reaction tank volume  can be uti-
lized.  This simulation showed a gypsum relative saturation of
1.19 in the scrubber effluent slurry.
                              -71-

-------
             TABLE 5-1.   SUMMARY OF RECYCLE/REUSE OPTIONS AT FOUR  CORNERS1
Existing
Condition
Case 1 Case 2
Weight Percent Solids
in Thickener Bottoms 10 30
Hold Tank Volume. 0 0
m3 (ftj)
Liquid to Gas Ratio, 4.7 4.7
I/in' @ STP (gal/scf) (35.2) (35.2)
% Recycle from the
Ash Pond 0 0
S02 Removal, % 30 30
Oxidation, 7. 98.6 98.6
Paniculate Removal
prior to scrubber, % None None
Scrubber Makeup Rate, 223 70.7
e/sec (CPU) (3540) (1730)
Costs :
Capital, 1976 $
Operating, 1976 $ 2
(mils/kWh)
Alternative
Two
Case 1

30
37,500
(1.33 x 10') (1.
4.7
(35.2)

0
50
98.6

None
70.7
(1120)
3,334,000 4
628,000 1
(.128)
Alternative
Three
Case 2

30
37,500
33 x 10')
10.0
(74.8)

0
50
98.6

None
70.7
(1120)
,275.000
,101,000
(.225)
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)
Allurnat ive
Four
Case 1

30
8900
(0.31 x 10b)
10.0
(74.8)

0
50
98.6

60
41.0
(650)
3,385,000
968.000
(.198)
'Rough cost estimates were made to compare technically feasible options and do not include a "difficulty to retrofit" factor.
2Includes capital cost amortization of 15% per year.

-------
          The fourth alternative shows that reaction tank vol-
ume 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 (60% of fly ash re-
moved prior to scrubber) .   Water makeup requirements were also
reduced to 41.0 2,/sec (650 GPM) .


          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.  The operating
costs shown in Table 5-1 include capital amortization at 15%
per year.


5.2       PSC Comanche

          The major water consumers at Comanche are the cooling
towers and the bottom ash handling systems.  There is no S02
scrubbing and the fly ash is presently trucked off site.  As
part of the study of the water  system at Comanche wet fly ash
sluicing was also considered.


          Table 5-2 presents a  summary of the three alternatives
which were examined for Comanche.  The first one involved using
cooling system blowdown from the towers operating at five cycles
of concentration to sluice fly  ash and bottom ash on a once-
through basis.  The effects of  C02 mass transfer 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 sluic-
ing cases, but potential scaling of CaC03 and Mg(OH)2 was pre-
sent.  Although gypsum relative saturation was less than the
critical value (1.3-1.4) variations in ash or makeup water qual-
ity may cause gypsum scale.  The calculated relative saturation
for gypsum was 1.24.  This alternative will result in an ash
pond overflow of about 32.7 I/sec (518 GPM) for each unit as
compared to the existing configuration pond overflow rate of
about 78 £/sec (1230 GPM) per unit.
                              -73-

-------
                          TABLE  5-2.    SUMMARY OF WATER RECYCLE/REUSE OPTIONS  AT COMANCHE1
 I
-^J
-F-

Cooling Tower Makeup Source
Cycles of Concentration in
Cooling Towers
Cooling System Treatment
Fly Ash Disposal Method
Type. % solids
Bottom Ash Disposal Method
Type, % solids
Recycle in Fly Ash
System, %
Recycle in Bottom Ash
System, %
Treatment in Ash Systems

Plant Makeup Requirements
I/sec (GPM)
Plant Discharge
I/sec (GPM)
Costs
Capital Investment, 1976 $
Operating Expenditures, 1976 §/yr 3
(mils/kW-hr)
Additional Cost to Treat Pond
Overflow for Zero Discharge
Capital , 1976 $
Operating, 1976 $/yr 3
(mils/kW-hr)
Total Cost for Zero Discharge
Capital, 1976 $
Operating, 1976 $/yr 3
(mils/kU-hr)
Existing Alternative
Conditions One
Softened River Water Softened River Water

5.0 5.0
Alternative
Two
Softened River Water

7.6
(Sulfuric acid and zinc polyphosphate used for all com

Dry Wet, 10%

Wet, 1% Wet, 47. a

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, 10%

Wet, 4% *

10%

100%
Brine Concentration
of Makeup (50%)

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)
Alternative
Three
Softened River Watur

8.4
litions)

Dry

Wet. 11

	

100%
None


450 (7120)

30.2 (480)

222.000
38,000
(0. 008)


3,883,000
989,000
(0.20)

4.105,000
1,027,000
(0.21)
          'Rough cost estimates were made  to compare technically feasible options and do not include a "difficulty to retrofit" factor.
          2Sluicing bottom ash at 4% solids may not be feasible due to hydraulic limitations of equipment.
          'Includes capital cost amortization at 15% per year.

-------
          The second alternative  involves  using  cooling system
blowdown at 7.6 cycles of concentration  as makeup to a recircu-
lating ash sluice system.  The makeup water was  used to sluice
fly ash, and recycled ash pond water was used  to sluice bottom
ash.   Only 1070 of the fly ash sluice water was recycled from
the pond.   Gypsum relative saturations in  the  fly ash sluice
line were calculated to be 1.54 -  1.74 depending on the level
of C0? 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 concen-
trations in the system.  Desupersaturation of  gypsum in the ash
pond will also not prevent scaling  since only  a  small portion
of the ash pond liquor is recycled  to the  fly  ash system.  This
alternative will produce an ash pond overflow  of about 14.4
2,/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 at 1% solids using cooling  tower blowdown  and pond re-
cycle with the towers operating at  8.4 cycles  of concentration.
This will provide 16.0 £/sec (260 GPM) of  cooling tower blow-
down per unit and will not alter  the boiler refractory cooling
requirements.  For this alternative about  15.1 2,/sec (240 GPM)
of ash pond overflow per unit is  obtained.  This water may be
discharged or recycled to the boiler and cooling tower makeup
systems after appropriate treatment.


          Rough cost estimates were make for the once-through
sluice system and the recirculating system (Alternatives  1 and
2) using cooling tower blowdown to  sluice  fly  ash with 507o 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 ex-
pensive alternative ($342,000 for capital  cost and about
$90,000/yr operating cost, including capital amortization at
157o per year) .  The third alternative is the least expensive
of the three.  Capital costs are  about $222,000  and operating
costs are about $38,000/yr, including capital  amortization at
15% per year.


          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
                               -75-

-------
require an initial capital cost of about $3.7 million and an
operating cost of about $863,000/yr, including capital amorti-
zation at 1570 per year.  These costs do not include the possi-
ble 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 soften-
ing/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).


          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 $l,027,000/yr for operating costs.
Since fly ash is currently disposed of in a dry fashion at
Comanche, no additional costs for the fly ash system are needed.
This makes the third alternative the least expensive for
achieving zero discharge.  These costs include brine concen-
tration, additional piping, additional pumping costs, and capi-
tal cost amortization at 15% per year.


5.3       GPC Bowen

          The major water consumers at Bowen are the cooling
towers and the ash handling systems.  Bowen does not employ any
S02 or particulate scrubbing.


          Table 5-3 presents a summary of the technically fea-
sible operations and the relative costs of each of these alter-
natives.  Two process alternatives were studied for the ash
sluicing system at Bowen.  The first case involved using cool-
ing tower Slowdown 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
                              -76-

-------
         TABLE  5-3.   SUMMARY OF TECHNICALLY  FEASIBLE OPTIONS  AT  BOWEN1

Cooling Tower Makeup Source
Cycles of Concentration In
Towers
Cooling System Treatment
Acid Addition Kate, kg/day2
(Ib/day)
Ash Sluice Makeup Source
1 Z Recycle in Fly Ash
~^j System
"^ % Recycle in Bottom Ash
System
Ash System Treatment
Plant Makeup Requirements,
e/sec (CPM)
Plant Discharge Rate,
e/sec (CPM)
Costs
Capital, 1976 $
Operating, 1976 $/yr J
(mtls/kw-hr)
Existing Condition
Makeup Pond ,
Service Water
1.7
None
0 (0)
Cooling Tower Slowdown
0
0
None
3250 (51,500)
1600 (25,000)
--
Alternative One
Makeup Pond ,
Service Water
5.7
U2SOi,
481 (1060)
Cooling Tower Slowdown
0
0
None
1880 (29,800)
255 (4050)
100.000
52,900
(.002)
Alternative Two
Makeup Pond ,
Service Water
15.
HjSOi,
608 (1340)
Cooling Tower Slowdown
60
100
Recycle Softening
1670 (26,400)
41 (650)
1,223,000
402,000
(.018)
Alternative Three
Makeup Pond, Service Water,
Brine Concentrator Uistlllatt
15.
H2SO,
608 (1340)
Cooling Tower Slowdown
60
100
Recycle Softening, Brine
Concentration of Pond
Overflow
1630 (25, BOO)
0 (0)
6,380,000
1,735,000
(.078)
1 Kouyh cost estimates were made to compare technically feasible options and Uo not Include a "difficulty to retrofit
2 As KHIX II2SO,,.
Includes capital cost amortization at 15% per year.

-------
5-3).  The effects of C02 mass transfer in the ash pond and sluice
tank on the system operation were investigated.  No gypsum scale
potential was identified in any of the cases with once-through ash
sluicing.


          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 recirculating
ash sluice system (Alternative 2 in Table 5-3).  A blowdown of
41 Jl/sec (650 GPM) is taken from the ash pond for this alternative.
If the pond recycle water remains supersaturated with respect to
gypsum, scaling will occur in this system. However,  this  situation
may be remedied by chemical treatment.  Sodium carbonate softening
of approximately 80% of the pond recycle water will maintain 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.


          Zero discharge from the cooling and ash sluicing sys-
tems (Alternative 3 in Table 5-3) may be achieved by installing
a softening/reverse osmosis/brine concentration unit to treat
the above ash pond overflow (41 2,/sec or 650 GPM) and recycling
approximately 50% of the clean water as boiler makeup and the
remainder as cooling tower makeup.


          Potential scaling of CaS03 is present in all cases
studied, both once-through and recirculating.  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 accurately the size of
reaction tank and frequency and quantity of acid washing re-
quired or if other measures are necessary.


          The rough cost estimates presented for the alterna-
tives in Table 5-3 indicate that reducing the ash pond over-
flow 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 expen-
sive option (about $100,000 capital cost with about $53,000/yr
operating costs) .  This option necessitates acid treatment in
the towers.  The operating expenses include capital amortiza-
tion at 15% per year.


          Reducing the ash pond overflow to about 41 s,/sec (650
GPM) by operating the cooling towers at 15.0 cycles of concen-
                              -78-

-------
tration (with acid treatment) and using  the  tower blowdown as
makeup to a recalculating ash sluice  system  (with Na2C03 soft
ening of 80% of the pond recycle) has an initial capital cost
of about $1,223,000 and operating costs  including capital cost
amortization (15% per year) of about  $402,000/yr.  The use of
a softening/reverse osmosis/brine concentrator unit to elimi-
nate the ash pond overflow discharge  (recycle to boiler and
cooling tower makeup)  for this alternative  would require a capi-
tal investment of about $6.38 million total.  The additional
operating costs would be about $1,222,000/yr, giving a total of
approximately $1,735,000/yr.


5.4       PP&L Montour

          The water system at Montour is similar to the one at
Bowen.  Once-through wet sluicing of  both fly and bottom ash is
employed and there is no equipment  to handle S02.  The cooling
towers at Bowen and at Montour are  natural draft.


          Table 5-4 presents a summary of the technically feasi-
ble options for the Montour water system as  compared to existing
operations and the relative costs of  each of these alternatives.
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 opera-
tion.  In one case, Alternative 4,  a  blowdown was taken from the
system to prevent CaSO^HaO scale.   The other three alterna-
tives did not discharge any liquid  streams and controlled
CaSOIt-2H20 scale with softening of  a  portion of  the pond re-
cycle water.


          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 re circulating  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.
                                -79-

-------
                   TABLE  5-4.    SUMMARY  OF  TECHNICALLY  FEASIBLE OPTIONS  AT  MONTOUR1
Cycles of Concentration
  In Cooling Towers
                          Existing Condition
                                                  Alternative 1
1.5 - 2.0
                                                                         Alternative 2
                                                20
                                                                                                 Alternative 3
                                                                       20
                                                                                                                         Alternative 4
                                                                                               20
Assumed Drift  Rate
  in Cooling Tuwers
  ll/sec (Cl'M)
                             62 (1,000)
                     62  (1,000)
                                                                       (650)
                                                                                            40 (650)
Ulowdown from Coaling
  Towers
  4/sec (Ul'M)
                            725 (11,500)
                      48 (760)
                                             40 (650)
Z Kecycle in Fly Asli

CO
o
1










Sluicing System 0
Sluice System Makeup Cooling Tower
Source Slowdown
Total Makeup Water Rate,
I/ace (CPU) 1,500 (24,000)
Ultimate Effluent Kate,
£/sec (CPM) 500 (7,900)
Treatment Required None

Costs
Capital, 1976 $
Operating, 1976 $/yr k
(mils/kW-hr)
89.
Cooling Tower
Slowdown

1,000 (16,000)

0
IliSU., (Cooling Tower) '
Ma2COs (Pond Recycle) a

640,000
173,000
(0.016)
89.
Cooling Tower
Blowdown

950 (15,000)

0
HjSOi, (Cooling Tower)2
HaH'Oa (Pond Recycle) s

668,000
187,000
(0.018)
89. 73.
River Water River Water

985 (15,600) 1,035 (16,400)

0 50 (800)
II 2 SO,, (Cooling Tower)2 II2S(\ (Cooling Tower) 2
NazCOj (Pond Recycle)3

622,000 485,000
169,000 103,000
(0.016) (0.010)
'liuugli (;ost  uslJniaLes were made to compare technically  feasible options  and do not include a "difficulty to retrofit" factor.
2Sulfuric acid treatment for CaCOj scale control.
3Na2CO3 softening for Ca removal.
4 Includes capital cosl amortization at  15% per year.

-------
          The other two alternatives  assume  that  the cooling tow-
ers can be operated at zero blowdown.   This  requires that the drift
be at least 65.4 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 discharge  similar to the
two previous alternatives.  Alternative 4 controlled the CaS(K-2H20
scaling potential by the use of a blowdown stream of about 50 Jl/sec
(800 GPM) from the ash pond (both units).


          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 mea-
sures are necessary.


          Rough cost estimates for the different  alternatives are
also presented in Table 5-4.  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 required for the
recirculating ash sluicing system.  It should  be  emphasized that
Alternatives 1 and 2 differ only  in the assumption 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.


5.5       MPC CoIstrip

          The major water systems at  the 350 Mw/unit Colstrip
plant are the cooling tower and combined SOz/particualte scrub-^
bing systems.  Colstrip is designed for and  is achieving zero dis-
charge through brine concentration of the cooling tower blowdown
and a disposal pond for the scrubber  sludge.


          Table 5-5 presents a summary of the  two combined system
alternatives for the Colstrip water system as  compared to existing
operating and the relative costs  of each alternative.  All_flows
reported in Table 5-5 refer to those  produced  from both units.


          The first alternative does  not involve  any changes in
operation of the cooling towers but uses cooling  tower blowdown
and untreated river water as scrubber makeup as opposed to soft-
ened river water and brine concentrator distillate as  is pre-
sently done.  A capital cost of $159,000 is  reported for piping
                               -81-

-------
                        TABLE 5-5.   SUMMARY OF WATER RECYCLE/REUSE  OPTIONS AT  COLSTRIP1
i
oo
         Cooling Tower Makeup
          Source

         Cycles of Concentration
          in Cooling Towers

         Cooling System Treatment

         Treatment Rate,
          «7sec (GPM)

         Cooling Tower Slowdown
           Rate, ll sec
Scrubber Makeup Source
         Plant Makeup Rate
          fc/sec (GPM)

         Plant Discharge Rate
          1/se.c (GPM)

         Costs: l
            Capital, 1976 $
            Operating, 1976 $/yr. 2
               (mils/kwh)
                                            Existing
                                           Conditions
                                                          Alternative
                                                              One
                                                        Alternative
                                                            Two
                             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  Blowdown,
Untreated River Water     Untreated  River Water
                                                            423 (6710)
                                  0.
                                                            159,000
                                                            -237,000
                                                             (-.046)
                                  423  (6710)
                                    0.
                                                           275,000
                                                          -217,000
                                                            (-.044)
           ]R@ugh cost estimates were  made  to compare technically feasible options  and do not include
           a, "difficulty to retrofit" factor.

           2Includes  capital  cost  cimortization at 15% per year.

-------
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 blow-
down 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 sys-
tem.  The increased capital  charges  result in  a lower operating
expense 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 dis-
 charge, 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 Table 5-5 do not include any  savings which could
 have been realized if the  Colstrip water  system had been de-
 signed for the most effective cascading of aqueous streams.  A
 savings in capital investment could  have  been  achieved by de-
 signing the cooling towers for slipstream treatment and by
 using only one 150 GPM  capacity  brine  concentrator as opposed
 to the two 200 GPM capacity  units presently  used.  It should be
 noted that MFC considered  sidestream softening but did not im-
 plement it since it was not  believed to be reliable technology.
 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) .


 5.6       Recycle/Reuse Alternatives in Combined Systems

          Minimization  of  makeup requirements  and aqueous dis-
 charges from coal-fired steam-electric generating stations re-
 quires efficient cascading of the aqueous streams within the
 whole plant.  The best  arrangement  for any plant is the water
 use scheme which attains the required environmental standards
 at the least possible cost.   The optimum  balance of these  con-
 flicting requirements is not always  obvious.   Therefore, several
 alternatives were presented  for each power plant.  Although
                              -83-

-------
conditions varied from plant to plant the alternative water sys-
tems presented were similar in their overall approach to water
utilization.


          The cooling towers require the largest amount of water
at most plants.  They also need water of reasonably high quality
to insure that scale will not form in the condenser.  The makeup
requirements to the cooling towers are usually supplied directly
from the water source for the plant, the highest quality large
supply of water.   The towers concentrate the dissolved solids in
this water because of the evaporation that occurs in the tower.
A blowdown from the tower is usually required to control the
level of dissolved solids and prevent scale formation.   The size
of the blowdown stream can be reduced if acid treatment and/or
softening are used to inhibit scale formation.   The size and
quality of this blowdown stream makes it a good candidate for
makeup to other plant water consumers such as ash sluicing or
scrubbing operations.


          All of the alternatives studied for the four power
plants employing cooling towers have the cooling towers serving
as the major recipient of the plant makeup water.  Smaller uses
of the plant makeup include feed to the demineralizers, for
boiler makeup, and general service water.   The remaining plant
water needs are supplied by the cooling tower blowdown.


          Two of the plants studied, Bowen and Montour, employ
wet fly and bottom ash sluicing.  All of the makeup to  these
systems is supplied from the cooling towers.  The rate  of
cooling tower blowdown at Montour is determined by the  water
requirement for the once-through sluicing system.  Excess blow-
down is discharged at Bowen.  In order to meet the once-through
ash sluicing water demand, the blowdown from the towers is much
larger than necessary to control condenser scale.  The  rate at
which water is delivered to the ash ponds in once-through sys-
tems is greater than the rate at which it is lost from the ponds
through evaporation and occlusion with the pond sludge.  This
excess water from once-through sluicing usually is discharged.


          Recirculating ash sluicing systems were investigated
as a potential recycle/reuse scheme.  The systems were designed
to use much less cooling tower blowdown by sluicing the ash
with a mixture of pond water and cooling tower blowdown.  In
some of the alternatives studied for Montour, the size of the
cooling tower blowdown was equal to the losses that occur in
                              -84-

-------
the pond,  thereby attaining zero discharge.  At the increased
cycles  of  concentration realized because of the reduced cooline
tower blowdown, acid treatment was needed to prevent CaC03 scale
in the  condensers .   Softening of the pond recycle stream was
needed  to  control CaS0lt-2H20 scale in the ash slurry line.  At
Bowen softening of 80% of the pond recycle stream was required
to control gypsum scale in a re circulating ash system with 60%
of the  fly ash sluice water recycled.  To attain zero discharge
with this  system would require an expensive treatment step,
such as brine concentration, for the pond overflow.  The quality
of the  cooling tower blowdown and the reactivity of the fly ash
at Comanche was such that a recirculating fly ash sluicing sys-
tem would probably scale the slurry lines at 10% recycle without
softening.  Softening of the recycle stream at 10% recycle
would not  be effective since the recycle represents only a small
portion of the total sluice water.  It was recommended that
Comanche continue to dispose of their fly ash by dry methods.


          The design of the Colstrip water system and alterna-
tives suggested in this study are similar to the other proposed
systems.  The cooling tower makeup is the major portion of the
water entering the plant and the blowdown from the towers serves
as makeup to other water consumers.  Some of the water is lost
in the S02 scrubbers via evaporation and occlusion with the
scrubber sludge and some of the water is lost in the recircula-
ting bottom ash system.


          The present operation of the Colstrip plant is at zero
discharge.  The alternatives proposed for the water system sug-
gest ways to attain zero discharge at reduced cost.  The major
savings occurs because the alternatives do not use as much brine
concentration as is presently used.  Only enough water to supply
boiler makeup is sent to the brine concentrators.


          In summary, the three major water consumers with in-
creased recycle/reuse opportunities at coal-fired power plants
are cooling towers, ash sluicing, and S02/particulate scrubbing.
Cooling ponds may also be considered as major water consumers
but the study of thermal dispersion in ponds was beyond the
scope of this project.  An investigation of increased recycle
in cooling ponds would require that the effects on the pond heat
dissipation be considered.  The cooling towers demand the high-
est quality water and produce a fairly large blowdown stream.
A desirable method to achieve zero discharge is to limit  the
cooling tower blowdown to a level no greater than  the total
makeup  requirements for the other water consumers.  The ash
                               -85-

-------
sluicing and S02/particulate scrubbing systems serve as good
water sinks because of the water losses occurring through evap-
oration and occlusion with sludges in these systems.  Cooling
tower blowdown can be used to sluice ash either in a once-
through or recirculating system or as makeup water to a scrub-
bing system (excluding demister wash).
                              -86-

-------
               GENERALIZED  IMPLEMENTATION  PLANS


1.0       INTRODUCTION

          Limited availability and rising  costs of water along
with reduced discharge requirements have placed an ever increas-
ing importance on water recycle/reuse at coal-fired power plants.
The first part of this project involved investigating water recycle/
reuse alternatives of five  typical power plants.  This section
of the final report presents generalized implementation plans
for the types of recycle/reuse possibilities identified in the
plant studies.


          Three types of major water  systems were  identified as
candidates for increased recycle/reuse  options  at  coal-fired
power plants.  These are cooling  tower, ash sluicing, and S02/
particulate  scrubbing systems.  At each of the  four plants studied
with cooling towers, increased recycle  was possible in the cooling
tower system through the treatment options of sulfuric acid addi-
tion for pH  control and/or  softening  for control of gypsum scale
potential.   For each of the two plants  studied  with once-through
wet fly ash  sluicing, a closed-loop recirculating  system may be
used with softening to control gypsum scale potential.  For scrub-
bing systems, the recycle/reuse options identified include:
1) using a normally discharged stream such as cooling tower blow-
down as scrubber makeup in  a cascaded water system, and 2) con-
verting from open-loop to closed-loop operation.


          The generalized implementation plans  for each of these
types of recycle/reuse options may be divided into four phases:

          Phase I   - System Characterization

          Phase II  - Alternative Evaluation

          Phase III - Pilot-scale Studies

          Phase IV  - Full-scale  Operation

The first phase,  system characterization,  is necessary  to  estab-
lish a data  base  for determining  existing  operating  characteris-
tics and identifying the types of recycle/reuse options possible.
This includes collecting all available  design  and  operating  "J"*-
mation and a sampling program to  supplement existing  data.   The
data base resulting from this phase  of the implementation  plan
can be used  to evaluate recycle/reuse options  for  both  extreme
                               -87-

-------
and average operating conditions in the second phase.  The most
important operating variables and their effects on system opera-
tion are presented in the characterization phase descriptions.


          The plant studies performed in this project used a
computer model to evaluate recycle/reuse opportunities.  The
results of the plant studies showed that computer models are
effective in identifying water recycle/reuse options in cooling
tower, ash sluicing, and S02/particulate scrubbing systems.   The
second phase of an implementation plan involves an analysis of
alternatives to evaluate the feasibility of those alternatives
under various operating conditions.   Computer simulation results
can be used to design necessary treatment step to prevent scale
formation, and provide a basis for designing any pilot-scale
equipment required for testing particular options.  The evalua-
tion discussions in this document include evaluation criteria
and a methodology for using a computer model to perform the
evaluations.


          Pilot-scale studies are listed as the third phase of
a generalized implementation plant.   These studies may be required
to determine parameters which may not be easily measured or cal-
culated from the characterization or simulation phases.  For ash
sluicing systems, these parameters include:  1) ash reactivity,
2) the effectiveness of including an ash reaction tank prior to
the sluice line to reduce scale potential, and 3) the amount of
C02 transfer occurring in the pond.   For scrubbing operations,
pilot-scale studies may be required to determine the operability
of a revised demister wash operation.  Pilot studies are not
required for cooling towers since they exist as recirculating
systems and may be changed gradually.  Care should be taken when
implementing a cooling tower system modification.  A sampling
program to monitor cooling tower operation will minimize the
risk of scaling during implementation.


          The final phase of an implementation plan is to make
full-scale modifications.  The pilot-scale test results may be
used to design the full-scale modifications.  In the case of
cooling towers, the full-scale modifications may be based on the
results of the evaluation phase of the implementation plan.   The
following sections discuss each of the phases of an implementa-
tion plan as  they apply to cooling tower, ash sluicing, and S02/
particulate scrubbing systems.
                              -88-

-------
2.0       COOLING TOWERS

          Wet cooling towers are generally  the largest water
consumers at coal-fired electric power plants.  The towers dis-
perse the waste heat from the condensers by contacting a recir-
culating water stream with the atmosphere.  Water losses occur
from the tower in three ways:  evaporation, drift, and blowdown
The sum of these three streams is  equal to  the makeup requirements
for the tower.  Evaporation accounts for the majority of the heat
lost from the tower.  Drift occurs because  fine droplets of water
are carried off as a mist due to the intimate contact of the air
and water.  Most cooling towers have mist eliminators designed to
minimize drift.  The blowdown is a purge stream which is used to
control the level of dissolved solids which will build up in the
recirculating water.  Figure 2-1 shows the  general flow scheme
for a cooling tower system including the tower and condenser.


          The evaporation and drift rates are set by the design
of the tower, the ambient conditions, and the cooling load.   The
blowdown rate is maintained at a level sufficient to prevent
scale formation in the condenser.  However, as shown by the plant
studies portion of this project, in many cooling towers the size
of the blowdown is much larger than required to prevent scale
formation.  Three of the four cooling tower systems investigated
could operate at increased recycle through  the use of acid addi-
tion to control pH.  By better defining the limits of scale for-
mation in recirculating cooling systems, the blowdown can be
reduced and thus the net water consumption  and discharge from
these systems can be reduced.


          This section presents a  methodology for defining the
limits of safe cooling tower operation and  implementing a demon-
stration plan.  This plan will allow increased recycle and
decreased makeup and blowdown requirements.  Only three of the
four implementation phases discussed in the introduction apply
to cooling towers.  These are:

          Phase I   -  Characterize Cooling Tower Operation

          Phase II  -  Evaluate Alternatives

          Phase III -  Design and  Implement Full-scale
                       Modifications
                               -89-

-------
                             EVAPORATION
                                       DRIFT
MAKEUP
                           COOLING
                            TOWER
                          CONDENSER
SLOWDOWN
   Figure  2-1.   General  cooling tower system flow scheme.
                               -90-

-------
          The first phase  involves  collecting  the  data base
which is required to identify  reasonable  system modifications
The second phase identifies  the  degree  of recycle  that can be
attained and the amount of treatment  that will be  required   A
detailed design of the modified  cooling tower  system'can be made
along with a detailed procedure  for implementation of the modifi-
cations from the results of  the  process calculations performed
in the second phase.  These  calculations  will  allow specification
of stream flows, temperatures, compositions, and treatment equip-
ment sizes.  Operation of  a  pilot  system  is not required for
cooling towers because cooling towers already  exist as recircu-
lating systems which can be  changed in  a  gradual manner.  The
cost of building a pilot system  would not be justified even though
there is a possibility that  if the  towers are  operated incorrectly
the main condenser could scale.  Care should be taken to safely
modify the operation of full scale  cooling towers.  A comprehen-
sive sampling program will allow the  cooling tower operation to
be monitored so that the risk  of scaling  is minimized.


          The plant studies  conducted in  this  program involved
the first two phases of an implementation plan.  Design and op-
erating data concerning four cooling  tower systems was collected
and used to evaluate operation both under existing and alterna-
tive conditions.  However, only  limited operating  data was avail-
able concerning these systems.


          Implementing the alternatives studied at the existing
plants could require some  additional  studies to better define
the variations in operating  parameters.  These studies would in-
volve a sampling program to  define  tower  operation and evaluation
of the system under extreme  conditions.  The plant studies per-
formed were based on results from  one set of grab  samples, exis-
ting makeup water quality  data,  and design information for tower
operating conditions  (air  flow,  circulating water  flow, clima-
tological data, etc.).


          The following discussions are written to apply to any
cooling tower facility.  All phases of  an implementation plan are
presented although the characterization and evaluation phases
have been addressed in the plant studies. Since virtually all
cooling tower systems are  unique with respect  to operating condi-
tions, a detailed characterization and  evaluation  phase should be
conducted before implementing  increased recycle at a particular
plant site.  The purpose of  the  plant studies  was  to identify and
evaluate the types of recycle/reuse alternatives achievable at
coal-fired power plants.   The  purpose of  this  document  is  to out-
line a procedure for implementing  the types of alternatives iden-
tified in the plant studies.
                               -91-

-------
2.1       Phase I:  Cooling Tower Characterization

          In the characterization phase the basic data which des-
cribes cooling tower operation is assembled.  These data should
include information about both average and extreme tower operat-
ing conditions.  This section discusses the important data for
characterizing tower operation.  First, the necessary data is
discussed and then a sampling plan to supplement available data
is presented.  If sufficient data is available, a sampling pro-
gram may not be required before implementing an increased
recycle/reuse option.


2.1.1     Identification of Process Variables

          Table 2-1 represents a data collection sheet which may
be used to assemble cooling tower system operating conditions.
This table includes data blanks for acid treatment, softening,
and other chemical treatments for cooling tower operations.  If
the treatment methods are not used, the data concerning them
will not be relevant.


          The most important stream to obtain long range data
about is the makeup water.  The makeup water quality determines
the quality of all of the other streams in the cooling system.
Small changes in the concentrations of dissolved solids in the
makeup water are magnified in the blowdown because of the con-
centrating effect of the cooling tower.  Any long range varia-
tions in the cooling water quality can be traced to changes in
the makeup water.  Seasonal variations in makeup water quality
may require variations in treatment levels required to prevent
scale formation in the condenser.  In order to insure that the
tower modifications do not cause scaling, the variations in
makeup water should be identified.


          The calcium, magnesium, sodium, chloride, sulfate,  ni-
trate, carbonate, silica, and TDS concentrations as well as pH
are the most important in defining the makeup water quality.   Cal-
cium carbonate, calcium sulfate, and silica are the primary scale
forming species in cooling towers.  The magnesium, sodium, nitrate,
and chloride concentrations are important since they affect scale
formation through chemical complexing or effects on activity.
High chloride concentrations may cause corrosion problems.  Phos-
phate scales may also form if the phosphate level is high.  The
results of the plant studies indicated that phosphate levels in
natural bodies of water are not high enough to cause any scaling
problems.   Phosphate levels should be checked to insure that
the makeup water does not have an unusually high level.  If high
                              -92-

-------
        TABLE 2-1.   GENERAL DATA SHEET FOR COOLING TOWERS

          Tower Parameters
          A.   Drift Rate
          B.   Approach
          C.   Cycles of Concentration
          D.   Circulating Water Temperature Change
 II.       Ambient Air
          A.   Dry Bulb Temperature                        °F
          B.   Wet Bulb Temperature                  _  °F
          C.   Flow Through Tower                    _ ACFM

III.       Cooling System Makeup Water
          A.   Flow                                  _ GPM
          B.   Composition
                  pH
                  total calcium as Ca^                   mg/£
                  total magnesium as Mg"^                 mg/1
                  total sodium as Na                      mg/Jl
                  total chloride as Cl~                   mg/5,
                  total nitrate as NO3                    mg/Jt
                  total carbonate as C0~^                  mg/£
                  total sulfur as SO^                     mg/£
                  total silica as Si02              	 mg/£
                  total dissolved solids            	 mg/H

 IV.       Cooling System Slowdown
          A.   Flow                                  	GPM
          B.   Composition
                  pH
                  total calcium as Ca"1"^             	 mg/£
                                       _l	i
                  total magnesium as Mg             	 mg/£
                              -93-

-------
   TABLE 2-1.   GENERAL DATA SHEET FOR COOLING TOWERS (Continued)
                   total sodium as Na
                   total chloride as Cl~
                   total nitrate as NOl
                   total carbonate as CO3
                   total sulfur as SO^
                   total silica as Si02
                   total dissolved solids
                                                mg/A
                                                mg/£
                                                mg/£
                                                mg/£
                                                mg/A
                                                mg/£
   V.
Condenser
A.  Flow Rate of Cooling Water
B.  Temperature of Exit Water
C.  Condensing Steam Temperature
                                                           GPM
                                                            °F
  VI.
Acid Addition
A.  Flow
B.  Wt. % K2S
                                                           Ib/day
                                                           7
                                                           /o
 VII,
VIII
Softening
A.  Additive Type (Lime, Soda Ash, C02)
B.  Amount of Chemical Addition
C.  Capacity of Treatment Equipment
D.  Size of Treated Stream
E.  Size of Waste Stream
F.  70 Solids in Waste Stream

Chemical Treatment (Scaling, Corrosion
Inhibitors)
A.  Type
B.  Amount
                                                           Ib/hr
                                                           GPM
                                                           GPM
                                                           GPM
                                                           7=
                               -94-

-------
levels (>01 mg/£) are detected, analysis for phosphate should
be included with the other analyses.


          Variations in both the individual species concentrations
and the total level of dissolved solids can be expected.  The full
range of possibilities should be identified.  This will"insure
that the treatments which are used  in the cooling tower operation
will be designed for all of the water qualities which the tower
will be expected to operate with.


          For characterizing cooling tower operation it is impor-
tant that climatological data be collected.  The wet bulb and dry
bulb temperatures are the most  important climatological data that
must be obtained for cooling tower  systems.  These data are the
determining factors affecting the air flow rate required to meet
the heat dissipation demand on  the  tower.


          The wet and dry bulb  temperatures affect the  evapora-
tive capacity of the air flowing through the tower.  Variations
in these temperatures will cause variations in the evaporation
rate and therefore will affect  the  makeup requirements  of the
system.  Increased evaporation  rates will cause a higher con-
centration of dissolved solids  in the circulating water if blow-
down and drift are constant.  Equation  2.1 shows the relation-
ship between evaporation, drift, and blowdown rates and the level
of concentration occurring in the tower.
                                B  + D + E
                                                           (2.1)
                           ^       B  + D

where

          C = cycles of  concentration

          B = blowdown rate

          D = drift rate

          E = evaporation  rate

From this equation it can  be  seen that changes  in the  evapora-
tion rate will directly  affect  the  cycles  of concentration  if
the blowdown is held constant.


          Expected variations in load placed on the cooling towers
should also be identified.   Changes in the load should directly
                               -95-

-------
affect the evaporation rate in the tower because most of the
cooling which occurs is due to evaporation (^90%).   The steam
condensing temperature is an important parameter in character-
izing a cooling tower system.   The highest temperature in the
system is in the condenser.  This is the most likely point of
CaC03 or CaSO<+ scale formation since these species  are less
soluble at higher temperatures.   The condenser temperature should
therefore be used when evaluating tower scaling potential.


          The data collection discussed so far is concerned with
data representing a reasonably long period of time.   This data is
very useful to characterize the operating conditions of the tower
under a variety of natural and man-made variations.   Of necessity,
this data must be collected over long periods of time and some
of it may not be available.  To supplement this data, direct
sampling of the cooling tower's major streams may be necessary.


2.1.2     Sampling Program

          In cases where tower operating data is incomplete, a
sampling program should be conducted at the plant for about two
weeks.  During this period it is recommended that about 3-5 sets
of grab samples per day be taken.  All of the samples should be
taken at approximately the same time to produce a series of "snap
shots" of the system operation which can be used to identify
deviations from steady-state.


          The data which are required include much  of the informa-
tion listed in Table 2-1.  The most important samples are of the
cooling tower makeup and blowdown.  If flow data is available,  it
should be noted, but the costs of magnetic, turbine, or orifice
metering devices for large streams are too expensive to use for a
short sampling period.  The wet bulb and dry bulb temperatures
and the inlet and outlet condenser temperatures should be recor-
ded at the time that the sampling takes place.


          The grab samples of cooling tower makeup  and blowdown
which are taken should be analyzed using the procedures which are
reported in the EPA's "Manual of Methods for Chemical Analysis  of
Water and Wastes".  Important parameters include pH, temperature,
total dissolved solids, calcium, magnesium, sodium,  carbonate,
sulfate, nitrate, chloride, and silica.


          If any treatment is used, such as acid for scale control
or chlorine for biological control, it should be monitored
                              -96-

-------
continuously.  It is more important  that average values of these
very small addition streams are known  than  the instantaneous
rate at the time when the grab sample  is taken.


          If softening is used for the cooling tower, separate
grab samples_of the inlet and outlet streams from the softeners
will be required.  Sample analyses of  the inlet and outlet stream
of the softener are required in the  same manner that the cooling
tower makeup and blowdown are.  The  outlet  stream from the soft-
ener should also be analyzed for  suspended  solids to determine
if the flow rate is too large for the  clarifier and sedimentation
system.  If a substantial level of suspended solids is noticed,
then the softener is being overloaded.


          Some of the data which  is  useful  to characterize a
cooling tower cannot be easily measured.  In this case design
information or values calculated  using indirect measurements must
be used.  In general, the water recirculating rate, the air flow
rate, and the drift are not well  known, especially in natural
draft towers.  Estimates of the water  flow  rate can be made using
data about the inlet and outlet condenser temperatures and the
electric load on the generator to estimate  the waste heat load.
The air flow rate can be estimated using data concerning the
load, the ambient conditions and  the approach on the tower.  The
drift is not easily estimated and can  only  be measured
approximately.


          To characterize the operation of  a cooling tower all
sources of information should be  consulted.  The main sources are
historical data kept by the power plant and government agencies,
and sampled  data which can be obtained over a short period of
time.  Any of the required data that cannot be found in these
two sources  can be supplemented by design information and in-
direct calculations.


2.2       Phase II;  Evaluation of Operating Alternatives

          The second phase of a cooling tower implementation
plan is to formulate and evaluate various modes of cooling tower
operation.  Formulation of alternative operating conditions will
depend on how the cooling tower fits into the overall plant
recycle/reuse scheme.  The plant  studies conducted in this pro-
gram showed  that cooling tower blowdown can be cascaded to other
power plant water systems such as ash  sluicing and S02/particulate
scrubbing.
                              -.97-

-------
          In a plant designed for minimum liquid discharge, the
cooling tower blowdown rate is set by the water requirements of
the systems which use cooling tower blowdown as makeup.  Since
evaporation and drift are determined by tower design, the blow-
down sets the cycles of concentration required in the tower (see
Equation 2.1).  The types of water systems used at a particular
plant site and therefore the quantity of water required by these
systems determine the cycles of concentration required in the
cooling tower.


          To determine the feasibility of operating the cooling
tower system at the cycles of concentration required by a parti-
cular recycle/reuse option, both potential scale formation and
the level of treatment required should be investigated.  The
plant studies portion of this project showed that a computer pro-
cess simulation package is a very useful tool with which to study
cooling tower systems.  The simulations can be used to identify
the operating characteristics of the tower under increased recycle
conditions.  The results of simulations can also be used to iden-
tify the treatment methods required to operate with increased
recycle and the degree of treatment necessary.   The following
sections describe the types of calculations required to evaluate
cooling tower operation and a methodology for using a computer
model to perform the evaluations.


2.2.1     Evaluation Criteria

          The cooling tower evaluation should involve mass and
energy balances to calculate the flows? compositions, and tem-
peratures of all of the process streams.   The calculations should
include predictions of the inorganic aqueous equilibria phenomena
that occurs in cooling towers, including solid-liquid and gas-
liquid equilibria.   The solid-liquid equilibria include all of
the important scale forming species:  CaCOs, Mg(OH)2, CaSO^-ZHzO,
Si02,  and other silicate scales.  The gas-liquid equilibria
predicitions are necessary to calculate the evaporation rate and
the degree of C02 transfer that occurs in the tower.  These cal-
culations allow cooling tower operation to be evaluated from both
the standpoints of scale potential and treatment requirements.


2.2.1.1   Scale Potential

          The most important factor which determines what cycles
of concentration can be reached in a cooling system is the poten-
tial for scale formation in the condenser.  If scale does form in
the condenser, the heat transfer can be reduced significantly.
This will raise the steam condensing temperature, increase the
                              -93-

-------
back pressure in the turbine  and reduce  the  efficiency of the
power generating cycle.


          In the plant  studies  portion of  this  study the con-
cept of relative saturation was used to  measure the scaling
potential of aqueous solutions.  For calcium carbonate, a common
scale, the relative saturation  is  defined  in Equation 2.2.
          R.S.
               (CaC03)
                          K
                           (CaC03)
                               (2.2)


                               (2.3)
                              K
                               (CaC03)
where
          R.S.
               (CaC03)
             K
              (CaC03)
                •c.-"  =
relative saturation of calcium
carbonate

solubility product of calcium
carbonate

activity of the calcium ion
                M- ++  =  moality of the calcium ion
                Y
                 Ca
                 aC03
                 M
                  col
                 Yco7
activity coefficient of the
calcium ion

activity of the carbonate ion


molality of the carbonate ion


activity coefficient of the
carbonate ion
 The activities must be used instead of the molalities of the
 ions in solutions to take into account the non-idealities which
 exist in aqueous solutions.  Equation 2.3 defines, relative satu
 ration in terms of ion molalities and activity cont
 activity coefficients account for deviations
 In the plant studies portion of this project, these
                                -99-

-------
were shown to become larger as the total dissolved solids of
the solution increased.


          When the relative saturation of a species is below 1.0,
no potential for scale formation exists.  When it is greater than
1.0, a potential for solids formation exists but scale will not
necessarily form.  When the relative saturation of a species is
above its "critical value", nucleation will occur and scale for-
mation is more likely.  The critical value for CaSOit-ZHzO is
1.3-1.4 and in this study, the critical values of CaC03 and
Mg(OH)2 were found to be about 2.5 and 3.4, respectively.


          The most common scale forming species in cooling towers
is calcium carbonate.   The relative saturation of CaCOs is depen-
dent on the concentration of the calcium and carbonate ions, as
shown in Equation 2.3.  The calcium concentration depends on the
cycles of concentration.  The carbonate concentration depends on
the pH and the liquid-gas equilibrium between gaseous carbon
dioxide and dissolved carbonate.  It is necessary that the C02
transfer which occurs between the air and the water in the tower
is accounted for to accurately predict the relative saturation of
CaCOa.  This calculation is especially important in cases where
there is no pH control because the pH of the circulating water
is affected by the amount of total carbonate species in solution.


          Figure 2-2 is a sample plot of the relative saturation
of CaC03 in the cooling tower blowdown as a function of the
cycles of concentration.  This data was obtained using the computer
model of cooling towers described in the plant studies portion of
the project.  This plot shows that the relative saturation of
CaCOs increases in a non-linear fashion, and the curve becomes
steeper as the cycles increase.  The relative saturation increases
at a faster rate at higher cycles of concentration because the
pH rises dramatically as a function of cycles, as can be seen in
Figure 2-3.   The increased pH causes less C02 to be desorbed in
the towers as well as causing a shift of bicarbonate ions to
carbonate ions.  The plots shown in Figures 2-2 and 2-3 apply to
cooling tower systems where the blowdown is adjusted to keep the
cycles of concentration at a low level to prevent scale formation.
However, the discussions concerning relative saturation and C02
transfer also apply to systems where treatment is used.


2.2.1.2   Treatment Alternatives

          In many cases blowdown alone is not sufficient to con-
trol CaC03  scale and pH control may be instituted.  Sulfuric
                              -100-

-------
   1.
  .75
 s

 O
 _J
 CD


 lil
 n
O

0.50


O
CE

3
V)
  .25
Ul
LU

OC
    1.0
                2.0
                            3.0
4.0
            5.0
                    [CYCLES OF  CONCENTRATION]
   Figure 2-2.  CaCO3 scale potential as a  function of

                 cycles of concentration in  cooling towers

                 without treatment (example).
                            -101-

-------
  7.9
  7.8
Z


§7.7
S
o
01
I
i-

u.
O
  7.6
  7.5
     1.0
2.0          3.0          4-°

     [CYCLES  OF CONCENTRATION]
                                                    5.0
  Figure 2-3.
Cooling  tower blowdown  pH as a function of
cycles of  concentration in towers  without
treatment  (example).
                            -102-

-------
acid treatment is used  to  control  the  pH  of  the recirculatine
water because it is the least  expensive of most commercially
available acids.  The gas-liquid-solid equilibria calculations
identify the pH required to  control  the relative saturation of
CaC03 to a specified level as  well as  the amount of sulfuric
acid needed to maintain this pH.   Mg(OH)2 and  some silicate
scales can also be controlled  with acid addition, but CaC03 is
usually the limiting scale.


          CaSOif*2H20 is another scale  that can be encountered
in cooling systems.  Since its solid-liquid  equilibrium is not
pH dependent, acid addition  will not reduce  its relative satura-
tion.  The addition of  sulfuric acid increases the sulfate ion
concentration and actually increases the  relative saturation of
CaSCK*2H20, although in most cases this effect is insignificant.
When problems are encountered  with gypsum scale, softening is
required to reduce the  calcium concentration,  either by pre-
treatment of the makeup water  or slipstream  treatment of the
recirculating water.  The  evaluation of a particular option
should involve  determining the chemical requirements for pre-
treatment and/or slipstream  treatment.


          The size of the  stream required to be softened must be
calculated.  This is especially important because it determines
the size of the softening  equipment  which will be required.  It
is imperative that the  softening equipment be  able to handle the
largest stream  that might  be necessary to soften.  If the equip-
ment capacity is exceeded, the flocculation  and sedimentation
times are reduced and CaC03  solids may carry over into the
cooling water.  The results  of this  carryover  could be a heavily
scaled condenser.  The  maximum stream  size should be as small
as reasonable though, because  the  capital cost of the softening
equipment is relatively high and is  a  direct function of the
size of the softened stream.


2.2.2     Model Application

          A fundamental computer model such  as the one used in
the plant studies is a  very  useful tool in performing the  cal-
culations discussed above  concerning evaluation of cooling tower
operation.  The cooling tower  model  should initially be verified
for the specific cooling system being  studied  using the data
collected in the characterization  phase.  This can best be done
by inputting data collected  during the sample  periods and  com-
paring the simulated results to measured  values.  If the data
base is extensive enough,  the  model  assumptions can be altered
to improve its  reliability.  In cooling  tower  systems parameters
                              -103-

-------
such as the air flow rate and the degree of C02 transfer are not
always known to a high degree of accuracy.  These parameters_may
be adjusted based on extensive operating data to more effectively
model the actual tower operation.


          The verified model can then be used to simulate alter-
native operating conditions.  The simulations should be performed
for the full variety of conditions which the tower can be expected
to operate under.  The analysis of the available historical data
is very useful in the identification of seasonal as well as non-
seasonal extremes in weather and water quality.  It is very
important that the worst case operation is simulated, so that the
cooling tower modifications, based upon these simulations, will
be able to handle all expected conditions.


          The model should not only be used to simulate cooling
tower operations which account for long range variations in oper-
ating conditions, but also for short-term upsets.   The effects of
drastic changes in load, steam condensing temperature, drift,
blowdown, and climate should be identified with simulations.
These simulations identify the kind of changes in treatment rates
that the modified system should be able to handle.   It is very
important that upsets in system operation do not cause the rela-
tive saturation of any scale forming species to rise above its
critical value, and scale the condenser.


          Several simulations should be performed to verify and
"fine tune" the model.  Alternative operating conditions may
require 10-15 simulations of varying ambient and operating con-
ditions.  Three to five additional simulations should be performed
to estimate the effects of short-term upsets.


          The results of these simulations identify the operating
conditions which can be expected with increased recycle.  They
specify the temperatures, flows and compositions of the important
streams under various operating conditions.  These results estab-
lish a sound basis upon which to design an acid treatment system,
if necessary, which can operate effectively under all foreseeable
situations.   If softening is required, the simulation results
identify the maximum flow through the softener and the maximum
chemical addition rate as well as the quality of the outlet
stream from the softener.  The results of these simulations should
establish a sound basis upon which to design system modifications
which will increase the cycles of concentration in the cooling
tower but will not reduce the reliability of the system.
                              -104-

-------
2.3       Phase III:  Cooling Tower Modifications

          After the optimum system operation is identified from
the results of the evaluations, a complete  design with the new
operating conditions can be prepared.  This design should include
specifications of the flows, temperatures,  and compositions of
the main process streams as well as all of  the new equipment and
instruments necessary to monitor the cooling system.


          A plan for the necessary changes  in operating conditions
from the initial system to the modified system can be developed
with the new design.  The plan will help insure safe operation
and minimize disruption to the main operating objectives of the
plant,  the production of electricity.


          This section discusses the kinds  of modifications that
can be made on a cooling tower to increase  the cycles of concen-
tration, and presents examples of how the operating changes could
be implemented.


2.3.1     Equipment

          In most cooling systems additional equipment will be
required to significantly increase the cycles of concentration.
This new equipment can include piping, pumps, controllers, feed-
ers, a clarifier, or a thickener.  This subsection discusses the
equipment necessary to increase recycle in  cooling towers.


          Additional piping and pumping capacity may be needed to
reroute the plant makeup to replace the water which is not supplied
by cooling tower blowdown because of the increased recycle in the
towers.  This will be a temporary situation required only during
implementation of the modifications when the cooling tower blow-
down is being reduced.  If pH control is added to the system, a
pump, a storage tank, and a pH controller will be required.  The
maximum capacity of the metering pump should be larger than the
flow rate calculated in the evaluations of  worst case conditions.
If slipstream or makeup softening is added  to the system, a
chemical feeder with variable feed rates, a flocculator, and a
sedimentation tank will be necessary.  The  residence time in the
flocculator should be at least 40-60 minutes.  Sedimentation re-
quires  a 2-4 hour residence time.  A slurry pump and piping will
also be required to pump the softener sludge to the ash pond.  The
cost of equipment for modifying cooling tower operation will be a
strong  function of the size of the system and the levels of treat-
ment necessary.
                               -105-

-------
          Any new treatments added to the system require addi-
tional monitoring.  The acid addition can be checked with pH
probes to insure that complete mixing of the acid and cooling
water occurs.  If only partial mixing occurs before the cooling
water enters the condenser, corrosion and scale problems might
occur.  The inlet and outlet streams from the softener should be
monitored continuously for conductivity, pH, and turbidity.  If
a significant increase in turbidity is noticed, the flow rate may
be too large for the flocculator and carryover is occurring, also
increasing the danger of CaC03 scale.  A common procedure for
softening operations is to perform "jar tests" to determine what
new operating conditions are necessary for safe operation.  How-
ever, if increased turbidity is seen, the blowdown from the tower
should be increased to avoid any scale formation.


          If slipstream treatment is used,  the flow through the
softener is an important operating variable.  The total amount
of calcium and magnesium removed in the softening step is
directly dependent on the flow and concentration of the calcium
and magnesium ions.   To effectively control the total amount of
hardness removed by the slipstream softener a flow meter should
be installed to monitor the flow on a continuous basis.


2.3.2     Implementation

          Implementation of cooling tower modifications which
involve increased recycle should be performed in a stepwise man-
ner to insure safe operation.   Each step should take about a week
of monitoring,  to insure that the tower is  operating as expected.
Samples should be taken of the makeup and the blowdown at least
twice a day and analyzed for calcium, magnesium, sodium,  sulfate,
chloride, and carbonate.  The TDS can be measured and correlated
to conductivity measurements which will give a continuous read-
out on the quality of the makeup and blowdown streams.  Changes
in the conductivity of these streams implies that changes in
the water quality have occurred and the towers are not operating
as expected.


          The manner in which increased recycle is implemented
can be best illustrated with examples.  There are four basic
modes of cooling tower operation:

          1)   Operation at low cycles of concentration
              so that no treatment is required.

          2)   Operation at higher cycles of concentra-
              tion with sulfuric acid addition.
                              -106-

-------
          3)   Operation at higher cycles of concentration
              with sulfuric acid addition and softening
              of the makeup water.

          4)   Operation at higher cycles of concentration
              with sulfuric acid addition and softening of
              a slipstream from the recirculating water.


          In the plant studies, two plants which operate their
cooling towers in the first mode were identified.  These plants
operate their natural draft towers at less than three cycles of
concentration.  These plants have a very high quality makeup to
their towers as do most plants which operate without treating
their cooling water for scale prevention.  Preliminary studies
of these cooling tower operations indicate that by changing to
the second mode of operation, the towers can operate safely at
cycles of concentration as high as fifteen.


          To add the capability of pH control to a cooling system
only requires a pump, a storage tank, and a pH controller.  After
this additional equipment is installed, the implementation proce-
dure can be initiated.  This implementation procedure should take
about five to ten weeks to complete the change from operation at
low cycles of concentration to a higher level.


          Each step which increases the cycles of concentration
by reducing the blowdown should be operated for at least a week.
During this period, the pH, temperature, and conductivity of the
makeup and blowdown will be monitored continuously.  The plant
operator should be instructed as to what the maximum acceptable
levels in these readings are and to increase the blowdown if any-
one exceeds the maximum.  It is recommended that complete chem-
ical analyses of the makeup and blowdown streams be performed at
least twice a day.  The results of these analyses can be used to
revise the maximum acceptable levels of the continuous read vari-
ables, if necessary.  The maximum temperature, pH, and conduc-
tivity pertain to a condition very close to the maximum accept-
able scale potential.


          The relative saturation of CaSO^HaO cannot be con-
trolled by acid addition because it is not pH sensitive.  If a
cooling tower is operating near the critical scaling value of
CaSO^HaO, softening is required to increase the degree of
recycle.   Softening can be used to reduce the calcium level in
the makeup or a slipstream from the recirculating water.  These
levels of operation represent the third and fourth modes listed
above.
                              -IQ7-

-------
          If softening is being added to an existing system,
testing operation of the softener is recommended before the
softened stream is fed into the cooling tower system.  Complete
chemical analyses of the effluent will insure that sufficient
calcium is being removed.  The pH should be low enough so that
CaC03 or another pH sensitive solid will not form where the make-
up is added to the cooling water.  The tubidity of the effluent
can be measured to insure that complete sedimentation is occurring
in the thickener.  Jar tests performed on the influent will deter-
mine the optimum lime addition rate.  Jar testing involves adding
various dosage rates to samples of the softener influent.   The
optimum dosage rate is the lowest rate which still produces a
precipitate that settles readily.


          After safe operation of the softening equipment is
assured, operation of the cooling tower with softening can be
attempted.  The tower can initially be run at the normal level
of recycle with the softening equipment in operation.  Complete
monitoring of the temperature, pH and conductivity of the make-
up, blowdown, and softener effluent is required to track the
system operation.  The plant operator should be instructed as to
the maximum acceptable levels of each of these variables.   Com-
plete analyses of these streams at least twice a day are recom-
mended.  Daily jar tests are necessary to determine the optimum
level of lime addition.  Operation at each step should be carried
out at least a week before the level of recycle is increased
further.
          Softening the makeup stream will control
scale and is easier to control that slipstream treatment, since
the softener is not in the cooling loop itself.   The results of
Phase I of this study have shown that slipstream treatment is
more efficient and can allow a greater degree of recycle in
cooling towers with smaller capital costs for equipment.  If a
plant wishes to change its cooling system from pretreatment to
side-stream treatment, this could be accomplished by repiping
the system.


          In the case of existing softening equipment, reliability
has been proven for softening the makeup stream.  Therefore, ini-
tial tests of the repiped system will be short.   The system
should be monitored continuously, as discussed earlier for soften-
ing in general, and the recycle should be increased in a stepwise
manner.  Because the quality of the slipstream tends to fluct-
uate more than the makeup water, more frequent jar tests are
recommended.   As with the other system modifications complete
change from initial operation to final operation should take
about five to ten weeks.
                              -108-

-------
          In summary, there are four basic modes of cooling tower
operation.   As the treatment level becomes more complex from acid
addition to slipstream softening, the degree of recycle achievable
also increases.  To change modes and increase the cycles of con-
centration in the cooling tower will usually require about five
to ten weeks after all new equipment is  installed.  This time is
required to insure safe  operation of the cooling towers at all
times and to educate  the operating personnel.
                                 -109-

-------
3.0       ASH SLUICING

          The ash sluicing system is a major water consumer at
many coal-fired power plants.  Most plants sluice the bottom ash
on an intermittent basis from the boiler to a pond where the ash
settles.  Fly ash is usually removed from the flue gas by an el-
ectrostatic precipitator, a mechanical collector, or a fabric
filter.  The fly ash collected by these means is often sluiced
from the collection mechanism to an ash pond for final disposal.
The bottom ash and fly ash can be sluiced to a common pond or
separate ponds.


          Bottom ash is usually sluiced in a slurry of 1-5% solids
and fly ash is sluiced at 5-10% solids.  Losses occur via evapora-
tion from the surface of the pond, occlusion with the ash in the
bottom of the pond (-50% solids), and seepage through the pond
liner.  The remaining sluice water must then be discharged or
used somewhere else in the plant.


          Because ash pond overflow is usually high in dissolved
solids it is not suitable for many other uses in the plant.  One
possible consumer of ash pond overflow is a scrubbing system.
However, the makeup requirements for a scrubbing system are
usually much smaller than the ash pond overflow from a once-
through sluicing operation.  Therefore, possible recycle/reuse
alternatives involving ash sluicing operations necessarily inc-
lude the possibility of closed-loop or partially closed-loop
systems.  The plant studies portion of this project investigated
both closed-loop ash sluicing and partially closed-loop opera-
tions.  In a closed-loop system, all of the ash pond overflow
is recirculated to sluice ash.  The only water losses from this
type of system are occlusion with the ash in the pond and evap-
oration.  In a partially closed-loop system, a portion of the
ash pond overflow is recycled and the remainder is either casca-
ded to a water consumer such as a scrubbing system, or it is dis-
charged.


          Figure 3-1 shows a general recirculating ash sluicing
system.  Some recirculating bottom ash sluicing systems presently
operate without significant problems.  Two of the plants studied
in this project operate recirculating bottom ash sluicing systems
without scale formation in the slurry lines.  Bottom ash is essen-
tially non-reactive and does not serve as a major source of dis-
solved solids to the slurry liquor.  Fly ash is generally much
more reactive and a significant amount of dissolved species can be
leached from the ash when it is slurried with water.  Most fly ash
sluicing systems are once-through operations, with the pond over-
flow discharged.
                              -110-

-------
                      ASH
                       i
                    EVAPORATION
       MAKEUP
        SLUICE
        WATER
SOLID/LIQUID
MIXING
f


POND




	 : 	 f SLOWDOWN
i
I
i
                                        T
                                      SLUDGE
                              POND  RECYCLE
      Figure 3-1.   Re circulating ash sluicing  flow scheme.


          This section presents  a methodology  for  implementing
a recirculating  ash sluicing system.   This  is  broken into four
phases:
          Phase  I

          Phase  II

          Phase  III

          Phase  IV
System Characterization

Alternative Evaluation

Pilot-scale Studies

Full-scale Operation
          The first phase  involves  collecting  a data base which
is required to identify  reasonable  system modifications.  The
second phase identifies  the  degree  of  recycle  attainable without
treatment to prevent scale formation.   If a greater level of re-
circulation is required, the second phase also identifies the
treatment requirements to achieve that level.


          The evaluation of  alternatives in Phase II will depend
heavily on the accuracy  of the  ash  reactivity  data collected in
Phase I.   The results of the plant  studies portion of this pro-
gram showed that ash reactivity is  a strong function of both the
pH and dissolved solids  content of  the sluice  water.  The trans-
fer of C02 between the atmosphere and  the ash  pond will affect
the pond pH and therefore affect the ash reactivity.  The_rela-
tionship  between ash reactivity and sluice water composition is
                               -111-

-------
a very complex one and deserves pilot studies to better  evaluate
recirculating ash sluice systems.  The third phase  in  the  imple-
mentation plan is to conduct pilot studies to better define  the
ash reactivity in a recirculating system and to evaluate treat-
ment options for preventing scale formation.


          The final phase of the implementation plan for recir-
culating ash sluicing is to design and startup the  full-scale
system.


3.1       Phase I:  Ash System Characterization

          In the characterization phase, the basic  data  necessary
to evaluate a recirculating ash sluicing system is  assembled.
These data include information about both average and  extreme
operating conditions.  This section discusses the important  data
for characterizing an ash sluicing operation.  First,  the  impor-
tant process variables are presented and discussed.  Next, a sam-
pling plan to supplement existing data is presented.   Finally,
bench-scale studies to further characterize the particular ash
to be sluiced are outlined.


3.1.1     Identification of Process Variables

          The major process variables which affect  the feasibility
of implementing a recirculating ash sluicing system are  listed
below:


          1)  Reactivity of the ash,

          2)  Rate of production of ash,

          3)  Surface area of the ash pond(s),

          4)  Local climatological data,

          5)  Sluice water flow and composition,

          6)  Degree of C02 transfer between the
              pond liquor and the atmosphere.


          The most important parameter is the ash reactivity.
The ash reactivity is defined as the amount of soluble species in
the ash.   The results of the plant studies portion  of  this pro-
gram showed that the major species leached from the fly  ashes
                              -112-

-------
studied are calcium, magnesium,  sulfate, and sodium   When the
ash is contacted with water, these  species dissolve and may cause
calcium carbonate or calcium sulfate  scale formation.


          The collection of accurate  data concerning ash reacti-
vity is therefore extremely important when evaluating the feasi-
bility of a recirculating ash  sluice  system.  The plant studies
conducted earlier in this program showed that the ash reactivity
varies considerably from plant  to plant and that the reactivity
of a particular ash is a function of  the sluice water composition.
The best way to collect accurate ash reactivity data is through
sampling and experimental studies.  These methodologies are pre-
sented in the following sections.


          The rates that fly and bottom ash are produced deter-
mine the amount of water required as makeup to a sluice system.
In a recirculating system, the  total makeup required is the sum
of the water lost through occlusion with the ash in the pond and
through evaporation.  The rate  of ash production and the percent
solids the ash will settle to  determine the amount of water lost
through occlusion.


          The pond surface area and local climatological condi-
tions determine the rate of water lost from the pond through
evaporation.  The makeup rate  required is therefore partially
determined by the pond area and climate.


          As 'previously stated,  the sluice water flow and compo-
sition affects the ash reactivity.  The makeup water will, to
some extent, determine the type and levels of dissolved salts
present in the recirculating system.


          The last parameter listed,  C02 transfer, will affect
the pH and carbonate concentration  of the pond recycle liquor,
and thereby affect the ash reactivity.  Characterizing the amount
of CO2 transfer occurring is a  very difficult task.  Not only
does the C02 transfer depend on the pond pH but it also depends
on the amount of surface agitation  resulting from wind.  The
plant studies portion of this  project investigated the two ex-
tremes of CO2 transfer:  no transfer  and complete equilibrium
with the atmosphere.  The actual level will be between these   _
two extremes.  Pilot studies to evaluate recirculating ash sluice
systems may provide data which  can  be used to better predict the
level of C02 transfer in full-scale systems.
                               -113-

-------
3.1.2     Sampling

          Sampling of an existing operation is one method to
determine ash reactivity.  By sampling the sluice water and the
ash slurry entering the pond, the amount of soluble species in
the ash may be calculated.   However, this calculation is depen-
dent on the assumption that no precipitation occurs in the sluice
line.  To determine if a potential for precipitation exists,
analyses of the ash slurry liquor can be used to predict scale
potential.  A sampling program can also provide information con-
cerning variations in ash reactivity.  A one to two week sampling
program should provide sufficient data to evaluate an existing
ash sluice system with respect to ash reactivity.


          Three to five samples from each slurry being sluiced
to the pond(s) should be taken each day.   It is recommended that
the inlet and outlet streams be sampled simultaneously,  if possi-
ble, to obtain a series of "snapshots" of the system operation.
These samples can be analyzed using procedures outlined in the
EPA's "Manual of Methods for Chemical Analysis of Water and
Wastes" for pH, temperature, total dissolved solids, total
suspended solids, calcium,  magnesium, sodium,  carbonate, sul-
fate, nitrate, chloride, and silica.


          If flow data is available, it should be noted, but the
costs of magnetic, turbine, or orifice metering devices for large
streams are too expensive to install for a short sampling period.
Ash flow rates data can be combined with the analysis for sus-
pended solids in the slurries to calculate the fly ash and bottom
ash slurry flow rates.


          Samples of the ash which are produced during this sam-
pling period should be obtained and labeled.  These samples can
then be studied under laboratory conditions and the results can
be compared to the results measured under actual operating condi-
tions.  This comparison can improve the usefulness of the labora-
tory findings.


3.1.3     Laboratory Studies

          In many systems the ash may serve as the largest source
of dissolved species in the sluice water.  This is especially true
when a recirculating fly ash system is operated where the pond
water is continually brought in contact with fresh ash and only a
small amount of "cleaner" makeup water.  Depending on the level
of ash reactivity, dissolved species concentrations may increase
                              -114-

-------
to the point where CaCO 3> CaSO,, Mg(OH)2> or other scales form
in the sluice line and restrict  the  flow of the slurry   A well
designed system will not operate for any significant period of
time in_the scaling mode.  In order  to design a recirculatine
ash sluicing system that will not  scale under normal operating
conditions it is important to predict the reactivity of the ash.


          Samples of both the bottom and fly ash which are pro-
duced at the plant should be obtained if recirculating systems
for both are being considered.   In most cases the reactivity of
the fly ash is much higher than  the  bottom ash.  Samples of both
should be studied initially because  the reactivity of ash is
affected by many parameters.  The  reactivity of ash is known to
be a function of the mine it is  from, the furnace it is burned
in, and the conditions at the time of combustion.  Therefore, it
is important that multiple samples of the ash from all of the
coals burned at the plant are obtained so that a reasonable
range of the ashes produced at the plant are studied.  The reac-
tivity of a given ash sample is  also a function of the pH and
composition of the sluice water.   This requires multiple tests
to obtain a reasonable estimate  of how the reactivity will change
with leachate composition.


          Two types of laboratory  studies were used to determine
ash reactivity in the plant studies  portion of this program.
Beaker leaching studies with deionized water were performed at
various pH's to determine the maximum level of soluble species
in the ash.  Recirculating bench-scale studies were also conduc-
ted.  The results of these two types of experiments indicated
that the amount of species dissolving from ash is much smaller
in a recirculating system than the level determined from the
beaker leaching studies.  The bench-scale results should more
closely reproduce actual conditions  and therefore should be
used to characterize ash reactivity.


          A flow schematic of the  bench-scale apparatus used
in the plant studies is shown in Figure 3-2.  This equipment
may be operated with varying degrees of recycle by adjusting
the blowdown rate.  A mixing tank  was used to contact the ash
with pond recycle and makeup water.  The ash slurry then flows
from the tank to the settling pond where the ash settles.  The
pond recycle and makeup water were transported with peristaltic
pumps and the ash was added via  a  worm screw feeder.
                               -115-

-------
        ASH STORAGE HOPPER
MAKEUP
WATER
WORM SCREW FEEDER

   ©
                8-
          MIXING TANK
                           -tXh
                    SLOWDOWN
                                SETTLING POND
                                                            OVERFLOW
                                                    - SAMPLE POINTS
         Figure 3-2.  Bench-scale model  of ash sluicing facilities

-------
          By sampling the makeup water,  pond  recycle water  and
the ash slurry, and monitoring  the  flows of ash, makeup water
blowdown  and pond recycle, material  balances  around the mixing
tank may be made to determine ash reactivity.  Analyses of the
aqueous samples taken should include  calcium,  magnesium, sodium
chloride, carbonate, sulfate, nitrate,  and silica   Variations '
in mixing tank residence time,  blowdown rate,  recycle rate  and
makeup water quality may be used to evaluate  ash reactivity and
potential scaling problems under different operating conditions.


          In order to facilitate the  completion of bench-scale
experiments within a reasonable time  period,  the size of the
equipment should be considered.  With the equipment used in the
plant studies, about 30 hours were  required for the system to
reach steady-state.  The pond liquor  residence time was about 6
hours.  Therefore, about five residence  times  were required for
steady-state operation.  The equipment must be sized before ash
samples are obtained so that an adequate amount of ash may be
available for the system to reach steady-state.


          The number of experiments required  for a particular
ash will depend on the levels of variations in operating para-
meters identified as being likely and the chosen equipment size.


3.2       Phase II:  Evaluation of  Operating Alternatives

          The second phase of a recirculating  ash sluicing imple-
mentation plan is to formulate  and  evaluate various modes of
operation.  Formulation of alternatives  will  depend on how the
ash sluice system fits into the overall  plant  recycle/reuse scheme.


          If the ash sluicing operation  is the final water con-
sumer in a minimum discharge cascaded plant water system, then
all of the pond overflow must be recycled.  This mode represents
completely closed-loop operation.   If some of  the pond overflow
is required as makeup to another water  consumer such as a scrub-
bing system, then only a portion of the  pond  overflow is recycled.
This mode represents a partially closed-loop  operation.


          To determine the feasibility  of operating a closed-loop
or partially closed-loop ash sluicing system,  both potential   _
scale formation and the level of treatment required should be in-
vestigated.  The plant studies  portion  of this project showed that
a computer process simulation package is a useful tool to study ash
sluicing operations.  A computer model  can be  used to evaluate
                               -117-

-------
scale potential and treatment alternatives based on the charac-
terizing data collected in the first phase.  It should be noted
here that the evaluation of alternatives will depend heavily on
the ash reactivity data collected and the amount of C02 transfer
which is assumed.   The following sections describe the types of
calculations required to evaluate recirculating ash sluicing sys-
tems and a methodology for using a computer model to perform the
evaluations.
3.2.1     Evaluation Criteria

          The recirculating ash sluice system evaluation should
include mass and energy balances to calculate the flows, composi-
tions, and temperatures of all of the process streams.  The evalu-
ations should involve predictions of solid-liquid and gas-liquid
equilibria for the species occurring in the system.   Important
solid-liquid equilibria include CaCOs, Mg(OH)2,  CaSOi>'2H20, Si02,
and other silicate scales.  Gas-liquid equilibria prediction is
necessary to account for C02 transfer between the ash pond and
the atmosphere.


          Any examination of a recirculating ash sluice system
will require that certain assumptions are made.   Steady-state
operation must be assumed so that an answer can be obtained with
a reasonable level of computational effort.  Dissolution and pre-
cipitation may be assumed to be instantaneous or calculated rates
which are consistent with measured data can be used.  The evalu-
ation should include predictions of system operation under vary-
ing conditions of C02 transfer between the pond and the atmos-
phere.  Plant measurements taken in Phase I of this project
showed that some C02 transfer occurs but complete gas-liquid
equilibrium between the ash pond and the atmosphere is not
achieved.  The evaluation must assume a reasonable percent solids
in the settled sludge in order to predict the loss of water from
the system due to occlusion with the solids.


          No real system operates under steady-state conditions
for a significant period of time, but a steady-state evaluation
under worst case conditions is very usefull for design of a sys-
tem which can operate safely under all conditions.  Instantaneous
dissolution and precipitation is a worst case assumption and may
cause overdesign.  Therefore, rate data should be used when avail-
able although this is not very easily obtained for many species.
Depending on the reactivity of the ash and the composition of the
pond liquor, C02 transfer can increase or reduce the scaling po-
tential of the slurry in the sluice line.  The sludge that nor-
mally settles in ash ponds is in the range of 4070 to 607, solids
                              -118-

-------
although it is not normally homogeneous.  The calculations and
assumptions outlined above allow  a proposed recirculatlnS ash
sluicing system to^be evaluated with  respect to scale pSLnlia
and treatment requirements.                     *^*^ Fu^encia
3.2.1.1   Scale Potential

        £ The most imPortant  factor which determines the feasi-
bility^of implementing a recirculating  ash  sluicing system is the
potential for scale formation in  the  ash slurry pipeline.  If
significant scale deposits  are formed in the pipeline, plugging
and subsequent interruption of the production of power may occur.


          In the plant studies portion  of this study, the concept
of relative saturation was  used to predict  scaling potential.
When the relative saturation  of a species is below 1.0, no po-
tential for scale formation exists.   When it is greater than 1.0,
a potential for solids formation  exists but scale will not neces-
sarily form.  When the relative saturation  of a species is above
a "critical value", nucleation and scale formation is likely.
The critical value for CaSQk'2R20 is  1.3-1.4.  In the plant
studies portion of this project the critical values of CaC03 and
Mg(OH)2 were experimentally determined  to be about 2.5 and 3.4,
respectively.  A detailed definition  of relative saturation was
given in Section 2.2.1.1


          Although the critical value of CaC03 was experimentally
determined to be about 2.5, samples taken of existing once-through
ash sluicing operations showed CaC03  relative saturations as high
as 38.8 without scaling problems.  This may be due to erosion by
the ash slurry or the ash particles may be  providing precipitation
sites for CaC03.  It should be emphasized here that if the evalu-
ation of an alternative predicts  a CaC03 relative saturation
greater than 2.5, a potential for scale formation exists but
operational problems resulting from heavy scale deposits may not
occur.  Pilot studies to better define  CaC03 scaling problems in
recirculating ash sluice systems  are  recommended before full-
scale operations are implemented.


3.2.1.2   Treatment Alternatives

          In an ash sluicing  system where excess scaling potential
is identified, treatment may  be required  to reduce,^e^"^ po-
tential.  The ash sluicing  evaluation should consider  treatment
technologies and the effects  of these technologies on  the  system
operation.  Potential treatment technologies for ash  sluicing  sys-
tems include:

                               -119-

-------
          1)   A reaction tank with sufficient
              residence time to allow the major
              portion of the reactive part of
              the ash to dissolve and to allow
              for controlled precipitation of
              any supersaturated species that
              might otherwise scale in the
              slurry line.

          2)   Softening to reduce the levels of
              calcium and magnesium in the water
              being recirculated to the sluicing
              operation.

          3)   Treatment to reduce the level of
              dissolved solids in the water being
              recirculated for sluicing the ash
              (brine concentration, reverse osmosis,
              etc.) .


          The first treatment that should be investigated is
the installation of a reaction tank at the point where the ash
and the water are contacted.  A reaction tank in the system will
allow a large portion of the soluble ions in the ash to dissolve
leaving the remaining ash essentially inert.  The tank also allows
any supersaturated species to precipitate in a controlled fashion
similar to the way CaSOit-2H20 and CaS03«%H20 precipitate in a
reaction tank in lime/limestone S02 scrubbing systems.  The re-
sults of bench-scale ash studies may indicate the rate of leach-
ing from the ash.  Precipitation rates of solids can be obtained
from the literature or laboratory studies.  The residence time
of the reaction tank should be large enough to allow sufficient
ash dissolution and solids precipitation to prevent scale forma-
tion in the sluice lines.  Pilot operations are recommended for
sizing a full-scale reaction tank.


          Other possible treatments that will reduce scaling po-
tential in the slurry are softening, brine concentration, reverse
osmosis, etc.  These treatments can best be employed on the water
which is recirculated from the ash pond.  Softening, the less ex-
pensive of the two options, can reduce the calcium and magnesium
levels in the recirculating water.  Other treatments can be used
to remove all types of dissolved solids from a portion of the
recirculating water,  or from the pond overflow.


          In many cases softening can be used to effectively re-
duce the level of calcium and magnesium to levels where the
                              -120-

-------
addition of the reactive species from thp
scaling situation in the slurry line   Ca
                    zs i^^                    <
^•™   ^-FI-^-   bu B^c?;£:Lea ror Pilot or xu^-scaie implementa-
tion.  Softening should be more expensive to build and operate
than a reaction tank but significantly less expensive than other
more elaborate treatments.


          Brine concentration can be used as a treatment option
for recirculating ash sluicing systems.  A discharge from the
pond can be treated and the clean water sent to the boiler de-
mineralizer and/or some other plant water consumer.  The major
effect on the ash sluicing system from this treatment option is
that more makeup water is required to operate the system.  Since
the makeup water is lower in dissolved solids than the pond water,
the scale potential in the slurry line is reduced.


3.2.2     Model Application

          A computer model such as the one used in the plant stu-
dies is a very useful tool in performing the evaluation of a re-
circulating ash sluice system.  However, since ash dissolution
kinetics are not well understood at this time, the results of the
computer calculations are based on instantaneous dissolution.  The
evaluations of ash sluicing systems are therefore heavily depen-
dent on the ash reactivity data input.  A computer model can be
used to perform multiple calculations which demonstrate the ef-
fects of variations in the input data.  The two factors which
are the most difficult to quantify accurately are ash reactivity
and C02 transfer in the pond.  For each alternative evaluated,
multiple simulations should be performed to determine the effects
of variations in both ash reactivity and the amount of C02 trans-
fer.


          Bottom ash is usually much less reactive than fly ash.
For this reason, most recirculating bottom ash systems will not
be prone to scale formation.  If the bottom ash system is going
to be operated completely independent of the fly ash system  a
few simulations will usually be sufficient to show that the bottom
ash can be sluiced in a recirculating system without scale.  If
the bottom ash is unusually reactive, more extensive study o±
recirculating systems may be required, in which case steps similar
to those proposed for recirculating fly ash systems would be
appropriate.


                              -121-

-------
          The key stream in an ash sluicing system is the slurry
which is transported to the ash pond.  If scaling problems are
encountered, they most likely will occur in the slurry where a
significant increase in total dissolved solids is usually exper-
ienced because of the dissolution of the reactive portion of the
ash.  When simulating ash sluicing systems, it is useful to assume
that all of the reactive solids identified in the leaching studies
enter the liquid phase instantaneously.  This assumption allows
the worst possible scaling potential to be identified.  If the
relative saturations of all of the potential scaling solids are
less than the critical values, this indicates that the system
can be safely operated under the simulated conditions.


          If scaling is indicated, treatment of the slurry or re-
circulating water may be required.  The results of computer simu-
lations may be used to size the necessary treatment equipment
for scale prevention.  However, pilot studies are recommended to
better define treatment requirements before full-scale implemen-
tation.


3.3       Phase III:  Ash Sluicing Pilot Studies

          For ash sluicing systems with essentially unreactive
ash, pilot studies may not be justified.  For recirculating sys-
tems sluicing only bottom ash, the results of evaluations may in-
dicate that the scale potential of the slurry is small, in most
cases.  These recirculating systems can be designed for full scale
operation as only a pond return line and additional pumps will be
required.


          For ash sluicing systems where scaling in the slurry
line is a concern because of a combination of reactive ash and
poor water quality, pilot studies would be very important, and
could produce significant savings in the full-scale design.  The
pilot system should include the treatment options identified in
the optimum system which resulted from the evaluations.  The
pilot system should be designed at a reduced scale which is di-
mensionally consistent with the expected full-scale design.


          Figure 3-3 is a diagram of a general pilot-scale facil-
ity.  The system employs a reaction tank, a softener, and a blow-
down capability necessary to implement the three major forms of
treatment discussed previously.  The clarifier serves the same
purpose as the ash pond in a full-scale system.  The clarifier
separates the slurry into a sludge and a solids-free liquid stream
which can be recirculated via the return line to slurry additional
                              -122-

-------
                     FLOW
                    METER
     MAKEUP WATER
         ASH
OJ
                    REACTION  TANK
          RETURN LINE
                           CHEMICALS
                                                              FLOW
                                                              .METER
SOFTENER
                                           SLUDGE -«—CT-K

                                                       2±3
                                                                                   SLURRY LINE
                          XFLOW
                          METER
                                                                                SLUDGE
                    SLOWDOWN
                      Figure 3-3.  General pilot scale ash sluicing facility.

-------
ash.  A clarifier has a much smaller residence time than an ash
pond, and thus allows the total system to attain steady-state in
a reasonably short time.  This allows a larger number of experi-
ments to be performed with the clarifier as opposed to a system
with a large settling pond.  If a clarifier cannot achieve the
solids concentration expected in a pond, a vacuum filter may be
added to the system.


          A pilot system such as this can best be used on-site at
a power plant.  The plant can serve as a source of ash and the
ash pond can be used as a receptacle for the sludge from the
clarifier and the softener.  The system should be built near the
ash pond where there generally is enough space for such a facili-
ty.  This will allow for short lines from the facility to the
pond and should keep the generally crowded area near the plant
itself free.  If this kind of location is chosen a dry method
must be used to transport the ash from the source,  near the plant,
to the pilot facility.


          Once the pilot facility is constructed on-site, a series
of experiments should be performed with the ash from the plant un-
der various operating conditions.  Each experiment should be per-
formed for a long enough period to attain steady-state operation.
The amount of time required will vary with the design and flow
rates chosen.  Important variables which should be studied
include the makeup water quality, the residence time in the reac-
tion tank, the degree of recycle, the weight percent solids in
the slurry, and the size of the softened stream.   Data may also
be taken to better define C02 transfer in the system.


          The makeup water should be similar to the water that is
expected to be used in full-scale operation.  If a number of make-
up water sources are being considered, experiments with the full
range of makeup water qualities should be performed.  In general,
the water used for these experiments should have a TDS level and
pH near the levels expected in the makeup water for full-scale
operation.


          The residence time in the reaction tank is a very impor-
tant parameter to study in the pilot studies.  The residence time
of a reaction tank can be changed by varying the level of the
slurry in the tank, and, thereby, the effective volume, or by
varying the flow through the tank.  Thus, with one piece of equip-
ment, a wide range of residence times can be studied.  The tank
must be large enough to attain the maximum desired residence time
at the minimum flow rate that will still produce turbulent flow
                              -124-

-------
in the slurry line^  All  smaller  residence  times can be attained
by increasing the flow rate or  reducing  the level in the reaction
tank.


          For a given ash flow  rate  the  makeup water requirement
is determined by the water lost from the system via evaporation
occlusion with solids in  the pond, and blowdown from the system!
The sum of the makeup water and the  recycled water is determined
by the weight percent solids desired in  the slurry.  Experiments
should be performed which vary  the percent  solids in the slurry
•and the blowdown from the system.  These both directly affect the
percent recycle that exists in  the system,  and, therefore, the
quality of the water entering the reaction  tank.  The weight per-
cent solids in an ash sluicing  system can usually vary from 0.5%
to 5% solids for bottom ash and 1% to 15% solids for fly ash.
The degree of recycle can be varied  from zero, in a once-through
system, to greater than 80% for systems  that operate with mini-
mum discharge.


          The softening equipment should be designed to operate
on a slipstream from the  recycle  line.   This is the most efficient
location for this operation because  the  highest concentration of
calcium and magnesium will exist  in  this stream.  For the pilot
design the softener should be large  enough  to handle all of the
largest recycle streams that the  system  can be expected to oper-
ate with.  In this way the level  of  softening can be varied from
0 to 100% of the recycle  stream for  all  conditions studied.


          Preliminary experiments should be performed under con-
ditions similar to those  simulated with  the computer model.  Sam-
ples of the ash slurry, clarifier underflow, clarifier overflow,
softener effluent and makeup water should be taken periodically
until steady-state is achieved.  Analyses of these samples for
suspended solids, calcium, magnesium, sodium, chloride, nitrate,
carbonate, sulfate, silica, pH, and  TDS  will allow the system to
be evaluated with respect to scale potential.  Material balances
may be performed around the reaction tank to calculate ash reac-
tivity.  These analyses will also allow  evaluation of the soft-
ening equipment performance.  These  preliminary results can be
compared to the simulated results and used  to design further ex-
periments under conditions closer to optimum operation.  These
preliminary experiments should  only  be performed long enough to
attain steady-state operation and obtain samples of the key
streams.
                              -125-

-------
          Experiments under conditions that have been identified
as desirable from the preliminary experiments should be performed
for longer periods of time to insure that scaling will not be a
problem.   These experiments can be used to identify optimum opera-
ting conditions.  They will identify the minimum reaction tank
volume, the minimum softening requirement and the maximum percent
recycle that can be attained with the ash and makeup water used
at the plant.  The total task should take about six months to
perform with 3-6 people working on the system.  Costs associated
with the pilot studies will depend on the number of experiments
to be performed as identified from the evaluation phase and pre-
liminary sluicing operations.


          The results of these studies will establish a strong
base upon which to design a full-scale system that will operate
with scale problems, minimize the discharge from the ash sluic-
ing system and be as economical as possible to build and operate.
The next subsection is devoted to a discussion of full-scale im-
plementation of a recirculating ash sluicing system at an existing
facility.


3.4       Phase IV:  Full-scale Implementation

          The final result of the pilot studies should be to
identify a recirculating ash sluicing system design which can
be operated at the power plant.  The system should be able to
operate for long periods of time without scale formation in the
slurry lines reducing performance.  The design should include
any monitoring equipment which may be necessary to minimize
equipment failure and operating costs.  The system should be
designed to facilitate regular inspections at times when the
power plant is undergoing routine maintenance.


          This section discusses the kinds of equipment that
may be used to build a recirculating ash sluicing system and
discusses an implementation procedure.


3.4.1     Equipment

          For some recirculating ash sluicing systems which
sluice very reactive ash, many pieces of equipment may be
necessary to safely operate the system.  For others where the
ash is essentially unreactive, only pumps and piping may be
required.   The former case will most likely exist in fly ash
systems while the latter may exist for some bottom ash systems.
                             -126-

-------
          The least expensive piece of equipment that may be
used to reduce the scaling potential of ash slurries is I reac-
tion tank.  The reaction  tank should be large enough to allow
the major portion of the  leachable ions in the ash to enter the
solution^and allow controlled precipitation of any supersatura-
ted species that may be formed due to the dissolution.  The
slurry can then be safely sluiced to the ash pond without danger
of scale formation.  The  pilot studies should determine the mini-
mum volume that will be required to obtain the necessary resi-
dence time.


          In^some systems the ash may be so reactive that con-
trolled _ precipitation  in  the reaction tank may not be sufficient
to eliminate scale and softening of the pond recycle may be
necessary.  The softener  will reduce the levels of calcium and
magnesium in the recycle  water and reduce the relative satura-
tion of CaSCU'2H20, CaCO 3 and Mg(OH) 2.  The size of the softened
stream should be determined in the pilot studies.  The residence
time in the flocculator should be at least 40-60 minutes and
sedimentation requires 2-4 hours.  The sludge produced by soft-
ening can be discharged into the ash pond.


          A third treatment option that may be identified as a
result of the pilot studies could be brine concentration.  The
level of dissolved solids in the slurry can be controlled with
a blowdown stream.  This  stream can be treated with a brine con-
centrator which can supply the treated stream to other plant
water consumers such as the boiler or cooling towers.  The size
of the blowdown should be determined from pilot studies which
will, in turn, determine  the size of the brine concentrator re-
quired.


3.4.2     Implementation

          Once the system is built, full scale implementation
should not present a problem.  The major portion of the sluice
water in a recirculating  system is usually the pond recycle.
The ash pond at most power plants has a very large residence
time, on the order of  one to four weeks.  Thus, the pond water
and, thus, the water quality of the sluice water will change
very slowly toward its new "steady-state" value.  Thus, the
dampening effect of the pond should keep the system from exper-
iencing drastic changes  in operation.  The system should be mon-
itored very closely through the  first month of operation to in-
sure no drastic upsets occur which will  cause  scale formation.
                               -127-

-------
4.0       SOa/PARTICULATE SCRUBBING

          Previous plant studies under this contract showed that
combined S02/particulate scrubbing systems may be operated in a
closed-loop mode to minimize makeup water requirements and elim-
inate aqueous discharges.  These studies also identified the
opportunity for reusing water discharged from other major water
systems as makeup water to a combined S02/particulate scrubbing
system.  To minimize plant water requirements and discharges,
the overall plant water system may be cascaded with the scrubb-
ing system being the final step.  Scrubbing systems typically do
not require the good water quality that a cooling tower does.
Cooling tower blowdown is therefore one candidate for use as
scrubber makeup in a cascaded plant water system.  For plants
with wet ash handling, ash pond overflow may be another candidate.


          The plant studies identified two potential problem
areas associated with using cooling tower blowdown or ash pond
overflow.  First, if the scrubber makeup water is near satura-
tion with respect to gypsum, scaling problems may be encountered
when this water is used in the demisters.  The amount of SOz
sorbed in the demister and the amount of CaCOs solids entrained
in the scrubber exit gas as well as the demister wash rate
determine if scaling is likely.


          The second problem area identified involves the dis-
solved solids concentration of the makeup water.  The evapora-
tion from the scrubbing system causes the makeup water to be
concentrated.  If the TDS level of the makeup is high, excessive
TDS levels (>20 wt.  70) in the scrubbing liquor may result.
Excessive TDS levels can increase the energy requirements for
pumping and decrease the mass transfer in the scrubber.   If the
alternate makeup source is high in chlorides, an excessive
chloride concentration may result in the scrubbing liquor which
may cause ocrrosion problems.


          These recycle/reuse opportunities were identified for
"throwaway" scrubbing systems where a waste sludge is produced
and disposed of in a pond system.  Some scrubbing systems are
used for S02  removal only.  These systems may be "throwaway" or
"regenerative".   Regenerative systems recover the sorbed S02 as
S02(2,) or H2S0lf(aq)  and recycle the alkali as opposed to dis-
psoing of the sorbed sulfur as calcium sulfite or calcium sulfate
sludge.  There is opportunity for recycle/reuse in regnerative
systems as well as throwaway systems.  In most cases regenera-
tive systems  require prescrubbers to prevent contaminant buildup
in the system,  and recycle/reuse options may be used in these
prescrubber sections.
                              -123-

-------
          In regenerative  S02  scrubbing  systems very little water
is required as makeup  to the main  scrubbing  loop.  This is becaule
no sludge is produced  in which water  is  occluded and little or no
water is evaporated in the scrubber.   The  gas entering the main
scrubber in most regenerative  systems  is near saturation with
respect to water since a prescrubber  or  quench is normally used
to remove particulates and chlorides  from  the flue gas upstream
of_the main_scrubbers.  Since  very little  water is lost in the
main scrubbing loop, the concentrations  of particulates and
chlorides will build to intolerable levels if they are not
removed by a prescrubber.


          The procedure to test  and implement the use of cas-
caded water in these prescrubber systems is  the same as for
throwaway systems  and  is discussed in  this section.  Emphasis
will be placed on  throwaway processes, however, since these
types of systems were  studied  in detail  in the previous studies
conducted under this contract.
          The  four phases  to  implement  a water recycle/reuse
shceme for  the major water systems  of a coal-fired power plant
were discussed in Section  1.0.   These are:

          Phase  I   -   System Characterization

          Phase  II  -   Alternative  Evaluation

          Phase  III -   Pilot-scale  Studies

          Phase  IV  -   Full-scale Operation


          The  first phase  involves  gathering design data to
characterize the scrubbing system.   It  also includes a sampling
program to  supplement  the  design data.  Since scrubbing systems
are subject to operating variations from design,  the sampling
will help identify the effects of these variations.  Sampling is
also recommended for determining demister operating variables.


          The  second phase involves evaluating the feasibility
of various  alternatives.   The design and operating data^collected
in the first phase can be  used as the basis for performing pro-
cess calculations to evaluate the feasibility of  changing from
open-loop to closed-loop or of changing makeup sources   These
evaluations involve determining demister scale potential and
scrubbing liquor composition.   Calculations should accurately
predict TDS levels in  the  system but pilot  studies in addition
to calculations  are recommended to  evaluate demister operation.
                               -129-

-------
          The third phase of implementation is to conduct pilot
studies to evaluate alternative feasibility.   Pilot studies will
allow an evaluation of scale potential in the demister.  Since
the factors which affect demister scale potential (S02 removal,
carryover, wash rate)  are not easily measured, pilot studies
should be conducted to assess the affects of process modifications
The fourth phase is to use the pilot study results to design and
start-up the full-scale modifications.


          The plant studies conducted in this program involved
the first two phases of an implementation plan.  Design and oper-
ating data for two combined S02/particulate scrubbing systems was
collected and used to evaluate operation under existing and alter-
native conditions.  Implementing the alternatives studied at the
existing plants would require additional studies to evaluate de-
mister operation and to better define variations in operating
parameters.


          The following discussions are written to apply to a
general throwaway scrubbing system.  All phases of an implementa-
tion plan are presented although the characterization and evalua-
tion phases were addressed in the plant studies.  Since^virtually
all scrubbing systems are unique with respect to operating con-
ditions, a detailed characterization and evaluation phase should
be conducted before implementing an alternative at a particular
site.  The purpose of the plant studies was to identify and evalu-
ate the types of recycle/reuse alternatives achievable at coal-
fired power plants.  The purpose of this document is to outline
a procedure for implementing the types of alternatives identified
in the plant studies.


4.1       Phase I:  System Characterization

          The first phase in implementing a recycle/reuse scheme
for a particular scrubbing system is to characterize the system
operation.  This includes collecting all design and operating
data available and supplementing this data by collecting samples
of the important process streams.  This section first  discusses
the design and operating data required to characterize the system
and then presents a sampling program to obtain additional data.


4.1.1     Identification of Process Variables

          The design and operating data necessary to characterize
a scrubbing system can be divided into three  areas.  These areas
represent the three major operations  in a typical S02/particulate
scrubbing operation:


                               -130-

-------
          1)  gas scrubbing  and  demisting,

          2)  solids precipitation,  and

          3)  solids concentration.


Figure 4-1 is a simplified block diagram which shows how these
three operations are combined.   The  gas is contacted with a
recalculating slurry in  the  scrubber section where the particu-
lates^and S02 are removed.   Demisters are generally used to
minimize entrainment of  the  scrubbing liquor in the clean gas.
The absorbed S02 is precipitated as  solid calcium sulfite and/or
calcium sulfate in the reaction  tank.  Concentration of the pre-
cipitated solids and the sorbed  ash  is achieved in the solid/
liquid separation section.   This concentration is made by a
clarifier, filter, pond  or a combination of the three.  This
section discusses the design data  for each of these portions of
an S02/particulate scrubbing systems that is required to char-
acterize the system.  All the  data requirements discussed in
this section should be easily  obtained from the design specifi-
cations submitted to the utility by  the scrubber manufacturer.


4.1.1.1   Scrubber and Demister  Data

          Table 4-1 is a data  sheet  listing the most important
design data needed to characterize the scrubber and demister
operations of a particular S02/ash scrubbing system.  The first
items listed in Table 4-1 for  the  scrubbers are the flue gas
flow, temperature, and composition.  The flue gas characteristics
affect the amount of S02  and ash removed and the amount of water
evaporated in the scrubbers.


          The makeup water required  for a scrubbing system is
determined by the evaporation  rate and the water lost by occlu-
sion with the solid waste (ash,  CaS03'%H20, CaSCU•2H20).  The
amount of solid waste generated, and, therefore, the occluded
water lost, decreases with decreases in the amount of S02 and
ash removed in the scrubbers.  As  the flue gas flow, S02 content,
or ash content is decreased  and  the  sludge percent solids
increased, the makeup water  required by the scrubbing system
will decrease.


          The design flue gas  characteristics for various plant
loads are therefore very important in determing the applica-
bility of a particular water recycle/reuse option involving the
scrubbing system.  The makeup  water  requirement of the  scrubbing
                              -131-

-------
MAKEUP
 WATER
   FLUE
   GAS
 ALKALI
                 STACK
                  GAS
DEMISTER
             SCRUBBER
              REACTION
                TANK
                                   SOLID/JJQUID
                                   SEPARATION
                                      WASTE
    Figure 4-1.  Typical  scrubbing  system flow scheme
                             -132-

-------
           TABLE  4-1.   SCRUBBER AND DEMISTER DATA

         Scrubbers
         A.  Design Flue  Gas
             Flow (wet  gas  excluding ash)
             Temperature
             Composition
               N2
               02
               CO 2
               S02
               HC1
               H20
             Ash loading into scrubbers
         B.  Fly Ash Removal Stage
             Liquid-to-gas ratio

             Design Removal
         C.  SO2 Removal Stage
             Scrubber Type (spray tower,
               marble bed, etc.)
             Liquid-to-gas ratio

             Design Removal
                                               Ib/hr
                                               °F

                                               vol %
                                              gr/scfd
                                              gal/1000 acf
                                              (outlet)
                                              7
                                              /o
                                              gal/1000 acf
                                              (outlet)
                                              7
                                              /o
II
Demisters
A.  Once-through  or recirculating loop
       if recirculating:
        makeup rate
        blowdown  rate
        tank volume
B.  Wash Rate
C.  Wash water source  (fresh makeup,
    clarifier overflow, etc.)
D.  Rate of scrubbing  liquor entrained
    with the gas  entering the demister
                                                       GPM
                                                       GPM
                                                       gal
                                                       GPM
                                                       GPM
                             -133-

-------
system will determine the mode of operation of the water system
supplying the scrubber makeup in a cascaded system.  This is
especially important in cases where zero discharge is desirable.
For example, if cooling tower blowdown is used exclusively as
scrubber makeup,  the flue gas characteristics determine the
tower blowdown rate and thus the cycles of concentration in the
towers.


          The design parameters for the fly ash and S02 removal
stages affect the amounts of ash and precipitated solids dis-
posed of and therefore the water lost by solids occlusion.
Although evaporation accounts for most of the water lost in
closed-loop scrubbing systems, the effects of ash and SOz
removal efficiency should not be ignored.  A more detailed dis-
cussion of the parameters affecting scrubber makeup water
requirements may be found in the section concerning the plant
studies conducted in this project.


          The design parameters in Table 4-1 for characterizing
the demister section of a scrubbing system include the system
type (once-through or recirculating), the wash rate, the wash
water quality, and the carryover rate.  Makeup water is normally
used wholly or in part as demister wash or demister wash makeup.
The feasibility of using cascaded water for scrubber makeup
water in the demister will depend on the demister characteristics
as well as the cascaded water quality.  The amount of SC>2 sorbed
and the degree of CaCOs dissolution that occurs in the demister
can be large enough to cause significant scaling, especially if
the wash water is already high in dissolved calcium and sulfate.
CaC03 solids are introduced into the demister loop by scrubbing
liquor entrainment in the gas entering the demister.


          Both once-through and recirculating demister wash
operations may be used depending on the wash water requirements
and the total scrubbing system makeup water requirement.  If
more water is required for demister wash than for the system
makeup, a recirculating system must be used.  A recirculating
system uses a catch tray to collect most of the demister wash
before it falls through the scrubber, whereas a once-through
demister wash system allows the demister wash water to fall
through the scrubber.  A recirculating system will be more
susceptible to scaling problems and therefore will require pilot
studies to determine if a cascaded water may be used as demister
wash.  A possible solution is to use a combination of cascaded
water and fresh makeup water as demister wash.
                             -134-

-------
4.1.1.2   Reaction Tank  Data

          Table>4-2 presents  the  design  data which  characterize
the solids precipitation section  of  a  lime  or  limestone based
SC5Ur o2S o£SHem- ,A reaction tank is  used  to  control CaS03«%H20
and CaSC\.2H20 scale potential by allowing  these  solids to preci-
pitate on recirculated seed crystals.  The  makeup alkali and some
makeup water are added to the reaction tank.   The alkali dissolves
and combines with the sorbed  sulfur  to form calcium sulfite and/or
calcium sulfate solids.


          The additive type,  composition, and  addition rate will
affect the liquid phase  concentrations of calcium and magnesium
in the system and therefore affect the calcium sulfite and cal-
cium sulfate relative saturations encountered.  The design para-
meters for the additive  are therefore  important in  characterizing
the solids precipitation section  of  a  scrubbing system.


          The type of water used  to  slurry  the additive as well
as the additive type will determine  the  scale  potential in the
alkali addition system.   The  use  of  waters  saturated or near
saturation with respect  to CaC03  or  03304-21120 may result in
scale formation of one or both of these  two species when some of
the additive dissolves.   This is  especially important for lime
systems where additive dissolution is  fast.


          The reaction tank volume and slurry  solids content are
included in Table 4-2 and affect  the precipitation rate.  In-
creases in slurry solids content  and the reaction tank volume for
a given precipitation rate lowers the  relative saturation of the
precipitating species.   The required precipitation rates are de-
termined by the amount of S02 sorbed in  the scrubber and the oxi-
dation occurring in the  system.   Higher  oxidation causes the re-
quired precipitation rate of  CaSO^HjO  to  increase but lowers
the required CaS03'%H20  precipitation  rate.


          The final item in Table 4-2  is the rate and composition
of makeup water added to the  reaction  tank.  The  quality of the
makeup water will affect the  dissolved solids  content of the li-
quor in the scrubbing system  but  was shown  to  have  little effect
on scale potential in the system  in  the  plant  studies performed
in this project.  The chloride content of the  makeup water is im-
portant since this affects the chloride  concentrations in the sys-
tem and may cause corrosion problems if  the level is too high.
The chloride content of  a cascaded water used  as  scrubber makeup
is therefore an important parameter  to consider in  a water recycle/
reuse scheme.


                                -135-

-------
                TABLE 4-2.   REACTION TANK DATA

 I.        Alkali Additive
          A.  Type (lime, limestone)                 	
          B.  Composition
              Lime
                CaO or Ca(OH)2                       	 wt. %
                MgO or Mg(OH)2                       	
                Inerts                               	
              Limestone
                CaCO 3                                	 wt. %
                MgCO 3                                	
                Inerts                               	
          C.  Addition rate                          	 Ib/hr
          D.  Water Source  for slurrying additive
              (makeup, clarifier overflow,  etc.)     	

II.        Tank
          A.  Volume                                 	 gal
          B.  Slurry suspended solids content        	 wt. %
          C.  Makeup water  added to tank
              Flow                                   	 GPM
              Composition (design basis)
                Calcium                              	 mg/£
                Magnesium
                Sodium
                Chloride                             	
                Sulfate (as SOl)                     	
                Nitrate (as
                              -136-

-------
4.1.1.3   Solids Concentration  Data

         ^Table 4-3 presents  the  design  information necessary to
characterize the solids  concentration  section of a scrubbing
system.  This data is  divided into two parts.  The first part
concerns the clarifier and  filter which  may  or may not be pre-
sent in every system.  In some  cases,  the blowdown from the
scrubbing loop is pumped directly to the pond or diluted with
pond liquor and pumped to the pond.  In  these cases, additional
data concerning mixing tank volume and solids concentrations of
all streams is required  to  characterize  the  system.  The second
portion of the data in Table  4-3  is for  the  pond system where
the waste solids (ash, CaS03'%H20, CaS(V2H20) are deposited.


          The solids concentration section of a scrubbing sys-
tem determines whether the  overall system is closed or open-loop.
In a closed-loop system  all water not  occluded with the settled
sludge or evaporated in  the pond  is recycled to the scrubbing
system.  The design flow rates  and suspended solids concentra-
tions of all of the streams in  the solids concentration section
of a scrubbing system  are necessary to fully characterize the
system.  The effects on  the operation  of this section of con-
verting from open-loop to closed-loop  system design should be
evaluated for this type  of  recycle/reuse opportunity.  Changes
in oxidation resulting from process modifications may have a
significant effect on  solid/liquid separation equipment.  Calcium
sulfite crystals are platelets  and do  not tend to settle as
easily as calcium sulfate solids.  Careful consideration of
these effects should be  made  before a  recycle/reuse alternative
is implemented.


4.1.2     Sampling Plan

          Although design data  provides  a great deal of informa-
tion concerning system characterization, a sampling program is
necessary to supplement  this  data.  Actual operation will
deviate from design condidtions with changes in coal composi-
tion, water composition, additive composition, and load.  Since
these four parameters  are continuously fluctuating, a sampling
program to characterize  system  operation under actual conditions
is necessary.  Also, demister operation  may  be more fully char-
acterized with a sampling program.  Material balances around  the
demister loop may be performed  using_ sample  data to determine
S02 removal and carryover in  the  demister.


          Two general  types of  measurements  should be made in a
characterization sampling program.  These are  analytical and
                              -137-

-------
                                                     'o
             TABLE 4-3.   SOLIDS CONCENTRATION DATA

 I.        Clarifier/Filter (if applicable)
          A.   Feed rate                             	 GPM
          B.   Bottoms suspended solids              	 wt. %
          C.   Overflow suspended solids             	 ppm

II.        Pond System
          A.   Number of Ponds                       	
          B.   Surface area of each pond
                                                          acres
C.  Settled sludge (bottom of pond)
      solids concentration                	 wt. 70
D.  Pond overflow rate                    	 GPM
E.  Pond overflow suspended solids        	 ppm
F.  Pond overflow recycled to
    scrubbing system                      	 GPM
G.  Pond overflow discharged                    GPM
                    -138-

-------
process measurements.  Analytical measurements  can be used to
characterize the chemistry  of  the scrubbing  system by determining
solids precipitation  rates,  S02  removal  rate, S02 oxidation  and
scaling potentials.   Process measurements  can be used to charac-
terize the system water  balance  for  different operating condi-
tions by determining  flow rates  for  key  process streams and
levels for major process tanks.


4.1.2.1   Analytical  Measurements

          Chemical  analysis  and  sampling schedules can be
divided into two categories, each serving  a  different function.
These categories are  system characterization measurements and
line-out measurements.   System characterization samples should
be taken after  sufficient time is allowed  for the system to
reach steady-state  operation.  Steady-state  operation is deter-
mined from the  line-out  measurements.  When  the system has
reached steady-state,  the stream compositions will not change
appreciably between sample  periods.


          The characterization sample  points and analyses for a
typical scrubbing system with  a  recirculating demister wash sys-
tem as shown in Figure 4-2  are presented in  Table 4-4.  The
analyses to be  performed on these samples  are divided into three
categories:  gas analyses,  liquid analyses,  and solids analyses.
Gas analyses on the flue gas and stack gas for  S02 , C02 , 02 ,
H20, and HC1 will determine the  S02  removal  rate, HCl removal
rate, and the evaporation rate in the  scrubbers.


          Liquid analyses are  shown  in Table 4-4 for the remain-
ing sample points.  It should  be noted that  some deviation from
the sampling plan shown  in  Table 4-4 should  be  expected for a
particular scrubbing  system.   If a  system  uses  a once-through
demister wash with  makeup water, sample  points  10  and 11 may be
eliminated.  Also,  if a  clarifier is not used,  sample points 7
and 8 may be eliminated.


          Not all of  the liquid  analyses listed in Table 4-4
necessarily need to be made for  each sampling period.  Some vari-
ables, such as  the  magnesium,  sodium,  nitrate,  and chloride
liquor concentrations, should  not change significantly under
normal operation.   Changes  in  makeup water quality,  alkali addi-
tive, coal composition,  or  load  are  the  major factors influencing
the magnesium,  sodium, nitrate,  and chloride levels.  Whenever
step changes in these variables  are  made,  complete liquor  analy-
ses should be made  until the system is at  steady-state  (line-out
measurements).
                              -139-

-------
-p-
o
i
                                                            •-STACK GAS



DEMISTER
WASH
TANK


- .^y

                                               SCRUBBER FEED(S)
                                                                                       MAKEUP

                                                                                       WATER
                                                                                      TO POND. CLARIFIER,


                                                                                      OR REACTION TANK
                                      SCRUBBER EFFLUENT(S)
                                                                                     CLARIFIER


                                                                                    UNDERFLOW
                                                                                       	"-DISCHARGE
                                                                               SLUDGE
                   Figure  4-2.   Sample points  for  scrubbing  system characterization.

-------
            TABLE 4-4,    SCRUBBING SYSTEM CHARACTERIZATION  MEASUREMENTS
                             SAMPLE  POINTS  AND  ANALYSES


Flue Gas
Stack Gas
Scrubber Feed(s)
Scrubber Effluent(s)
Alkali Slurry
Makeup Water
CJarifler Underflow"
Clarifler Overflow"1
Pond Return
Demister Wash 5
Catch Tray Overflow6
Sample Gas Analyses
S02 €02 O2 |I20 MCI
No .
1 X X X X X
2 X XX
3
4
5
6
7
8
9
10
11
Liquid Analyses
Ca


X
X
X
X
X
X
X
X
X
MB'


X
[X
X
X
[X
[X
IX
X
[X
Na1


X
X
X
X
X
X
X
X
X
Cl1


X
XI2
X
X
X)2
X]2
X]2
X
X]3
C02


X
X
X
X
X
X
X
X
X
N03'


X
[X]2
X
X
IX]2
[X]2
[X]2
X
[X]3
S02


X
X
X
X
X
X
X
X
X
Total
s


X
X
X
'X
X
X
X
X
X
Solids Analyses
%
Solids Ca Mg 2 2 '


X X X X X X
X [XXX X XJ2
X XXX

X [X X X X X]2
X
X
X X X X X X
X X X X X X
 These analyses should be performed periodically  sln^e they should not  change rapidly.

2These analyses can be deleted if differences between these points and  sample point 3 are negligible.

3These analyses can be deleted in differences between this point and sample point 10 are negligible.

'"'Ihese sample points may be eliminated for systems which dn not use clarlfiers or thickeners.

 srhese sample points may be eliminated for systems which use once-through demlster wash with makeup water.

-------
          Solids analyses are shown in Table 4-4 for the scrubber
feed, scrubber effluent,  alkali slurry, clarifier overflow,
demister wash, and catch tray overflow streams.  In addition,
total suspended solids analyses are shown for the clarifier
overflow and pond return streams.


          To characterize a scrubbing system, a sampling program
should last from 1-2 weeks for a given set of operating condi-
tions to allow complete system line-out.   The actual time
required will depend on the stability of the operating variables
and the process design.  Larger reaction tank residence times
will require a longer period of time for the system to reach
steady-state.  Line-out samples to determine if the system is
at steady-state should be taken twice daily.


          The cost for conducting such a sampling program will
vary directly with the length of the sampling program and the
number of sampling points and analyses to be made.   Meaningful
cost estimates for a characterization sampling program can only
be made for a specific system after a detailed program is out-
lined.  The addition of sampling ports and/or laboratory capa-
bilities at a specific site should be considered as well as the
manpower requirements and the number of samples and analyses to
be made.
4.1.2.2   Process Measurements

          In order to properly characterize the performance of
a scrubbing system, certain process measurements must be gathered
in conjunction with the chemical analyses discussed in the pre-
vious section.  These process measurements include such vari-
ables as liquor flow rate, flue gas flow rates, various tank
levels, and important stream pH's and temperatures.  Table 4-5
indicates which process measurements are required for system
characterization purposes.


          Several of the desired flow rates may be monitored on
a regular basis by plant personnel.  These may include scrubber
feed, clarifier feed, and alkali additive.  Before starting a
sampling program all existing flow meters should be calibrated
by using variations in tank levels with time.  Periodic calibra-
tions of flow meters should be made throughout a sampling
program.
                             -142-

-------
TABLE 4-5.  PROCESS DATA REQUIREMENTS
Stream Name
Gas Streams
Flue Gas
Stack Gas
Liquid Streams
Scrubber Feed
Scrubber Bottoms
Clarifier Feed
Clarifier Overflow
Reaction Tank
Effluent
Pond Return
Alkali Slurry
Makeup Water
Demist er Wash
Catch Tray Overflow
Demister Wash Loop
Slowdown
Tank Level Data
Vessel
Additive Tank
Reaction Tank
Demister Loop Tank
Surge Tanks
Flow Rate
X
-
X
-
X
X
X
X
X
X
X
-
X
Level
X
X
X
X
Pressure pH
X
X
X
X
X
X
X
X
X
X
X
X
X




Temperature
X
X
X
X
X
X
X
X
X
X
X
X
X




                  -143-

-------
          Important streams whose flows are not already moni-
tored should be included in a comprehensive characterization pro-
gram.  The flue gas rate can be measured with a pitot tube
placed in the scrubber gas feed duct.   Necessary correlations of
flow rate versus pitot reading can be generated by performing
periodic pitot traverses of the gas duct.  This measurement
should be made as close to the scrubber inlet as possible to
minimize errors due to leakage.


          Slurry stream flow rates may be monitored by magnetic
flow meters.  Additional flow meters should be obtained for
slurry streams whose flows are not already monitored in the
scrubbing loop (scrubber feed, clarifier feed, reaction tank
effluent, demister wash loop blowdown) .  The flow meters can be
used for the pilot studies discussed in Section 4.3 as well as
for the system characterization phase.


          Water makeup streams (to pump seals or demister wash
tank) should be equipped with rotameters.  The rotameters (both
existing and new) should be calibrated before starting the
characterization sampling.  Intermittent flows such as periodic
washing of the catch tray may be characterized by noting the
frequency of washing.


          The flows around the clarifer and pond are not as
critical as the streams previously discussed for purposes of
characterizing scrubbing system operation.  Solids balances and
pump characteristics can be used to estimate clarifier under-
flow, clarifier overflow, and pond return rates.  The combined
flow of these three streams is normally small in comparison to
other streams entering the reaction tank.  For this reason, any
errors resulting from imprecise flow measurement of these
streams will have a negligible effect on material balances per-
formed around the reaction tank.
          Temperature and pH measurements are listed for all of
the liquid streams in Table 4-5.   This should be done routinely
as a part of the sampling procedure.  Stream temperatures within
the scrubbing loop should remain relatively constant.  Some heat
losses occur through pipe and open vessels, however.  Initially
all stream temperatures should be recorded.  As the sampling
program proceeds, it may be possible to omit some temperature
measurements if only small differences exist.
                             -144-

-------
      _    Tank levels should be recorded routinely.   In  combina-
tion with water makeup rates, the slurry levels of the larser
vessels will determine the magnitude of the system's  water re-
quirements.  Increases or decreases in tank slurry levels will
affect the calculation of water evaporation and addition rates
through material balances.


4.2       Phase II:  Alternative Evaluation

          The second phase of a scrubbing implementation plan
is to formulate and evaluate various modes of operation.  For-
mulation of alternative operating conditions will depend on how
the scrubbing system fits into the overall plant recycle/reuse
scheme. ^The plant studies conducted in this program  showed that
alternatives include converting from open-loop to closed-loop
operation and changing makeup water source.   Alternative makeup
water sources may include cooling tower blowdown or ash  pond
overflow.
          To determine the feasibility of implementing a recycle/
reuse alternative, both potential scale formation in the demister
and the TDS level  (including chlorides) of the scrubbing liquor
should be investigated.


          The plant studies portion of this project  showed that
an effective tool  for performing alternative evaluations is a
process simulation computer model package.  A process simulation
package can also be used to evaluate the consistency of the data
collected in the system characterization phase of an implementa-
tion plan.  This section first discusses the calculations required
to evaluate scrubber alternatives and then presents  a methodology
for using a model  to perform the evaluations.


4.2.1     Evaluation Criteria

          The scrubbing evaluation should involve mass and energy
balances to calculate flows, compositions, and temperatures of
all process streams.  An overall material and energy balance in-
cludes a humidification calculation to determine the amount of
water evaporated by the flue gas, and a prediction of solid-liquid
equilibrium for the solid waste.  The important solid-liquid equi-
libria include CaC03, CaS0^2H20, and CaS03'%H20.  Of the remain-

23ESE-S: SH£S-£S 5;
                               -145-

-------
overall balance and the suspended solids concentrations of all
the streams.  Calculations around the scrubbing vessel include
the prediction of gas-liquid equilibria between the C02 and S02
in the gas and liquid streams in the scrubber.  A common assump-
tion for CO2 is that C02 is transferred between the scrubbing
liquor and the gas so that the liquor equilibrium partial pres-
sure of C02 is equal to the C02 partial pressure in the gas.  A
scrubbing process evaluation should include this equilibrium so
that the amount of C02 entering or leaving the system through
the scrubber may be determined.


          The evaluation should also include prediction of the
equilibrium partial pressure of S02 above the scrubbing liquor.
Calculation of this parameter will allow an evaluation of the
ability of a given scrubbing liquor to remove a specified amount
of S02 from the flue gas.  Any changes in liquor composition re-
sulting from the implementation of a recycle/reuse scheme may be
evaluated in terms of S02 removal capability by comparing liquor
equilibrium S02 partial pressure to the outlet gas S02 partial
pressure.  The prediction will allow an estimate of the S02 re-
moved in the demister to be evaluated.  The partial pressure of
S02 above the demister wash should be less than the partial pres-
sure of S02 in the stack gas.


          The calculations outlined above will allow a particular
scrubbing recycle/reuse option to be evaluated in terms of scale
potential and dissolved solids content of the scrubbing liquor.
The following sections present discussions of the prediction of
scale potential and TDS levels in a typical scrubbing system.


4.2.1.1   Scale Potential

          The most important factor which determines the feasi-
bility of implementing a scrubbing recycle/reuse alternative is
the potential for scale formation in the demister.  If significant
scale deposits form,  the system must be shut down for cleaning.


          In the plant studies portion of this study, the concept
of relative saturation was used to predict scaling potential.
When the relative saturation of a species is below 1.0, no poten-
tial for solids precipitation exists.  When the relative satura-
tion is greater than 1.0, the species is supersaturated and scale
may form.  In the presence of seed crystals, the relative satura-
tion may exceed 1.0 without scale formation.  However, as relative
saturation increases above 1.0, a critical value is reached where
nucleation occurs.   When this happens, scaling may occur since
conditions favorable to the creation of new crystal nuclei also
tend to produce scale.

                              -146"-

-------
          Past experience has  shown  that  calcium sulfate  dihy-
drate will form chemical scale when  its relative saturation
reaches 1.3-1.4.  A scaling  limit  also  exists  for calcium sul-
fite hemidydrate, but  it is  not  as well defined as the  limit
for gypsum.  A value of at least three  to four times  the  sulfite
saturation level  (R.S. = 1.0)  is necessary to  initiate  sulfite
scaling.  Some systems have  been reported to operate  at six to
seven times the sulfite saturation value  without scaling.  A
detailed definition of relative  saturation was given  in Section
2.2.1.1.
          If a recirculating  demister wash  system  is used, the
evaluation of potential  scale formation in  the  demister should
include a prediction  of  solid precipitation and dissolution rates
in the demister wash  tank.  The  concept of  relative saturation
may be used to calculate rates as  follows:

                       R  = KafCV  (R.S.  - 1)

where

            . R =  precipitation or  dissolution rate

             K =  a  temperature dependent constant

             a =  crystal interfacial  area

             f =  weight  fraction of the considered
                  species in the  solid phase

             C =  total suspended solids concentration

             V =  tank volume

          R.S. =  relative saturation  of considered
                  species


          It is important that the scrubbing process evaluation
include rate calculations and determination of  the relative sat-
urations of important species in the  system.  The  effects of pro-
cess modifications  must  be evaluated  in terms of these parameters
to insure that implementation of a water recycle/reuse option
will not result in  scale formation in the demister.
                              -147-

-------
4.2.1.2   Dissolved Solids Concentration

          If the scrubber makeup water is high in dissolved
solids, excessive TDS levels in the scrubbing system may be en-
countered.  The water evaporated in the scrubbing system causes
the dissolved solids in the makeup water to be concentrated in
the scrubbing loop.


          Excessive TDS levels in the scrubber liquor can in-
crease the energy requirements for pumping the liquor and can
decrease the mass transfer characteristics in the scrubber.  As
the TDS level increases, the viscosity of the scrubbing liquor
will increase.  For example, the viscosity of a 25 wt. % NaCl
solution is about twice that of pure water while a 5 wt.  °L solu-
tion has a viscosity only about 10% higher than pure water.  As
the viscosity of the scrubbing liquor increases, the energy re-
quirements for pumping the liquor will increase.


          The TDS level will also affect the surface tension of
the liquor.  Excessive TDS levels may cause larger droplet sizes
in a spray tower.  This may reduce the mass transfer characteris-
tics of the droplets and adversely affect scrubber performance.


          The level of chloride present in the makeup water may
also limit the use of an alternate makeup source in a scrubbing
system.  If the chloride level in the scrubbing liquor is too
high, corrosion problems may be encountered.


          The material and energy balances discussed previously
will allow prediction of the TDS and chloride levels expected
under alternate operating conditions.   The limitations discussed
here should be considered in evaluating the feasibility of con-
verting to closed-loop operation or of changing makeup water
source.


4.2.2     Model Application

          As part of a recycle/reuse implementation plan, a pro-
cess model may be used for two purposes:

          1)  check consistency of data obtained
              in characterization phase,  and

          2)  evaluate and optimize alternative
              operating conditions.
                              -148-

-------
The data collected both from design information and the  sampling
program can be used  as  inputs to the model to calculate  makeup
water requirements,  precipitation rates,  and scaling potentials
in the system.  The  values  determined by  the model  may then be
compared to the measured values obtained  from the sampling pro-
gram to check the data  consistency and the validity of the pro-
cess model.  In addition to comparing stream compositions and
pH's, the calculated equilibrium S02 partial pressure above the
scrubber effluent stream should be compared to the  measured S02
partial pressure  in  the stack gas.  If the measured stack gas S02
concentration does not  exceed the calculated S02  equilibrium
partial pressure, a  gas or  liquid phase anlaytical  error is in-
dicated.  In this case, the analytical data should  be re-examined.


          Once the data consistency checks have been made and the
process model has been  verified, simulations of the scrubbing sys-
tem under alternative operating conditions should be made.  Alter-
native simulations should be performed for a variety of  conditions
under which the scrubbing system is expected to operate.  These
should include different loads and different coal compositions as
well as variations in makeup water quality.   The  design  and opera-
ting data collected  in  the  characterization phase of the implemen-
tation plan should be used  to identify the magnitude of  variations
expected for the  scrubbing  system.


          Simulations for both typical operation  in the  alterna-
tive mode and for worst case operation should be  made.   In the
case where cooling tower blowdown or a combination  of fresh makeup
water and cooling tower blowdown is being considered as  scrubber
makeup water, several simulations should  be performed to optimize
the system configuration or ratio of fresh water  to cooling tower
blowdown.  This is particularly important for scrubbing  operations
with recirculating demister wash systems.   The level of  S02
removal  oxidation,  and carryover in the  demister section should
be  investigated with simulations to determine the long-term
effects of these  parameters on system operation.


          In addition to investigating long range variations in
operating conditions, simulations to determine the  effects of
short-tera upsets should be made.  Short-term upsets may include
changes in S02 removal, oxidation, flue gas temperature, and
makeup water Quality.  For  example, in a  cascaded water  system
where cooling'tower  blowdown is used as scrubber makeup, any up-
sets which occur  in  the cooling system will cause a change in
the scrubber makeup  water quality.
                              -149-

-------
          The results of these simulations should identify opera-
ting conditions which can be expected under a recycle/reuse alter-
native.  Stream temperatures, flows, and compositions will be
specified by the model and should allow evaluation of the feasi-
bility of the recycle/reuse option with respect to system scaling
potential and TDS levels.  The simulation results may be used as
a basis for designing and process modifications or for designing a
detailed test program for further investigations on a pilot or
prototype scale.  The number of simulations required and there-
fore the cost of performing the simulations will depend upon the
type of modifications considered and the stability of the parti-
cular system under consideration.  If the scrubbing system is
subject to wide variations in operating conditions,  more simula-
tions should be performed to fully evaluate the effects of insti-
tuting a water recycle/resue option.  Meaningful cost estimates
can only be made for a particular system undergoing a specific
recycle/reuse modification(s).


4.3       Phase III:  Pilot-Scale Studies

          After the characterization and simulation phases of an
implementation plan have been completed, pilot-scale testing
should be performed to further analyze the effects of the water
recycle/reuse scheme on scrubber operation.  Two types of scrub-
bing system modifications have been identified as parts of a plant
water recycle/reuse scheme.  These are converting from open-loop
to closed-loop operation and using a cascaded water stream as
scrubber makeup.  This section discusses each of these two types
of scrubber modification separately.


4.3.1     Converting to Closed-Loop Operation

          One method identified for minimizing the water require-
ments of a SOa/particulate scrubbing system is to operate the
system in a closed-loop mode.  In a closed-loop system all the
excess water leaving with the solid waste is recycled to the
scrubbers.  The only water lost with the solids will be that
occluded with the sludge.  If a pond is used, the sludge will
be about 50% water.  If a vacuum filter is used for solids con-
centration, the sludge may be as low as 3578 water for systems
with high oxidation (either forced by air sparging or naturally
occurring).


          Since scrubbing systems are not designed with a sepa- '
rate clarifier, filter, or pond for each module, the isolation
of an individual module is not feasible for testing the effects
of converting from open-loop to closed-loop operation.  Either a
pilot scale or one-module prototype installation is required to
                             -150-

-------
test closed-loop operation.   Additional piping will  be  required
to return the pond overflow  for systems using a pond for  solid/
liquid separation.  For  scrubbing systems  with high  oxidation
a vacuum filter may be installed to effectively dewater the
sludge to about 65% solids.   Piping would  then be  necessary to
return the filter supernatant to the system.   Converting  to
closed-loop operation will lower the system makeup water  require-
ments and, therefore, affect the demister  water balance.  If a
once-through demister is used,  additional  water to supplement
the decreased makeup can be  taken from the clarifier overflow.
If the evaluation phase  showed this option not to  be feasible,
a recirculating demister system may be installed.  This would
involve adding a catch tray  to the scrubber internals,  a  demister
wash tank, and additional piping and pumps.   The cost of  the new
equipment will depend on the size of the system and  the particu-
lar process option chosen (once-through or recirculating  demister)


          Once the process modifications have been made,  the
pilot or prototype system may be started up.   As with the char-
acterization sampling scheme,  line-out samples  should be  taken
until the system has reached steady-state.  The time required to
reach steady-state will  depend on the particular system design.
Larger reaction tanks and the use of a pond will increase the
system residence time and, therefore,  require longer line-out
time.


          The sampling strategy outlined in Section  4.1.2 also
applies to pilot testing. Closed-loop operation should be tes-
ted under a variety of conditions to insure safe operation with
the modified system.  These  conditions will have been identified
by the characterization  phase as being likely to occur.   Changes
may involve variations in flue gas flow, S02  concentration, and
makeup water composition.


          As was illustrated in Table 4-4  of  the characterization
section  eas analyses are recommended for  the flue gas  entering
the system and the stack gas leaving the scrubbing system.  Com-
plete liquor analyses are recommended for  the makeup water, clar-
ifier overflow, and pond return   Liquid and  solid analyses are
recommended for the scrubber feed  scrubber effluent alkali
slurry, clarifier underflow, and demister  loop  slurries^Liquid
analyses should include  calcium  magnesium sodium  chloride,
carbonate, nitrate, sulfite, and total sulfur   Solid P^ase
analyses should include  calcium, magnesium, carbonate,  sulfite,
and sulfate.
                              -151-

-------
          In many cases the soluble species (magnesium, sodium,
chloride,  and nitrate)  may not change appreciably throughout the
system since system liquid phase residence times are much shorter
than solid phase residence times.   If preliminary results indi-
cate this, then some of the liquid phase analyses may be elimi-
nated or performed less frequently.  Notes to this effect are
listed at the bottom of Table 4-4.


          Operation of the pilot or prototype facilities in a
closed-loop mode over the range of operating conditions to be
expected on a full-scale will establish a strong basis upon which
the full-scale modifications may be based.  Effects on the scrub-
bing operation as well as the optimum demister configuration for
closed-loop operation can be assessed with the results from the
pilot or prototype studies.


4.3.2     Changing Makeup Water Source

          As discussed previously for converting to closed-loop
operation, isolation of an individual module of a scrubbing sys-
tem may not be practical to study system modifications.  Again,
a one-module prototype or pilot-scale installation is ideal for
testing this water recycle/reuse scheme.


          Additional piping and a makeup water mixing tank will
be required in order to study the use of a mixture of existing
makeup water and an alternative source (cascaded water) as scrub-
ber makeup.  Using a mixture of makeup waters will allow the op-
timum level of use of cascaded water as makeup to be determined.
It is recommended that four or five different ratios of makeup
water sources be used and the scrubbing system chemistry be char-
acterized under each condition.  Scaling potential should be
evaluated for the demister wash loop as well as for the main
scrubbing loop of the system.


          The sampling strategy outlined for the characterization
phase of implementation may be used in this phase also.  Sample
points and analyses were tabulated in Table 4-4 of the character-
ization section.


          The results of these pilot or prototype studies will
allow the design of a full-scale modification to use cooling
tower blowdown (or other cascaded water) or a mixture of the
tower blowdown with fresh makeup water.  The pilot studies will
identify any process modifications necessary to prevent scaling
in the demisters.  They will also identify the optimum mixture
                              -152-

-------
of makeup waters  to  allow reliable system operation if the  new
makeup water  source  may not be used exclusively.


4.4       Phase IV:   Full-Scale Operations

          The  final  result of the pilot studies will be a system
design which will allow closed-loop operation or substitution of
an alternate makeup  water source.  Either one of these process
modifications  may require changing the demister loop design to
prevent  scale  formation in the demister.   Converting a once-
through  demister  system to a recirculating system will involve
installing  a  catch tray under the demister so that most of  the
wash water  may be recycled.  Additional tanks,  pumps,  and piping
will also be  required.


          After the  equipment modifications have  been  made,  it is
recommended that  the system operation be  monitored closely  until
steady-state  is achieved.  The same sampling strategy outlined in
the characterization phase should be used to perform a final eval-
uation of the full-scale modified system.   The  frequency of sam-
pling may be  decreased somewhat since the line-out time required
for a full-scale  system using a pond will be significantly  longer
than a pilot  or prototype installation using a  vacuum  filter.
After reliable operation is established,  a regular sampling pro-
gram should be established to monitor system performance.   This
program  will  be especially important when the plant is undergoing
a step change such as changing the type of coal fired  or the
system load.
                              -153-

-------
5.0       SUMMARY


          Generalized implementation plans have been presented
for recycle/reuse alternatives identified in the plant studies
conducted in the first part of this program.  Recycle/reuse op-
tions were identified for cooling tower,  ash sluicing, and SOz/
particulate scrubbing systems at coal-fired power plants.


          Options for cooling towers included increased recircu-
lation resulting in lower blowdown and makeup rates.   Increased
recirculation in cooling towers may require the addition of pH
control with sulfuric acid or softening (either makeup or slip-
stream) to control gypsum scale potential.  The implementation
plan discussed was divided into three phases:  1) system charac-
terization, 2) alternative evaluation, and 3) full-scale modifi-
cations.  This plan presented methodologies for identifying
operating variables and using these variables to evaluate scaling
potential and treatment requirements for alternate operating
conditions.  The equipment and sampling requirements for imple-
menting full-scale changes in cooling tower operation were also
presented.


          Recycle/reuse options identified for ash sluicing
operations include converting to completely or partially closed-
loop operation.  These options may require treatment to prevent
scaling.  Possible treatments identified include 1) the use of a
reaction tank to allow ash dissolution and solids precipitation
to prevent scaling in the sluice line and 2) the use of softening
to reduce the dissolved calcium in the pond recycle.   A plan was
presented whereby a recirculating ash sluice system may be imple-
mented.  Four phases were identified 1) ash characterization,
2) alternative evaluation, 3) pilot studies, and 4) full-scale
operation.  The first phase discussion presented a methodology
for identifying ash reactivity with bench-scale experiments.  The
second phase presented a methodology for evaluating the feasibi-
lity of a recirculating system based on scale potential and treat-
ment requirements.  Pilot studies (third phase) were recommended
for ash sluicing to better define operating parameters which are
not easily predicted and to evaluate treatment options.  The full-
scale phase discussion included descriptions of equipment required
and important operating variables.


          For S02/particulate scrubbing systems, two potential
recycle/reuse options were identified in the plant studies.
These included converting to closed-loop operation and using an
alternate makeup water source such as cooling tower blowdown or
                              -154-

-------
ash pond overflow.  A general  implementation plan has been pre-
sented for instituting these types of modifications.  As with
ash sluicing, four phases were identified:  1) system character-
ization, 2) alternative evaluation, 3) pilot studies, and 4) full-
scale modifications.  The first phase discussion presented the
important operating variables  to  be considered and a sampling
plan for obtaining operating data.  The  second phase presented a
methodology for evaluating  the feasibility of recycle/reuse op-
tions.  The pilot studies  (third  phase)  presented a plan for
testing operation with a recycle/reuse option.  Potential prob-
lem areas are demister scale formation and TDS level in the
scrubbers.  The full-scale  discussion described the equipment
necessary and recommended a sampling strategy to be used to
monitor system performance.
                                -155-

-------
                           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).
                             -156-

-------
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).
                              -157-

-------
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,  B.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.
                             -158-

-------
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    Palmer-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.

                              -159-

-------
UH-007    Uhlig,  Herbert H.,  Corrosion and corrosion control,
          an introduction to  corrosion science and engineering.
          N.Y... Wiley, 1963.
                             -160-

-------
                                TECHNICAL REPORT DATA
                         If lease read Inuntctions on the reverse before completing!
 EPA-6QO/7-78-055a
4. TITLE AND SUBTITLE water Recycle/Reuse Alternatives in
Coal-fired Steam-electric Power Plants: Volume I.
Plant Studies and General Implementation Plans
                                                    5. REPORT OATS
                                                      March 1978
                                                    6. PERFORMING ORGANIZATION CODE
                                                      . RECIPIENT'S ACCESSION NO
7. AUTHOR(S)
James G. Noblett and Peter G. Christman
                                                    8. PERFORMING ORGANIZATION REPORT NO.
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 REPORT AND PERIOD COVERED
                                                      Final; 6/75-2/78	
                                                     14. SPONSORING AGENCY CODE
                                                      EPA/600/13
 is. SUPPLEMENTARY NOTES jERL-RTP project officer is Frederick A. Roberts, EPA/ERC,
 200 S.  35th St. , Corvallis, OR  97330  (503/420-4715).
is. ABSTRACT
              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  recirculation 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 discusses the
 recycle /treatment/reuse opportunities for cooling, ash sluicing, and SO2/particu-
 late scrubbing systems as well as combined systems. It also includes the implemen-
 tation plans .
 17.
                             KEY WORDS AND DOCUMENT ANALYSIS
                 DESCRIPTORS
                                          b.lDENTIFIERS/OPEN ENDED TERMS
                                                                 c. COSATI Field/Group
  Pollution             Combustion
  Water Treatment     Cooling Water
  Filtration            Scrubbers
  Circulation           Sulfur Oxides
  Electric Power Plants  Dust
  Coal                 Mathematical Model
                                         Pollution Control
                                         Stationary Sources
                                         Water Recycle/Reuse
                                         Ash Sluicing
                                         Particulate
13B

07D

10B
21D
21B
13A
07A
07B
11G
12A
  3. DISTRIBUTION STATEMENT

  Unlimited
                                         19. SECURITY CLASS (This Report I
                                         Unclassified
21. NO. OF PAGES
  188
                                          20. SECURITY CLASS (This page)
                                          Unclassified
                                                                  22. PRICE
 EPA Form 2220-1 (9-73)
                                    -161-

-------