U.S. Environmental Protection Agency Industrial Environmental Research EPA~600/7~ 77-052
Office of Researcn and Development Laboratory ._ 4_
Research Triangle Park. North Carolina 27711 Mdy 1977
DISPOSAL OF BY-PRODUCTS
FROM NONREGENERABLE FLUE
GAS DESULFURIZATION SYSTEMS:
Second Progress Report
Interagency
Energy-Environment
Research and Development
Program Report
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S.
Environmental Protection Agency, have been grouped into seven series.
These seven broad categories were established to facilitate further
development and application of environmental technology. Elimination
of traditional grouping was consciously planned to foster technology
transfer and a maximum interface in related fields. The seven series
are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
A. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
This report has been assigned to the INTERAGENCY ENERGY-ENVIRON>ENT
RESEARCH AND DEVELOPMENT series. Reports in this series result from
the effort funded under the 17-agehcy Federal Energy/Environment
Research and Development Program. These studies relate to EPA's
mission to protect the public health and welfare from adverse effects
of pollutants associated with energy systems. The goal of the Program
is to assure the rapid development of domestic energy supplies in an
environmentallycompatible manner by providing the necessary
environmental data and control technology. Investigations include
analyses of the transport of energy-related pollutants and their health
and ecological effects; assessments of, and development of, control
technologies for energy systems; and integrated assessments of a wide
range of energy-related environmental issues.
REVIEW NOTICE
This report has been reviewed by the participating Federal
Agencies, and approved for publication. Approval does not
signify that the contents necessarily reflect the views and
policies of the Government, nor does mention of trade names
or commercial products constitute endorsement or recommen-
dation for use.
This document is available to the public through the National Technical
Information Service, Springfield, Virginia 22161.
-------
EPA-600/7-77-052
May 1977
DISPOSAL OF BY-PRODUCTS FROM
NONREGENERABLE FLUE GAS
DESULFURIZATION SYSTEMS:
Second Progress Report
by
J. Rossoff, R.C. Rossi, R.B. Fling,
W.M. Graven, and P.P. Leo
The Aerospace Corporation
Environment and Energy Conservation Division
P.O. Box 92957
Los Angeles, California 90009
Contract No. 68-02-1010
Program Element No. EHE624A
EPA Project Officer: Julian W. Jones
Industrial Environmental Research Laboratory
Office of Energy, Minerals, and Industry
Research Triangle Park, N.C. 27711
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Washington, D.C. 20460
-------
ABSTRACT
This report provides results of the first three years (1973-1975)
of a study by The Aerospace Corporation to determine environmentally sound
methods for the disposal of wastes from nonregenerable flue gas desulfuriza-
tion systems. Characterizations of untreated and treated wastes from seven
different scrubbers at eastern and western plants, using lime, limestone, or
double-alkali absorbents, have been conducted. The data presented identify
concentrations of salts and trace elements related to potential environmental
pollution for both treated and untreated wastes. Physical properties, e. g. ,
bulk density, compression strength, permeability and viscosity, are given.
Ponding of untreated wastes in impermeable impoundments appears to be a
viable process for pollution control; however, the ability to reclaim the land
has not been determined. Chemically fixed sludges in landfills have been
shown to be structurally adequate. Although zero pollutant discharge of
leachates from chemically fixed sludges may not be attained in all cases, the
reduction of the mass release of constituents to the subsoil through the reduc-
tion of solubility, permeability, and surface water can be accomplished. Cost
estimates (1976 dollars) for fixation disposal equate to $2 to $3 per ton of
eastern coal burned (depending on the process selected); lined pond disposal
costs of untreated materials are approximately $1. 60 to $2. 20 per ton of coal
(depending on the type of material used); and the cost for ponding on natural
clay soils is $1. 00 per ton of coal.
This report was submitted in partial fulfillment of Contract Num-
ber 68-02-1010 by The Aerospace Corporation under the sponsorship of the
U.S. ^Environmental Protection Agency. Work completed through 24 Novem-
ber 1975 is reported; however, selected updates have been made during the
publication period.
111
-------
CONTENTS
ABSTRACT iii
ACKNOWLEDGMENTS xvii
CONVERSION TABLE xix
I. INTRODUCTION ~~ i
1. 1 Report Coverage 1
1. 2 Objectives and Study Approach 2
1. 3 Organization of Report 3
II. CONCLUSIONS AND OBSERVATIONS 5
2. 1 Disposal Criteria 5
2. 2 Effect of Scrubbing Process Variables on
Sludge Chemistry 5
2. 3 Trace Element Content 6
2. 4 Physical Properties 6
2. 5 Chemical Properties 6
2. 6 Disposal Cost Estimates 7
III. RECOMMENDATIONS 9
3. 1 Disposal Criteria 9
3. 2 Field Evaluations 10
IV. SUMMARY 11
4. 1 Disposal Criteria 11
4. 2 Physical Properties 12
4. 2. 1 Wet Bulk Density 13
4. 2. 2 Coefficient of Permeability 14
v
-------
CONTENTS (Continued)
4. 2. 3 Viscosity 16
4.2.4 Compaction 19
4. 2. 5 Compressive Strength 19
4. 3 Chemical Characterization 20
4. 3. 1 Generation of Trace Elements in
FGD Wastes 24
4. 3. 2 Process Variables 24
4.3.3 Chemical Solubility Analysis 26
4.4 Environmental Acceptance of FGD Waste 27
4.4.1 Polluta.nt Access to the Environment 27
4.4.2 Alternative Disposal Techniques 28
4.4.3 Assessment of Pollution Potential 29
4. 5 Disposal Cost Estimate 30
V. DISPOSAL CRITERIA 33
5. 1 Disposal Alternatives for FGD Waste 33
5. 2 Environmental Effects of FGD Waste
Disposal Alternatives 33
5. 2. i Rainwater Seepage 33
5. 2. 2 Landfill Runoff 40
5. 2. 3 Land Reuse 41
5. 3 Selection of Disposal Criteria 41
VI. PHYSICAL PROPERTIES DETERMINATION 43
6. 1 Background 43
6. 2 Solids Characterization 44
6. 3 Viscosity 45
6. 3. 1 Experimental Procedure 45
6. 3. 2 Viscosity Test Results 46
6. 4 Wet Bulk Densities 48
6. 4. 1 Experimental Procedure 49
6. 4. 2 Bulk Density Test Results 49
6. 5 Coefficient of Permeability 54
6. 5. 1 Experimental Procedure 56
VI
-------
CONTENTS (Continued)
6. 5. 2 Permeability Test Results ............ 58
6. 6 Compaction ............................ 61
6. 6. 1 Experimental Procedure ............. 62
6. 6. 2 Compaction Test Results ............. 62
6. 7 Compression ........................... 66
6.7. 1 Experimental Procedure ............. 68
6. 7. 2 Compression Test Results ............ 68
VII. CHEMICAL CHARACTERIZATION OF UNTREATED
FGD WASTE ................................ 71
7. 1 Overview and Assessment .................. 71
7. 2 Power Plant Scrubbing Facilities, Sampling,
and Chemical Analysis .................... 73
7. 3 Chemical Analyses ....................... 73
7. 3. 1 Experimental Procedure ............. 73
7. 3. 2 Results of Chemical Analyses ......... 75
7.3.3 Determination of Material and Ionic
Charge Balance ................... 77
7. 4 Effect of Scrubbing Process Variables
on Scrubber Liquors ...................... 81
7.4. 1 In-Process Variations in Scrubber
Liquor Composition ................ 81
7. 4. 2 Variation in Composition of Sludge
Liquors with Time ................. 89
7. 4. 3 Effect of pH and Ionic Strength on
Concentration .................... 91
7. 4. 4 Comparison of Selected Solute Trace
Elements in Sludge Liquors from
Limestone, Lime, and Double-Alkali
Scrubbing Systems ................. 92
7. 4. 5 Relationship of Trace Element Content
Between Input Ingredients and
Scrubber Sludge ................... 96
7. 4. 6 Distribution of Trace Elements in Coal
Through Combustion ................ 99
7.4.7 Comparison of Trace Elements in Eastern
and Western Coals ................. 100
VII
-------
CONTENTS (Continued)
VIII. CHEMICAL SOLUBILITY ANALYSES 103
8. 1 Basis of Program Model 103
8. 2 Operational Description 105
8. 3 Test of Program 106
8. 3. 1 Comparison of Experimental and
Calculated Major Species 106
8.3.2 Comparison of Experimental and
Calculated Trace Species 107
8. 4 Summary and Conclusions Ill
IX. EVALUATION OF THE ENVIRONMENTAL
ACCEPTABILITY OF FGD SLUDGE 113
9. 1 Overview '. . . 113
9. 2 Routes of Accessibility of Potential Chemical
Pollutants to the Environment 114
9. 2. 1 Pollution by Vaporization 115
9. 2. 2 Pollution by Wind Whipped Spray 115
9.2.3 Pollution by Runoff 116
9. 2. 4 Pollution by Rainwater Leaching of
Untreated Wastes 116
9. 2. 5 Pollution by Rainwater Leaching of
Chemically Treated Wastes 129
9. 3 Alternative Disposal Techniques 139
9.3.1 Ponding 139
9.3.2 Ponding with Underdrainage 140
9. 3. 3 Dry Sludge Disposal 141
9.3.4 Chemically Treated Sludge Disposal 141
9. 4 Assessment of the Chemical Pollution Potential
on the Environment by Alternative Disposal
Methods 142
X. DISPOSAL APPLICATIONS AND ESTIMATED COSTS 147
10. 1 Disposal Methods 147
10. 2 Disposal Cost Estimates 152
Vlll
-------
CONTENTS (Continued)
10. 2. 1 Ponding Cost Estimates 152
10. 2. 2 Cost Comparisons: Ponding Versus
Fixation 154
REFERENCES 165
APPENDIXES
A. CHARACTERIZATION OF SLUDGE SOLIDS 1 69
B. DESCRIPTION OF SCRUBBER SYSTEMS 185
C. CHEMICAL ANALYSIS TECHNIQUES USED BY
AEROSPACE 211
D. CHEMICAL CHARACTERIZATION DATA SHEETS 217
E. EFFECT OF SCRUBBING PROCESS VARIABLES ON
WASTE SYSTEM CHEMISTRY DATA PLOTS 245
F. MATHEMATICAL DETERMINATION OF PROGRAM
SCRUB 261
G. FGD WASTE TRANSPORT COSTS 273
IX
-------
TABLES
1. Flue Gas Desulfurization Systems Sampled as Data Base 3
2. Wet Bulk Densities of FGD Wastes as a Function of
Dewatering Method 15
3. Unconfined Compressive Strength Test Results Summary 21
4. Chemical Constituents in Leachate from Untreated and
Chemically Treated FGD Waste after 1 and 50 PVD 22
5. Phase Compositions of FGD Solids in Weight Percent 23
6. Input Data for Study Cases 30
7. Sludge Disposal Cost Ranges 32
8. Environmental Effects of Disposal Alternatives 34
9. Chemical Constituents in Leachate from Untreated and
Chemically Treated FGD Waste after 1 and 50 PVD 39
10. Untreated Sludge Bulk Densities 55
11. Permeability of Untreated and Chemically Fixed
FGC Sludges 59
12. Unconfined Compression Test Results 69
13. Flue Gas Desulfurization Systems Sampled as Data Base 74
14. Range of Chemical Constituents in FGD Sludges 76
15. Material Balance 78
16. Ionic Charge Balance 79
-------
TABLES (Continued)
17. Net Change in Scrubber Liquor Composition of Major,
Minor, and Trace Constituents Between Initial and Final
Stages in Scrubber System ......................... 82
18. Relative Proportions of Trace Elements in Limestone,
Lime and Double -Alkali Systems ...................... 95
19. Comparison of Chemical Constituents in Sludge Liquors
with Leachate after 1 and 50 PVD: TVA Shawnee Lime-
stone Sludge ................................... 123
20. Comparison of Chemical Constituents in Sludge Liquors
with Leachate after 1 and 50 PVD: SCE Mohave Lime-
stone Sludge ................................... 124
21. Comparison of Chemical Constituents in Sludge Liquors
with Leachate after 1 and 50 PVD: LPC Phillips Lime
Sludge ....................................... 125
22. Comparison of Chemical Constituents in Sludge Liquors with
Leachate after 1 and 50 PVD: APS Cholla Limestone Sludge ... 126
23. Comparison of Chemical Constituents in Sludge Liquors
with Leachate after 1 and 50 PVD: GM Parma Double-Alkali
Sludge ....................................... 127
24. Comparison of Chemical Constituents in Sludge Liquors
with Leachate after 1 and 50 PVD: TVA Limestone Sludge
(Chemfix) ..................................... 134
25. Comparison of Chemical Constituents in Sludge Liquors
with Leachate after 1 and 50 PVD: TVA Shawnee Lime-
stone Sludge (IUCS) .............................. 135
26. Comparison of Chemical Constituents in Treated Sludge
Leachate after 1. and 50 PVD: SCE Mohave Limestone
Sludge (IUCS) .................................. 136
27. Comparison of Chemical Constituents in Treated Sludge
Leachate after 1 and 50 PVD: DLC Phillips Lime Sludge
(Calcilox by Dravo) .............................. 137
28. Input Data for Study Cases
XI
-------
TABLES (Continued)
29. Significant Features of Disposal Methods 148
30. FGD Disposal Methods Operating by End of 1976 149
31. Significant Environmental Factors Associated with
Alternative Waste Disposal Methods 153
32. Total Ponding Costs: PVC-20 Liner; 10-ft Sludge Depth 158
33. Total Ponding Costs: PVC-20 Liner; 20-ft Sludge Depth 159
34. Total Ponding Costs: PVC-20 Liner; 40-ft Sludge Depth 160
35. Total Ponding Costs: Hypalon-30 Liner; 10-ft Sludge
Depth 161
36. Total Ponding Costs: Hypalon-30 Liner; 20-ft Sludge
Depth 162
37. Total Ponding Costs: Hypalon-30 Liner; 40-ft Sludge
Depth 163
XLl
-------
FIGURES
1. Permeability of Treated and Untreated Sludge: Aerospace
Analyses 17
2. Viscosity of Desulfurization Sludge 18
3. Compaction of FGD Sludge 20
4. Relationship Between Trace Element Content in Coal and
Sludge Solids 25
5. Relationship Between Trace Element Content in Coal and
Sludge Liquors 25
6. Leachate TDS from Treated and Untreated Sludges as a
Function of PVD 28
7. Mass Loading of TDS to Subsoil for Various Disposal
Modes of Treated and Untreated FGC Wastes 31
8. Viscosity of Desulfurization Sludge 47
9. Wet Bulk Density of TVA Shawnee Limestone Sludge and
Dewatered Samples: 1 Feb 1973 50
10. Wet Bulk Density of TVA Shawnee Limestone Sludge and
Dewatered Samples: 15 Jun 1974 50
11. Wet Bulk Density of TVA Shawnee Lime Sludge and
Dewatered Samples: 19 Mar 1974 . 51
12. Wet Bulk Density of DLC Phillips Lime Sludge and
Dewatered Samples: 17 Jun 1974 51
13. Wet Bulk Density of GM Parma Double-Alkali Sludge and
Dewatered Samples: 18 Jul 1974 52
14. Wet Bulk Density of APS Cholla Limestone Sludge and
Dewatered Samples: 1 Apr 1974 52
Xlll
-------
FIGURES (Continued)
15. Wet Bulk Density of UPL Gadsby Double-Alkali Sludge and
Dewatered Samples: 9 Aug 1974 53
16. Wet Bulk Density of SCE Mohave Limestone Sludge and
Dewatered Samples: 30 Mar 1973 53
17. Permeability of Treated and Untreated Sludge: Aerospace
Analyses 60
18. Compaction Test Results for TVA Shawnee Limestone Sludge
Solids 63
19. Compaction Test Results for TVA Shawnee Lime Sludge
Solids 63
20. Compaction Test Results for GM Parma Filter Cake Solids .... 64
21. Compaction Test Results for UPL Gadsby Double-Alkali
Sludge Solids 64
22. Compaction Test Results for DLC Phillips Lime Sludge
Solids 65
23. Compaction Test Results for APS Cholla Limestone
Solids 65
24. Compaction of FGD Sludge 67
25. Relationship Between Liquor System pH and Ionic
Strength 93
26. Median Values for Trace Element Concentrations in Sludge
Liquors of Three Absorbent Systems 94
27. Average Trace Element Content of Sludge Solids 97
28. Average Trace Element Content of Sludge Liquor 98
29. Comparison of Coal Analyses with Literature Data 101
30. Comparison of Experimental Calcium Concentration and
Concentrations Calcjlated with Two Computer Programs 108
31. Comparison of Experimental Sulfate Concentration and
Concentrations Calculated with Two Computer Programs 109
xiv
-------
FIGURES (Continued)
32. Analysis of Leachate from TV A Shawnee Limestone Sludge:
Aerobic Conditions 118
33. Analysis of Leachate from SCE Mohave Limestone Sludge:
Aerobic Conditions 119
34. Analysis of Leachate from DLC Phillips Sludge: Aerobic
Conditions 119
35. Analysis of Leachate from APS Cholla Sludge: Aerobic
Conditions 120
36. Analysis of Leachate from GM Parma Sludge: Aerobic
Conditions 120
37. Analysis of Leachate from TV A Shawnee Limestone Sludge:
Anerobic Conditions 121
38. Analysis of Leachate from SCE Mohave Limestone Sludge:
Anaerobic Conditions 122
39. Analysis of Leachate from DLC Phillips Sludge: Anaerobic
Conditions 122
40. Analysis of Leachate from TVA Shawnee (Chemfix) Sludge:
Aerobic Conditions 130
41. Analysis of Leachate from TVA Shawnee (Chemfix) Sludge:
Anaerobic Conditions 130
42. Analysis of Leachate from TVA Shawnee Limestone (IUCS)
Sludge: Anaerobic Conditions 131
43. Analysis of Leachate from SCE Mohave (IUCS) Sludge:
Aerobic Conditions 131
44. Analysis of Leachate from SCE Mohave (IUCS) Sludge:
Anerobic Conditions 132
45. Analysis of Leachate from DLC Phillips (Calcilox) Sludge:
Aerobic Conditions 132
xv
-------
FIGURES (Continued)
46. Analysis of Leachate from DLC Phillips (Calcilox) Sludge:
Anerobic Conditions 133
47. Leachate TDS from Treated and Untreated Sludge as a
Function of PVD 143
48. Mass Loading of TDS to Subsoil for Various Disposal
Modes of Treated and Untreated FGC Wastes 144
49. Installed Liner Costs 155
50. FGD Waste Ponding Disposal Costs: 20-mil PVC 156
51. FGD Waste Ponding Disposal Costs: 30-mil Hypalon 157
xvi
-------
ACKNOWLEDGMENTS
Appreciation is acknowledged for the guidance provided by the
EPA Project Officer, Julian W. Jones, Industrial Environmental Research
Laboratory, Research Triangle Park, North Carolina. His assistance has
appreciably contributed to the cohesiveness of this program, which is particu-
larly broad in technical scope.
Valuable contributions were also made by the following Aerospace
personnel: F. D. Hess, R. W. Laube, and the late Lawrence J. Bornstein.
Henry M. ^aesp, Head <--
Materials Analysis Department
Materials Sciences Laboratory
Director
Office of Stationary Systems
Environment and Energy
Conservation Division
Approved by:
Toru lura, General Manager
Environment and Energy
Conservation Division
xvii
-------
CONVERSION TABLE
A list of conversion factors for British units used in this
report is as follows:
British
1 ac re
1 British thermal unit
per pound
1 foot
1 cubic foot
1 inch
1 gallon (U.S.)
1 pound
1 mile
1 ton (short)
1 ton per square foot
1 gram per square foot
1 part per million
1 pound per square inch
1 cubic yard
Metric
4047 square meters
2.235 Joules per gram
0. 3048 meter
28. 316 liters
2. 54 centimeters
3. 785 liters
0. 454 kilogram
1. 609 kilometers
0. 9072 metric tons
9765 kilograms per square meter
10.76 grams per square meter
1 milligram per liter (equivalent)
0.0703 kilogram per square
centimeter
0.7641 cubic meter
xix
-------
SECTION I
INTRODUCTION
1. 1 REPORT COVERAGE
This interim report presents results obtained over a three-year
period of work (1973-1975) performed in a broad-based study of the charac-
terization and disposal of the by-products from nonregenerable flue gas
desulfurization (FGD) systems. This work was performed under EPA Con-
tract No. 68-02-1010, which began on 23 November 1972 and is scheduled to
continue through 23 January 1977. The first-year results have been reported
in an initial report (1). This report, with some exceptions identified below,
incorporates the first-year data with results obtained during the following two
years and specifically covers the period of study from 23 November 1972 to
23 November 1975. Data obtained during the period of publication (after
23 November 1975) are incorporated wherever possible. A final report will
summarize all data to completion as of 23 January 1977.
Exceptions to the incorporation of all results are as follows:
During the second and third years of this study, the contract was modified to
incorporate two new tasks so that current technological advancements could
be assessed in light of the basic data produced in the general study. These
tasks are as follows: (a) the Environmental Protection Agency (EPA) field
disposal evaluation of treated and untreated wastes at the Tennessee Valley
Authority (TVA) Shawnee Steam Plant, and (b) the assembly, assessment,
and reporting of all flue gas cleaning (FGC), waste-related, research and
development (R&D) activities of EPA, TVA, and private industry. These
two tasks relate to specific events and, for maximum visibility into large
quantities of data, are reported in separate documents: the Shawnee field
evaluation project in (2), and the FGC R&D task in (3). They will also be
reported in separate documents at the conclusion of the current contract
period on 23 January 1977.
In addition to the reports just discussed, formal presentation of
study data have been made periodically throughout the program at three EPA
Flue Gas Desulfurization Symposiums (4-6), the American Society of
Mechanical Engineers (ASME) Annual Technical and Environmental Symposi-
um (7), the National Ash Association Ash Utilization Symposium (8), and two
EPA Lime /Limestone Scrubbing Industry Briefings (9 and 10). Additionally,
closely related data have been reported in (11).
-------
The basic study, which is reported herein, covers the following
subjects: disposal criteria, chemical and physical characterization of
untreated wastes and wastes chemically treated in various laboratories,
environmental effects, chemical solubility, and safe disposal methods and
costs. The materials analyzed during the first three years of this study
were obtained from seven different nonregenerable scrubbing systems at
plants burning eastern and western coal and using lime, limestone, and
dual-alkali absorbents (Table 1). Additionally, numerous sam-
ples have been analyzed from the EPA Shawnee field disposal project and
reported in (2). Samples from many of these sources have been chemically
fixed by three different commercial processors and analyzed in the Aerospace
laboratories. (During the fourth year of this study, additional untreated
samples to be analyzed will include two different types of sulfite wastes that
have been oxidized to gypsum, one Shawnee scrubber waste that is ash-free,
one waste produced in a carbide sludge absorbent system, and one double-
alkali waste. )
1. 2 OBJECTIVES AND STUDY APPROACH
The specific objectives of this study are as follows:
a. To identify environmental problems associated with FGC waste
disposal by determining FGC waste chemical and physical
characteristics.
b. To assess current FGC waste disposal methods, including feasi-
bility, performance, and costs, by conducting laboratory studies
of wastes under conditions associated with waste disposal; by
providing engineering support and conducting chemical and physi-
cal analyses for the Shawnee field disposal evaluation; by evalu-
ating other available data; and by conducting engineering cost
studies of disposal methods.
c. To identify alternative disposal methods that provide an accept-
able means of FGC waste disposal by cost-effective methods.
d. To make recommendations regarding alternative disposal
approaches.
e. To assembly, integrate, assess, and report all FGC waste-
related activities sponsored by EPA and industry.
The objectives of this study are to be met by (a) reviewing water
quality standards and waste management regulations, and correlating this
study's technical evaluations with limitations imposed by these regulations
as appropriate; (b) performing chemical and physical characteristics, as
appropriate, of FGC wastes, coals, makeup water, fly ash, and chemi-
cally treated wastes from EPA-specified sources; (c) surveying and
-------
TABLE 1. FLUE GAS DESULFURIZATION SYSTEMS
SAMPLED AS DATA BASE
Power Plant
Tennessee Valley Authority
(TV A) Shawnee
Steam Plant
TVA Shawnee
Steam Plant
Arizona Public
Service (APS) Company,
Cholla Power
Plant
Duquesne Light
Company (DLC),
Phillips Power
Station
General Motors (GM)
Corporation,
Chevrolet- Parma
Power Plant
Southern California
Edison (SCE) Mohave
Generating Station
Utah Power and
Light (UPL) Company,
Gadsby Station
Scrubber
System
Venturi and
spray tower,
prototype
Turbulent
contact
absorber,
prototype
Flooded-disk
scrubber,
wetted film
absorber
Single- and
dual-stage
venturi
Bubble-cap
tower
Turbulent
contact
absorber,
pilot plant
Venturi and
mobile bed.
pilot plant
Scrubbing
Capacity,
MW (equiv)
10
10
120
410
32
< 1
< 1
Coal
Source
Eastern
Eastern
Western
Eastern
Eastern
W e s te r n
Western
Absorbent
Lime
Limestone
Limestone,
fly ash
Lime
Soda ash.
lime
Limestone
Soda ash,
lime
-------
analyzing technical and economic data pertaining to FGC waste disposal;
(d) planning, coordinating, evaluating, and reporting the EPA field disposal
evaluation at TVA Shawnee (2); (e) continually meeting with or collecting
FGC waste disposal R&D data from approximately 18 EPA-sponsored study
contract teams and industry as appropriate (3); and (f) making recommenda-
tions for environmentally sound FGC waste disposal techniques. Final
recommendations will be made at the conclusion of the study.
1. 3 ORGANIZATION OF REPORT
Sections II and III provide conclusions and recommendations,
respectively. Section IV summarizes the study findings and provides discus-
sions and appropriate tables and figures to support the findings. Sections V
through X comprise the total technical discussion.
Appendices A through E provide descriptions of scrubber systems
from which samples were taken, laboratory analytical techniques, and all
untreated data.
-------
SECTION II
CONCLUSIONS AND OBSERVATIONS
These interim results are based upon current data from the
ongoing Aerospace study. Updating as appropriate will be conducted in the
final report in January 1977.
2. 1 DISPOSAL CRITERIA
Federal criteria do not exist specifically for the disposal of flue
gas desulfurization (FGD) wastes; however, EPA criteria for groundwater
resulting from federal land application of waste water are the EPA "National
Interim Primary Drinking Water Regulations" (12), issued in December 1975.
Groundwater criteria can be met by ponding of untreated wastes in an imper-
meable basin or by fixation landfilling that prevents the collection of surface
water. The acceptability of fixation landfilling that allows the collection of
surface water generally requires a mixing of seepage with groundwater or
streams if drinking water criteria are to be met.
In the absence of structural criteria, a conservative, unconfined
compressive strength of 25 psi was selected in this study for sites which are
to be reclaimed. Compression strength criteria can be met by (a) chemical
fixation, (b) wastes converted to gypsum, or (c) wastes dewatered and mixed
with fly ash and compacted.
Technical and economic studies of disposal methods as a function
of waste characteristics are continuing. The results should form a signifi-
cant portion of the data base from which disposal criteria can be determined.
2. 2 EFFECT OF SCRUBBING PROCESS VARIABLES
ON SLUDGE CHEMISTRY
Process variables affect the concentrations of soluble chemical
species in system liquors through changes in process chemistry:
a. The concentration of major chemical species and trace elements
in FGD waste decreases as the sludge passes from the scrubber
to the disposal point.
b. The concentration of major chemical species increases with
time from startup until a steady-state condition is reached for
-------
all species. Trace element concentrations reach steady state
rapidly and are not affected by major species steady-state
conditions.
c. The scrubber pH is responsible for trace elements leaching from
fly ash; the pH of the systerp downstream of the scrubber does
not affect the concentration of these trace elements in the scrubber
liquor.
2. 3 TRACE ELEMENT CONTENT
The trace element content in FGD sludge is a direct function of the
combustion products of coal:
a. A direct correlation exists between the trace element content of
coal and the trace element content in FGD wastes.
b. Fly ash represents the major source of trace elements in all but
the most volatile elemental species (e.g. , mercury and selenium)
that are scrubbed from flue gases.
2. 4 PHYSICAL PROPERTIES
The behavior of FGD wastes in a disposal site is a function of the
unique physical properties of the wastes:
a. The permeability coefficients of untreated FGD wastes are typi-
cally 10~4 crn/sec and of treated wastes are 10"^ cm/sec or less
[based upon sample materials fixed by Chemfix, Dravo, and IU
Conversion Systems (IUCS)].
b. Pumpability (<20 poise) was found for untreated wastes having a
solids content that ranged between 32 and 70 percent.
c. Bulk densities of untreated wastes as a function of dewatering
techniques and material characteristics varied between 1. 30
and 1. 87 g/cc.
d. Compaction of untreated sludges dewatered to about 80 percent
solids produced permanent displacement of 1 to 4 percent.
e. Treated wastes have unconfined compression strength greater
than 1. 8 tons per square foot (25 psi).
2. 5 CHEMICAL PROPERTIES
The soluble chemical content of FGD waste liquors typically have
approximately 10,000 mg/£ total dissolved solids (TDS)(steady state), except
-------
for double-alkali scrubber systems, whose liquors have a much higher TDS.
Trace elements lie typically between 0.01 and 1 mg/f depending on coal con-
tent and fly ash collection techniques:
a. The leachate quality of rainwater percolated through untreated
FGD waste attains a nearly-constant, TDS content of 2000 mg/£,
primarily sulfate salts, after passage of 5 pore volume displace-
ments (PVD). Initial leachate content is as high as the soluble
chemical content and is dependent upon the type of FGD system.
Leachate through treated FGD waste has initial dissolved solids
content of nominally one-half that of untreated waste and after
5 PVD attains a level of one-half to one-fourth that of the leachate
from untreated wastes.
b. Mass loading to subsoil having a permeability coefficient of
10~5 cm/sec from FGD disposal sites by rainwater leaching
varies from a high value of >1 g/cm^/yr for ponding of
untreated waste that is continually covered with water to a low
value of 0. 01 g/cm^/yr for dry disposal of treated waste. Chemi-
cal treatment makes possible a disposal method that minimizes
seepage and solubility, thereby producing pollution loadings that
are 20 to 40 times less than that of the disposal method for
untreated sludge in which a liner is used and supernate is recircu-
lated to the scrubber.
2. 6 DISPOSAL COST ESTIMATES
Total disposal cost estimates based upon 30-year averages are
as follows:
$/Ton Sludge $/Ton Eastern
(Dry Basis) Coal Burned Mills/kWh
Ponding with
Natural Clay Base 3.50 1.00 0.43
Ponding plus Liner 5. 70 - 7. 80 1. 60 - 2. 20 0. 7 - 1. 0
Chemical Fixation 7.30-11.40 2.10-3.20 0.9-1.4
Ranges for each category shown result from variations in types of sludges
and methods available for disposal. (Limiting conditions for these costs are
given in Section IV, Table 7.) Costs are in January 1976 dollars.
-------
SECTION III
RECOMMENDATIONS
In this study, two current areas of study have been identified in
which a more complete understanding and more accurate definition of envi-
ronmentally sound disposal methods can be attained through a greater
emphasis of effort. These are (a) the correlation of disposal data with dis-
posal criteria, and (b) field evaluations of operational-type, commercial
disposal systems. EPA projects in both of these areas are in progress, but
experience gained in the Aerospace study indicates that a broader scope of
effort now is advisable and can produce the necessary data in a reasonably
short time period. Recommendations are as follows.
3. 1 DISPOSAL CRITERIA
Because no federal criteria exist for the disposal of flue gas desul-
furization (FGD) sludges and because of the wide variations in sludge characteris-
tics, disposal methods, and site characteristics, a data base is being developed
by EPA from which disposal standards can be developed. Additionally, the
neutralization of effluents from a disposal site by soil attenuation and mixing
with groundwater and streams before the water reaches the consumer tap
or other points of uptake indicate that a high degree of conservation may not
be necessary in the control of sludge disposal in most localities. This is
highlighted by the usually low concentrations of trace elements in seepage
liquors, although higher than allowed in drinking water, and the fact that the
contained salts of interest, chlorides and sulfates, are not toxic compounds.
Assuming that the site is managed so that a direct discharge, seepage, or
runoff is not directly consumed, it may be possible to demonstrate that any
of the following is possible:
a. Regulate control of treated or untreated sludge by definition only
of the disposal conditions, procedures, and site arrangement.
b. Allow untreated sludges, limited by site conditions.
c. Allow treated or conditioned sludge leachate constituent concen-
trations in excess of drinking water criteria, but less than in an
untreated liquor for all sites or as limited by site conditions.
d. Allow defined treated or conditioned sludge constituent mass
release for all sites or as limited by site conditions.
-------
It is recommended, therefore, that detailed disposal case studies
be conducted for selected representative sites, with consideration given to the
type of sludge that may be produced at each site; the site-specific conditions,
e. g. , climatological, hydrological, and geological; and the disposal method
or methods appropriate for each site. Data produced in the ongoing 19 EPA
FGC disposal projects (3) including material characterizations, soils studies,
alternative disposal methods studies, and criteria data base development
studies should be used to determine the environmental acceptability of these
disposal conditions for human, animal, marine, and agricultural consumption,
as appropriate. These analyses with few, if any, exceptions should provide
a major and possibly conclusive answer to the criteria necessary for FGC
disposal.
3. 2 FIELD EVALUATIONS
The EPA is currently conducting two field disposal evaluation
projects using treated and untreated sludges, i. e. , the TVA-Aerospace
Shawnee project, which has been in operation for approximately 18 months,
and a planned Louisville Gas and Electric project scheduled to begin in the
summer of 1976. A valuable assessment tool can be made available through
minor modifications at the Shawnee site. That project includes the study of
potential environmental impact conditions of three different commercially
treated sludges disposed under worst-case conditions in which rainwater is
trapped on the surface in such a way that seepage is maximized. Those
studies have already shown that the seepage from treated sludges contain
constituent concentrations, though considerably less in quantity than in
untreated sludge seepage, are initially in excess of drinking water criteria.
It is also known that no seepage will occur if no hydraulic head exists.
It is therefore recommended that after monitoring has been con-
ducted for approximately two years, at which time a reasonable understanding
of those materials will have been attained, at least two of the three treated
disposal ponds should be modified to prevent the occurrence of standing water.
The primary purpose will be to determine whether seepage is prevented, by
monitoring the existing leachate wells, and if it is not totally prevented, then,
the degree of reduction can be assessed. A secondary benefit will be the
determination of suspended solids in the runoff. Dissolved solids in the run-
off from such sites are not expected to be a major problem, but this factor
would be determined in simple laboratory tests because the age of these ponds
precludes a meaningful measurement of dissolved solids.
This modification to the Shawnee project will provide valuable
data regarding the benefit of proper site management.
10
-------
SECTION IV
SUMMARY
A summary highlighting results of the total effort to date is
presented in this section. Brief discussions of each area of investigation are
given, and selected figures and tables are included to either simplify the
summary or to provide the data necessary to complete the discussions. Fig-
ures and tables are repeated as appropriate in the following sections when
necessary to complete the detailed technical descriptions therein.
4. 1 DISPOSAL CRITERIA
The concern in this study regarding disposal criteria is (a) for the
quality of groundwater or surface waters after the introduction of flue gas de-
sulfurization (FGD) waste liquors by means of seepage or runoff, and (b) for
the structural quality of the waste materials if the disposal site is to be
reclaimed. General conditions of wastes being studied for disposal in ponds
or landfills include untreated wastes, dewatered and compacted wastes, gyp-
sum, and wastes fixed by chemical processing. Systems that have a super-
nate or underdrainage operate in a closed loop, thereby preventing a direct
discharge problem. The environmental effects to be considered, then, are
leachate migration to groundwater in all cases, runoff from fixed wastes in
landfills, and strength of all fixed and all conditioned (e. g. , dewatered)
wastes. One exception is a fixed material placed upstream of a dam from
which surface water (supernate and rainfall) is occasionally discharged to a
stream.
No federal criteria apply specifically to these disposal conditions;
however, at this time, the EPA is conducting a study to establish a data base
for the development of standards for the disposal of flue gas cleaning (FGC)
wastes (3). Eventually, federal or state regulations will apply as a result of
the "Resource Conservation and Recovery Act of 1976, " Public Law 94-580,
Zl October 1976 (13). Because of the wide variations in the characteristics
of wastes, weather, soils, topography, and groundwater from site to site,
permits are currently being written and applied on a site-specific basis. At
this time, it appears that local regulations attempt to attain a zero pollution
discharge and that all FGC wastes must be either fixed or impounded in an
impermeable basin. The federal regulation that most nearly relates to a
limit on seepage water quality is the EPA "Alternative Waste Management
Techniques for Best Practicable Waste Treatment," issued in the Federal
Register (14). However, these criteria apply to public-owned treatment and
land application of waste water and not to disposal of FGC wastes.
11
-------
These federal criteria define that the groundwater, resulting
from land applications of waste water, shall be limited to the maximum con-
taminant levels contained in the "National Interim Primary Drinking Water
Regulations" (NIPDWR) (12) or to the existing concentration, whichever is
greater, "if the groundwater criteria should be established by the Regional
Administrator ... . " The NIPDWRs apply, however, to the concentrations
at the consumer tap and are similar to the United States Public Health
Service (USPHS) 1962 standards (15), but in contrast to USPHS no limits are
given in the NIPDWR for sulfate or chloride. Because the concentration of
constituents within any leachate seepage that may occur from untreated and
treated wastes (during the filling period and early stages of seepage, there-
after) exceeds drinking water criteria (Table 4) it appears that some form of
attenuation or prevention is needed. Possible approaches, either individu-
ally or in combination, are as follows:
a. Eliminate seepage by using lined basins.
b. Reduce seepage by using lined basins.
c. Attenuate constituents via cation exchange and adsorption in the
soil.
d. Provide a means of mixing seepage with the groundwater.
e. Provide a means of mixing seepage and groundwater with con-
necting streams.
f. Chemically reduce solubility and permeability and provide a
means of mixing seepage with groundwater.
g. Chemically reduce solubility and permeability and reduce or
eliminate seepage via runoff of rainwater.
Additionally, certain structural qualities must be attained if the site is to be
reclaimed. In the absence of specific criteria, a selection in this study of
25-psi minimum unconfined compressive strength has been made.
Currently, studies are continuing in this program, other EPA
projects, and in industry to define and evaluate environmentally sound, least-
cost methods of waste disposal, covering the range of (a) ponding; (b) dewater-
ing, mixing with fly ash and compacting; (c) conversion to high solids content
gypsum; and (d) fixation. Data from these studies are expected to provide a
data base from which an appropriate disposal mode can be selected for any
given site.
4. 2 PHYSICAL PROPERTIES
The physical properties measured in this study for the charac-
terization of FGD sludges include the wet bulk densities of the sludges as a
12
-------
function of dewatering processes, the coefficient of permeability, the
viscosity as a function of water content, the compactibility, and the uncon-
fined compressive strength as a function of moisture content. Further char-
acterization includes the identification of the crystalline phases in each sludge,
the relative quantity of each phase, and the morphological description of each
phase. In addition, sludge samples chemically treated by commercial pro-
cessors are evaluated with respect to their coefficient of permeability and
unconfined compressive strength. An interpretation of the resultant behavior
of all experimental tests performed is made relative to the physical or chemi-
cal characteristics of each sludge, and an evaluation is made of these behav-
iorial effects on the requirements for environmentally safe disposal.
4. 2. 1 Wet Bulk Density
The dewatering characteristics of FGD sludges are important to
the various disposal techniques in that they affect the volume of the disposal
basin, the methods of handling the sludges, and the condition of the sludges
in their state of ultimate disposal. All of these variables affect the economies
of disposal as well as the pollution potential of the resultant waste. The
ability of a sludge to be dewatered is primarily a function of the size and dis-
tribution of particles and the morphology of the particles determined by crys-
talline phase. In this study, four methods of dewatering were investigated in
the laboratory: settling, settling by free drainage, vacuum filtration, and
centrifugation. These methods are those most often used or considered
because of their relative cost effectiveness.
The results of this study reveal that all dewatering behavior
responds to a relationship that relates the wet bulk density Pg of a material
to the weight fraction fs, the true density ps of the solid phase, and the
density of water pw, by the equation:
P P
s w
P
B pwf + p (1 - £ )
This relationship corresponds to a condition in which the water content of the
dewatered sludge exceeds the amount just necessary to fill the void spaces
when the particles are closely packed together. The dewatering method and
the characteristics of each sludge establish particle packing that can be
determined by measuring the dry bulk density of the sludge when all water
has been dried from the sample. If water is incrementally added to the dried
sample, the wet bulk density will increase relative to the dry bulk density PD,
by the equation:
PB -
13
-------
When sufficient water is added to just fill all void spaces between particles, p,
the maximum density for that sludge, is reached and is defined as the coinci-
dence of the two equations.
If the dewatering efficiency of each method is defined as the per-
cent of the measured wet bulk density relative to the maximum density attain-
able, a means of comparing the efficiency of each method is available. In all
cases, the highest dewatering efficiency was obtained by vacuum filtration
(93. 6 average percent), and the lowest efficiency was by centrifugation (81.9
average percent); settling (86. 2 average percent) and free drainage (87. 2
average percent) efficiencies lie in between. However, when actual wet bulk
densities are compared as in Table 2, it is seen that there is little difference
between vacuum filtration and centrifugation and, in some cases, centrifuga-
tion produces the most dense sludge among the four methods being evaluated.
The discrepancy is explained by the difference in particle packing
that results from each method. Centrifugation is a high-force method that
packs particles closely together, but when performed in a glass vial water is
not efficiently removed from between the particles. Vacuum filtration is a
method that uses considerably less force and may result in poor particle
packing relative to the applied force. The vacuum-assist is most effective in
removing the water from between the particles, and this is the primary source
of the high dewatering efficiency. Settling with or without drainage is a low-
force method, and particle packing .is generally poorest. The difference
between these two methods is a consequence of the loss of some water by
drainage.
When industrial filters and centrifuges are used rather than labor-
atory equipment, filtration is generally slightly less effective with a resultant
decrease in dewatering efficiency. On the other hand, centrifugation is usu-
ally less effective in particle packing and more effective in removing water
because of a difference in mechanism for separating water from solids.
Thus, in field conditions, the efficiency of dewatering by these two methods
is more nearly equivalent, with a general tendency for centrifugation to pro-
vide more consistent results. Settling methods are always less effective than
either filtration or centrifugation, but in many cases their use constitutes the
most cost-effective dewatering method.
4. 2. 2 Coefficient of Permeability
The physical parameter that most significantly affects the pollu-
tion potential of a sludge is the permeability of the sludge to rainwater since
this parameter governs the amount of leaching water passing through the
sludge. Disposal techniques are developed that can minimize the recharge
of rainwater to subsoils, but in all cases the permeability coefficient defines
the maximum limit to the amount of liquid that can pass through the sludge.
The coefficient of permeability was measured on both untreated
sludges and sludge chemically treated by several processors. Permeation
columns were constructed by generally accepted methods (Section 6. 5. 1), and
14
-------
TABLE 2. WET BULK DENSITIES OF FGD WASTE AS A FUNCTION OF DEWATERING
METHOD
Sample
Source
and
Date
Shawnee
Limestone,
2/1/73
Shawnee
Limestone,
6/15/74
Shawnee
Lime,
3/19/74
CM
Double Alkali,
7/18/74
Utah
Double Alkali,
8/9/74
Duquesnc
Lime,
6/17/74
Cholla
Limestone,
9/1/74
Mohave
Limestone,
3/30/73
Dcwatering Method
Settled
Percent
Solids
49.0
52.9
41.5
40.0
37.2
47.6
46.7
66.6
Density,
g/cc
1 .45
1.46
1.34
1.31
1.30
1.40
1.39
1.65
Settled and
Drained
Percent
Solids
55.7
58.3
43.4
43.9
41.4
53. 1
50.9
67.2
Density,
g/cc
1. 51
1.53
1.36
1 . 35
1.33
1.48
1.44
1.67
Centrifuge
Percent
Solids
59.8
63.3
49.9
50.9
62.2
57.2
60.9
77.0
Density,
g/cc
1.56
1.60
1.44
1.43
1.62
1.52
1.58
1.86
Filter
Percent
Solids
65.0
65.9
56.0
57.8
54.6
57.0
53.4
80.3
Density,
g/cc
1.65
1.64
1. 51
1. 52
1. 50
1.52
1.48
1.78
-------
the rate of water passage through the columns was measured with time and as
a function of hydraulic head. In some cases, packed sludge columns were
compacted to greater particle packing densities, and the measurements were
then repeated. The measured permeability coefficients are plotted as a
function of void volume in Figure 1 .
The results of these measurements show that the permeability
coefficients of all the untreated sludges are approximately 1 x 10~4 cm/sec,
and generally between 2 x 10" and 5 x 10 cm/sec, as shown in the shading
in Figure 1. These results also show that for any specific sludge, the
expected relationship of decreasing permeability with decreasing void volume
appeared. A correlation with particle size as well as particle packing frac-
tion explained the permeability behavior of FGD sludges, but exceptions to
the general observation prevent establishing a direct correlation.
Chemically treated sludges are plotted in Figure 1 as specific
datum points, and it is seen that the permeability of these materials is nearly
identical to the permeability of untreated sludge. It must be noted, however,
that all of these measurements were conducted on treated sludges that were
pulverized in order to pack the permeation column. In this condition, the
columns represented material that has been placed and cured, and subse-
quently moved and replaced. Although this method of disposal is intrinsically
expensive, it is and has been practiced at several power plants and represents
a state-of-the-art technique.
Alternatively, some treated sludges were measured in a solid
condition that precluded fracturing at any time. The values of permeability
measurements were as large as 10"' to 10"° cm/sec. However, permeabil-
ity measurements of sludges treated under field conditions (2) have shown
that the permeabilities of the treated sludges range from 10"^ to 10"' cm/sec.
4. 2. 3 Viscosity
The viscosity of FGD waste slurries is a direct measure of its
pumpability. Thus it affects handling procedures, transporting, and waste
disposal methods. Moreover, basic system design considerations are
affected by the relative ease of pumping the waste to desired .locations. FGD
wastes contain finely divided particulate matter suspended in a water system.
This particulate matter tends to range from colloidal size to about 100 jim
and consists of three major phases having markedly different morphologies.
It is both the particle size distributions and phase morphologies that are
believed to influence the viscosity of the sludges. Measurement procedures
are described in Section 6. 2. 2. 1.
Viscosity measurements were performed on sludges at varying
water contents in ambient temperatures with a commercial viscometer having
a measuring range from 3 to 150 poise. Measurement procedures are
16
-------
ID'3
1
E
o
sf
F PERMEABILI
o
jb
O
l
5
0
j£
ID'5
\
\
\
\
v V
;f vi
\
..... o* \
\
\ y
\
Wf. POND\
: ~1
_
POND
E
5 x 10"? -1-
5.5xlO"8j-
^-T-^ APPROXIMATE RANGE
iU FOR UNTREATED MATERIALS
* PULVERIZED
n: FAMILIES OF DATA HAVING SAME SLOPES
POND B
POND C
POND E
TVA SHA'MEE/DRAVO (Ref. 2)
TVA SHAWNEE/IUCS (Ref. 2)
TVA SHAWNEE/CHEMFIX (Ref. 2)
SAMPLE SOURCES DATE
TVA SHAWNEE LIMESTONE 6/15/74
TVA SHAWNEE LIME 3/15/74
SCE MOHAVE LIMESTONE 3/30/73
GM PARMA DOUBLE ALKALI 7/18/74
UPL GADSBY DOUBLE ALKALI 8/9/74
DLC PHILLIPS LIME 6/17/74
APS CHOLLA LIMESTONE 4/1/74
CHEMFIX LAB A IUCS LAB
DRAVO LAB
0.20
0.30
0.40 0.50
VOLUME FRACTION OF SOLIDS
0.60
0.70
Figure 1. Permeability of treated and untreated sludge: Aerospace analyses.
-------
described in Section 6. 3. 1. The results of the viscosity tests are presented
in Figure 2 for the seven sludges measured and reveal that easily pumpable
mixtures (< 20 poise) range from a high solids content of 70 percent to a low
solids content of 32 percent. In addition to the position of each viscosity
curve on the graph, these results further reveal that the slope of the viscosity
curves can be separated into two groups.
120
100
o
O-
1/1
o
o
I/I
^ 60
40
20
CURVE SLUDGE
GM PARMA DOUBLE ALKALI
UPL GADSBY DOUBLE ALKALI
TVA SHAWIMEE LIME
DLC PHILLIPS
TVA SHAWNEE LIMESTONE
TVA SHAWNEE LIMESTONE
TVA SHAWNEE LIMESTONE
SCE MOHAVE LIMESTONE
APS CHOLLA LIMESTONE
2 4
FLY ASH.%
7.4
8.6
40.5
59.7
20.1
40.1
40.9
3.0
58.7
30
40 50
SOLIDS CONTENT, WEIGHT %
60
70
Figure 2. Viscosity of desulfurization sludge.
18
-------
Whereas particle shape, particularly the platey sulfite particles
have been credited with the change in the rheological nature of sludge, in the
viscous-fluid behavior of these sludges it is not apparent that the sulfite phase
plays a decisive role. On the other hand, these data clearly suggest that fly
ash increases the fluidity of a sludge and high pH may increase its viscosity.
Whereas particle shape, size, and distribution appear to influence viscosity
behavior, the precise effect each may have is not clear from these limited
data.
4. 2. 4 Compaction
The compaction capability of FGD sludges can be an economic
asset by providing a means of increasing the mass of waste disposed of
within a specified volume. Moreover, if compacted sludge behaves like fly
ash, significantly reduced permeabilities and increased strength can be
expected, and these may increase the environmental acceptability of the
sludge. The compactability of a sludge is dependent upon particles rearrang-
ing under an applied force by sliding past each other. This process is
enhanced by the presence of some pore water, which acts as a lubricant, but
there must also exist air voids into which the particles can move. The parti-
cle size distribution and crystalline morphology are the two most important
parameters influencing the compactability of FGD sludges.
The same nonvibratory compaction test was performed on each of
six FGD sludges at nominally 100 psi. Dried sludge samples were loaded
into pellet dies for the tests, and the change in sample heights were noted.
While constrained by the die, the samples deflected as much as 15 percent
under the load; however, when removed from the die, the actual compaction
ranged between 1 and 4 percent as shown in Figure 3. In contrast, related
studies on fly ash have been shown to compact from 7 to 20 percent (16).
The lack of compaction for FGD sludges was explained by the
crystal structure of the calcium sulfite platelets and the interwoven lath-like
structure of the calcium sulfate crystals. In many cases, the presence of
water in the pores increased the deflection under load but contributed little
or none to the permanent deformation. The absence of significant permanent
deformation resulted in no measurable increase in strength and in a limited
decrease in permeability to water percolation. These tests suggest that,
while some benefit may be gained by compaction, it is not likely to be a cost-
effective operation in untreated sludge disposal.
4. 2. 5 Compressive Strength
Strength measured in an unconfined condition was compared
with unconfined strength data of soils. These tests provide a means of evalu-
ating the suitability of FGD wastes in landfill and structural fill situations.
Unconfined compression tests were conducted by standard test
methods for treated and untreated FGD wastes. The results of these tests
are summarized in Table 3 for samples tested in a dry condition (no free
19
-------
1.4
1.3
"i
E
o
CT>
>-- 1.2
.
UJ
O
1.1
1.0
0.9
0.8
FLY ASH
(Ref. 16
APS CHOLLA
TVA SHAWNEE LIMESTONE
TVA SHAWNEE LIME
_
UPL GADSBY
60
70 80
SOLIDS CONTENT, WEIGHT %
90
100
Figure 3. Compaction of FGD sludge (dry basis).
water), a damp condition (partial pore water), and wet (as placed operation-
ally). The results indicate that, on an overall basis, untreated FGD wastes
in a damp condition (~8% water) are stronger than samples that are dried of
all water. Nevertheless, these data indicate that untreated sludges, having a
solids content of approximately 85 percent or greater, have unconfined com-
pressive strengths similar to those of natural soils.
Treated FGD wastes have strengths that are capable of structural
landfill applications. The data do not show an appreciable difference between
the strengths in the dry or wet conditions.
4. 3
CHEMICAL CHARACTERIZATION
Chemical analyses from all sampling points from lime, limestone,
and double alkali scrubber waste streams are summarized in Table 4, where
the range of concentrations for each chemical species is given. The
20
-------
TABLE 3. UNCONFINED COMPRESSIVE STRENGTH TEST
RESULTS SUMMARY
FGD Waste
Untreated
Treated
Range of Water
Content, %
5.6 - 14.4
0. 0
>30a
2.4 - 5. 6
0. 0
(35 - 60)*
Strength
Range, Ton/ft2
0. 9 - 2. 4
0. 6 - 1. 5
0. 0
25. 4 - 48. 5
22. 3-52.3
1. 8 - 39
Typical conditions in-place.
distribution of trace elements in system liquors tends to lie between 0.01 and
1 mglS. for all elements except mercury, which has a concentration distribu-
tion about one-tenth that of other trace elements. In the solids portion, the
trace element distribution is approximately two orders of magnitude greater
than the liquor.
Major chemical species concentrations in liquors depend strongly
on the scrubber system parameters and tend to have total dissolved solids
content of about 10, 000 mg/j(! except during startup and in selected exceptions.
The chloride concentration in the liquor depends primarily on the chloride
content of the coal. Major chemical species content in the solids also depends
upon system parameters, primarily oxidation conditions and fly ash collec-
tion methods.
The crystalline phases present in the solids portion of each FGD
waste are presented in Table 5. These data reveal that gypsum and calcium
sulfite hemihydrate are the principal sulfur products and that a broad range
of fly ash contents (3 to 60 percent in this study) might be expected from a
survey which includes both eastern and western coal, with either separate or
simultaneous fly ash collection. Several soluble phases were present as a
consequence of salt formation during drying of occluded water. The presence
of limestone in all samples is a consequence of both unreacted limestone
absorbent and carbonate formation by adsorption of carbon dioxide from the
atmosphere.
The chemical pollutants in FGD sludges originate from process
ingredients, coal, combustion products, absorbents, and process makeup
water. The chemical characterization of FGD sludges included analytical
measurements of both the solid and liquid fractions of sludge sampled at
various positions along the scrubber circuits. Process ingredients from
which the chemical pollutants originate were also analyzed in order to provide
21
-------
TABLE 4. CHEMICAL CONSTITUENTS IN LEACHATE FROM
UNTREATED AND CHEMICALLY TREATED FGD
WASTE AFTER 1 AND 50 PVD (CONCENTRATIONS
IN
Chemical
Constituent
Arsenic
Beryllium
Cadmium
Chromium
Calcium
Lead
Mercury
Selenium
Zinc
Chloride
Fluoride
Sulfite
Sulfate
Total Dissolved
Solids
Chemical Oxy-
gen Demand
pH
Untreated Waste
1st PVD
0. 02
0. 01
0. 015
0.045
0.045
0. 25
0. 0005
0.055
0. 65
2600
4. 0
50
6500
10, 000
10
6.6
50th PVD
0. 004
<0. 004
0. 002
<0. 003
0. 010
0. 010
<0. 00005
0. 006
0. 04
100
0.8
30
1200
2200
6
5.5
Treated Waste
1st PVD
0. 015
0. 008
0. 02
0. 06
0. 025
0. 15
0. 0007
0. 040
0.27
600
1. 1
35
3500
6000
7
6.8
50th PVD
<0. 004
<0. 002
0. 002
0.003
0. 005
0. 03
0. 0002
0.008
0. 03
75
0. 4
20
450
1000
4
8.0
22
-------
TABLE 5. PHASE COMPOSITION OF FGD WASTE SOLIDS IN WEIGHT PERCENT
Atomic
Formula
CaSO4'2H2O
CaSCy 1/2H2O
CaSCy 1/2H20
CaCOj
MgSO4'6H2O
Na2SO4-7H2O
NaCl
CaS04a
Fly Ash
Total
TVA Shawnee
Limestone,
2/1/73
21.9
18.5
38.7
4.6
20. 1
103.8
TVA Shawnee
Limestone,
7/12/73
, 15.4
21.4
20.2
3.7
40.9
101.6
TVA Shawnee
Limestone,
6/15/74
31.2
21.8
4.5
1.9
40. 1
99.5
TVA Shawnee
Lime,
3/19/74
6.3
48.8
2.5
1.9
/
40. 5
100.0
SCE Mohave
Limestone,
3/30/73
84.6
8.0
6.3
1. 5
3.0
103.4
GM Parma
Double Alkali,
7/17/74
48.3
12.9
19.2
7.7
6.9
7.4
101.4
APS Cholla
Limestone,
4/1/74
17.3
10.8
2.5
58.7
89.3
DLC Phillips
Lime,
6/17/74
19.0
12.9
0.2
59.7
91.8
UPL Gad 9 by
Double Alkali,
8/9/74
63.8
0.2
10.8
17.7
8.6
101. 1
UO
Phase not explicitly measured; presence deduced from x-ray study.
-------
an understanding of the relationship between the source and fate of these
potential pollutants within the FGD scrubbing processes.
4. 3. 1 Generation of Trace Elements in FGD Wastes
Chemical analyses of FGD sludges identified trace elements in both
the solid and liquid fractions of the sludges (Table 4). When the specific
analyses are plotted against the analyses for coal (corrected for other pro-
cess ingredients), a relationship is shown, as in Figure 4 for solids and
Figure 5 for liquids (see Section VII, Figures 27 and 28 for details). This
relationship shows the direct relationship between the trace elements in sludge
and those in coal. In general, t-hese data indicate that approximately one per-
cent of the trace elements in the sludge is distributed in the liquid fraction.
The correlation that exists between the trace element content in
coal and the trace element content in fly ash further suggests that fly ash is the
principal source of these trace elements in the sludge. Whereas the data base
in this study is not adequate to unequivocally explain fully the source of trace
element content in the sludge, the contribution due to fly ash as described in the
literature sources must be recognized.
Since the direct relationship between trace elements in sludge and
in coal is clearly indicated, the trace metals from western coal, having typi-
cally lower concentrations of arsenic, cadmium, mercury, and zinc than
eastern coal are expected to produce sludges having lower concentrations of
these elements. This behavior was found as expected.
4. 3. 2 Process Variables
The chemistry of the scrubbing liquor is affected by the absorbent
used in each system. It is found that, in general, the trace element content in
the liquor is highest in the limestone systems, intermediate in the lime sys-
tems, and lowest in the double-alkali systems. The only readily recognized
systematic parameter that can affect these differences is the pH within the
scrubber. Among them, the limestone scrubber operates at the lowest pH
and the double-alkali system at the highest pH. When specific systems are
compared on a one-to-one basis, the apparent relationship between scrubber
pH and trace element content is confirmed in every case. This effect is inter-
preted as a consequence of trace elements leaching from fly ash during the most
acid part of the recycle system.
An evaluation was made of the trace element content in the system
liquor at various positions in the scrubber process. Chemical analysis indi-
cated that the system liquor pH increased and trace element content showed a
slight decrease enroute from the scrubber to the disposal site. The results
for trace elements may be interpreted as a response to system pH or a
response to the changes taking place in the concentration of major chemical
species. The in-process analyses showed, for the major species, that a
rapid oxidation of sulfite ion and the precipitation of calcium sulfate also take
place enroute to the disposal site. The trace element content in the liquor may
24
-------
400
100
< 10
o
0.1
0.01
0.1 1 10 100
AVERAGE TRACE Elf WENT CONTENT OF SLUDGE SOLIDS, ppm
1000
Figure 4. Relationship between trace element content in
coal and sludge solids.
100
£
s
10
0.001
0.01 0.1 1.0 10
AVERAGE TRACE ELEMENT CONTENT OF SLUDGE LIQUOR, mqli
100
Figure 5. Relationship between trace element content in
coal and sludge liquors.
25
-------
be decreasing by precipitation in response to decreasing ionic strength, by
coprecipitation resulting from the scavenging action of the calcium sulfate, by
adsorption onto newly created crystal surfaces of the calcium sulfate phase,
or by the pH changes previously discussed.
An evaluation of all chemical analyses of trace elements in the
liquor was made as a function of both pH and ionic strength. No correlation
was found for any element when chemical data from all systems were simul-
taneously compared, and it must be concluded that neither pH nor ionic
strength are primary variables. However, when chemical data were com-
pared within a single system, a pH correlation was found. Thus, it is con-
cluded that the primary source of trace elements is the coal, and these elements
enter the scrubber system primarily by leaching from fly ash in the scrubber.
The pH of the scrubber affects the trace element content in the liquor, but the
pH of the system downstream of the scrubber is shown to have only a minor
effect on this value.
When the chemical species in system liquor are evaluated over a
period of time, the data indicate that all species increased from startup. The
calcium sulfate saturation level is established quickly as is the concentration
of trace elements. Chloride ion concentrations appear to build up at a slower
rate and continue to build up until a steady state is established. As the chlor-
ide ion concentration increases, it effects an attendant increase in ionic
strength that increases the calcium sulfate equilibrium levels. Trace element
contents do not reflect these changes either from a lack of analytical sensi-
tivity at very low concentration, from a lack of accuracy of analytical data,
or from a lack of response to changes in major species concentrations.
4. 3. 3 Chemical Solubility Analysis
A comparison of experimental scrubber liquor compositions with
compositions calculated from the assumption of equilibrium solubility was
made for each analysis presented in Appendix D by using two different com-
puter programs. One program was specifically designed to predict maximum
concentrations of trace chemical species as a function of the concentration of
major species from empirical solubility data. The other program was
designed to predict major species concentrations in a liquor from thermo-
dynamic data, but the trace chemical species concentrations were not
predicted.
When the results of the two programs are compared for major
species, it is shown that the two programs are in virtual agreement with
respect to the saturation of calcium sulfate in every analysis and saturation
of calcium sulfite in only a few cases. Moreover, the agreement between the
two independent approaches relative to the experimental data provides sup-
portive evidence to the accuracy of the experimental measurements.
For the nine trace elements measured, there was no evidence
that trace element saturation was the controlling parameter for the trace
26
-------
metal concentration in the liquor. This finding is consistent with the finding
of an absence of trace element response to changes in major chemical species
within the liquor system. For several trace elements, anomalies exist rela-
tive to the measured versus expected concentrations and require further
experimentation to resolve.
4. 4 ENVIRONMENTAL ACCEPTANCE OF FGD WASTE
An evaluation of the environmental impact of FGD waste was made
as a consequence of the alternative routes of pollution to the environment
and the methods by which the wastes may be disposed of. The accessibility
of pollutants to the environment can be minimized by (a) decreasing the perme-
ability of the wastes to reduce the seepage of water, (b) reducing the leach-
ability of the wastes through a reduction of the solubility of the material, and
(c) managing seepage and runoff to limit the excess of waste constituents to
groundwater or surface waters. Disposal concepts include (a) chemical treat-
ment of waste to affect alternation of its permeability and teachability proper-
ties, (b) total impoundment of untreated or treated wastes to isolate them
from the environment, and (c) several methods of impounding untreated but
dewatered wastes. Many of these concepts produce a. structurally soluble
material which makes site reclamation possible. The assessment of these
methods shows that although intrinsic pollution potential of each method is
different from the other, the protection of the environment is best assured
only through the implementation of appropriate disposal site management
techniques.
4. 4. 1 Pollutant Access to the Environment
An evaluation of the chemical pollutant potential to the environ-
ment by vaporization has shown that this hazard is essentially nonexistent.
Pollution by wind-whipped spray or runoff are determined by the quality of
the supernate and can be as contaminated as system liquor or as benign as
rainwater. In the case of runoff, suspended solids may be a more critical
pollutant than dissolved solids, especially for untreated water.
By far, the most persistent pollution potential is by water perco-
lation through the sludge and subsoil into the subterranean water table. The
severity of the pollution is dependent upon both the quantity and quality of the
leachate from the disposal sites. Leachate quality was determined by labora-
tory experimentation by leaching column tests.
Results from these tests show that the concentration of all chemi-
cal species decreases continually in the leachate, initially at a high rate and
after the displacement of 3 to 5 pore volumes, the rate of decrease is typi-
cally very low or the concentration becomes constant. The initial high rate
of decrease is believed to be a consequence of a flushing mechanism; the con-
stant or low-rate segment appears to be the consequence of the solubility of
crystalline phases.
27
-------
Chemical treatment of a waste by fixation processing produces a
leachate with concentration values of major species nominally one half of
untreated sludge at all equivalent leachate volumes. Highly soluble chemi-
cal species like sodium and chloride ions are removed effectively from the
system in the initial pore flushing. After approximately 5 PVD, the TDS con-
sist primarily of calcium sulfate ions. The results from these tests are pre-
sented for TDS in Figure 6, in which all values are normalized to the initial
concentration of leachate for untreated sludge. The shapes of these leachate
curves are representative of all major chemical species, i.e., calcium,
sodium, chloride, sulfate, and sulfate ions. The chemical oxygen demand is
directly related to sulfite ion concentration. It has shown an initial value as
high as 70 mg/jg and attains a steady-state value at about 7 m.g/1 in anaerobic
leachate. The oxygen demand criteria for fisheries in most states is in the
range of 4 to 6 mg/£.
Chemical treatment did not reveal a discernible difference from
untreated material for trace elements. In most cases, all trace element content
in the leachate dropped below detection levels after the initial flushing. In
the remaining cases, only lead, zinc, and selenium were found to persist at
concentration levels significantly above background, and this finding was
independent of chemical treatment.
4.4.2
Alternative Disposal Techniques
The methods by which FGD waste may be disposed will be deter-
mined by the environmental acceptability of the waste and the cost of waste
1 i-
IS
§0
o >
^ LU
il0-1
12
0.01
I
I
10 20 30 40
PORE VOLUME DISPLACEMENTS
50
60
Figure 6.
Leachate TDS from treated and untreated
sludge as a function of PVD.
28
-------
disposal. In addition, methods already used in fly ash disposal or in the
disposal of other industrial wastes may be used in preference to new or untried
techniques. Thus, sludge sluicing to a pond is considered a viable method but
may require a pond liner in the event that environmental pollution may other-
wise occur. An alternative to ponding is the technique whereby the sludge is
sluiced to the disposal basin, but instead of supernate liquor collection the
disposal basin is underdrained so that excess liquor can be returned to the
scrubber while enhancing structural properties and minimizing seepage by
(a) eliminating the supernate hydraulic head and (b) interrupting the internal
head via the underdrain mechanism (see Section 9. 3. 2 for other effects).
This method may be preferred in some cases with respect to pollution, costs,
and ultimate disposal site disposition.
In the case of chemically treated waste, it is possible to dispose
of the waste in a manner resembling ponding, similar to the Pennsylvania
Power Company's Bruce Mansfield Station disposal operation. Monitoring of
local streams is necessary because of the seepage caused by the constant
hydraulic head of the impounded water.
An alternative method for disposal of untreated waste uses a filter
or centrifuge to provide a material that can be transported and placed by truck
transfer. While transportation costs for this method are intrinsically high,
the method offers advantages that offset trucking costs. Particular attention
to structural stability may be necessary in this method.
Chemical treatment of FGD wastes provides an additional
assurance of environmental acceptance, but the material must be placed in a
disposal site in a manner consistent with sound disposal practices. This
method is the most expensive among the various disposal alternatives but may
be necessary in certain cases where structural stability or environmental
protection might otherwise be compromised.
4. 4. 3 Assessment of the Pollution Potential
On the basis of the laboratory data obtained in this study, an
assessment was made of the relative pollution potential of the various alterna-
tive methods for FGD waste disposal. Of the routes by which pollutants can
enter the environment from a disposal site, it is considered that the most
serious method is by percolation of rainwater through the waste and the con-
tamination of groundwater from the dissolved chemical species carried by
the recharge.
From the quality of leachate and the permeability rate of water
through the waste, the yearly mass loading to the subsoil below each type of
disposal site was determined from the site conditions described in Table 6,
and these data are plotted in Figure 7. Cases 1 and 3 are for untreated FGD
29
-------
TABLE 6. INPUT DATA FOR STUDY CASES
Case
1
2
3
4
5
Disposal
Method
Lake
Lake
Pond
Pond
Landfill
Surface
Water
Constant
supernate
Constant
supernate
10. in. /yr
recharge
10. in. /yr
recharge
1 in. /yr
recharge
FGC Waste, 5-Year Fill
Waste
Condition
Untreated
Treated
Untreated
Treated
Treated
Depth,
ft
30
30
30
30
30
Permeability,
cm/sec^
io-4
.o-5
io-4
io-5
.o-5
Fractional
Pore
Volume
0.67
0.67
0.67
0.67
0.67
Assumed maximum hydraulic head of 6 ft during filling, including depth of
wastes; 1 ft constant water cover thereafter.
For all cases, subsoil permeability = 10 cm/sec.
waste disposed in a pond. Case 1 has a constant hydraulic head, whereas in
Case 3 the water that is available for recharge is only the net from rainfall
after evaporation. Cases 2 and 4 are for treated FGD waste disposed and
are managed as in Cases 1 and 3, respectively. Case 5 is for treated waste
disposed in a landfill in a manner such that only one inch per year is allowed
to recharge because of the disposal site design that maximizes runoff. Cases
similar to Case 5 are also possible for untreated sludge which may or may
not be oxidized to gypsum, dewatered, and placed in a landfill. In these
cases, mass loading may be as great as that in Case 3 or as low as that in
Case 5, depending upon disposal methods. Generally, the mass of TDS that
passes through the subsoil varies by a factor of 100 between ponding untreated
waste and landfilling chemically treated waste.
4. 5
DISPOSAL COST ESTIMATES
Cost estimates for ponding and fixation landfilling have been made
and reported by Aerospace on several occasions. During recent studies
associated with the EPA Shawnee field disposal evaluation project, Aerospace
cost estimates were made of fixation disposal and were reported in the initial
report on that study (2). The Aerospace estimates for lined-pond costs,
which were presented in the initial report on our sludge disposal study (1) and
30
-------
E
u
on
co"
o
CO
o
CO
CO
1.0
0.1
0.01
0.001
CASE 1
CASE 2
A END OF 5th PORE VOLUME
I
I
I
I
20 40
60 80
YEARS
100 120 140 1300
Figure 7. Mass loading of TDS to subsoil for various
disposal modes of treated and untreated
FGC wastes.
at the EPA 1974 Flue Gas Desulfurization Symposium in Atlanta (5), are
updated in this report to January 1976 costs. A summary of these costs
estimates is presented in Table 7, and a discussion of the study is given in
Section 10. 2.
31
-------
TABLE 7. SLUDGE DISPOSAL COST RANGES (UNTREATED AND
CHEMICALLY TREATED LANDFILLS, 1000-MW
STATION. 50 PERCENT LOAD FACTOR, 30-YEAR
AVERAGE, JANUARY 1976 DOLLARS)
Disposal
Method
Untreated
Pond
Pond
Chemi-
cally
Treated^
Base
Material
Natural Clay6
Liner*
Indigenous
Soil
$/Ton Sludgea'b
(Dry)
3.50
5.70-7.80
7. 30-11.40
$/Ton Coal
1. 00
1. 60-2. 20
2. 10-3. 20
Mills /kWhb'C'd
0. 4
0. 7-1. 0
0. 9-1.4
510,000 short tons/year average (dry basis) including fly ash.
bCoal burned at rate of 0. 88 Ib/kWh, 3% sulfur, 12% ash, 85%
removal, 1.2 CaCO3/SO2 mole ratio.
f»
Land costs at $1000/acre are included (equivalent to $0. 25/ton sludge,
dry).
Disposal within 5 miles of power plant.
e 6
Assumes coefficient of permeability of clay is 1 x 10" cm/sec or
better.
Ponding costs cover range based on low-to-high material costs,
i.e., PVC-20 (low) to Hypalon-30 (high).
"Chemical fixation costs vary, depending on characteristics of the waste
and the disposal process chosen.
32
-------
SECTION V
DISPOSAL CRITERIA
This study is directed toward the determination and evaluation of
environmentally acceptable methods for land disposal, with emphasis on
landfilling and ponding. Ocean and mine disposal alternatives are being in-
vestigated in other EPA-sponsored studies (3) using applicable materials
characterization data from this and other programs. Therefore, disposal
criteria applicable to this study are restricted to the quality of groundwater
and surface water affected by flue gas desulfurization (FGD) waste waters by
means of seepage or runoff and to the structural quality of the waste material
as it applies to subsequent disposal site reclamation. Currently, there are
no criteria that are specifically applicable to land disposal methods that could
provide direct guidelines to the acceptability of FGD waste disposal methods.
However, at this time, EPA is conducting studies to establish a data base for
the development of appropriate standards for the disposal of these wastes (3).
5. 1 DISPOSAL ALTERNATIVES FOR FGD WASTE
Whereas the lack of appropriate disposal criteria prevents a
direct evaluation of specific disposal methods relative to limiting standards,
it is possible, nevertheless, to evaluate the relative merit of currently used
and proposed methods of disposal. Although many variations of disposal may
be used, the most practiced generalized alternatives currently available can
be summarized as disposal of wastes in (a) ponds, (b) basins, or (c) landfills.
In all cases, closed-loop scrubber systems are assumed (1 1), and direct
discharge of scrubber liquors or untreated wastes to streams is not con-
sidered. Ponding is a method by which the waste is sluiced to the disposal
site and supernate returned to the scrubber. Basin disposal is defined here
as a method of disposal in a closed site where runoff is contained; additionally,
it is presumed that an underdrainage system is incorporated so that excess
waters can be returned to the scrubber. A landfill is defined here as an open
site where runoff is not contained within the site. The general categories of
disposal and the considerations required for environmental control are shown
in Table 8.
5.2 ENVIRONMENTAL EFFECTS OF FGD WASTE
DISPOSAL ALTERNATIVES
5.2.1 Rainwater Seepage
In each case, seepage of rainwater through the waste and eventual
contamination of groundwater pose an environmental concern for all disposal
33
-------
TABLE 8. ENVIRONMENTAL EFFECTS OF
DISPOSAL ALTERNATIVES
Type of
Disposal
Pond
Basin
Landfill
Condition
of Waste
Untreated
or
chemically
fixedb
Untreated
or
conditioned
Conditioned
or
chemically
fixedb
Primary
Drainage
Supernate
Supernate
Under drainage
Runoff
Environmental Effects
Seepage
Yes
Yes
Yes
Yes
Runoff
No
No
No
Yes
Land Reuse
No
Yes
No
Yes
Untreated waste refers to FGD sludges as emitted from primary or
secondary dewatering equipment
Chemically fixed sludges refer to the waste treated by one of several com-
mercial processes that make these wastes suitable for landfill disposal.
CConditioned waste refers to sludge treated by techniques other than chemical
fixation and includes oxidation to gypsum and dewatering by mixing with dry
fly ash or other agents that allow the material to be handled in a manner
similar to that for soils.
methods. Runoff is a potential source of environmental pollution for landfill
sites because, by definition, these sites are open and do not necessarily re-
turn water to the scrubber. Only in the case of ponding is it clear that the
disposal site is not directly amenable to land reclamation efforts, although
even in some of these cases it may be possible upon retirement to air-dry,
cap, and vegetate the site.
Ultimately, federal or state regulations will apply as a result of
the Resource Conservation and Recovery Act, PL 94-580, 21 October 1976
(13), which gives EPA the authority to regulate wastes from air pollution con-
trol devices. In the absence of federal standards for FGD waste disposal,
existing criteria were examined to establish a basis on which appropriate reg-
ulations may be promulgated. The federal regulation that most nearly relates
to a limit on seepage water quality is the EPA's "Alternative Waste Manage-
ment Techniques for Best Practical Waste Treatment" (14). These criteria
apply to public-owned treatment and land application of waste water. Relevant
excerpts from these criteria are as follows:
34
-------
The ground water resulting from the land application of wastewater,
including the affected native ground water, shall meet the following
criteria:
Case I: The ground water can potentially be used for drinking water
supply.
(1) The maximum contaminant levels for inorganic chemicals and
organic chemicals specified in the National Interim Primary Drinking
Water Regulations (40 CFR 141) (Appendix D) for drinking water supply
systems should not be exceeded except as indicated below (see Note 1).
(2) If the existing concentration of a parameter exceeds the maxi-
mum contaminant levels for inorganic chemicals or organic chemicals,
there should not be an increase in the concentration of that parameter
due to land application of wastewater.
Case II: The ground water is used for drinking water supply.
(1) The criteria for Case I should be met,
(2) The maximum microbiological contaminant levels for drinking
water supply systems specified in the National Interim Primary Drink-
ing Water Regulations (40 CFR 141) (Appendix D) should not be ex-
ceeded'in cases where the ground water is used without disinfection
(see Note 1).
Case III: Uses other than drinking water supply.
(1) Ground water criteria should be established by the Regional
Administrator based on the present or potential use of the ground water.
The Regional Administrator in conjunction with the appropriate
State officials and the granted shall determine on a site-by-site basis
the areas in the vicinity of a specific land application site where the
criteria in Case I, II, and III shall apply. Specifically determined
shall be the monitoring requirements appropriate for the project site.
This determination shall be made with the objective of protecting the
ground water for use as a drinking water supply and or other desig-
nated uses as appropriate and preventing irrevocable damage to ground
water. Requirements shall include provisions for monitoring the
effect on the native ground water.
APPENDIX D
GROUND WATER REQUIREMENTS
The following maximum contaminant levels contained in the National
Interim Primary Drinking Water Regulations (40 CFR 141) are re-
printed for convenience and clarity. The National Interim Primary
Drinking Water Regulations were published in final form in the Federal
Register on December 24, 1976. In accordance with the criteria for
best practicable waste treatment. 40 CFR 141 should be consulted in
its entirety when applying the standards contained therein to waste-
water treatment systems employing land application techniques and
land utilization practices.
Maximum contaminant levels for inorganic chemicals. The follow-
ing are the maximum levels of inorganic chemicals other than fluoride:
35
-------
Contaminant:
Level
(milligrams
per liter)
Arsenic 0.05
Barium 1.
Cadmium - 0.010
Chromium 0.05
Lead 0.05
Mercury 0. 002
Nitrate (as N) - 10.
Selentum - -- - 0.01
Silver -- -- 0.05
The maximum contaminant levels for fluoride are:
Temperature Level
degrees Degrees Celsius (milligrams
Fahrenheit1 per liter)
53.7 and below 1 2 and below 2.4
53.8 to 58.3 - 12.1 to 14. 6 2.2
58.4 to 63.8 14. 7 to 17. 6 2.0
63. 9 to 70. 6 17. 7 to 21. 4 1.8
70. 7 to 79. 2 21. 5 to 26. 2 1.6
79. 3 to 90. 5 26. 3 to 30. 5 1.4
1
Annual average of the maximum daily air temperature.
Essentially, these criteria state that the groundwater, resulting
from land applications of waste water, shall be limited to the maximum con-
taminant levels contained in the "National Interim Primary Drinking Water
Regulations" (NIPDWR) (12) or to the existing concentration if the latter is
greater. If the groundwater is to be used for other than a drinking water
supply, "the ground water [sic] criteria should be established by the Regional
Administrator. ..." In contrast to the "United States Public Health Service
[USPHS] Drinking Water Standards, 1962" (15), which limit sulfate and chlo-
ride to 250 mg/£ each, no limits are given in the NIPDWR for sulfate or
chloride; however, these and other substances may be included in secondary
standards, when issued.
It is apparent that water quality criteria are not finalized when it
is considered that (a) sodium, sulfate, chloride, copper, zinc, and other
substances do not appear in the NIPDWR and are undergoing further study
(see following excerpts) and (b) the groundwater resulting from land applica-
tion of waste water is required to meet the criteria for NIPDWR, whereas
the maximum contaminant levels (MCL) specified in the NIPDWR (except for
turbidity) apply to water at the consumer tap. The following excerpts provide
some insight into relevant background considerations in the development of
water criteria.
36
-------
a. Excerpts from NIPDWR (12):
13. Sodium. Several comments suggested the possibility of an MCL
for sodium and the National Drinking Water Advisory Council recom-
mended that consideration be given to monitoring for sodium so that the
public can be informed of the sodium content of available water. These
concerns result from the fact that many persons in the United States
suffer from diseases which are influenced by dietary sodium intake. In
addition, persons may wish to limit their sodium intake for other
reasons. However, EPA has not proposed an MCL for sodium, and the
Advisory Council did not recommend an MCL, because the data available
do not support any particular level for sodium in drinking water, and
because regulation of sodium by an MCL is a relatively inflexible, very
expensive means of dealing with a problem which varies greatly from
person to person.
EPA has requested the National Academy of Sciences to include sodium
in its study of the health effects of inorganic chemicals. In the mean-
time, the Agency recommends that the States institute monitoring pro-
grams for sodium, and that physicians and consumers be informed of
the sodium concentration in public water systems so that they can take
action they may consider appropriate.
14. Sulfate. Comments also were submitted urging the adoption of an
MCL for sulfate. As in the case of sodium, the National Drinking Water
Advisory Council recommended monitoring for sulfate levels, but did
not recommend the adoption of a maximum contaminant level.
The sulfate question is similar to the sodium question in that available
data do not support the establishment of any given level. A relatively
high concentration of sulfate in drinking water has little or no known
effect on regular users of the water, but transients using high sulfate
water sometimes experience a laxative effect. Whether this effect will
occur, and its severity, varies greatly with such factors as the level of
sulfate in the water being consumed and the level of sulfate to which the
transient is accustomed. EPA recommends that States institute moni-
toring programs for sulfates, and that transients be notified if the
sulfate content of the water is high. Such notification should include an
assessment of the possible physiological effects of consumption of the
water.
The National Academy of Sciences has been asked to consider sulfate
in its study. An MCL for sulfate will be proposed if it is supported by
c'ne available data.
POINT OF MEASUREMENT
Other comments on maximum contaminant levels focused on the pro-
posed requirement that such levels be tested at the consumer's tap.
Concern was expressed over the inability of the public water system to
control potential sources of contaminants which are under the control
of the consumer.
The promulgated definition of "maximum contaminant level, "
§ 141. 2(d), retains the requirement that the maximum contaminant
level be measured at the tap except in the case of turbidity, which
should be measured at the point of entry to the distribution system.
However, the definition has been expanded to make clear that contami-
nants added to the water by circumstances under the control of the con-
sumer are not the responsibility of the supplier of water, unless the
contaminants result from corrosion of piping and plumbing resulting
from the quality of the water supplied. It should be noted, however,
that this requirement should not be interpreted as to discourage local,
aggressive cross connection control measures.
37
-------
b. Excerpt from McKee and Wolf (17):
CHLORIDES
a. Domestic Water Supplies. Chlorides in drinking water are
generally not harmful to human beings until high concentrations are
reached, although chlorides may be injurious to some people suffering
from diseases of heart or kidneys. Restrictions on chloride concentra-
tions in drinking water are generally based on palatability requirements
rather than on health.
Chlorides in water may impart a salty taste at concentrations as low
as 100 mg/jj , although in some waters 700 mg/i may not be noticeable
(see also Tastes). According to the Kettering Laboratory, the most
taste-sensitive people can detect chloride from calcium salts at 96 mg/jf
and from sodium salts at 121 mg/Jt (based on work by Whipple). For
average individuals, the taste threshold is about 400 mg/jj . Lockhart
et al. reported the taste thresholds of NaCl KC1, and CaCl2 to be 345,
650, and 347 mg/jf, corresponding to chloride concentrations of 210,
310, and 222 mg/jf respectively. The fact that a taste is detectable,
however, does not mean that it is objectionable; indeed, many people
prefer water that is partly mineralized to the flat taste of distilled
water.
The tolerance of chlorides by human beings varies with climate and
exertion, and chlorides lost through perspiration may be replaced by
chlorides in either the diet or drinking water. From hot dry areas have
come reports that chloride concentrations up to almost 900 mg/jt have
not been harmful, and Rudolfs maintains that even 1000 mg/jt is harm-
less. According to a Kettering Laboratory survey, laxative effects of
chloride are generally apparent only when an individual has been accus-
tomed to a lower concentration of the salt. Within a few days, an indi-
vidual adjusts physiologically to a higher level of salt intake. In
general, it is the cation (calcium, magnesium, sodium, or potassium)
associated with the chloride that produces a harmful effect. It has been
reported in Russia that under conditions of artificially induced thirst,
test subjects drank 40-60 percent less water containing 1000 mg/jf of
chloride than they did of tap water.
Essentially, these excerpts as applied to land application of waste
water state that for those chemical species which FGD waste is most likely to
generate as a pollutant, i. e. , sodium, sulfate, and chloride, there currently
exist no primary maximum contamination levels, but suggest that such limits
may be subsequently proposed, probably as secondary standards, if they can
be supported by the available data. It is presumed that similar criteria will
be promulgated for control of FGD waste waters and that sodium, sulfate, and
chlorides will be subject to similar considerations. In the absence of federal
regulations for FGD waste, it may be prudent to anticipate that at least one of
these species will be regulated and that, in controlling the maximum contami-
nant level of one, the other two are likely to be similarly controlled.
Because sludge is not discharged to streams and since chemical
oxygen demand (COD) is significant only for fresh liquor (Table 9), it is not
considered here. One of the difficulties in applying existing water standards to
FGD waste waters is that they apply to the quality of the groundwater or tap
water and do not directly define the quality of seepage waters from a FGD waste
38
-------
TABLE 9. CHEMICAL CONSTITUENTS IN LEACHATE FROM
UNTREATED AND CHEMICALLY TREATED FGD
WASTE AFTER 1 and 50 PORE PVD (CONCENTRA-
TIONS IN mg/l)
Chemical
Constituent
Arsenic
Beryllium
Cadmium
Chromium
Calcium
Lead
Mercury
Selenium
Zinc
Chloride
Fluoride
Sulfite
Sulfate
Total Dissolved
Solids
Chemical Oxy-
gen Demand
pH
Untreated Waste
1st PVD
0. 02
0. 01
0. 015
0. 045
0. 045
0. 25
0. 0005
0. 055
0. 65
2600
4. 0
50
6500
10, 000
10
6.6
50th PVD
0.004
<0. 004
0. 002
<0. 003
0. 010
0. 010
<0. 00005
0. 006
0. 04
100
0. 8
30
1200
2200
6
5.5
Treated Waste
1st PVD
0.015
0. 008
0. 02
0. 06
0. 025
0. 15
0. 0007
0. 040
0. 27
600
1. 1
35
3500
6000
7
6.8
50th PVD
<0. 004
<0. 002
0. 002
0. 003
0. 005
0.03
0. 0002
0. 008
0. 03
75
0.4
20
450
1000
4
8.0 _
-------
disposal site. Nevertheless, possible approaches to meet existing standards,
either individually or in combination, are as follows:
a. Eliminate seepage by using impermeable liners or by collecting
and recycling all seepage.
b. Account for attenuation via adsorption or cation exchange in
underlying soil.
c. Force-mix seepage with groundwater for dilution.
d. Mix seepage and groundwater with connecting streams.
e. Chemically reduce solubility and/or permeability and mix with
groundwater.
f. Chemically reduce solubility and/or permeability and reduce or
eliminate seepage by maximizing runoff.
The selection of techniques to accommodate anticipated standards
for groundwater quality as a consequence of FGD waste seepage will depend
on climatological, geological, and hydrological considerations of the disposal
site as well as the quality of leachate that emits from the site. Typical con-
centrations of leachate from untreated and chemically treated sludges are
given in Table 9. Total mass loading to groundwater as determined from
concentration and permeability measurements will determine the impact on
the groundwater. Permeability coefficients can vary over several orders of
magnitude and depend on the specific properties of sludges and the methods
by which the sludge is placed into the disposal site.
5. 2. 2 Landfill Runoff
The previous discussion has addressed the prevention of pollution
from rainwater seepage through a disposal site. For the landfill method of
disposal, runoff is a major concern for environmental control. For those
cases where runoff is allowed to inundate adjacent flatland through which it
may permeate, the regulations promulgated for seepage would be applicable.
For other cases where runoff is directed to receiving streams, the most
applicable regulations would be state stream standards. Thus, where the
receiving stream size is small relative to water supplied to the stream by
landfill funoff, the prevention of stream contamination may be a primary
environmental concern. As in the case of seepage, the method selected for
the prevention of pollution will be determined by the quality of runoff waters.
Possible approaches for the prevention of stream contamination are as follows:
a. Collect all runoff in lined siltation ponds.
b. Cover and vegetate landfill so that runoff water quality is not
contaminated by the FGD waste.
c. Force-mix runoff waters with stream waters.
40
-------
d. Divert runoff to flLatland where evaporation and seepage can take
place.
e. Dilute landfill runoff with runoff from other sources before emit-
ting to streams.
Selection of any of these alternatives will be dependent upon site-
specific considerations.
5.2.3 Land Reuse
In addition to chemical contamination prevention, environmental
concern also includes the attainment of certain structural qualities if the site
is to be reclaimed. In the absence of specific criteria, a conservative value
of 1. 8 kg/cm^ (25 psi) minimum unconfined compressive strength was con-
sidered because it could be met on the basis of the following observations:
a. Heavy equipment on tires carrying 45 psi have repeatedly driven
onto a saturated, fixed pond at the TVA Shawnee field evaluation
site (2), maintained traction, and left only a mild track in the
material. This material has an unconfined compressive strength
of 27 psi in a wet condition.
b. All chemically treated wastes studied attained and usually
exceeded this strength.
c. Sulfite wastes oxidized to gypsum attain this strength at a solids
content of about 65 percent.
d. Sulfite wastes attain this strength at a solids content of about
70 percent.
e. The standard test for unconfined compressive strength is simple
and inexpensive.
Studies are continuing in this program to relate disposal site strength to con-
fined compressive strength criteria and to include untreated sludges that
attain a degree of strength through dewatering.
5.3 SELECTION OF DISPOSAL CRITERIA
It has been demonstrated that disposal methods exist by which
anticipated chemical and structural criteria can be met. Simply stated, if
reclamation of the land is not required, ponding in a naturally impermeable
or lined basin is adequate. If structural characteristics are required,
chemical fixation of the material and placement in an impermeable basin
would certainly provide adequate strength and would not allow seepage to
groundwaters. Moreover, a lined siltation pond can assure environmental
41
-------
protection from any runoff waters. However, these solutions may be overly
conservative and incur costs far beyond that necessary to meet expected
criteria.
Demonstrations, both small-scale and operational, of fixed and
conditional waste landfills (without liners) are now in progress and should
verify the environmental acceptability of such sites when managed to prevent
the collection of water on the surface and, thereby, the 'elimination of hydrau-
lic heads and subsequent seepage. Runoff is collected in a siltation pond
where suspended solids settle and are periodically dredged and returned to
the fixation process. One operational fixed waste disposal site allows surface
water. Its environmental acceptability is partially based upon the mixing of
supernate with a nearby stream and a connecting river. All sites are moni-
tored for water quality.
It is apparent that each disposal site and the material placed in it
have individual characteristics different from most others. These include
waste material properties, weather, topography, soil characteristics, and
nearby stream quality and flow characteristics. Therefore, the disposal
method chosen for any site will generally be selected on site-specific condi-
tions. Because of this, the establishment of a single criterion for all cases
may be impractical.
At this time, studies are continuing in this program, other EPA
projects, and in industry to define and evaluate environmentally sound, least-
cost methods of waste disposal, covering the range of ponding; dewatering,
mixing with fly ash and compacting; conversion to gypsum and dewatering;
and chemical fixation. Data from these studies are expected to provide a
base from which an appropriate disposal mode may be selected for any given
site. In addition, the EPA is reviewing local regulations on a national scale
to identify applicable regulations and to determine whether they will permit
the disposal of treated and untreated flue gas cleaning (FGC) wastes. Because
of the wide variations in the characteristics of wastes, weather, soils, topo-
graphy, and groundwater from site to site, permits are currently being
awarded on a site-specific basis. Eventually, state regulations will apply as
a result of the "Resources Conservation and Recovery Act of 1976" (13), but
these regulations will not be enacted until federal standards are promulgated.
A detailed discussion of disposal and environmental relationships
and the methodology of alternative disposal techniques is given in Section IX.
42
-------
SECTION VI
PHYSICAL PROPERTIES DETERMINATION
6. 1 BACKGROUND
The disposal, handling, and transportation techniques that will be
applied to flue gas desulfurization (FGD) wastes will be strongly dependent
upon the physical behavior of these wastes and the resultant costs. The
physical properties of FGD wastes that limit or restrict the applicability of
certain techniques include the degree of solids content, crystalline phase
composition, particle size, and distribution. Experimental tests were con-
ducted to characterize the FGD wastes of seven power plant scrubbing facili-
ties, v including pilot, prototype, and full-scale scrubber units ranging from
1 - to 125-MW equivalent capacities. In each case, the FGD waste studied
was the waste material for each facility, received in its normal state of dis-
posal. Although recognition was given to the potential that some waste
materials may be suitable for use in specific commercial applications (1),
the tests were directed toward the use of FGD wastes in landfilling and land
reclamation applications. The potential for commercial application of FGD
waste at present represents a limited quantity of material, and the required
knowledge of specific physical properties is strongly dependent upon the
particular application. In contrast, the major applications for these wastes
are in landfilling and land reclamation, where the physical properties
required are well known.
The technologies for transporting, handling, and disposal of fly
ash have evolved over many years of electrical power production; to the
extent possible, it appears that these technologies will be extended to FGD
wastes as well. Two important differences exist, however: (a) The quantity
of FGD wastes that will be produced exceeds fly ash production; and (b)
whereas environmental controls on the disposal of fly ash have been some-
what limited the disposal of FGD wastes must be done in an environmentally
acceptable manner. Thus, the technologies used for fly ash must be investi-
gated relative to their environmental impact when applied to FGD wastes.
Moreover, since FGD wastes characteristics differ, both physically and
chemically from fly ash characteristics, their suitability in specific applica-
tions must be evaluated under the various conditions in which they are
expected to be disposed.
A complete description of the power plant scrubbing facilities of each unit
and test samples is given in Appendix B.
43
-------
The physical parameters investigated include specific gravity,
bulk density as a function of solids content, water retention for particular
dewatering techniques, viscosity of slurries at various solids content, per-
meability as a function of particle packing fraction (converse of void frac-
tion), compactability as a function of solids content, and strength as a
function of solids content. The following subsections describe the tests
conducted and present the results for each of the FGD wastes investigated.
Wherever possible, the effects of fly ash are shown.
In addition to FGD waste disposal in an untreated state, the option
exists to condition or chemically treat FGD waste materials by a number of
alternative methods. Such treatment necessarily increases the unit disposal
cost, but renders these materials more environmentally acceptable. Selected
tests of physical properties have been conducted on chemically treated FGD
wastes to determine both environmental acceptability and physical behavior
as related to handling and structure in land reclamation applications.
6.2 SOLIDS CHARACTERIZATION
The physical properties of any liquid-solid mixture are dependent
upon the characteristics of both the liquid and the solid constituents as well
as the interaction between them. The FGD wastes are such mixtures and
contain four principal crystalline phases: calcium sulfite hemihydrate, cal-
cium sulfate dihydrate, fly ash, and unreacted lime or limestone (usually
appearing as calcium carbonate by reaction with atmospheric CO2). These
solid phases exist as small crystals suspended in an aqueous liquor which is
usually in equilibrium (saturated) with these solids. In addition, sodium
chloride or calcium chloride is present as totally dissolved salts. A com-
plete chemical characterization of each FGD system liquor is given in
Appendix D.
The relative amounts of each of the solid crystalline phases is
dependent upon the system design parameters: the sulfur content of the coal
used, efficiency of SC>2 scrubbing, fly ash in the flue gas entering the scrubber,
efficiency of the system to remove fLy ash, stoichiometric ratio of absorbent
relative to sulfur content, absorbent utilization efficiency, and amount of oxi-
dation of the sulfur products that takes place in the scrubber system. The
efficiency of scrubbing SO2 and fly ash from the flue gas and the amount of
sulfur product oxidation that takes place are primarily functions of system
design, but other factors enter into these processes, e. g. , combustion
behavior in the boiler, nature of mineral phases in the coal and its total ash
content, reactivity of lime or limestone, and particle size and distribution of
the sulfite particles. The independent and interdependent variables that affect
the composition and characteristics of the resulting waste product are so
numerous and varied that it is not likely that any two sulfur waste products are
identical. Further, each crystalline phase with its specific characteristics
has an influence on the behavior of the sludge. Characterization of the solids
portion of each sludge studied relative to phase identification and particle
characteristics is discussed in Appendix A. To the extent possible, correla-
tions between the particular characteristics of these phases and behavior of
the sludge are made in subsequent subsections.
44
-------
6.3 VISCOSITY
The existing technology for transporting fly ash to a disposal site
often includes sluicing the fly ash with waste waters. An option for trans-
porting FGD wastes to a disposal site is similar in method; i.e. , the waste
is pumped in an as-produced condition as a slurry or sludge.
The viscosity of the liquid waste is a direct measure of its pump-
ability, which affects the cost of transportation as well as system design.
For example, if upon dewatering the liquid waste the viscosity is affected
adversely in a way that precludes the pumping of the waste' to the disposal
site, the system may be designed so that the dewatering operation is per-
formed at the disposal site. This approach can be applied to either primary
dewatering (clarifier, thickener) or secondary dewatering (filter, centrifuge).
Waste materials produced in FGD systems contain finely divided
solids suspended in a water system. Sulfur waste products (both sulfate
and sulfite) tend to have particle sizes typically in the same range as fly ash,
between 100 and 1 |o.m. Whereas fly ash is produced as spheres, sulfur
wastes are either platey or rosettes, as sulfite, or blocky, as sulfate. In
addition, unreacted limestone (or lime) is usually present in the waste and
contributes an additional shape parameter. The particle shape, particularly
for the sulfite, is the cause of the water retention nature previously observed-
in FGD waste (1). In addition, fly ash when present in FGD wastes can pro-
vide a marked measure of fluidity that the sludge would not otherwise have.
(One of several reasons for adding fly ash to concentrate mixes as a replace-
ment for cement is to increase pumpability.) The sulfite and fly ash work in
opposition to each other in promoting pumpability, but factors other than
shape affsct viscosity. These include solids content, particle size, and par-
ticle size distribution. Thus, the property of viscosity is a function of a
multitude of parameters and is not sufficiently understood so that behavior
can be predicted from a knowledge of these parameters; therefore, measure-
ment of this property becomes necessary.
6.3.1 Experimental Procedure
k
The viscosity measurements were performed at room tempera-
ture using a rotating cup viscometer'1" having a cylindrical sleeve immersed
in the FGD fluid waste and rotating at 64 rpm. The 2. 4-cm (15/16 in.) diam-
eter rotating sleeve generates a shear rate of 7.9 cm/sec (15.6 ft/min) at its
outer surface. The range of viscosities that can be measured with this
sleeve is 3 to 150 poise. (Water has a viscosity of 0. 01 poise.)
Prior to measurement, the solids content of the sludge was
adjusted to the desired value, and the mixture was homogenized by hand stir-
ring. The cylindrical sleeve was then immersed in the sludge and left undis-
turbed for 15 sec, during which time the stirring motion ceased. The vis-
cometer was then turned on, and the viscosity was noted.
'"Model VT-02, S and A Products Co. , Brooklyn, N. Y.
45
-------
6.3.2 Viscosity Test Results
The viscosity of FGD sludges from seven power plants as a
function of water content is presented in Figure 8. Sludge from the
Tennessee Valley Authority (TVA) Shawnee Station turbulent contact absorber
(TCA) scrubber was tested from samples taken on three occasions. The
results show that, among the sludges tested, pumpable mixtures (<20 poise)
ranged from a high solids content of 70 percent for the Arizona Public Ser-
vice (APS) Company Cholla sample to a low solids content of 32 percent for
both the Utah Power and Light (UPL) Company Gadsby Station and the General
Motors (GM) Chevrolet-Parma Power Plant samples. From this wide
spread, the importance of experimentally determined data for system design
parameters becomes clearly evident.
The relative position of each curve presented in Figure 8 may be
rationally explained, but certain anomalies nevertheless exist for nearly
every attempt at fully understanding this behavior. The position of the APS
Cholla sludge at high solids content can be attributed to the high concentration
of fly ash (nearly 60 percent) in the sludge and also possibly to the presence
of large blocky sulfate particles, which constitute the major fraction of the
remaining solids. The UPL Gadsby and GM Parma sludges both had a low
fly ash content ( 7 to 8 percent) and a low solids content. Only the Southern
California Edison (SCE) Mohave Generating Station sludge was lower in fly
ash content (3 percent), but its position as the second-highest in solids
content among the sludges tested appears anomalous. The SCE Mohave
sludge solids consisted of large, blocky equiaxial sulfate particles, and it
may be the similarity of these particles to spherical fly ash that determined
the viscosity of this sludge in relationship to that of other sludges.
The sludge from the TVA Shawnee TCA scrubbing facility was
tested on three sampling occasions; the curves 5, 6, and 7 in Figure 8 are
nearly parallel but are displaced over a 7 percent solids range (at 20 poise)
in the order of highest viscosity for the lowest solids content, again reveal-
ing the role of fly ash in increasing fluidity. This sludge was characterized
by large, thin, stacked plates of sulfite, and the fly ash particle-size range
was broad. The Shawnee venturi sludge (curve 3) was chemically similar to
the TCA sludge and had the same particle-size range of fly ash, but the
sulfite did not exist as large plates but, predominantly, as rosettes of sulfite
platelets or small stacks of plates. It is not understood whether the change
in sulfite particle shape was responsible for its relative viscosity since by
fly ash content this sludge would be expected to lie between curves 5 and 7.
The Duquesne Light Company (DLC) Phillips Power Station sludge
contained predominately large lath-like crystals of sulfate and had a high fly
ash content (~60 percent), much like the APS Cholla sludge. On the basis of
these variables, the DLC Phillips sludge would be expected to have a viscosity
much like the Cholla sludge. The relative position of the Phillips sludge is
probably also a consequence of other forces acting within the sludge material.
46
-------
120
100
o
o.
l/)
o
o
=: 60
40
20
CURVE SLUDGE
CM PARMA DOUBLE ALKALI
UPL GADSBY DOUBLE ALKALI
TVA SHAWNEE LIME
DLC PHILLIPS
TVA SHAWNEE LIMESTONE
TVA SHAWNEE LIMESTONE
TVA SHAWNEE LIMESTONE
SCE MOHAVE LIMESTONE
APS CHOLLA LIMESTONE
2 4
FLY ASH.%
7.4
8.6
40.5
59.7
20.1
40.1
40.9
3.0
58.7
30
40 50 60
SOLIDS CONTENT, WEIGHT %
70
Figure 8. Viscosity of desulfurization sludge.
47
-------
The slopes of the viscosity curves as a function of solids content
lie in two groups: The lower slope group includes all sludges from both TVA
Shawnee scrubbers and the UPL Gadsby facility; the higher slope group
includes all other sludges. Presently, no parameter has been identified to
explain this behavior.
Thus far, the results of viscosity tests suggest that fly ash
decreases the viscosity of a sludge. In addition, particle shape, size,
and distribution appear to influence viscosity behavior.
6.4 WET BULK DENSITIES
Dewatering of power plant sludge is advantageous for several
reasons which are associated with disposal costs and environmental con-
siderations. Because land areas will be used for FGD waste disposal,
the decrease in volume that accompanies effective dewatering will mini-
mize the land required for disposal. In many western states where water
scarcity often dictates maximum water reuse, the water returned to the
scrubbing system reduces the need for fresh makeup water. Further,
dewatering minimizes the amount of liquor associated with the solids and
thus reduces the pollution potential associated with the potentially high
mobility of the liquid phase of wastes; it facilitates handling and reduces
fixation costs.
The ability to dewater a sludge is a function of many variables
but primarily of the dewatering method used, the size and distribution of
particles, and the structure of the particles as determined by crystalline
phase. Primary dewatering by clarifiers or thickeners separates water
from a slurry principally through the action of gravitational forces. The
secondary methods of dewatering typically used are vacuum filtration and
centrifugation. In some cases, natural settling may also be considered if
the resultant properties and economics are favorable.
The effectiveness of a dewatering operation is best measured by
the relative quantity of water that remains with the solid after performing
the dewatering operation. The wet bulk density of a sludge is the weight of
a unit volume of dewatered sludge containing both liquid and solid phases.
The wet bulk density Pg of a material can be calculated from the weight frac-
tion solids content fs, the true density of the solids ps, and the density of
water pw by the following equation:
P P
r
P
B p f
rw s
48
-------
When the amount of water within the sludge samples is lower than the value
that represents a point of maximum density where all particles are in contact,
the density of the sludge can be calculated by the following equation:
The solids content at the maximum density of the sludge is determined by the
coincidence of these two expressions and is given as follows:
f - ? t1^}
s(max) - 1 +pw[(l/pD) - (l/ps)] ( '
Thus, the relationship of wet bulk density with solids content can be calcu-
lated from only two values, the true density of the solids pg and the dry bulk
density of the solids pn-
6.4.1 Experimental Procedure
The wet bulk density was determined on eight FGD sludge samples
after each was dewatered by settling, settling with free underdrainage, vacuum
filtration, and centrifugation. These determinations were made by measuring
the weight and volume of a dewatered sample. Subsequently, the samples
were dried to constant weight, and their weight and volume were again deter-
mined. From the latter measurements the dry bulk densities were deter-
mined, and from the relative change in weight the solids content of the wet
sample was determined.
The true density of each solid was measured by a method of air
displacement, using a Beckman air pycnometer. The air pycnometer oper-
ates on the principle of Boyle's law in which the volume of air displaced by
the sample is accurately measured.
6.4.2 Bulk Density Test Results
The wet bulk density of the eight FGD sludges as determined by
Equations (3) and (4) is presented in Figures 9 through 16 as solid lines. In
addition, the measured wet bulk density of dewatered samples is plotted as
datum points on these curves and represent the density obtained by (a) set-
tling in a volumetric flask, (b) settling in a Buchner funnel where the sludge
is allowed to drain freely, (c) filtration in a Buchner funnel with vacuum
assistance, and (d) centrifugation in a glass vial with a laboratory centrifuge.
The right-hand portion of the figures are calculated from Equation (4) and
are different for each dewatering method, because the dry bulk density of the
solids differs as a consequence of the method of dewatering. This is so
49
-------
2.0
1.9
1.8
1.7
1.6
"e 1.5
1.4
1.3
1.2
1.1
1.0
0.9
u.8
TVA SHAWNEE LIMESTONE SLUDGE
1 FEB 1973
CENTRIFUGE-
FILTER,
FILTERED-,
CENTRIFUGED-]
DRAINED-i
SETTLED-]
I , I
2.0
1.9
1.8
1.7
1.6
E 1.5
1.4
1.3
1.2
1.0
0.9 -
20 40 60 80
SOLIDS CONTENT. WEIGHT %
100
0.8
TVA SHAWNEE LIMESTONE SLUDGE
15 JUN 1974
CENTRIFUGE .
FILTERED-,
CENTRIFUGEDn
DRAINED!
SETTLED-
I . I
20 40 60 80
SOLIDS CONTENT. WEIGHT %
100
Figure 9. Wet bulk density of
TVA Shawnee limestone
sludge and dewatered
samples: 1 Feb 1973.
Figure 10. Wet bulk density of
TVA Shawnee limestone
sludge and dewatered
samples: 15 Jun 1974.
50
-------
2.0
1.9
1.8
1.7
1.6
1.5
1.4
1.3
1.2
1.1
1.0
0.9
0.8
TVA SHAWNEE LIME SLUDGE
19 MAR 1974
CENTRIFUGE
FILTERED-,
CENTRIFUGED-!
SETTLED-i
DRAINED,
20 40 60 80
SOLIDS CONTENT. WEIGHT %
100
2.0
1.9
1.8
1.7
1.6
1.5
1.4
1.3
1.2
1.1
1.0
0.9
0.8
DLC PHILLIPS LIME SLUDGE
17 JUN 1974
CENTRIFUGE
FILTERED
CENTRIFUGED
DRAINED,
SETTLEI
I , I
20 40 60 80
SOLIDS CONTENT, WEIGHT %
100
Figure 11. Wet bulk density of
TVA Shawnee lime
sludge and dewatered
samples: 19 Mar 1974.
Figure 12. Wet bulk density of
DLC Phillips lime
sludge and dewatered
samples: 17 Jun 1974.
51
-------
2.0
1.9
1.7
1.6
6 1.5
1.4
1.3
1.2
1.1
1.0 -
0.9 -
0.8
GM PARMA DOUBLE-ALKALI SLUDGE
18 JUL 1974
CENTRIFUGE-
FILTERED-,
CENTRIFUGED-i
DRAINED
SETTLED
ED-i
1
I , I
2.0
1.9
1.7
1.6
1'5
1.4
1.3
1.2
1.1
1.0 -
0.9 -
20 40 60 80
SOLIDS CONTENT, WEIGHT'S.
100
0.8
APS CHOLLA LIMESTONE SLUDGE
1 APR 1974
CENTRIFUGE
CENTRIFUGED-i
FILTERED-)
DRAINEO-i
SETTLED
I , I
20 40 60 80
SOLIDS CONTENT, WEIGHT %
100
Figure 13. Wet bulk density of GM
Parma double-alkali
sludge and dewatered
samples: 18 Jul 1974.
Figure 14. Wet bulk density of
APS Cholla limestone
sludge and dewatered
samples: 1 Apr 1974.
52
-------
2.0
1.9
1.8
1.7
1.6
''5
1.4
1.3
1.2
1.1
1.0
0.9
0.8
UPL GADSBY DOUBLE ALKALI SLUDGE
9 AUG 1974
CENTRIFUGE-
CENTRIFUGED-,
FILTERED-,
SETTLED-|
DRAINED-i
2.0
1.9
1.8
1.7
1.6
1.5
1.4
1.3
1.2
20 40 60 80
SOLIDS CONTENT. WEIGHT*
100
1.0
0.9
0.8
SCE MOHAVE LIMESTONE SLUDGE
30 MAR 1973
CENTRIFUGE .
FILTERED
CENTRIFUGED-!
DRAINED-i
SETTLED-!
20 40 60 80
SOLIDS CONTENT. WEIGHT*
100
Figure 15. Wet bulk density of
UPL Gadsby double-
alkali sludge and
dewatered samples:
9 Aug 1974.
Figure 16. Wet bulk density of
SCE Mohave lime-
stone sludge and
dewatered samples:
30 Mar 1973.
53
-------
because the packing of the particle is a function of the forces placed on the
particles during the dewatering process.
From the experimental points in Figures 9 through 16, it is
revealed that the highest density of sludge is obtained by vacuum-assisted
filtration in most sludges and by centrifugation in others. In all cases, rela-
tively small density differences resulted from these two dewatering methods.
Moreover, when industrial filters and centrifuges are used rather than lab-
oratory equipment, the difference between the two methods are expected to
be indiscernible. In every case, the force of laboratory centrifugation was
sufficiently greater than filtration, as revealed by the packing density of
particles and by the bulk density of dry sludge (right-hand border of figure),
that the expected maximum density obtainable was always greatest by centrifu-
gation. In several sludges, the particle packing obtained by vacuum filtra-
tion was not as great as by settling or free drainage even though the actual
measured density by filtration was always greater. It is believed that this is
because particle packing by vacuum filtration takes place so rapidly that
particles do not have the opportunity of assuming minimum volume configu-
rations .
In most sludges, there was little difference in the density of
sludges dewatered by settling or by free drainage, nor was there much dif-
ference between the expected maximum density as determined by the particle
packing density.
A compilation of the dewatering behavior of these sludges is given
in Table 10 and includes, for each experimental method, the wet bulk density
and solids content, the dry bulk density, the calculated maximum density,
and the solids content as given by the intersection of Equations (3) and (4).
These data show that the sludges with the best overall dewatering
characteristics, i.e., the TVA Shawnee (limestone), the SCE Mohave (lime-
stone), and DLC Phillips (lime) sludges, were those with the coarsest
particle size distributions. The double-alkali systems produced the finest
particle size distributions and had the worst dewatering characteristics. The
most effective method of dewatering in most cases was by vacuum-assisted
filtration. However, centrifugation was nearly as effective and was superior
in several instances.
Settling and free drainage did not reveal a significant difference
in dewatering characteristics and reflected the similarity of these two
methods observed by experimentation. The inference of this observation is
that the sludge settles to the same density whether water drains or rises to
the surface.
6. 5 COEFFICIENT OF PERMEABILITY
The potential pollution capability of sludge liquor is governed by
the mobility of leaching waters and is determined by the permeability coeffi-
cient of the various media through which these waters must pass. Liquor
mobility can be controlled either by (a) containment of the waste in a basin
having an impervious liner, (b) chemically treating the waste in such a
54
-------
TABLE 10. UNTREATED SLUDGE BULK DENSITIES
Sample Source
TVA Shawnee Limestone
(2/1/73);
ps = 2.48
TVA Shawnee Limestone
(6/15/74);
ps = 2.45
TVA Shawnee Lime
(3/19/74);
ps = 2.53
CM Parma Double Alkali
(7/18/74);
ps = 2.45
UPL Gadsby Double Alkali
(8/9/74);
ps = 2.60
DLC Phillips Lime
(6/17/74);
ps = 2.50
APS Cholla Limestone
(9/1/74);
ps = 2.53
SCE Mohavc Limestone
(3/30/73):
ps = 2. 53
Settle*3
PB
PD
pM
.45
. 10
.65
.46
. 11
1.66
1.36
0.93
1. 56
1.31
1.20
1.70
1.33
0.96
1.60
1.40
0.96
1. 57
1. 39
1. 12
1.68
1.65
1. 30
1.76
fB
fD
fM
49.0
100
66.4
52.9
100
66.8
43.4
100
59.4
40.0
100
69.9
41.4
100
60.5
47.6
100
60.8
46.7
100
67. 1
66.6
100
72.5
Drain
PB
pD
pM
. 51
.09
.65
.53
.08
1.64
1.34
0.97
1. 58
1.35
1.22
1.71
1.30
0.89
1.55
1.48
1.09
1.65
1.44
1.22
1.73
1.67
1.35
1.79
fB
fD
fM
55.7
100
65.9
58.3
100
65.9 .
41.5
100
60.6
43.9
100
71.0
37.2
100
57.6
53. 1
100
65.8
50.9
100
70.2
67.2
100
74. 5
Filter*3
pB
pD
pM
1.65
1. 16
1.71
1.64
1. 15
1.69
1. 51
0.87
1. 53
1. 52
1. 10
1.69
1.50
1. 16
1.74
1. 52
0.98
1. 59
1.48
1. 10
1.66
1.78
1.48
1 . 86
fB
fD
fM
65.0
100
69.5
65.9
100
68.5
56.0
100
57.2
57.8
100
69.3
54.6
100
69. 1
57.0
100
62.0
53.4
100
66.0
80.3
100
77.2
Centrifuge
pB
pD
pM
1.56
1.45
1.86
1.60
1.48
1.86
1.44
1.30
1.77
1.43
1.46
1.84
1.62
1.61
1.99
1.52
1.43
1.84
1.58
1.61
1.95
1.86
1.71
2. 11
fB
fD
fM
59.8
100
78. 5
63.3
100
78.3
49.9
100
72.2
50.9
100
78.4
62.2
100
81.2
57.2
100
77. 1
60.9
100
81.4
77.0
100
84.5
ps: true density of dry solids (particle density).
pB, fB: wet bulk density (g/cm^l and solids content ('
pD, fD: dry bulk density (g/cm?) and solids content ('
pM, fM: wet bulk density (g/cm^) and solids content ('
of dcwatercd sludge;
of dry sludge;
at maximum density.
55
-------
manner that the waste itself is impermeable, or (c) preventing or minimizing
rainwater from recharging the disposal site. Each of these methods of con-
trol requires a unique disposal technology and specific attention to disposal
management. In the first case, the control by liner containment is indepen-
dent of the permeability of the waste itself. However, with such phenomena
as a breach in the liner as a consequence of earth movement, liner deterio-
ration, or burrowing animals, the control of pollution may then be dependent
upon the properties of the waste and soil. Chemically treated waste is
expected to produce a material with a lower permeability than untreated
waste. In this case, intended control is provided by a reduction in the con-
centration of chemical constituents that can leach from the treated waste
and/or a reduction in the amount of liquid that passes through the waste. In
the case of rainwater, control of the amount of liquid passing through the
waste is determined primarily as a result of disposal site design and mainte-
nance, and this control can be applied independently of or in conjunction with
the other two control methods.
In all three cases, permeability of leaching waters through the
waste provides a controlling parameter that serves as a maximum limit to
the amount of liquid that recharges the subterranean water supply. The
amount of liquid and the level of contamination of this liquid are jointly
responsible for the pollution potential of any given waste disposal site. For
this reason, the coefficient of permeability of each of the treated and un-
treated FGD sludges was measured.
6.5.1 Experimental Procedure
The permeability of sludge from seven separate scrubbing facili-
ties was determined by passing deionized water through sludge columns by
the constant head technique. Experiments were conducted by packing 100
grams of dried, powdered sludge solids into 4-cm diameter glass columns
or 200 grams of dried sludge solids into 6-cm diameter glass columns. The
sludges were oven-dried either at 80 C in air or at 70°C in a vacuum until
constant weight was achieved. Generally, the finer sludges were dried at
the lower temperature, in a vacuum, in order to accelerate the drying
process. In all cases, leaching water was added to the column from the top.
The procedure used required a determination of the pore volume
by calculating the particle packing density in the column from the weight of
the sludge and its height in the packed column. Then a volume of deionized
water was added to the column that exceeded the pore volume by nominally
3 percent. The columns were monitored as water slowly worked its way, by
gravity, through the packed sludge solids. The excess water was collected
in a volumetric cylinder and compared with the calculated excess. In all
cases except one, the measured excess and calculated excess were within
1 percent. This was true because all sludges, except one, retained all pore
water and, in measuring excess water, it was apparent not only by the
absence of supernate water but also because the rate of water dripping from
the column abruptly went to zero.
56
-------
In the evaluation of whether this technique fully wetted the sludge
solids, the close correlation between calculated and measured excess water
indicates that less than 1 percent unwetted pore space could have remained
in any of the columns. Moreover, since the difference between calculated
and measured excess water was equally distributed both negatively and posi-
tively, it is reasonable to conclude that the difference was a consequence of
a random error in measurement rather than a systematic error in pore
wetting.
The one exception to this behavior was observed with the SCE
Mohave sludge, where more than 2 percent additional excess water was
measured over the calculated value. This sludge was also the only one found
in the dewatering study that created open pore space by a mechanical dewater-
ing technique. It was further observed that a gradual decrease in the rate of
water dripping from the column took place before the rate went to zero. For
this sludge, the particle packing density and pore volume were recalculated
from measured excess water taken at the point where the drip rate decreased.
The permeation measurements were begun when an amount of
deionized water was added to a known height above the top of the sludge; the
amount of water percolating through the column was then collected and mea-
sured as a function of time.
The procedure just described was also used on chemically treated
sludges. Preparation of the samples included crushing and pulverizing the
fixed material after drying. The powder was packed into the column in the
same manner as the untreated sludge. Although this technique does not
provide an accurate value for treated sludge placed in an undisturbed manner,
it should provide an upper bound as well as a reasonable description of the
permeability of a sludge that is moved after placing and curing.
Additionally, the permeability of undisturbed fixed sludge was
determined by the following method. A cylindrical core from a block of fixed
material was cut slightly smaller than the inside diameter of the permeation
column. A layer of a silicone-based adhesive was applied to the circumfer-
ential area of the sample and allowed to cure. A second layer was then
applied; the sample was slid into the column; and the adhesive was allowed to
cure. Both glass and clear plastic columns were used in order to assure a
proper seal between the sample and column wall. In low permeation samples,
it was necessary to apply a slight gas overpressure (2 to 6 psig; 4 to 12-ft
H_O) to accelerate the flow of water through the sample.
An additional test procedure was applied to the untreated sludge
at the conclusion of the permeation measurement. The supernate water was
decanted from the surface of the sludge, and the column was allowed to air-
dry for 24 hr. The sludge was then compacted with a plunger, but the thixo-
tropic nature of the material and the high level of retained water did not
permit as much compaction as desired. A new column height was measured,
and the particle packing density and pore volume were recalculated. The
permeation rate was again measured, as previously described.
57
-------
6.5.2 Permeability Test Results
The results of the permeability tests are presented in Table 11
and in Figure 17. Datum points in the figure are connected by solid or dashed
lines representing two different slopes. The lower slope was established by
the best fit through the TVA Shawnee limestone permeation data and was
applied to the DLC Phillips lime and SCE Mohave limestone data. In all three
cases, the same slope adequately represents the results. The steeper slope
was selected as the best fit through the TVA Shawnee lime, the APS Cholla
limestone, and the GM Parma and UPL Gadsby double-alkali systems. There
is no theoretical basis for selection of two slopes rather than individual
slopes for each system.
The implication of these data is that permeation through sludge
responds in one of two behavior patterns as a function of the amount of com-
paction of the material. Generally, the sludges that correspond to the lower
slope had higher permeation coefficients and compacted to higher particle
densities (lower void volume) than the sludges with higher slopes. Moreover,
the change in particle density was typically greater for the lower slope sludge,
as a consequence of identical compaction conditions.
A probable explanation for the observed behavior may be the dif-
ference in particle size between the two groups of sludges. The TVA Shawnee
lime system and both double-alkali systems produced a sludge with finer
particle sizes than the three sludges showing a low slope and the APS Cholla
sludge showing a high slope. If it were not for the Cholla sludge, it might be
concluded that the change in slope was solely a function of particle size.
An alternative explanation may be offered that is not intrinsically
dependent on particle size but, rather, is a function of the technique of com-
paction. In all cases, the columns were allowed to air dry for 24 hr. In the
coarser particle sizes, the effect of surface tension on the loss of water is
less pronounced, and the possibility for greater drying takes place. There-
fore, the increase in particle density upon subsequent compaction is a conse-
quence of the sludge being more capable of reaching higher particle density
values than the finer particle-sized sludge in which water would not evaporate
as readily. Again, the APS Cholla sludge, being coarser in particle size,
would be expected to behave as the other coarser particles. However, the
Cholla sludge contained a bimodal distribution of particles in which fly ash
constituted the finer fraction. It is likely that this distribution, which allowed
the finer material to fill the void space between the coarser particles,
accounts for the significantly lower permeability of the Cholla sludge at
equivalent values of particle density. It may also be that the presence of the
finer particles provided the resistance to drying attributed to the finer
particle-sized sludges such that, upon compaction, this sludge then behaved
like the finer sludges.
Whether either of the mechanisms postulated to explain the
observed behavior is correct may be insignificant relative to the observation
that the fine-particle sludges reduced permeability coefficients significantly
as a result of very small changes in particle density. This behavior may
suggest that a more efficient compaction might decrease the permeability
58
-------
TABLE 11. PERMEABILITY OF UNTREATED AND CHEMICALLY
FIXED FGC SLUDGES
Sample
Source
TVA Shawncc
Limestone
TVA Shawncc
Lime
SCE Mohavc
DLC Phillips
CM Parma
Double Alkali
AI'S Cholla
I'PL Cadsby
Double Alkal i
TVA .Shawnei-
Li me stone
TVA Shawnee
Limestone
SCE Mohavc
( ll'CSlk
TVA Shawncc
Lime
lll'CS)b
TVA Shawnec
Limestone
(Chemfix)
TVA Shawnce
( Dravo)
DLC Phillips
(CaJcilox)
Sample
Date
2/1/73
6/15/74
6/15/74
6/15/74
6/15/74
3/l"/74
3/ 10/74
3/CV74
'./30/73
5/30/73
3/30/73
6/17/74
6/17/74
6/17/74
7/IS/74
7/1S/74
7/18/74
4/1/74
4/1/74
4/1/74
s/y/7-i
S/9/74
S/9/74
V2-VT5
6/12/75
2/27/75
6/12/75
Replications
1
3
1
3
3
,
1
(2)
(3)
3
3
(2)
3
2
(21
1
1
(21
1
1
1
1
12)
(2)
(5)
(2)
1
1
1
1
(2)
1
1
1
1
(2)
1
1
Fractional
Void
Volume
0.69
0.60
0. 58
0. 58
0. 55
0.75
0.74
0.72
0.47
0.43
0. 34
0.68
0. 58
0.40
0.71
0. 60
0.65
0. 56
0. 54
0. 54
0.75
0.73
0. 70
0.6'!
0. 54
0. 55
0. 65
0. 53
0. 57
-
0.68
0.72
0.70
0.78
0.75
0.70
0.78
0.76
Permeability
Coefficient.
cm/sec
2. 3 v I0~^
1 .0 x 10"
9.6 x 10"
8. 5 x 10"'
5.9 x 10''
1 .7 x 10'^
5.3 x 10"^
6.0 x 10"'
5.0 x 10"^
7.5 x 10"^
1 .6 x 10
1 .2. x I5"j
1 . 3 - 10"^
7.4 v 10
8.2 , 10"5
2. 5 x 10";!
8. 1 .. 10
2.7 x IO"5
1 .8 < 10"'
1.1 . 10
".K ,s \0"\
1 . 3 x 10"
1 . 2 t 10
2.2 - IO""1
5.5 x IO"8
7.9 y IO"5
7.3 x 10"
1 . 9 A 1 0
5.5 , 10".
5. 5 x 10"
4. 1 x IO"5
1 . 5-2. 1 x 10
4.7 x 1C'5
3.2 y \Q-4
6.9 ^ 10"'
3.8 . I0"^|
4 . 9 x 10" .
2. 1 y 10"
Remarks
Column packed as slurry
Compacted wet
Compacted wet
Column packed as slurry
Compacted wet
Compacted wel
Compacted wel
Pu] ve rized
Solid, undisturbed
Pulvc- rizcd
Pul ve ri/.ecl
Pulverized, compacted wet
Solid, undisturbed
Solid, undisturbed
Pul ve rized
Solid, undisturbed
Pulverized
Pul ve rized
Solid, undisturbed
Pulverized
Pulverized
Pulverized, compacted wet
aRcplications of those in parenthesis refer to multiple measurements on a single column using varying
hydraulic heads.
IU Conversion Systems Laboratory.
59
-------
io~3
5 ID'4
Ou
l_l_
o
10
-5
5.5x10 -1-
rggi APPROXIMATE RANGE
isill FOR UNTREATED MATERIALS
* PULVERIZED
ZT FAMILIES OF DATA HAVING SAME SLOPES
POND B
POND C
POND E
TVA SNA'/NEE/DRAVO (Ref. 2)
TVA SHAWNEE/IUCS (Ref. 21
TVA SHAWMS/CHEMFIX (Ref. 2)
SAMPLE SOURCES DATE
TVA SHAWNEE LIMESTONE 6/15/74
TVA SHAWNEE LIME 3/15/74
SCE MOHAVE LIMESTONE 3/30/73
GM PARMA DOUBLE ALKALI 7/18/74
UPL GADSBY DOUBLE ALKALI 8/9/74
DLC PHILLIPS LIME 6/17/74
APS CHOLLA LIMESTONE 4/1/74
CHEMFIX LAB A IUCS LAB
DRAVO LAB
0.30
0.40 0.50
VOLUME FRACTION OF SOLIDS
0.60
0.70
Figure 17. Permeability of treated and untreated sludge: Aerospace analyses.
-------
coefficient even further and, given only a 5 percent reduction in void volume,
would produce an order-of-magnitude reduction in permeability coefficients.
On the other hand, the permeability of the compacted sludge may
be only a consequence of a spatial redistribution of particles that results in
only a small change in particle density. In this case, more efficient compac-
tion may not further improve particle distribution, and the result may be
similar to the coarse-particle behavior. Further work is required to clarify
this uncertainty.
The data presented in Figure 17 strongly suggest that the perme-
ability coefficient of untreated sludges generally fall in the range of 2 X 10"*
to 5 X 10~5 cm/sec (shaded area of Figure 17) and can be described as typi-
cally having a value of 1 x 10""* cm/sec independent of void volume. This
behavior was similarly observed for treated sludges that had been pulverized
in the laboratory and which simulated sludges that are moved after placing
and curing. The only exception to this general behavior was observed on
treated sludges that were not pulverized. In most cases, the permeability
was low, similar to concrete materials, and suggests that treated sludges
placed and cured without subsequent moving may form a highly impermeable
waste basin. Whether all fixation methods will produce similar results is
uncertain, but it is reasonable to conclude that the permeability of treated
sludge that is placed and not subsequently moved will be less than that of
sludge which is moved after placement.
6.6 COMPACTION
One of the considerations in the disposal of FGD sludges is the
volume or land area required for placement of the sludge. A previous
study (1) has indicated that for pond disposal a placement depth of 30 ft is
optimal. Thus, by using a 30-ft thick disposal depth and knowing the rate of
sludge production and wet bulk density of the sludge, it is possible to calculate
the volume or land area required for disposal. The size of the required land
area directly affects the economics of disposal such that the cost of secondary
dewatering and mechanical compaction of the sludge enters into the total con-
sideration of sludge disposal technology.
Section 6.4 discusses and compares the results of dewatering
techniques; from these data, it is possible to calculate the cost benefits of
each method. This section discusses the results of compaction tests con-
ducted on six different FGD sludges relative to the size, shape, and distribu-
tion of particles and to the water content of each compaction sample. The
compaction tests can be performed only on samples having a moisture content
less than the maximum density value defined by Equation (5) and shown in
Figures 9 through 16. In this partially dried state, air voids partially fill the
free volume between particles, and the function of the compaction process is
to rearrange particles in a manner such that these air voids are eliminated.
It is erroneous to assume that a maximum density could be obtained at a zero
moisture content because the presence of water is necessary to provide lubri-
cation to the particles for facilitation of their motion past each other. The
61
-------
resultant change in sludge density is a consequence of both the available
air-void volume and the ability of the particles to slide past each other under
the applied load.
6.6.1 Experimental Procedure
The compaction of six FGD sludges was performed by applying a
load onto the samples constrained from lateral displacement, recording the
change in sample height directly on the loading ram, and comparing the height
before and after loading by micrometer measurements. The samples, in all
cases, were the dried cake from the wet bulk density study that had been dewa-
tered by settling in a free draining condition. From each sludge cake three
disks were hand-carved, 1.25-in. in diameter and nominally 0. 05-in. in height,
such that they fit snugly into a pellet die. The dry disks were placed in a humid-
ity chamber for various periods of time and were withdrawn when they
reached a selected moisture content based on dewatered sludge properties
shown in the bulk density curves in 6.4.2. These were stored individually
in sealed plastic containers where they were allowed to equilibrate for one
week before being tested.
The tests were conducted by first weighing the samples to deter-
mine the precise moisture content, measuring the sample height, and care-
fully inserting the sample into the pellet die. The plungers were inserted
into the die and placed into a constant strain test machine, and a 125-lb load
(~ 100 psi) was applied. From the displacement of the ram on the test
machine and from the height of the sample after being withdrawn from the
pellet die, volume changes were calculated.
6.6.2 Compaction Test Results
The stress-strain behavior of the sludges during the compaction
tests are presented in Figures 18 through 23. The solids content was
selected at three nominal values (79, 86, and 93 percent levels), represent-
ing three levels of air-dried sludge. As shown, it is apparent that the amount
of water in the samples had a marked effect on the compaction that occurred
under load. In general, the most compaction took place at the highest mois-
ture content, and the least compaction took place at the lowest moisture con-
tent. However, several exceptions to this generality are noted, specifically
in the APS Cholla sludge. The compaction of the sludge under maximum load
ranged from 15.5 percent for the TVA Shawnee limestone sludge down to
slightly over 2 percent for the DLC Phillips sludge. Except for the TVA
Shawnee sludges, maximum compaction was generally only 5 to 7 percent.
When the compacted sludges were removed from the die and
measured with a micrometer, the sample heights were all found to be greater
than the heights indicated on the testing machine. The percent volume change
by compaction calculated from these heights is indicated in the figures by
arrows. These results indicate that upon removal of the die wall constraints,
the samples returned to nearly their original height. In other words, these
62
-------
TVA SHAWNEE LIMESTONE SLUDGE
15 JUN 1974
CURVE SOLIDS. %
5 10
VOLUME CHANGE, %
94.2%
86.3%
78.9%
ARROWS INDICATE
END POINT AFTER
REMOVAL OF
SAMPLE FROM MOLD
Figure 18. Compaction test results for TVA Shawnee lime-
stone sludge solids.
100
l/l
I/)
UJ
Of.
Q.
Q_
O
60
40
20
TVA SHAWNEE LIME SLUDGE
19 MAR 1974
1 2
5 10
VOLUME CHANGE,
CURVE
1
2
3
SOLIDS. %
93.5
83.8
75.6
ARROWS INDICATE
END POINT AFTER
REMOVAL OF SAMPLE
FROM MOLD
15
Figure 19. Compaction test results for TVA Shawnee lime
sludge solids.
63
-------
100
R 80
60
£ 40
Q_
O
20
GM PARMA FILTER CAKE
CURVE
1 1
3
SOLI OS, %
90.5
84.0
78.5
ARROWS INDICATE
END POINT AFTER
REMOVAL OF
SAMPLE FROM MOLD
2 3 4
VOLUME CHANGE, %
Figure 20. Compaction test results for GM Parma
filter cake solids.
100
z 80
c/n
£ 60
OL
Q.
O
o
o
40
20
UPL GADSBY DOUBLE-ALKALI SLUDGE
CURVE
1
2
3
ARROWS INDICATE
END POINT AFTER
REMOVAL OF
SAMPLE FROM MOLD
2 3 4
VOLUME CHANGE, %
Figure 21. Compaction test results for UPL Gadsby
double-alkali sludge solids.
64
-------
100
= 80
a.
z.
O
5 40
o
o
20
DLC PHILLIPS LIME SLUDGE
2 1
ARROWS INDICATE
END POINT AFTER
REMOVAL OF
SAMPLE FROM MOLD
2 3 4
VOLUME CHANGE, %
Figure 22. Compaction test results for DLC Phillips
lime sludge solids.
100
a.
LLJ"
oc.
a.
O
o
o
60
40
20
,_ APS CHOLLA LIMESTONE SLUDGE
1 APR 1974
1
CURVE SOLIDS.%
1 92.6
2 85.3
3 77.0
ARROWS INDICATE
END POINT AFTER
REMOVAL OF
SAMPLE FROM MOLD
2 3 4
VOLUME CHANGE, %
Figure 23. Compaction test results for APS Cholla
limestone sludge solids.
65
-------
results indicate that the compaction observed in the test was nearly all
elastic compaction and little repacking of particles under load occurred.
The permanent deformation that was experienced was as small as 1 per-
cent and did not exceed 4 percent in any of the sludges.
The wet bulk density at each value of solids content, when nor-
malized to a dry bulk density basis, is an indication of the permanent change
in particle packing. These values are plotted in Figure 24 and show that
little compaction takes place. For contrast, a compaction envelope taken
from seven different fly ashes compacted and sampled in the field (15) was
plotted. For fly ash, minimum compaction was about 7 percent, and maxi-
mum compaction was nearly 20 percent. Thus, the results experienced from
compacting fly ash are not to be expected in field sludge compaction on the
basis of these data.
An explanation of this behavior is to be found in the crystalline
morphology of the sludges. For sludges having platey calcium sulfite crys-
tals, it is easy to envision that particles packed together bend and deflect
when stress is applied such that the particles again return to their original
configuration when the stress is removed. For lath-like gypsum crystals,
the growth trend is to form intergrown crystals, with the result being a
lossely packed, open network of crystals; in some cases, the presence of fly
ash in the sludge will fill the interstices between the crystals. Force applied
to this arrangement of particles flexes the gypsum crystals and possibly
breaks the intergrown clusters, but increased particle packing is difficult.
Only in those cases where particles are reasonably blocky does effective
packing take place, and this took place only in the GM Parma and APS Cholla
samples, among the six sludges tested.
The consequence of limited permanent deformation is that an
increase in strength and decrease in permeability coefficient expected from
compaction is not realized. Because particle rearrangement is evidently not
affected by compaction, strength is not measurably increased. However, the
small permanent deformation that is experienced is sufficient to reduce water
passage, and the permeability of the compacted material is reduced by a
factor of two (Section 6.5).
6.7 COMPRESSION
The compression test is used as a general measure of the strength
of FGD material and is judged by comparison with soil material. The value
of this test is to provide a relative measure of the strength of a material in a
simple and straightforward manner. It is not useful as an engineering test
because it does not account for the strengthening effect of contiguous material
in landfill placement. The lateral constraints afforded by adjacent material
can produce strengths that are greater, by 3 to 5 times, than the strength of
freestanding material.
For FGD waste disposal in landfill applications, a comparison of
soils and sludge provides an effective means for evaluating the acceptability
of waste in the useful reclamation of a waste disposal area. Moreover, the
66
-------
1.4
1.3
r\
E
u
cr>
>-' 1.2
1.1
o
1.0
0.9
0.8
FLY ASH
APS CHOLLA
DLC PHILLIPS
TVA SHAWNEE LIMESTONE
TVA SHAWNEE LIME
UPL GADSBY
I
I
60
70 80
SOLIDS CONTENT, WEIGHT %
90
100
Figure 24. Compaction of FGD sludge (dry basis).
67
-------
additional strength provided by chemical treatment processes can be evaluated
for potential structural fill applications.
6.7.1 Experimental Procedure
Compression strength tests were conducted essentially in accor-
dance with the standard test methods (18) defined by ASTM C39-64. Test
samples of untreated FGD wastes were prepared by casting slurries of the
waste material into containers nominally 2 in. in diameter and 4 in. in length.
Excess water was removed by decanting from the surface and through weep
holes in the base of the mold. Samples of chemically treated FGD wastes
were tested either as supplied by the manufacturer or as cut and machined by
conventional machining tools.
Samples were tested under two conditions, in a full dry state and
in a state of about 10 percent moisture (or at the moisture content as received
in the case of treated sludge). Drying of samples was by air, accelerated by
the heat from the upper surface of a conventional laboratory drying oven.
6.7.2 Compression Test Results
The results from the compression tests of dried untreated sludges
were generally unsatisfactory because of the inherent difficulties in preparing
samples. The major problem with untreated materials was caused by crack-
ing during drying and was possibly a consequence of large shrinkage in some
cases. In other cases, samples chipped or crumbled during the simple han-
dling procedures required to prepare flat, parallel loading surfaces. Because
of the generally difficult preparation, samples were accepted for testing that
are not acceptable by normal standard test method procedures. Whereas
three samples were cast for each representation of a single sample of sludge,
only in one case was it possible to test all three. In several cases, no sam-
ples were available for testing; in other cases, one or two samples were
tested, but the samples were not all acceptable because of dimensional defects.
The results of compression tests are given in Table 12 where the
strength of each sample is recorded as a function of moisture content within
the sample. In many cases, the strength of the samples as measured was
lower than that which would be expected if the sample had not experienced
dimensional defects. In other cases, very low strength is suspected as a
consequence of incipient shrinkable cracks.
The test results show that untreated sludge is not readily compared
with soil (1. 5 to 7 tons/ft^). The difficulty in preparing samples for testing is
generally indicative of the problems expected on the surfaces of air-dried
sludges in the field. Air-drying, possibly assisted by solar heating, will pro-
duce a material that shrinks and cracks such that when loaded the material will
crumble.
The chemically treated laboratory samples had strengths that indi-
cated more than adequate stability for structural fill purposes. Values are
68
-------
TABLE 12. UNCONFINED COMPRESSION TEST RESULTS
Sample Source
Moisture
Content, %
Strength,
tons/ft2
Remarks
Oven-Dried Samples'
TVA Shawnee Limestone
(2/1/73)
TVA Shawnee Limestone
(6/15/74)
TVA Shawnee Lime
(3/19/74)
DLC Phillips Lime
(6/17/74)
CM Parma Double Alkali
(7/18/74)
APS Cholla Limestone
(4/1/74)
UPL Gadsby Double Alkali
(8/9/74)
TVA Shawnee Limestone
(Chemfix)
TVA Shawnee Limestone
(IUCS)
DLC Phillips Lime
(Calcilox)
8.9
0.0
7.4
10.3
0.0
0.0
0.0
14.4
9.9
8.7
10.4
0.0
0.0
0.0
5.6
7.2
8. 1
0.0
0.0
3.2
2.4
0.0
5.6
0.2
1.6
0.6
2.4
1.4
0.8
1. 1
1.3
0.9
2.0
2. 1
1.7
1.2
0.9
1.4
.4
. 5
.6
.3
. 5
46.7 )
48.5 >
52.3)
25.4
22.3
Shrinkage cracks
Chipped edge
--
Loading surface chipped
Cracks on surface
Cracks on surface
_ _
Slight shrinkage cracks
Slight shrinkage cracks
Slight shrinkage cracks
No samples retained
integrity
All samples retained
shape and edges
All samples cracked
Material in granular form.
Samples prepared in IUCS
Laboratory
Material fixed by DLC,
using Calcilox additive
at the Phillips station
Cured-in-Field Samples'
TVA Shawnee Limestone
(Chemifix)
TVA Shawnee Limestone
(Dravo)
TVA Shawnee Limestone
(IUCS)
51
58
37
7.4 to 9.6 \
1.9 to 2.4 >
29. 5 to 36.7 )
Analyses of core samples
taken from the EPA
Shawnee Field Disposal
Evaluation Project
(Ref. 2)
Oven-dried samples had strengths that were 6 to 48 percent higher than those of
the cured-in-field samples.
69
-------
also given in Table 12 for samples that were obtained from the TVA Shawnee
field evaluation project (2) and analyzed in the as-cured condition. These
provide confirmation that chemically treated sludges attain adequate structural
strength.
70
-------
SECTION VII
CHEMICAL CHARACTERIZATION OF UNTREATED FGD WASTE
The environmentally acceptable techniques for ultimate disposal
of FGD wastes are those that do not permit excessive quantities of chemical
pollutants to enter the environment. The sources of chemical pollutants are
the process ingredients consisting of combustion products of coal, lime or
limestone absorbents, and process makeup water. Soda ash is a potential
pollutant source in the double-alkali system. For this study, chemical pol-
lutants are defined as those elemental or compound species that, if allowed
to enter the environment, may be detrimental to the health or well being of
man, animals, or natural plant life for reasons of their toxic, or otherwise
undesirable, nature. Trace elements found in coal and limestone include
heavy metal elements that are known to have a toxic effect on living organ-
isms when their concentration exceeds an acceptable tolerance limit. In
addition, major chemical species involved in the scrubbing process, although
not hazardous to man, can nevertheless be a noxious irritant if allowed to
enter the environment in an uncontrolled manner. The objective of the task
reported in this section is to characterize flue gas desulfurization (FGD)
sludges and the process ingredients from which they originate so as to provide
an understanding of the relationship between the source and fate of potential
chemical pollutants within the various FGD scrubbing processes.
7. 1 OVERVIEW AND ASSESSMENT
Chemical analyses of FGD waste were performed for differing
scrubbing facilities under various operating conditions and design parameters.
The evaluation of these data clearly reveals that the concentrations of trace
elements found in sludge solids are directly dependent upon the concentrations
of those metals found in the input ingredients, primarily the combustion prod-
ucts (fly ash) of coal. The concentrations of these elements in sludge liquors
are shown to be similarly dependent upon their content in the coal and are
presumed to result from the leaching of fly ash.
As a consequence of the direct relationship between trace element
content in sludge and in coal, it is expected that trace elements originating in
coal will be found principally in the sludge. When comparing eastern and
western sludges, it was found that western sludges had concentrations slightly
less than eastern sludges, particularly for arsenic, cadmium, mercury, and
zinc, which are found in significantly lower concentrations in western coal.
71
-------
When the chemistry of the scrubbing liquor is investigated as a
function of scrubber absorbent, i.e., limestone, lime, or double alkali, it
is found that, in general, the trace element content in the liquor is highest in
the limestone systems, intermediate in the lime systems, and lowest in the
double-alkali systems. The most probable systematic parameter that affects
these differences is the pH of the scrubber liquor. The limestone scrubbing
systems operate at the lowest pH, and the double-alkali systems operate at
the highest pH. When a limestone and a lime system both operating on the
same coal (TVA Shawnee facilities) or two systems operating at equivalently
high sulfate concentrations and ionic strengths (SCE Mohave and GM Parma)
are compared on a one-to-one basis, the same relationship between scrubber
pH and trace metal content is found. This relationship suggests the role of
pH in the leaching of fly ash.
An evaluation was made of the trace metal content in the system
liquor at various stages in the disposal process from the scrubber to the final
disposal site. The analyses indicated that the pH of the system liquors
increased and trace element concentrations metals showed a decrease enroute
through the system. These results may be interpreted as a response to sys-
tem pH or a response to the changes taking place in the concentration of major
chemical species. The in-process analyses revealed a rapid oxidation of sul-
fite ion and a precipitation of calcium sulfate. The trace metal content in the
liquor may have decreased by precipitation in response to decreasing ionic
strength, by co-precipitation resulting from the scavenging action of the cal-
cium sulfate, by adsorption onto newly created surfaces of the calcium sulfate
phase, or by pH changes.
However, when the chemical analyses of trace elements in all
liquor samples from every power plant sampled are evaluated as a function
of pH or of ionic strength, no correlation is apparent. It was determined
from these data that changes in ionic strength do not affect the measured con-
centration of trace metals and that pH changes affect these concentrations
only within a system where major system parameters are invariant. From
the positive correlation of liquor analyses with coal combustion product anal-
yses, it is shown that trace metal content in system liquor is primarily con-
trolled by the content of these metals in the coal and only secondarily
controlled by the pH of the system. Moreover, the pH control can be seen
only when coal, the primary controlling parameter, is invariant as observed
in in-process changes or when comparing the processes which use the same
coal source, e. g. , the TVA Shawnee facilities.
When the chemical species in system liquor are evaluated over a
period of time, the data indicate that all species increase from startup. The
calcium sulfate saturation level is established quickly as is the concentration
of trace metals. Chloride ion concentrations appear to build up at a slower
rate and continue to build up until a steady state is established. As the chlo-
ride ion concentration increases, a concomitant increase in ionic strength
72
-------
causes higher calcium sulfate saturation levels. Trace element contents in
the liquor do not reflect these changes, and this result is consistent with the
previous observation of invariance with changes in ionic strength, indicating
that trace metals are not saturated in system liquors and concentrations are
primarily dependent on the coal content (through fly ash) and on the removal
from the system of these elements through occluded liquor in the sludge.
This section presents a brief description of the power plant
scrubbing facilities that provided the data base for this report, descriptions
of the sampling procedure and subsequent chemical analyses, and a discus-
sion of the various relationships that were observed between scrubber chem-
istry and the many variables of the scrubbing process and process ingre-
dients. Comparisons and contrasts of the various scrubber designs, eastern
and western coal, and lime and limestone absorbents are given, and the role
of fly ash in the trace element content of FGD sludges is identified.
7.2 POWER PLANT SCRUBBING FACILITIES, SAMPLING,
AND CHEMICAL ANALYSES
The data base for this study was generated from samples taken
from the power plant scrubbing systems listed in Table 13. A complete
chemical characterization of the scrubbing process was performed on six of
these systems, and a more limited analysis was performed on the remaining
system. These FGD systems represent full-scale units having greater than
100-MW equivalent generating capacity, as well as prototype and pilot-scale
systems of only 1-MW equivalence of generating capacity. Lime, limestone,
and double-alkali systems are represented, as are both eastern and western
coals. For the purpose of this study, the effect of these system design vari-
ables was investigated relative to the scrubber waste chemistry and the po-
tential of these wastes as an environmental pollutant source.
Appendix B describes each scrubbing facility from which samples
were obtained. Periodic samples were obtained from many of these facilities,
and the specific conditions existing for each sampling are described. In
addition, a schematic diagram of the scrubbing facilities is presented, with
an identification of the sample source locations.
7.3 CHEMICAL ANALYSES
7.3.1 Experimental Procedure
As a general procedure, samples were taken by the scrubber sys-
tem operating personnel. Some samples were filtered at the site to arrest
reactions between liquor and solid reagents. All samples were packaged in
containers and shipped to The Aerospace Corporation.
Upon receipt of these samples at the laboratory, all slurry sam-
ples were filtered through Whatman No. 40 filter paper, and the pH and
sulfite content of the filtrate were determined at once. Liquid samples were
73
-------
TABLE 13.
FLUE GAS DESULFURIZATION SYSTEMS
SAMPLED AS DATA BASE
Power Plant
Tennessee Valley Authority
(TV A) Shawnee
Steam Plant
TV A Shawnee
Steam Plant
Arizona Public
Service (APS) Company,
Cholla Power
Plant
Duquesne Light
Company (DLC),
Phillips Power
Station
General Motors (CM)
Corporation,
Chevrolet- Par ma
Power Plant
Southern California
Edison (SCE) Mohave
Generating Station
Utah Power and
Light (UPL) Company,
Gadsby Station
Scrubber
System
Venturi and
spray tower,
prototype
Turbulent
contact
absorber,
prototype
Flooded-disk
scrubber,
wetted film
absorber
Single- and
dual- stage
vcnturi
Bubble -cap
towe r
Tu rbiilcnl
contact
absorbc r.
pilot plant
Venturi and
mobile bed.
pilot plant
Scrubbing
Capacity,
MW (equiv)
10
10
120
4 10
32
< 1
< 1
Coal
Source
Eastern
Eastern
Western
Easte rn
Eastern
W e s te r n
Western
Absorbent
Lime
Limestone
/ Limestone,
fly ash
Lime
Soda ash,
lime
Limestone
Soda ash,
lime
74
-------
stored in small polyethylene bottles and frozen. Solids samples were dried
at 70°C and stored in sealed polyethylene storage bags.
All chemical analyses were performed by standard chemical anal-
ysis techniques, in accordance with, at the very least, the minimum criteria
identified by the EPA Analytical Quality Control Laboratory (19-21). A
description of the chemical analysis methods are presented in detail in Appen-
dix C for both liquids and solids. In every case, both the accuracy and pre-
cision of the method were determined with respect to analyses of the scrub-
ber liquors.
The results of the chemical analyses are presented in tabular
form in Appendix D; slurry analyses are given in separated form, i. e., liquid
and solid. The effluents identified as "separated" are those samples which
were filtered at the sampling site. Samples identified as "retained" are those
that are filtered at Aerospace, as were those not identified.
7.3.2 Results of Chemical Analyses
The results of the chemical analyses from the sources presented
in Appendix D are summarized in Table 14. The range of chemical con-
stituents were taken from the analyses of the sludge samples after the final
stage of dewatering, when leaving to enter the disposal basin.
Several observations can be made from these data that reveal
the chemical constituent concentrations that may be expected from FGD
scrubber systems. Among the trace elements in sludge liquors, the higher
range of concentrations were generally less than 1 mg/i with only three ex-
ceptions, aluminum, arsenic, and selenium; none of these were significantly
greater than 1 mg/£ . The highest values for all others lay between 0. 1 and
1 mg/t except for mercury, which had a maximum value of 0. 07 mg/f . In
nearly each case, the total range of liquor concentration for any element lay
within two orders of magnitude. This trend appears similar to the concen-
tration for trace elements in the solids, except that the magnitude of the
concentrations were generally greater by a factor of 100 relative to the
liquor.
A correlation exists between the liquor and solids concentrations
of beryllium, cadmium, and mercury, in that the values were generally
lower in both phases, which appears to indicate that these elements were in
the sludges in relatively minor amounts. On the other hand, concentrations
in the solids were generally high for chromium and zinc (possibly, also, cop-
per), but these higher values were not reflected in the range of liquor concen-
trations. Whereas in the first case the correlation strongly suggests a con-
centration limitation based on input amounts, in the second case, the lack of
correlation more nearly suggests that solubility may have been controlling
the liquor concentrations of these elements. An apparent exception to these
correlations was selenium, which had the broadest concentration range in
the liquor (nearly four orders of magnitude) but had the narrowest concentra-
tion range in the solid (factor of eight).
75
-------
TABLE 14. RANGE OF CONCENTRATIONS OF CHEMICAL
CONSTITUENTS IN FDG SLUDGES
Scrubber
Constituent
Aluminum
Arsenic
Beryllium
Cadmium
Calcium
Chromium
Copper
Lead
Magnesium
Mercury
Potassium
Selenium
Sodium
Zinc
Chloride
Fluoride
Sulfate
Sulfite
Chemical Oxy-
gen Demand
Total Dissolved
Solids
PH
Sludge Concentration Rangea
Liquor,
(except
0.03
<0. 004 -
<0.002 -
0.004 -
180
0.015 -
<0.002 -
0.01
4.0
0.0004 -
5.9
<0.0006 -
10.0
0.01
420
0.6
600
0.9
<1
2800
4.3
mg/|
PH)b
2. 0
1.8
0. 18
0.11
2600
0. 5
0.56
0. 52
2750
0.07
100
2.7
29,000
0.59
33,000
58
35, 000
3500
390
92, 500
12.7
Solid, ppmc
_
0.6
0.05
0.08
105, 000 -
10
8
0.23
-
0.001
-
2
-
45
-
-
35,000 -
1600
-
_
_
52
6
4
268,000
250
76
21
-
5
-
17
(4.8)
430
(0.9)
-
473, 000
302,000
-
_
-
Data derived from Appendix D
Liquor analyses were conducted on 13 samples from 7 power plants burning
eastern or western coal and using lime, limestone, or double-alkali
absorbents.
%
'Solids analyses were conducted on 6 samples from 6 power plants burning
eastern or western and using lime, limestone, or double alkali.
76
-------
Among the major species, the concentrations of sulfate and
calcium are dependent upon the solubility of gypsum. When relatively high
values of sodium or chloride are present, they result principally from
additions of either absorbent or makeup water. As the major FGD waste,
sulfur can exist either in the sulfate or sulfite phase. As may be expected,
the solids sample having the highest sulfite content also had the lowest sulfate
content.
One of the most important parameters in the liquid portion of the
sludge is the total dissolved solids (TDS). By the nature of the chemical
reactions that take place in the scrubber, the value for TDS can be expected
to be reasonably high. For the lime and limestone systems, TDS generally
approximates 10, 000 mg/£. Higher values were observed in the double-alkali
sample and in a sample in which salt was added to the makeup water. Lower
values were observed in two systems during startup and in a third system
operating in a partial open-loop operation.
7.3.3 Determination of Material and Ionic Charge Balance
To ascertain whether liquor analyses are complete and internally
consistent, determinations of material and charge balances are of consider-
able importance. Material balance is obtained by comparison of the sum of
the ion concentrations with the concentration of TDS. A deficiency of the
former relative to the latter may be indicative of the presence in the liquor
of soluble species for which analyses have not been conducted. However,
because of limitations concerning the accuracy of analyses, discrepancies
may exist between the results of these measurements for individual samples.
Examination of the results of measurements for a number of similar samples
is often necessary to determine whether true deficits exist. If the sum of the
ion concentrations is significantly greater than the TDS, it is almost certain
that an analytical error has occurred.
An independent test is obtained by comparison of the sum of the
(equivalent) concentrations of positive ions (cations) with the sum of the
(equivalent) concentrations of negative ions (anions). Imbalances of ionic
charge alone do not unequivocally indicate where the analytical results are
in error. In conjunction with TDS data, the charge balance data are much
more useful. The signs of the imbalances help to pinpoint the discrepancy.
For example, if there is a TDS surplus and an excess of negative ionic
charge, there must be a deficit among the cation analyses. Similarly, with
a TDS deficit and an excess of positive ionic charge, one or more of the
cation analyses must be too high. If the ionic charge is in balance and there
is a material imbalance, the TDS measurement is probably in error.
Material balances and charge balances have been computed from
analyses of 40 samples of sludge liquor, and the results are shown in
Tables 15 and 16. Only two liquor samples were excluded from the available
data because analyses had not been conducted for several components. The
results have been grouped according to source and date of sampling for com-
parison of average imbalances. Six sources are represented, four of which
were sampled from two to four times each. For each sampling date, from
77
-------
TABLE 15. MATERIAL, BALANCE
-vj
oo
Sample Source
TV A Shawnee Limestone (2/1/73)
TVA Shawnee Limestone (7/12/73)
TV A Shawnee Limestone (11/27/73)
TVA Shawnee Limestone (6/15/74)
SCE Mohave Limestone (3/30/73)
TVA Shawnee Lime (3/19/74)
TVA Shawnee Lime (5/16/74)
TVA Shawnee Lime (6/27/74)
APS Cholla Limestone (4/1/74)
APS Cholla Limestone (11/7/74)
DLC Phillips Lime (10/4/73)
DLC Phillips Lime (6/17/74)
CM Parma Double Alkali (7/18/74)
All Samples
No. of TDS
Surpluses
1
2
2
0
3
4
2
4
3
2
1
3
3
30
No. of TDS
Deficits
1
0
0
4
0
1
3
0
0
0
1
0
0
10
Average Net
TDS Surplus
(or Deficit), %
- 1.0 ± 2.8
2. 3 ± 2.9
31.0 ± 14.9
- 3. 1 ± 2.5
15.6 ± 1.2
3.0 ± 6.8
2.8 ± 4.4
5.2 ± 1.8
4.2 ± 3.6
7. 3 ± 9. 3
1. 6 ± 4.0
3.8 ± 2. 5
7.6 ± 1.3
4.9 ± 9.9
Average Absolute
TDS Imbalance, %
2.0 ± 1.4
6.0 ± 3. 5
4.1+2.7
2.9 ± 2.2
--
7.0 ± 8. 5
Results of TDS measurements may be high for one or more sample.
-------
TABLE 16. IONIC CHARGE BALANCE
Sample Source
TV A Shawnee Limestone (2/1/73)
TVA Shawnee Limestone (7/12/73)
TVA Shawnee Limestone (11/27/73)
TVA Shawnee Limestone (6/15/74)
SCE Mohave Limestone (3/30/73)
TVA Shawnee Lime (3/19/74)
TVA Shawnee Lime (5/16/74)
TVA Shawnee Lime (6/27/74)
APS Cholla Limestone (4/1/74)
APS Cholla Limestone (1 1/7/74)
DLC Phillips Lime (10/4/73)
DLC Phillips Lime (6/17/74)
GM Parma Double Alkali (7/18/74)
All Samples
No. of Cation
Deficits
1
2
0
4
0
5
4
0
1
2
1
3
0
23
No. of Anion
Deficits
1
0
2
0
3
0
1
4
2
0
1
0
3
17
Average Net
Cation (or Anion)
Deficit, %
4. 1 + 11.5
20. 5± 4.4
- 0.9 ± 0. 3
9. 5 ± 5.2
- 1.7 ± 0.6
14.8 ± 9.9
3.0 ± 5.0
- 8.8 ± 8.0
- 2.0 ± 5. 3
41. 3 ± 17.6
- 3.9 ± 6.6
29.7 ± 5.9
-12.0 ± 9.4
6.4 ± 15. 3
Average
Absolute
Ion Imbalance, %
8.2± 5.7
--
--
--
--
--
4. 7 ± 3.0
--
3.7 + 3.6
--
4.7 ± 5. 5
--
--
11. 7 ± 11.9
Comments
a
-- -
--
.-
..
b.c
--
--
a,c,d
--
e
--
-vl
vO
Results of chloride analyses may be high for one or more samples.
Results of sulfate analyses may be high for one or more samples.
"Results of calcium analyses may be low for one or more samples.
Results of sodium analyses may be low for one or more samples.
'Results of magnesium analyses not included.
-------
two to five samples are included. In Table 15, the average net TDS surplus
or deficit is given, together with the corresponding standard deviation. Av-
erage absolute (disregarding sign) TDS imbalances are also given, together
with the standard deviation where both surpluses and deficits occurred within
a sample group. For all 40 samples, the average net TDS surplus was 4.9
percent with a standard deviation of 9.9. The corresponding average abso-
lute TDS imbalance for all 40 samples was 7.0 percent with a standard devi-
ation of 8.5. Only 2 of the 13 groups showed net TDS deficits, although 10
of the 40 individual samples showed deficits, for a ratio of TDS surpluses to
deficits of 3:1. For one group of two samples, with an average TDS imbal-
ance of 31 percent, TDS measurement errors are indicated, as shown in the
comments column of Table 15, since the ionic charge was in balance for both
samples. For one group of samples, the measured TDS ranged from 92, 000
to 95,000 mg/^. Since the analyses give, at most, three significant figures,
the observed imbalances of 14 to 17 percent for the three samples of this
group constitute better agreement than justified by the accuracies of the data.
Only for 1 sample, of the remaining 35 samples, was the TDS imbalance
greater than 10 percent. Therefore, it may be concluded, tentatively, that
no major soluble species has been overlooked in the analyses.
The results of ionic charge balance calculations are shown in
Table 16. For the 40 samples, the average absolute ion imbalance was 11.7
percent with a standard deviation of 11.9. There were 23 samples with cation
deficits (anion surpluses) and 17 samples with anion deficits (cation sur-
pluses). The average net cation deficit (anion surplus) for all samples was
6.4 percent with a standard deviation of 15.3. Seven groups showed average
net cation deficits, and six groups showed average net anion deficits. Thus,
no systematic analytical errors are likely. Nine of the 40 samples showed
charge imbalances greater than 20 percent. These nine samples were mem-
bers of the five groups with average absolute imbalances greater than
10 percent.
The results of analyses of each of these nine samples were ex-
amined and compared with the results for other samples within the same
groups or other groups representing the same source sampled at different
dates. On this basis, each of the imbalances greater than 20 percent can be
attributed to one or more of six explanations shown in the "Comments"
column of Table 16.
Additional analyses for magnesium and repetition of questionable
analyses to complete the listed corrections have not been possible, princi-
pally because of the depletion of sample supplies. However, it may be con-
cluded from the results which have been given that satisfactory ionic charge
balances have been obtained for 85 percent of the liquor samples that were
analyzed. Together with the satisfactory material balances shown for 90 per-
cent of the samples analyzed, the results of the charge balance calculations
demonstrate an excellent validation of the analytical data which have been
reported.
80
-------
7.4 EFFECT OF SCRUBBING PROCESS VARIABLES ON
SCRUBBER LIQUORS
The effects of scrubbing process variables on the chemistry of the
scrubber system was determined by evaluation of the results of the chemical
analyses presented in Section 7.3. The variables that were investigated are
(a) in-process variations in the scrubber liquor composition as the liquor
leaves .the scrubber (or scrubber recirculation loop) to the point of discharge
from the system; (b) variation in composition of system liquor as a function
of time, as determined from periodic sampling; (c) effects of liquor pH or
ionic strength on the liquor composition; (d) effect of differences in chemistry
of limestone, lime, and double-alkali scrubber systems on the trace element
content of sludge liquors; (e) influence of the trace element content in coal on
the trace element content in the sludge; (f) effect of coal source (i.e., eastern
or western) on the trace element content in the sludge; and (g) relationship
between the trace element content in fly ash and the trace element content in
the sludge.
The discussion in the following subsections are based on the
chemical analyses performed on specific samples. In most cases, the
explanations offered are believed to be generally applicable to all like systems,
However, it is recognized that these data are limited and interpretation of
these data could change in the light of a larger data base.
7.4.1 In-Process Variations in Scrubber Liquor Composition
An evaluation was made of the changes that take place in the
process stream of scrubber liquors, from the scrubber effluent to the final
filtering or dewatering. The questions to be answered are as follows:
a. What chemical constituents undergo systematic variations,
increases or decreases, in the scrubber liquor system
circuit?
b. Do in-process variations depend on the scrubber system
involved (limestone, lime, or double-alkali) ?
c. To what mechanism may changes be ascribed (synergistic
change, coprecipitation, or solution of constituents from
fly ash)?
Data were compiled for all major, minor, and trace constituent
analyses for five scrubber systems, in the form of graphs (presented in
Appendix E) showing the concentration of each constituent, from the scrubber
through intermediate process steps to the final filtration or clarifying opera-
tion. Figure E-I in Appendix E presents data for the TVA Shawnee (turbulent
contact absorber (TCA) and venturi systems, and for the GM, Parma, the
DLC Phillips, and the SCE Mohave systems. The APS Cholla scrubber
system was not included in this analysis, as it is actually a parallel, two-
stage process, and further sampling would have been required to character-
ize either stage.
81
-------
TABLE 17.
NET CHANGE IN SCRUBBER LIQUOR COMPOSITION OF MAJOR,
MINOR, AND TRACE CONSTITUENTS BETWEEN INITIAL AND
FINAL STAGES IN SCRUBBER SYSTEM
Constituent
Aluminum (Al)
Antimony (Sb)
Arsenic (As)
Beryllium (Be)
Boron (B)
Cadmium (Celt
Calcium (Ca)
Chromium (Cr)
Cobalt (Co)
Copper (Cu)
Iron (Fc)
Lead (Pb)
Magnesium (Mg)
Manganese (Mn)
Mercury (1 lu.)
Nickel (Nil
Potassium (K)
Selenium (Se)
Silicon (Si)
Silver (ApJ
Sodium INal
/.int C/.nt
Chloride (Cl)
Fluoride (F)
Sulfate (SO4)
Sulfite (SOj)
TDS
pH
Limestone
Increase
X
XX
XX
XX
X
XXX
XX
X
XX
XX XX
Decrease
XXX
X
X
XX
X
XXXX
X
X
X
XX
X
XX
X
X
XXX
XXXX
XXXX
XXXX
No Sipnifi-
cant Change
t- ZO"',.)
XXX
XXX
XXX
XX
XXX
X
XX
XX
XXXX
X
X
XXXX
Lum-a
Inc rease
X
X
X
XX
XX
XXXX
Decrease
XX
X
XXX
X
XX
XX
XX
XXX
X
X
X
X
XXX
XX
XXX
XX
XX
XXX
XX
No Siqnifi-
cant Chani:e
I 20"'..)
X
XX
X
XXXX
XX
xx-
xx
XX
X X X X
XX
XXX
X
XXX
X
XXX
X
XX
XX
XX
X
Double Alkali3
Inc rease
X
X
X
X
X
X
X
X
No SigniTi-
cant Change
« 20"'.,)
X
X
X
X
X
X
X
X
X
00
ro
Each "X" represents a separate sample set.
-------
From the graphic array of data in Appendix E, Table 17 was
prepared, showing the overall increases or decreases in each constituent
passing through the circuit. It is useful to define how the chemistry of solu-
tions is expected to affect each element or chemical species in the scrubber
liquor. These fall into four groups:
a. Metals which are at least moderately soluble in the scrubber cir-
cuit, defining this term to include the alkaline earths calcium and
mercury, despite the relatively low solubilities of their sulfate
and carbonate compounds. The alkali metals potassium and
sodium are in this category.
b. The greater number of heavy metals, including the transition
metals, which form insoluble oxides, hydroxides, and sometimes
carbonates. These are most likely to respond to changes in pH,
becoming less soluble as pH increases.
c. A few metalloids, arsenic, boron, selenium, and silicon, which
form weakly acidic oxides and at high pH, soluble sodium salts.
At low pH, only silicon oxide is highly insoluble; the others,
arsenic, boron, and selenium would not be pH sensitive.
d. The anions chloride, fluorine, sulfate, and sulfite whose salts are
generally soluble but, except for chloride, form insoluble calcium
compounds. A few trace metals (silver, molybdenum, and mer-
cury) form insoluble chlorides, but their quantity is insufficient
to reduce the chloride level substantially. Conversely, the
presence of these anions might affect certain cations; for example,
a high SC*4 level may reduce the lead content by formation of
PbSO4.
The following discussion presents on an element-by-element basis
the patterns of depletion or enrichment that occurred for each scrubber system.
Aluminum
A uniform decrease in aluminum occurred; this correlated with
a rising pH and indicates the precipitation of A1(OH)3- Aluminum actually
reaches its minimum solubility at a pH of about 6. 5 and at a higher pH may
be slowly redissolved by hydrolysis. This effect may be present in the TVA
Shawnee lime process samples for 16 May 1974, which show.ed an initial
decrease in aluminum content, followed by a smaller increase.
Arsenic
Arsenic showed an increase in one sample set from the limestone
series and a decrease in four sample sets from both the lime and limestone
systems, with four sets remaining unchanged (within 20 percent). The ar-
senic in the GM Parma double-alkali system, which operates at a high pH,
was below the detection limit of 0.004 ppm, even though arsenic at 14 ppm
83
-------
was substantial in the GM Parma coal. It is likely that a good deal of
escaped from the stack, and the amount that dissolved in the scrubber
underwent little change in concentration.
Antimony
Antimony showed a decrease in one sample each of the lime and
limestone series and no change in two other sample sets from the lime sys-
tems. Noted decreases were small and indicates a slight tendency toward
removal in the system.
Beryllium
In the lime and limestone systems, no sample set showed an
increase; two showed small decreases; and seven remained unchanged. Thus,
the overall process underwent no significant change in beryllium. The beryl-
lium content of the double-alkali system was below detection (<0.005) although
the coal content, 1 ppm, was the highest in beryllium of those analyzed.
This may reflect precipitation of the oxide at high pH, although no pH effect
could be discerned for the lime and limestone systems.
Boron
Boron decreased in the lime system, but this was not a response
to a change in pH. It may be evidence of the removal of an insoluble calcium
borate.
Cadmium
Cadmium showed no pattern of changes for the limestone system
samples, but in the lime system samples, there were two decreases in
cadmium and two with no change. This is attributable to the precipitation
of the insoluble hydroxide, Cd(OH)2> with increasing pH. The fact that it
did not take place at all times may indicate the importance of removal pro-
cesses, i.e., coprecipitation and scavenging.
Calcium
There was a general pattern of decrease in the calcium content
for the lime and limestone systems and a slight rise in calcium for the
double-alkali system. The decrease was in accordance with the slow growth
of crystals of insoluble calcium salts to the size where separation occurs.
It was not paralleled, however, by a decrease in sulfate content and can not
be fully explained without reference to the role of bisulfite in controlling the
calcium content.
Chromium
There were two increases in chromium content in the limestone
system samples, one decrease, and two with no change. In lime system
samples, the proportions were one increase and two with no change; in the
84
-------
double-alkali system, the chromium content was below detection (<0.02 ppm).
This element, then, showed no evidence of a systematic in-process change
in any of the three systems.
Cobalt
' There was a decrease in the cobalt content of the only sample set
representing the limestone system. The lime system samples showed an
increase in cobalt in one sample set, with a second sample set remaining at
the same level. It is possible that, for the limestone system, the in-process
addition of a limestone slurry depressed the cobalt level by formation of the
insoluble CoCOo.
Copper
There was no general change in the copper content for any of the
three systems.
Iron
Iron decreased in the single analysis of a limestone system sam-
ple set and in three lime system sample sets. Forming an insoluble hydrox-
ide, it would be expected to show a decrease with the downstream increase
in pH.
Lead
For the limestone system, two sample sets showed an increase
in lead content and two showed a decrease. The lime system samples showed
no change in the lead content in four sets; consequently lead was considered un-
changed. For the double-alkali system, there was an increase in the level
of lead, although it would seem that, in promoting the formation of the basic
carbonate, the addition of lime and soda ash would have had the opposite
effect. There is no apparent explanation for this increase in lead unless it is
introduced in the lime or soda ash.
Magnesium
Analyses of magnesium were run for four limestone system
sample sets, and three lime sets; no appreciable change was observed in any
of these.
Manganese
Analyses of manganese were made for two lime system sample
sets; one of them showed an increase and the other a decrease in manganese
content. Overall, the manganese was considered unchanged.
85
-------
Mercury
Two decreases in Hg were noted for the limestone system, three
unchanged values for the lime system, and one decrease for the double-
alkali system. The pattern, therefore, is one of a uniform decrease in mercury
content. This element forms insoluble oxides, carbonates, and chlorides,
all of which may have contributed to purging it from the liquor.
Nickel
Nickel likewise decreased for the limestone and one lime system
sample set. A second lime system set showed no change in nickel, but the
net effect was a decrease. This is accountable as a function of pH, repre-
senting removal of insoluble nickelous oxide. The carbonate was also insolu-
ble, as was a basic carbonate, 2 NiCOg'S Ni(OH)2( which would have formed
at a higher pH.
Potassium
Changes in potassium were the same for all three systems, i.e,
a low to moderate increase to levels no greater than 20 percent except for
the double-alkali system clarifier No. 2 underflow. This increase is believed
to represent the slow dissolution of potassium from silicates present in the
fly ash carried over into the scrubber and, possibly, as a contaminant in the
soda ash of the double-alkali system.
Selenium
Selenium in four limestone system sample sets showed two
increases and two unchanged. The levels, for four lime system sample sets,
showed one decrease and three unchanged. One selenium sample set in the
double-alkali system decreased about 50 percent as it passed through the
No. 1 clarifier.
Silicon
Three lime system analyses showed a reduction in silicon content,
but only one of these was substantial (about 33 percent); the others were
slight. The decrease was predicted on the use of the scrubber-effluent-
retained sample at 4-ppm silicon, as the initial point. The scrubber-effluent-
separated sample was only 2-ppm silicon; the difference is believed to be
silicon that was dissolved from fly ash in the period of approximately two
weeks after the sample was taken. Therefore, it is likely that in the scrub-
ber circuit itself there was no real change in silicon content for this system.
Silver
Silver decreased in one limestone and two lime system sample
sets and showed no change in a third lime system set. This may be con-
strued as a net decrease ascribable to removal of insoluble chloride, oxide,
or carbonate of silver with increasing pH.
86
-------
Sodium
There was no change in sodium in two sample sets, one a lime-
stone process and the other a lime process set. A third lime process sam-
ple set showed a loss of 50 percent of the sodium content of the in-process
liquor. This is considered unaccountable.
The double-alkali system, to which sodium was added in the form
of carbonate in midprocess, was not included in this evaluation because its
value in the final waste is dependent upon the efficacy of the filter cake wash-
ing operation.
Zinc
In the limestone system sample sets, the zinc increased slightly
in one instance, decreased in threer and remained the same in another. Like-
wise for the lime system, it decreased in three sample sets, the net effect
being a decrease in zinc. It may represent the precipitation of a zinc car-
bonate, ZnCO3. For one sample set in the double-alkali system, there was
an increase in zinc content as it passed through the No. 1 clarifier.
Chloride
A decrease in Cl" occurred for all four sample sets for the TVA
Shawnee limestone system. There is no obvious mechanism for this decline
in a readily soluble ionic species, but it is noted that Cl" forms many weak
complexes with metals, including calcium, and these may be scavenged from
solution by fly ash, by precipitating calcium salts or by unconsumed calcium
sulfite, CaCOs. The lime systems showed only two small decreases in
chloride, with two sample sets remaining unchanged; evidently the chloride
removal process is much less effective in this system. Finally, the double-
alkali process indicated an in-process gain in chloride. This might result
from the introduction of small amounts of NaCl with the soda ash. On the
other hand, the systems trend from a decrease to an increase paralleled the
level of pH of the systems. For an alternative explanation, the concentration
of chloride may decrease as these ions or complexes are adsorbed on surfaces
of newly forming precipitates as in the limestone process, but at a higher
system pH the hydroxyl ion competes for these sites with no net gain or loss
as in the lime process. However, in the double-alkali system, the increase
in circuit pH may favor the adsorption of the hydroxyl ion sufficiently that
chloride ions or complexes are desorbed, for an increase of chloride in the
system liquor.
Fluoride
In the limestone system, the fluoride content showed an increase
for the two sample sets analyzed, presumably due to the slow solution of the
fluoride present in the fly ash. The lime and double-alkali systems showed
no change in chloride content, probably because the higher pH of these sys-
tems is less effective in leaching fluoride from the fly ash.
87
-------
Sulfite and Sulfate
In all cases, sulfite oxidized rapidly downstream of the scrubber,
and the resultant sulfate either increased to saturation or remained at
saturation levels (see Section 8.3.1).
Total Dissolved Solids
The in-process TDS content would be expected to parallel those
of its major constituents, calcium chloride and sulfate, and this was, in fact,
the case. TDS in the limestone systems showed decreases in all sample sets.
The lime system had two sample sets showing a decrease in TDS and one
other remaining unchanged. TDS in the double-alkali samples showed a
considerable in-process increase between the mixing Tank No. 2 underflow
and the filter effluent, correlating closely with the sulfate content.
pH (Hydrogen Ion Concentration)
Since the overall scrubber process is one of neutralization of a
sulfurous acid liquor, the pH is expected to rise from scrubber to filter. It
did so for all eight sample sets of the limestone and lime systems. For the
double-alkali system, the initial upstream point is at the No. 2 Mixing Tank,
after the addition of the neutralizing lime. Consequently, no further change
in pH may be traced in the circuit.
From the graphic presentation of data in Table 17, overall in-
creases or decreases of constituents can be observed. Summing up all
changes in concentration, it is noted that there were 27 increases, 72 de-
creases, and 78 without significant change (<20%). On an element-by-
element basis, it appears that only potassium and fluoride showed a general
increase in concentration; that aluminum, boron, calcium, iron, mercury,
nickel, silicon, silver, chloride, sulfate, and sulfite showed a decrease; and
that the remainder were unchanged or responded in a manner too subtle to be
clearly discernible as they passed through the process circuit. In all cases
(except the double-alkali process), the TDS decreased and the system pH
increased from the scrubber to the point of disposal.
On the whole, the scrubber bleed system is one in which many
elements appear to be removed by coprecipitation or scavenging, probably by
the formation and growth of sulfite or sulfate because of the rapid oxidation
of the bisulfite ion. Also, most of the remaining constituents are unchanged
in concentration as they pass through the circuit. Except for the increases
in potassium and fluoride, both attributed to leaching of fly ash, the concen-
tration of major, minor, and trace elements in the scrubber liquors tends to
decrease along the process circuit.
System dependence is apparent for magnesium for the limestone
and lime process, possibly because of the competing mechanisms involving
dissolution and precipitation as a function of pH change in the process.
Chloride ion centent is seen as a system variant for all three scrubbing
88
-------
systems, possibly also a consequence of absorption or coprecipitation. The
changes in calcium, sodium, and sulfate concentration in the double-alkali
systems are attributed to intentional in-process chemistry. However,
selenium and .zinc decreased and increased, respectively, in the double-alkali
system for which no clearly understood mechanism is apparent.
7.4.2 Variation in Composition of Sludge Liquors with Time
It is predictable that during the startup period of a scrubber
system, an increase in dissolved solids takes place. Thereafter, if the
liquid circuit is completely closed-loop, there would be a slow increase in
the more soluble constituents such as salts of sodium and magnesium, while
the less soluble ones, including salts of most of the trace elements, would
level off at their saturation values. If the circuit is not entirely closed,
allowing some solubles to escape in the pond slurry or filter cake, soluble
constituents in the recirculating liquors would reach steady state levels
determined by the bleedoff rate. Thus, to determine whether constituents
of sludge liquors undergo any significant changes in concentration during
extended periods of operating time, plots were made (Appendix E, Figure
E-2) of the major and trace constituent levels in the disposal liquor of each
set of samples as a function of sampling date. The scrubbing systems
selected for this evaluation were those for which multiple sample sets had
been analyzed over time periods ranging from 4 to 1 6 months. These plants
were the TVA Shawnee TCA process, the Shawnee venturi process, the DLC
Phillips venturi process, and the APS Cholla venturi absorption tower process.
Sample series included a set taken shortly after startup from the
TVA Shawnee lime process and one from the TVA Shawnee limestone process.
For the lime process, the first sample was taken on the third day after startup,
and an interval of about two months elapsed between the initial sampling and
the second sampling. During this period, the concentration of 15 constituents
including all major species increased in the liquors; the concentration of
3 trace constituents decreased; and the concentration of 5 trace constituents
remained practically unchanged. There was a general buildup of ionic species
between the first and second sampling shown in a 50 percent increase in TDS.
A third sample set was taken approximately two months after the
second sample set. The TDS increased an additional 15 percent, reflecting
the increasing concentration of all the major chemical species, but the num-
ber of trace constituents that increased totaled 10 compared to 9 trace con-
stituents that showed a decrease and 3 trace constituents that showed no
change in concentration. Moreover, of the 10 trace elements that showed
an increase between the second and third sampling, only 5 of these showed
an increase between the first and second sampling. Thus, among the trace
species, only lead, potassium, sodium, silicon, and fluoride ions showed a
consistent increase over the four-month period, whereas all major chemical
species increased significantly during this time.
The other scrubber system sampled in the startup period, the TVA
Shawnee limestone process, was first sampled about six weeks after an inter-
mittent period of operation in an open-loop mode. The second sample was
89
-------
taken approximately five months later. In this time period, all the major
chemical species showed an increase in concentration, reflecting an increase
in TDS of 300 percent. Among the trace species, seven increased; five
showed no appreciable change; and none showed a decrease. A third sample
was taken after 10 months from the initial sample. The TDS increased an
insignificant amount reflecting chemical stability during this time period.
However, in this time period eight trace constituents decreased; one remained
virtually unchanged; and only two trace constituents increased, chromium and
copper, both continuing their increase from the initial sampling.
The fourth sampling was taken about six months later at a point
in the TVA Shawnee experimental program when large quantities of magne-
sium were introduced into the scrubber loop. The result was a significant
increase in TDS, primarily as a consequence of soluble magnesium sulfate.
All trace metals decreased in this period, except beryllium and lead. This
latter sampling may not be appropriate for consideration in this evaluation
because of the large increase in sulfate ion concentration.
The DLC Phillips station was sampled twice, about eight months
apart. All of the trace elements analyzed in both sets of samples either
showed a decrease or were virtually unchanged. The TDS dropped sharply
during this period as a consequence of operation in a more open-loop mode.
The trace elements would be expected to decrease for the same reason.
The APS Cholla station was sampled twice about seven months
apart. In this time period, the TDS increased more than 50 percent pri-
marily as a consequence of chloride buildup. During this time, nine trace
elements increased in concentration; five decreased; and four were unchanged.
For the four power plants evaluated, it has been observed that
from startup there is a rapid buildup in the concentration of major species
that can reach relatively stable conditions at a concentration where the rate
of constituent loss in the solid waste product is exactly equal to the rate at
which that constituent is scrubbed from the flue gas. Among the trace ele-
ments, with the exception of lead, it is deduced from these data that after a
buildup period during startup there are no systematic time-related trends in
concentration levels for the trace constituents of the liquor. Small trends,
of course, may have occurred for a few elements, notwithstanding the
generally inconclusive data. The fact that trace element concentrations do
not change systematically with time nor with changes in major species con-
centration indicates that they are controlled by some factor other than
saturation chemistry.
The variation in concentration of a given element from one set of
samples to the next is ascribed principally (lead possibly being an exception)
to the effects of differences in input (coal content) to operating conditions,
and in the case of low-concentration-level trace metals to low analytical
precision. It should be noted that in the case where TDS increased because
of an increase in sulfate content, the trace elements decreased, but when
90
-------
the TDS increase was caused by chloride ion increase, the trace elements
increased. This behavior suggests that sulfate ion may be exercising some
chemical control on the content of trace elements. Insufficient data are
available to provide confirmation of this observation.
The relation of trace element concentrations of sludge liquors
with scrubber operating time, when measured in terms of several months,
is summarized as follows:
a. Trace element concentrations are not time-dependent,
barring an initial startup period.
b. Fluctuations in concentration are ascribed principally to
variations in trace element content of process ingredients
(fuel, reactant, and process water) and operating conditions.
7.4.3 Effect of pH and Ionic Strength on Concentration
The evaluation of the chemical analyses presented in Sections
7.4.1 and 7.4.2 revealed a general decrease in trace element content as pH
increased. In Section 7.4.2, it was observed that TDS (and therefore ionic
strength) increased with the time of scrubber operation. No correlation
could be detected for trace elements, except possibly for the case of lead.
On the basis of these observations, correlations between trace element con-
centrations, pH, and ionic strength were attempted to determine whether
more precise correlations of these parameters with the concentrations of
trace elements could be determined.
A plot of the concentrations of eight trace elements as a function
of system liquor pH is presented in Appendix E, Figure E-3. It is expected
that if hydroxyl ion concentration were controlling the trace element concen-
trations, a maximum concentration of the trace element would be reached at
each value of pH as a decreasing function with increasing values of pH. No
indication of direct pH control of the trace chemical species can be detected
from these data (see Section 8).
A plot of the solute concentration of trace elements as a function
of the ionic strength of the system liquors is presented in Appendix E,
Figure E-4. In this case, it would be expected that the maximum concentra-
tion of a trace element would increase as ionic strength increased. Again,
no correlation could be established.
In most scrubber systems, pH is used as a system control. In
Figure 25, ionic strength is plotted as a function of pH. The purpose was to
91
-------
determine whether the hydroxyl ion concentration is influential in controlling
the concentration of major species. (The double-alkali systems were not
included in these analyses. ) As before, no correlation is apparent from
these data.
The conclusion that must be drawn from the data, presented in
Appendix E, Figure E-3 and E-4, is that trace element concentrations within
the scrubber liquors are not controlled by either system pH or the ionic
strength of the liquor. This conclusion is in agreement with the results,
discussed previously, which ascribe trace element concentrations in system
liquors to the variations in trace element content of process ingredients and
to major system design variables as demonstrated in Section 8. It is not
apparent in any of these analyses that a level of saturation had been reached
by any of the trace elements.
7.4.4 Comparison of Selected Solute Trace Elements in Sludge
Liiquors from Limestone, Lime, and Double-Alkali
Scrubber Systems
Overall concentration levels of solute trace elements in scrubber
liquors using the limestone, lime, and double-alkali processes were evaluated
to determine whether levels differed appreciably for the three processes.
A graphic representation of the variation among specific scrubber
systems, showing median values for individual systems, is presented in
Figure 26. For example, the bar for arsenic-lime stone shows the range of
median values of arsenic within each of the three scrubber systems: APS
Cholla, SCE Mohave, and TVA Shawnee, using the limestone process.
The graph shows that in most cases there is considerable overlap
in range of trace element levels between power plants using different scrubbing
processes and different sources of coal. Comparing the lime and limestone
systems in Figure 26 for six elements shows that the range for the lime
system lies within the range for the limestone system; for three elements,
the range for the lime system lies below, but overlaps, the range for the
limestone system.
Comparing the single set of double-alkali system samples (GM
Parma) to the lime and limestone systems shows that for at least one element
the double-alkali system lies below the range for the limestone systems and,
for one other element, the double-alkali system lies above the range for the
lime and limestone systems. For all other elements, the concentration lies
always less than midway within the range for the lime and limestone system.
92
-------
o
2.
LU
o:
0.30i-
0.25
0.20
^ 0.15
o
0.10
0.05
10 11
PH
Figure 25. Relationship between liquor system pH and ionic strength.
-------
ARSENIC
BERYLLIUM
CADMIUM
CHROMIUM
COPPER
MERCURY
LEAD
SELENIUM
ZINC
+O
" " o-* ~ ABSORBENT:
*~^ LIMESTONE
0 n__A D"~ " u IVA ^HAWNbb
^<>0~~* SCE MOHAVE
rni ADC r\ \r\\ 1 A
,_, ini «"j LMULLA
O
O LIME
TVA SHAWNEE
"" . o DLC PHILLIPS
0
g_ ,_..,w . , .^ DOIIRIF AIKAI 1
«D O GM PARMA
O
1-1 Kfl
O^^^^.^^^^^
O
H_^_^^.H_ ^^..^^^n
O
1 1 1 1 1
0.001 0.01 0.1 1.0 1(
CONCENTRATION,
Figure 26. Median values for trace element concentrations in sludge liquors of three
absorbent systems.
-------
If a direct comparison of individual facilities is made, other
observations are possible. Comparing the GM Parma and SCE Mohave
systems, both operate with electrostatic precipitators (resulting in low levels
of fly ash in the sludge) and comparably high ionic strengths. The concen-
trations of five trace elements are clearly greater in the SCE Mohave liquor,
and for two trace elements the concentrations are clearly greater in the GM
Parma liquor. For two other trace elements, the difference in concentrations
is not discernible. In a comparison of the TVA Shawnee limestone and
Shawnee lime, for all but one trace element the limestone liquor has higher
concentrations than the lime liquor, but for at least two of these trace ele-
ments the difference is not great.
The effect of system absorbent is summarized in Table 18, which
shows that, in nearly every case, the trace metals were highest in the lime-
stone system, intermediate in the lime system, and lowest in the double-
alkali system. The only obvious system parameter that can account for this
order of trace metals is scrubber pH, which was lowest in the limestone sys-
tems and highest in the double-alkali system.
TABLE 18. RELATIVE PROPORTIONS OF TRACE ELEMENTS IN
LIMESTONE, LIME, AND DOUBLE-ALKALI SLUDGES
As
Be
Cd
Cf
Cu
Hg
Pb
Se
Zn
Scrubber System
Limestone
1
High
High
Medium
High
High
High
Low
High
High
Lime
Medium
Medium
High
Medium
Low
Medium
Medium
Medium
Low
Double
Alkali
Low
Low
Low
Low
Medium
Low
High
Low
Medium
These specific comparisons support the general observation that
suggests that scrubber pH has an effect on the solute trace metal content but
that system pH is not generally influential. In the specific comparisons of
the two TVA Shawnee facilities, the predominantly higher trace concentra-
tions in the limestone system is particularly significant because less makeup
water is used in that specific lime system.
These comparisons show that while, in general, the trace metal
levels in limestone systems were high and those in the double-alkali system
were low, the variance among individual scrubbing facilities was as great as
or greater than the variance among different systems. Through the
95
-------
comparisons of scrubbing systems, there is support for the conclusion that
the determining factors in trace metal levels in liquors are primarily system
variables that control the input, principally the amount entering the system,
and the output that leaves the system as water. System chemistry, in particu-
lar scrubber pH, has a recognizable influence, but the effect of other system
variables, such as residence time and percent solids in the sludge in circuit
and system efficiency as determined by mechanical design, may contribute to
the establishment of the solute trace metal concentration level.
7.4.5 Relationship of Trace Element Content Between Input
Ingredients and Scrubber Sludge
In previous subsections, it has been suggested that process input
ingredients, i. e., coal, alkaline absorbent, and makeup water, may directly
affect the concentration of trace metals in the scrubber sludge. In this sub-
section, the comparison is made of the trace element content in process
ingredients and in the solid and liquid portion of scrubber sludges originating
from these materials. In determining the content of trace metals in process
ingredients, the contribution of the absorbent and process water was deter-
mined by adding to the coal a value based on the stoichiometric ratio of the
absorbent to the sulfur content of the coal. In all cases, the contribution
made by process waters was found to be insignificant relative to the contri-
bution made from coal, and the correction due to the absorbent, in most cases,
modified the coal analysis only slightly. A graphical representation of these
comparisons are made for sludge solids in Figure 27 and for sludge liquor in
Figure 28, where the trace element content is determined from the average
or median value representing a single sample set taken on a given date.
The relationship between the concentration of trace elements for
coal (corrected for absorbent) and sludge solids in Figure 27 shows excellent
agreement. The band as drawn has a slope of unity of over more than three
orders of magnitude, which indicates a directly proportional relationship
between the amount of a trace element present in the coal and that found in
the sludge.
The comparison for sludge liquors in Figure 28 is supportive of
a directly proportional relationship with the process coal, but the confidence
of this relationship is far less clear. In this case, the bandwidth is nearly
two orders of magnitude, and the point scatter is much greater than for
solids. The trace element concentration in the liquor is approximately two
orders of magnitude less than that for the solids, and some of the uncertainty
associated with precision at lower concentrations may be reflected in this
scatter. Nevertheless, allowing for this scatter, it is evident that the slope
of the data is identical to that for the solids. In general, the trace element
content of the liquor is approximately 0. 01 the trace element content of the
solids. There is no evidence in these plots for control of trace elements by
solubility limitations. If this were so, the datum points in Figure 28 for
specific elements would be expected to be along a vertical line representing
a saturation concentration independent of coal content greater than that value.
96
-------
400 r
100
Q.
10
a
UJ
"J-1 1
1
0.1
0.01
LEGEND:
O ARSENIC MERCURY
D BERYLLIUM 0 COPPER
A CADMIUM 9 LEAD
CHROMIUM B SELENIUM
^ ZINC
i i i
I
i i i i
0.1 1 10 100
AVERAGE TRACE ELEMENT CONTENT OF SLUDGE SOLIDS, ppm
1000
Figure 27. Average trace element content of sludge solids.
-------
200
100
E
o.
o.
O
O
10
00
a
LJJ
o 1
0.001
e
n
ARSENIC
BERYLLIUM
CADMIUM
CHROMIUM
COPPER
MERCURY
LEAD
SELENIUM
ZINC
0.01 0.1 1.0 10
AVERAGE TRACE ELEMENT CONTENT OF SLUDGE LIQUOR, mqll
100
Figure 28. Average trace element content of sludge liquor.
-------
The correlation between coal and sludge is especially significant
when evaluated with respect to the range of 3 to 60 percent fly ash in the
sludge sampled. Since fly ash is a major constituent in most of the' sampled
sludges, it would be expected that sludges having higher values of fly ash
would have higher values of the trace elements within them. When evaluated
point for point, it is revealed, as a general rule, that those sludges with high
fly ash content tend to lie on the right side of the band, whereas low fly ash
sludges are more generally found on the left side of the band. It is presumed
that if the trace element content for each sludge were corrected for relative fly
ash content, the data bandwidth would measurably narrow.
In addition to the above considerations, the trace element content
in sludge liquors has already been shown to be affected within a system by
scrubber pH. If the trace element content for the sludge liquor could be cor-
rected for the differences in pH, a further reduction in data scatter would be
evident, and the relationship with coal content would be much more apparent.
In summary, the trace element content of sludge solids and liquors
are directly dependent upon the trace element content in the process ingre-
dients, particularly coal; data scatter is explained as a function of the varia-
tion in fly ash content in the sludges. In addition, scatter in the liquor data
is expected because the scrubber pH has been shown, Section 7.4.4, to in-
fluence (but not control) the concentration of trace metals in the system
liquo r s.
7.4.6 Distribution of Trace Elements in Coal Through Combustion
In Section 7.4.5, a correlation was observed between the trace
elements concentrated in the FGD sludge and those originating in the coal.
The relationship shows that the concentration of trace elements in the sludge
had the same relative concentration as that originating in the coal. The band-
width of the data scatter for solids is about one order of magnitude, and the
range of fly ash in the sludges sampled also varied by about one order of
magnitude, suggesting that fly ash may be the major source of trace metals
in the sludge. Moreover, in evaluating the data bandwidth in Section 7.4.5,
it was revealed that sludge with high fly ash content tends to lie on the high
side of the band and low fly ash sludge was more prevalently on the low side
of the data band.
Chemical analysis of both coal, bottom ash, and fly ash are pre-
sented in Appendix D. These data do not provide a sufficiently large data
base to perform a materials balance. However, extensive work has recently
been reported in the literature (2225) that can be used for comparison with
this study. Since the contribution of trace elements in the sludge are seen to
be a consequence of the combustion products of coal, only those elements in fly
ash or those existing as vapor in the flue gas can enter the scrubber. No
available data even remotely suggest the scrubbing efficacy of the different
element vapors from flue gas by a desulfurization scrubber. Thus, only the
concentration of trace elements in the fly as'h can be used in comparison with
trace element concentrations in sludge.
99
-------
Trace elements are also found in bottom ash and some elements
will pass through the system as vapor in the flue gas. Many of the nine trace
elements of primary interest in this study (because of limits to them in drink-
ing water criteria) show an inproportionately high contribution to the flue gas
stream, either as vapor or ultrafine particulates (typically <0. 1 [am), based
on measured concentrations in the wastes relative to their total concentration
in the coal. Of these, mercury in particular is found preferentially in the
flue gas (as are sulfur, chlorine, and fluorine). Additionally, arsenic,
selenium, cadmium, and chromium appear to be preferentially in the flue gas.
Those elements that as oxides are easily and readily found as glass modifiers,
such as lead, beryllium, zinc, and copper, are concentrated in the fly ash
fraction.
The results of this study are consistent with the trace element dis-
tribution studies found in the literature and show that the major source of trace
elements found in the sludge originate in the fly ash and possibly, in some
cases, from the flue gas. The previous section reveals that a small fraction of
the trace element content is in an unsaturated state in the sludge liquor. It is
presumed that these metals originate by leaching from the fly ash during the
more acid cycle of the scrubbing operation and that the higher concentrations
of trace elements in low pH systems reflect this leaching process.
7.4.7 Comparison of Trace Elements in Eastern and Western Coals
The correlation between trace elements in sludge and coal demon-
strated in Section 7.4.5 suggests that the trace element content expected in
eastern and western scrubbing liquors may differ as a consequence of differ-
ences in trace element content of the coals. Figure 29 presents a comparison
of the trace element content between the coals analyzed in this study with
the analysis of coals by region performed by the U.S. Bureau of Mines (26),
U.S. Geological Survey (27), and Illinois State Geological Survey (ISGS)(24).
The bar graphs for coal are in the 90+ percent range as presented in a recent
EPA report (25); the dots represent the average values from the ISGS study.
Individual coal analyses from this study are grouped by region and identified
by letter.
The comparison of literature analyses in Figure 29 show that the
trace element content from the literature for the Appalachian region coal and
interior region coal are nearly identical for most elements in both average
value and range, with the exception of cadmium and zinc, which are signifi-
cantly higher in the interior region coals, and lead where disagreement
among investigations is apparent. Comparing the literature values for
eastern coal (combined Appalachian and interior regions) with western coal,
it is seen that western coals tend to have lower trace element contents in
general and significantly lower contents for arsenic, cadmium, mercury,
and zinc.
When comparing coal analyses of this study with the literature
analysis in Figure 29, it is seen that, with few exceptions, the range of mea-
sured values falls within the range for each element as reported in the litera-
ture. Moreover, in most cases, the average value from this study lies within
100
-------
AEROSPACE DATA
0 * n
ARSL-NIC l
fm "* )
BERYLLIUM < * *
X. A °
r AIMVAI i i/v/i ]- Bafl H n
CADMIUM 1 .f._. « 4 T
C j <
CHROMIUM r~
r
COPPER rJ
r; ._
..m.Piiru/ ₯ n ea Hn
MERCURY '
L_£J_ T
i r A Pk .
LtAU [
Q
SEIEHIUM * V
SrltNIUM L_
_-. .. ' 1
C.,. ,
ZINC <
1
1 1 1
0.01 0.1 1
COAL SOURCE:
n nncfln a TVA SHAWNEE
1 * * ' DLC PHILLIPS
HOA/I DA DA/I A
blVl rAKIVlA
<> APS CHOLLA
err AAnuA\/c
but ivlvJnAVt
D 1 ITTDATIIDC r\ATA
LITERATURt DAIA
f> + KANbt Ur LUAL
R n" m ANALYSES-
o "I APPAIA^HIA^
1 1** ^^
n_.., <> ct^r INTERIOR
1 1 KB i n . ^ \rt/F^TFRM RFfi
' (Rpfs si ihro
MEAN VALUE (
° MEAN VALUE
pan CPnCB
_.., - _JL. j
i j
Si
* BM °n«n nnn
o r T
1 O
1 1 1
10 100 1000
CONCENTRATION, ppm
Figure 29. Comparison of coal analyses with literature data.
-------
a factor of two of the average value reported in the ISGS (24) or EPA (25)
study. These data show that measured values from western coal for arsenic,
cadmium, mercury, and zinc were lower than the measured values from east-
ern coals. No other element showed a significant difference between regions
except for copper, which had high values in the western coal. These high
values may be explained by the colocation of the coal fields studied in this re-
port with a major copper-producing region.
Through a comparison of eastern and western coals, one may ex-
pect trace element concentrations in the sludge from these coals to differ
only in the content of arsenic, cadmium, mercury, and zinc among the ele-
ments studied. This study has confirmed this expectation and suggests that
the concentration of any element would be found in the sludge at a value re-
lating directly to its concentration in the coal.
102
-------
SECTION VIII
CHEMICAL SOLUBILITY ANALYSES
The composition of the effluent from the lime and limestone wet
scrubbing system is determined by the interactions of species originating
in the lime or limestone, flue gas, fly ash, process water, and air.* In the
liquid phase of the recirculating system, solubility is the primary factor
which limits the concentration of each component. From a knowledge of the
interaction of the chemical species in each scrubber system and their solu-
bility product constants, it is possible to predict the maximum concentration
of each species as determined by its solubility under the specific chemical
conditions of the scrubber liquor. A minimum of analytical data is needed,
consisting of the pH, the ionic strength, and the concentration of only one
principal species with which the liquor is saturated. Although the calcula-
tions are not complex, it is advantageous to use the computer to obtain itera-
tive solutions to the equations. The Aerospace Corporation computer pro-
gram SCRUB was designed to predict maximum concentrations of trace
chemical species as a function of the concentrations of major species from
solubility data. A program written by the Radian and Bechtel corporations
was similar in that it was designed to predict major species concentrations
in a scrubber system from thermodynamic data, but the trace chemical
species concentrations were not predicted.
8. 1 BASIS OF PROGRAM MODEL
Computer program SCRUB was written specifically to calculate
the composition of a liquor of known pH and ionic strength and a known con-
centration of either calcium, or sulfate, or total sulfite, or total carbonate.
The liquor was presumed to be saturated with these ions as well as with nine
other elements selected because of their potential adverse health effects:
beryllium, cadmium, chromium, copper, mercury, lead, zinc, arsenic,
and selenium. For sulfite and carbonate, two options were available.
Either the liquor was assumed to be saturated with one or both of these ions,
or the liquor was in equilibrium with a specified partial pressure of either
SOz or CO2 gas, or both. At high pH the saturation assumption is necessary,
while at low pH gas phase equilibrium is required to avoid an unrealistic ally
high total CO2 or total SO2 content of the liquor. In addition, the liquor may
contain magnesium, sodium, or chloride at concentrations which must be
Atmospheric ©2 and CO2 interact with the system liquors where there is
prolonged exposure such as in reaction tanks, holding tanks, and clarifiers.
103
-------
specifically measured. Either potassium or nitrate is supplied by the
program in an amount necessary for electrical charge neutrality. These
latter five species do not enter into the ionic equilibria, except for magnesium
at high pH, but contribute to the total ionic strength of the liquor and must be
considered for electrical charge balancing.
In order to limit the size and complexity of the program, a number
of simplifying assumptions have been made in modeling the liquor system. It
is known that some of these assumptions will result in systematic inaccuracies,
but the intent is to determine the approximate upper limits on the amounts of
the trace elements set by their individual solubilities in the liquor. Therefore,
it is considered unnecessary to include those factors which would have only
minor effects on the solution equilibria but which would add to the complexity
of the program. For example, the solubilities of the four calcium compounds
of primary interest in the scrubber liquor, namely carbonate, hydroxide,
sulfite, and sulfate, have been shown to be only slightly dependent on tempera-
ture in the 0 to 100°C temperature region (28). Neither is the water solubility
of SC>2 strongly dependent on temperature. Since the scrubber liquor will be
within this temperature range and the exact temperature at the time of
analysis is usually not recorded, it is unnecessary to include temperature
corrections to the solubility product constants. Another assumption
is that intermediate ion-product equilibria can be ignored in establishing
solubility product constants. This assumption was also made in most of the
older measurements found in the literature*; therefore, as long as the total
amount of each soluble ionic species is being considered, the calculated
results will be valid. By using experimentally determined expressions to
correct the solubility product constants for the effect of ionic strength of the
liquor, it is unnecessary to use activities and activity coefficients in the
solubility calculations. The mathematical demonstration of the validity of
this simplified treatment of the thermodynamics of these systems is pre-
sented in Appendix F.
Program SCRUB contains the additional assumption in that the
expected maximum concentration of the cations is the minimum value soluble
in the presence of the given concentrations of the anions: hydroxide,
carbonate, sulfite and sulfate. Similarly, for the above four anions, as well
as the four anions involving the elements arsenic and selenium, concentra-
tions would be calculated using the solubilities of the calcium and magnesium
compounds, and the smaller of the two values would be taken as the maximum
solubility. This iterative procedure was necessary because the literature
does not contain all of the solubility product constants Kgp necessary to com-
plete the 10 X 10 matrix of anions and cations. Furthermore, the wide range
of Ksp values would result in a very "ill-conditioned" matrix which, upon
inversion would produce useless solutions to the equations, e.g., negative
values of ionic concentrations. In every case, account was taken of the
effects of pH on the anion concentration and of ionic strength on the solubility
product constant.
-I.
Most ionic interactions are the sources of deviation from ideal behavior of
electrolytes, which are accounted for empirically by the mean activity
coefficient.
104
-------
8.2 OPERATIONAL DESCRIPTION
A brief operational description 6f program SCRUB is given
herein starting with a typical case in which the pH, the ionic strength, the
partial pressures of CO£ and SO^, the calcium concentration, and the con-
centrations of chloride, magnesium, and sodium are specified input data.
All input and output concentrations are in units of milligrams per liter.
The partial pressures of the gaseous species are expressed in atmospheres.
After checking the selected options, all concentrations are converted to
moles per liter; the hydrogen and hydroxide ion concentrations are obtained
from the pH; and the calcium and hydroxide concentrations and calcium and
chloride concentrations are compared so that the input data will not be in
conflict with the respective solubility products. Provisional values of the
total sulfite and total carbonate are computed from the partial pressures
of CO2 and SC>2, respectively. Sulfate, total sulfite, total carbonate, total
arsenic, and total selenium concentrations are calculated from the calcium
and again from the magnesium concentrations using the respective values
of K° . The smaller of the two results for each anion is retained in each
case. The concentration of each of these anions, including hydroxide, is
then used to calculate the solubility of each of the eight cations (excluding
calcium and magnesium), and the smallest value for each of the cations is
retained. The charge is balanced by the introduction of sufficient nitrate
or potassium. If the ionic strength has not been specified, it is computed so
that acid dissociation and solubility product constants corrected for the effect
of ionic strength can be determined. These values are now used in place of
the K° values in a repetition of the calculations beginning with the calcu-
lation of H+ and OH" concentrations from the pH. If the ionic strength has
not been specified, the calculation of each ion concentration (except for the
input data for calcium, magnesium, chloride, and sodium) is iterated until
the calculated ionic strength is changed by less than 2 percent, whereupon all
the ion concentrations are printed.
The operation is quite similar if the concentration of one of the
anions, i.e., sulfite, carbonate, or sulfate, is the specified input quantity
rather than calcium, except in this case the calcium concentration is calcu-
lated from the solubility product constant of the given anion together with its
corresponding input concentration. The remaining calculations are the same
as for the preceding illustrative case; however, in this case the input anion
concentration is the only value that is not altered by the iterative calculations
of solubility.
If experimental values of the concentrations of calcium and one
or more of the appropriate anions, as well as the pH, are available, each
input concentration is treated as a separate case, and the results are
averaged to obtain a calculated composition which may be compared with the
experimental data.
105
-------
8.3 TEST OF PROGRAM
A test of the validity of the described program was made by
comparing the experimentally determined analytical data for 47 samples of
scrubber liquors taken from various locations at 6 power generating stations
with the maximum concentrations of major chemical species computed with
SCRUB by using experimental values for calcium and for sulfate as input.
It was assumed that each liquor was saturated with calcium sulfate (gypsum),
although the compositions of the liquors varied considerably. Calcium con-
centrations ranged from 200 to 3100 mg/jg ; sulfate concentrations ranged
from 700 to 35,000 mg/Jt ; total sulfite concentration ranged from a negligible
amount to 3500 mg/i; pH range from 2.7 to 12.8; and ionic strength ranged
from 0.06 to 1.6. Agreement between experimental and calculated compo-
sitions over these wide ranges of conditions indicates that the program
method and thermodynamic constants are valid.
A second test consisted of comparing the calculated concentra-
tions of major species for the same 47 samples with similar calculations of
composition resulting from a programvoriginally written by The Radian
Corporation (29) and later modified by the Bechtel Corporation. This pro-
gram only calculates the concentrations of the major species, and the
methods used differ markedly from those of the Aerospace program SCRUB.
However, by comparing the experimental values for calcium and sulfate with
the results calculated with the two independent programs, it is possible to
corroborate the assumption of saturation and to check the validity of the
approximations made in SCRUB. The Radian-Bechtel (R-B) program makes
few approximations and solves a system of nonlinear simultaneous equations
to obtain the best-fit composition of the sample. For this reason, more
input data are required. Corrections are made for temperature, pH, ionic
strength, activity coefficients, and water of hydration. As a result, the
program requires considerably more internal theoretical information.
8.3.1 Comparison of Experimental and Calculated
Major Species
Calculated calcium and sulfate concentrations were compared
with measured values for each of the 47 liquor samples. For the 47 samples,
the mean ratio of the experimental calcium ion concentration to the calcium
ion concentration calculated with the R-B program was 1.01. This ratio
ranged from a high value of 2. 19 to 3. low of 0.30, and the standard deviation
was 0.33. The mean ratio of the experimental sulfate ion concentration to the
sulfate ion concentration calculated with the R-B program was 1.02. The
high value of this ratio was 2.00; the low value was 0.44; and the standard
deviation was 0.28.
With the Aerospace program SCRUB using the data from the
same 47 samples, the mean experimental-calculated concentration ratio was
1.11. The high and low values of this ratio were 2.37 and 0.39, respec-
tively, and the standard deviation was 0.43. With the Aerospace program,
the experimental calculated concentration ratios were equal for both calcium
106
-------
and sulfate ions since both ratios are identical to the product of the
experimental calcium ion concentration and the experimental sulfate ion
concentration divided by the corrected solubility product constant.
The measured concentration of either calcium or sulfate has a
direct influence on the calculated concentration of the other. If the measured
concentration of either calcium or sulfate is seriously in error, then the
calculated concentrations of both ions will not agree with their experimental
counterparts. Therefore, it is possible to identify analytical data that are
likely to be in error. For six samples, results of calculations with both
computer programs showed that the analyses of six species were probably
in error. When the values of these six concentrations were arbitrarily
altered to be consistent with the data base, the mean concentration ratios
were decreased to 1.04 (±0.29) with the Aerospace program and to 1.01
(±0.26) with the R-B program.
The constancy of these ratios and their close approximation to
unity furnish good evidence that the liquors were saturated with calcium
sulfate. In only a few samples was there similar evidence of saturation with
calcium sulfite. No comparable analytical data were available to ascertain
whether the liquors were saturated with calcium carbonate. For liquors
with low pH values, saturation with either calcium sulfite or calcium car-
bonate is improbable.
*
Figures 30 and 31 are pH plots of the experimental concentra-
tions of calcium and sulfate and the values calculated with the two programs
for direct comparison of the data for each of the 47 samples. The concen-
tration ratios were plotted against pH, ionic strength, calcium concentration,
and sulfate concentration in a search for correlations. These ratios did not
appear to depend on any of these parameters within the ranges tested, as
shown by the data in Appendix F (Figures F-1 through F-3).
8.3.2 Comparison of Experimental and Calculated
Trace Species
A comparison of experimental and calculated concentrations of
minor species may not constitute a suitable test of the validity of the
Aerospace calculations because there is no assurance that the liquors were
saturated with all, or any, of the nine trace elements. Nevertheless, these
comparisons have been made using the experimentally determined major
species concentrations as input data. In Appendix F (Figures F-4 through
F-12), the logarithms of the concentrations of the nine elements (in units of
milligrams per liter) are plotted against pH for both experimentally deter-
mined concentrations of trace elements and for concentrations calculated
with the Aerospace program. It is evident that for all nine elements the pH
of the liquor was the factor that determined the calculated solubility, at least
over a portion of the pH range. For arsenic (Figure F-4 in Appendix F),
It is not intended to demonstrate a correlation of the data with pH. The
solution pH is used as a more informative identifier than a sequential
sample number.
107
-------
o
00
3500
3000
S 2500
2000
8 1500
1000
500
CALCULATED (Aerospace)
O CALCULATED (Radian-Bechtel)
O EXPERIMENTAL
e
O
>8
35000
30000
25000
20000
15000
10000
5000
7
PH
10 11 12 13
Figure 30. Comparison of experimental calcium concentration and concentrations
calculated with two computer programs.
-------
3500
3000
s; 2500
B
o
& 2000
1500
00
1000
500
CALCULATED (Aerospace)
O CALCULATED (Radian-Bechtel)
O EXPERIMENTAL
O
35000
30000
25000
20000
15000
10000
5000
7
PH
10 11 12 13
Figure 31. Comparison of experimental sulfate concentration and concentrations
calculated with two computer programs.
-------
this pH range was below 9.5. Above 9.5, all of the arsenic was
available for precipitation as the arsenate (or arsenite) of calcium. A
similar explanation exists for the form of the calculated curve for selenium
(Figure F-ll in Appendix F). Above pH 6, the selenite concentration was
not reduced by the formation of biselenite ions. For beryllium, chromium,
copper, mercury, and zinc (Figures F-5, F-7 through F-9, and F-12 in
Appendix F), the applicable pH range extended over the entire range of
experimental measurements from pH 2.7 to 12.8. Cadmium (Figure F-6 in
Appendix F) was precipitated by hydroxide ions above pH 10 and at a lower
pH by carbonate ions originating from the atmospheric CC>2 or limestone.
Below pH 7.5, the concentration of carbonate was reduced by the formation
of bicarbonate. The calculated curve for the solubility of lead (Figure F-10
in Appendix F) was complicated by its precipitation by three agents. Above
pH 9, hydroxide ions controlled the lead concentration; between pH 9 and
pH 7, carbonate ions originating from atmospheric CO^ or limestone were
controlling. Similar to cadmium, below pH 7.5, the formation of bicar-
bonate decreased the available carbonate. Below pH 7, lead was precipi-
tated by sulfate ions, the concentration of which was reduced by the formation
of bisulfate below pH 3.
The experimental concentrations of the nine elements indicate no
apparent dependence on pH. With several exceptions which will be cited, the
experimental data for all nine elements scattered within a band which spanned
approximately two orders of magnitude from 0.4 to 2 X 10~3 mg/j?, except
for selenium (2 to 2 X 10"2 mg/Jl. ) and zinc (1 to 1 X 10~2 mg/^). For lead
and mercury, the bandwidth was approximately three orders of magnitude,
from 2 to 4 X 10~3 mg/^ for lead and from 0.2 to 2 X 10~4 mg/4 for mercury.
One sample (pH 7. 15) was disregarded since the analyses for 3 of the 9 ele-
ments were higher, by a factor of 10, than any of the remaining values for
those elements. This sample was the first (chronologically) to be analyzed,
so the results may have been in error because of uncertainties in experi-
mental procedures.
For arsenic and selenium, the experimental data are all lower
than the calculated solubility by at least three orders of magnitude. This
most probably was the result of insufficient material to exceed the solubility
product constant so that no precipitation occurred. For the other seven
elements, the experimental data are lower than the calculated curves below
a characteristic pH but are above the calculated curves at a higher pHs.
This characteristic pH varies from a low value (for mercury) of 4.5 to a high
value (for cadmium) of 10.5.
Since solubilities greater than the theoretical values do not
correspond to a condition of thermodynamic equilibrium, it is well to con-
sider alternative explanations of the data. The observed invariance of the
experimental data over this broad pH range may cause one to suspect that
the elements were not in solution. Instead, the ions of these elements may
have been adsorbed on the surfaces of dispersed colloidal particles and were
analyzed as though they were in solution. The invariance of the data with pH
would then be the result of approximately constant colloidal particle densities
for all the samples. Additional experiments are necessary to test this
hypothesis.
110
-------
Other explanations such as systematic errors in analysis, Ndue
to instrument bias or due to contamination seem unlikely since these results
are a correlation of analyses from four independent laboratories
(Appendix C).
8.4 SUMMARY AND CONCLUSIONS
Data from 47 samples of effluent from lime and limestone
scrubbers at 6 power generating stations have been analyzed and compared
with calculations based on solubility equilibria. In every case, there is
evidence that the liquor was saturated with calcium sulfate. In only a few
samples, is there evidence of saturation with calcium sulfite, presumably
because in the other samples the sulfite had ample time to be oxidized by
atmospheric oxygen. No analytical data were available to resolve the
question of calcium carbonate saturation because saturation levels were
typically lower than detection levels of carbonate.
Furthermore, for both calcium sulfite and calcium carbonate,
the question of saturation of the scrubber liquors cannot be completely
resolved by analytical data unless the initial composition of the gaseous phase
in contact with the liquor is maintained. Both sulfur dioxide and oxygen
partial pressures, along with pH and ionic strength, determine the equilib-
rium concentration of total sulfite in a saturated liquor, and carbon dioxide
has a similar role in determining the equilibrium concentration of total
carbonate.
For the nine trace elements evaluated, the experimental data give
inconclusive evidence on the question of saturation. Specifically, trace ele-
ment concentrations in liquors were invariant with changes in pH and major
anion species. Further experiments are needed to reconcile apparent
discrepancies between experimental and calculated results.
With computer program SCRUB, one can calculate the composi-
tion of scrubber liquor from the pH, the ionic strength, and the concentration
of one ion involved in saturation equilibrium. An operational description of
this program has been given. The results of its application to the 47 liquor
samples demonstrate its usefulness for predicting maximum concentrations
of components of scrubber liquors based on solubility considerations.
Ill
-------
SECTION IX
EVALUATION OF THE ENVIRONMENTAL
ACCEPTABILITY OF FGD SLUDGE
The laboratory data presented in this report reveal the phenom-
enological behavior of flue gas desulfurization (FGD) sludges to be expected
in field disposal. The behavior characteristics have been separated into
chemical and physical properties, and experiments have been conducted in a
manner so as to simulate closely the environmental conditions expected in
actual field disposal. Thus, the range of experiments have included every
type of disposal considered likely from an economic point of view, and most
properties have been determined on the basis of environmental acceptance.
9. 1 OVERVIEW
It would be desirable to evaluate all disposal conditions on an
absolute basis. However, the current lack of appropriate criteria regulating
FGD sludge disposal does not allow an absolute definition of "environmental
acceptability. " Moreover, the environmental impact of FGD sludge disposal
will be dependent not only on the properties of the sludge, but also on the
geological and hydrological conditions of the disposal sites. Thus, an evalu-
ation of the environmental impact of FGD sludge must be made solely on the
anticipated behavior of the sludge under the various alternative methods of
disposal, without specifically defining the environmental acceptability of the
sludge. The following sections discuss the expected impact of FGD sludges
on the environment with respect to the range of properties observed in this
study. It is presumed that the range of properties observed in the sludges
sampled are representative of the sludges that are being and will be produced.
In the small sampling of this study, an attempt was made to provide repre-
sentation for eastern and western coals; lime, limestone, and double-alkali
systems; sludges with and without fly ash; and systems operating at both high
and low pH. Although not every combination of variables could be included,
it is believed that the following discussions will be valid for a majority of the
sludges that will require disposal.
In order to evaluate the environmental impact of alternative
disposal techniques for FGD wastes, it is necessary to assess the various
routes by which chemical pollutants may enter the environment from a dis-
posal site and to determine the relative mobility of the various chemical
species with respect to their environmental accessibility. Thus, in addition
to chemically characterizing FGD sludges, this task effort includes experi-
mental determination of the potential pollution hazard that is inherent in each
113
-------
of the various disposal techniques. The assessment of the potential pollutant
hazard includes considerations of the chemical state of the pollutant and the
route by which it may enter the environment.
9.Z ROUTES OF ACCESSIBILITY OF POTENTIAL CHEMICAL
POLLUTANTS TO THE ENVIRONMENT
The disposal of FGD wastes makes possible several alternative
routes by which chemical constituents in the desulfurization wastes may enter
the environment. The facilities generating these wastes are described in
Appendix B. As previously discussed, the quantitative identification of
chemical constituents in the wastes has been determined (Section VII), and
correlations between the chemical constituents and the variables of each
system have been made such that the pollutant potential of each waste can,
for the most part, be anticipated (Section 7.4). In this section, the various
routes by which these pollutants can enter the environment are explored so
that the actual pollutant potential can be assessed relative to the techniques
by which the waste is disposed. Additionally, through experimentation, the
capability of a waste for creating a pollution problem is determined. The
environmental conditions simulated by each experiment are presented, as
well as the environmental consequences of each simulated method of disposal.
The route of access of chemical pollutants to the environment
from FGD waste can be through either air or water. The untreated waste is
disposed of as sludge, a mixture of solids and occluded liquor. The introduc-
tion of potential pollutants into the environment through air may occur either
by vaporization of the pollutant from the solid or liquor phase by mass trans-
fer of spray from the surface of a wind-whipped disposal pond or by blowing
dust from an air-dried site. Pollutant introduction (through water) into the
environment from managed disposal sites can occur by leaching the sludge
with rainwater, by both displacement of the occluded liquor and dissolution
of the solid phase.
In the case where the sludge is treated by one of several techniques
to convert it into a solid material, the opportunities for a chemical pollutant
to enter the environment is more limited. Pollutant introduction into the air
by vaporization is minimized, and by spray transfer it is avoided except where
water is allowed to collect on the surface of a solid mass. If weathering has
produced a powdery surface which is left exposed, blowing dust could be a
problem. Pollutant introduction through water by leaching is minimized if
the solidified sludge offers greater resistance to water passage. Pollutant
introduction by runoff water can also be reduced if the solidified waste does
not release chemical constituents as readily as untreated waste. In every
potential route for environmental pollution by the chemical constituents of
FGD wastes, the disposal of solidified waste offers the opportunity to mini-
mize the pollution potential relative to disposal of untreated waste. The cost
effectiveness of solidification relative to untreated waste disposal by alterna-
tive methods will be dependent upon site-specific conditions.
114
-------
The following sections describe the experiments and results
conducted to evaluate the pollution potential of various methods of disposal.
9.2.1 Pollution by Vaporization
The potential for vapor loss of a particular chemical species is
dependent upon a number of factors, but primarily the vapor pressure of the
species and the effect of temperature on vapor pressure. The most severe
condition that could exist under normally occurring environmental circum-
stances is that in which a sludge dries and increases in temperature under
direct solar heating. Under extreme conditions, the surface of the sludge
may reach temperatures as high as 150 F (66 C), and these temperatures may
be adequate to vaporize certain chemical species.
For determination of whether volatile chemical species may
sublime from the surface of a disposal basin while being heated by solar
energy, the clarifier underflow from the TVA Shawnee limestone system
(sampled on 1 February 1973) was dried thermally in such a way that the vol-
atiles were analyzed with a quadrupole mass spectrometer. At temperatures
to 180°F (82°C), only water was observed; at increasing temperatures, the
water signal dropped off; at 650°F (334°C) strong signals from SC>2 and CO2
were present from dissociation of the sulfate and carbonate phases. No
other volatile species was observed.
The sensitivity of the quadrupole mass spectrometer was deter-
mined from the measurement of free vaporization of elemental mercury. It
was found that a partial pressure of 2 x 10~9 atm of mercury could be de-
tected. Thus, vaporization of mercury (bp 357°C) or mercuric chloride
(bp 302°C), if it had occurred, was below this level of detection. The sensi-
tivity for other chemical species present in the sludge by mass spectrometry
was comparable or even higher. Thus, the lack of signal from any chemical
species indicated that its presence in the vapor phase over the sludge was at
a value of less than 10-9 atm. This value is many orders of magnitude lower
than that required to create an environmental hazard. It was concluded,
therefore, that environmental pollution by vapor phase loss to the atmosphere
does not occur even under extreme climatic conditions.
9.2.2 Pollution by Wind-Whipped Spray
An alternative method of disposal of untreated sludge may be one
in which the sludge is ponded such that water collects on the sludge surface;
another method of disposal is that in which the sludge is chemically treated,
disposed of in a basin, with rainwater allowed to collect on its surface. In
either case, strong winds blowing across the surface of the pond or basin
may produce a spray containing the chemical constituents of the surface
water. Under most conditions, the fine spray particles fall to the earth
within a mile of the basin. Environmental pollution in this case, then, may
be restricted to contiguous agricultural areas, with the extent of pollution
particles dependent upon the concentration of chemical species in the pond
surface water. Values for pond supernate quality can be obtained from the
115
-------
TVA Shawnee Field Disposal Evaluation Project (2). Wind-spray or
wind-blown dust (at sites where supernate has evaporated or where fixed
sludge landfills are exposed to the air) will be a problem only in certain
specific areas. In these areas, proper site management should be main-
tained to prevent environmental pollution.
9.2.3 Pollution by Runoff
The potential environmental pollution by rainwater runoff from
the surface of FGD sludges exists through the action of surface leaching of
chemical constituents and from erosion of particulates from the surface.
Subsequently, it can affect either surface waters if the runoff flows directly
into streams, or groundwater if the runoff is allowed to percolate through
adjacent land. This applies to sludges that are disposed of as a landfill,
either above or below grade. At this time, only chemically treated sludges
are being disposed of operationally as described above. When treated sludge
is disposed of by this method and the site is managed to facilitate landfill
operations, such as placement and compaction, rainwater is not allowed to
stand on the surface and, therefore, runs off. The potential environmental
effects of this runoff would be of concern.
The environmental concern for these sites results from the fact
that freshly treated sludges could produce relatively high concentrations of
suspended and dissolved solids in the runoff compared with those from a
cured material. Since freshly treated material is added to the site continu-
ously, the concern for the quality of the runoff would exist throughout the
development period of the landfill. Currently, a runoff collection ditch
leading to a siltation pond is used to prevent pollution from suspended solids;
the collection of water from a major portion of the landfill area allows for
dilution of excessive dissolved solids from local areas of the site con-
taining freshly treated sludge. Freeze-thaw cycling may periodically create
runoff problems by exposing new surface areas to rainwater; therefore, a
compacted overburden of soil is desirable for completed portions of the
landfill.
These observations represent only a limited examination of runoff
conditions. Current programs by private firms and the EPA are expected
to produce a better understanding of these potential problems.
9.2.4 Pollution by Rainwater Leaching of Untreated Wastes
The most serious form of pollution potential exists as a conse-
quence of the action of rainwater leaching through a disposal basin. In the
case of untreated sludge, this pollution potential can exist as a consequence
of most disposal techniques because the disposal basin is designed to im-
pound rainwater whether or not the disposal technique is by ponding. Water
116
-------
percolates through sludge whenever freestanding water collects on its surface,
The amount of leachate that percolates through the sludge depends upon the
permeability of the sludge, the permeability of the basin subsoil, and the
amount of water made available to the basin either by rainfall or through
ponding. The quality of the leachate depends on the availability of chemical
species to the percolating waters.
Sludge normally contains occluded liquor from the scrubbing pro-
cess in an amount representing 35 to 80 percent by weight of the total sludge,
depending upon the extent of dewatering. Most of this occluded liquor is
flushed from the sludge by the first three pore volume displacements (PVD)
of percolating water; thereafter, the quality of leachate is dependent upon the
solubility of the solid phase of the sludge. The determination of the quality
of leachate from untreated sludge as a function of the amount of water pass-
ing through it is described in the following experiments.
Experimental Procedure
Two sets of experiments were conducted. The first set used
100 grams of dried sludge packed into 4-cm glass colums through which 3.6
liters of water were allowed to pass. A set of three columns was packed for
each sludge tested, representing the TVA Shawnee limestone sludge, the
Arizona Public Service (APS) Cholla limestone sludge, the Duquesne Light
Company (DLC) Phillips lime sludge, the General Motors (GM) Parma double-
alkali sludge, and the Southern California Edison (SCE) Mohave limestone
sludge. Leaching water (unbuffered) adjusted to pH 4, 7, and 9, with either
HC1 or NaOH, was used on each set of columns. The leaching water and the
leachate were allowed to react with air so as to simulate aerobic conditions.*
On the basis of the results of the first experimental set, the
second set of leaching experiments were conducted, simulating anaerobic
conditions.* These experiments duplicated the first except that the leaching
water and leachate were protected from interaction with the atmosphere;
200 grams of dried sludge was used in 6-cm columns, and the leaching water
of pH 7 only was used.
In the first experimental set, leachate analysis was performed
on the first 600 m& of leachate collected and at increments of 750 mS. there-
after. In the second set of experiments, analysis was performed on the 1st,
'Whether leaching water or leachate will be aerobic or anaerobic is dependent
upon the sludge and site conditions. Aerobic leaching water can result from
supernate equilibrium with the atmosphere; however, subsequent percolation
through a sulfite sludge can create an anaerobic condition in the leaching
water. The resultant leachate can remain anaerobic in the subsoil; however,
if the soild is plugged by filtering of fine and colloidal particles from the
sludge and penetration through this filter zone is slower than permeation
through the soil, the opportunity exists for the creation of air voids in the
soil. In this case, the leachate may become aerobic. In field conditions,
all four combinations of aerobic and anaerobic leaching water and leachate
are possible.
117
-------
2nd, and 3rd PVD, and then on the 15th, 30th, and 60th PVD. Pore volume
was calculated from the mass and volumes of each packed column. In both
experimental sets, analyses were performed for arsenic, beryllium, cad-
mium, chromium, copper, lead, mercury, selenium, zinc, chloride, fluoride,
sulfate, pH, and TDS.
Results of Experiments
The results of the aerobic experiments are presented in
Figures 32 through 36. A single curve is presented for each species, repre-
senting the median value of concentration from the three values of pH for
leaching water used on each sludge type. The analytical results indicate no
discernible difference between the columns as a result of differences in the
starting pH of the leaching waters, except for certain trace elements as
shown in the curves. Moreover, the median value from the three columns
used as replicates provide a high level of confidence to the accuracy of the
data. The curves are normalized to the concentration of the first pore volume
which essentially reflects the concentration of occluded mother liquor. Major
species and trace metals existing at high concentration (>0. 1 mg/S. ) are
presented from which typical behavior can be observed. For most trace
elements existing in low concentrations in the liquor, the concentration in
the leachate dropped below the detection limit of the test method after only
a few PVD.
1.0
cr
O
a.
t
i/>
2 0.1
o
0.01
O
g
0.001
2
LITERS
10 20 30 40
PORE VOLUME DISPLACEMENTS
50
Figure 32. Analysis of leachate from TVA Shawnee
limestone sludge: aerobic conditions.
118
-------
cc
o
O-
0.1
<
UJ
EK
o
i
0.01
o
o
2
LITERS
I I I
10 .20 30 40 50 60
PORE VOLUME DISPLACEMENTS
70
Figure 33. Analysis of leachate from SCE Mohave
limestone sludge: aerobic conditions.
10 20 30 40 50 60
PORE VOLUME DISPLACEMENTS
so.
70
80
Figure 34. Analysis of leachate from DLC Phillips
sludge: aerobic conditions.
119
-------
cc
2
at
u_
e
UJ
>
0.1
o
<
a
- o.oi
z
o
o
z
o
SO,
TDS
Cl
3 1
1 1
3 10 20
2
LITERS
|
30
1
40
3
1
50
4
1
60
PORE VOLUME DISPLACEMENTS
Figure 35. Analysis of leachate from APS Cholla
sludge: aerobic conditions.
s
_J
o
=>
UJ
o:
O
a.
g 0.1
^
I
£
£
<
i
o
a
-o.oi
<
cc
o
o
Cl
- \\\As
" \\ Be
- \Cd,Cr
Cu, Pb
1
0 - 1
1
1 1 1
2 3 4
LITERS
1 1 1
10 20 30
PORE VOLUME DISPLACEMENTS
40
Figure 36. Analysis of leachate from GM Parma
sludge: aerobic conditions.
120
-------
The results of the anaerobic leaching experiments are presented in
Figures 37 through 39. These curves are from single columns and obtain their
reliability from sequential pore volume analyses. A comparison of results
from both the aerobic and anaerobic leaching columns is given in Tables 19
through 23. The concentration of major and trace species is given after the 1st
and 50th PVD and is compared with the concentration of the mother liquor of
each scrubbing system.
From an overall assessment of the leaching data, it was determined
that for each of the major species, represented by sulfate, chloride, and TDS
analysis, the concentration decreased rapidly during the first few pore volumes,
and in most cases 90 percent of the decrease in concentration of the leachate
between the 1st and 50th PVD had taken place at the completion of the 3rd PVD.
A similar observation can be made for trace elements with the exception that
some elements did not show the sustaining concentration level after the initial
pore flushing, as was seen for major species. This behavior may be due to the
fact that trace amounts are more difficult to flush from the system, or the
observed behavior may be only the result of greater uncertainty in the analytical
data because of their low concentration, or perhaps some trace elements are
present as fine particulate matter which continues to be flushed.
When comparing aerobic and anaerobic leaching, it would appear
from the figures that a greater drop in concentration occurs in the anaerobic
1.0
§
Q.
£ 0.1
UJ
Of
0.01
§
0.001
Cl
I
6
LITERS
10
10 20 30 40 50 60
PORE VOLUME DISPLACEMENTS
70
Figure 37. Analysis of leachate from TVA Shawnee
limestone sludge: anaerobic conditions.
121
-------
o
>
o
a.
Oi
LL.
O
a:
£
0.1
;o.oi
o
o
LITERS
I
10 20 30 40 50
PORE VOLUME DISPLACEMENTS
60
SO,
70
Figure 38. Analysis of leachate from SCE Mohave
limestone sludge: anaerobic conditions.
a:
O
0.1
o.oi
o
CJ
1
6 8
LITERS
Pb-
10
10 20 30 40 50 60
PORE VOLUME DISPLACEMENTS
SO,
12
70
80
Figure 39. Analysis of leachate from DLC Phillips
sludge: anaerobic conditions.
122
-------
TABLE 19. COMPARISON OF CHEMICAL CONSTITUENTS IN
SLUDGE LIQUORS WITH LEACHATE AFTER 1
AND 50 PVD: TVA SHAWNEE LIMESTONE SLUDGE
Constituent
As
Be
Cd
Cr
Cu
Pb
Hg
Se
Zn
Cl
F
S03
so4
pH
TDS
Sludge
Liquor,
mg/£a
0.14
0.054
0.003
0.09
0.01
0.25
<0.05
--
--
2250
6.2
80
10,000
8.3
15,000
Aerobic Conditions, mg/S.
1st PVDb
0.06
0.008
0.002
0.025
0.007
0.12
0.05
0.03
0.85
1350
2.7
18
6500
4.7
10,500
50th PVDb
0.01
0.004
<0.001
0.003
0.010
0.01
<0. 00005
0.006
0.045
120
<0.2
25
1200
5.0
2400
Anaerobic Conditions, mg/^
1st PVD
0.06
--
0.035
0.07
0.05
0.47
0.005
0.20
0.60
1700
10.8
--
9000
8.5
12,500
50th PVDb
<0.004
--
0.003
0.015
<0.02
0.055
0.0004
0.01
0.01
70
1.1
--
1000
7.4
1600
Liquor analysis on clarifier underflow.
Values taken from curve.
123
-------
TABLE 20. COMPARISON OF CHEMICAL CONSTITUENTS IN
SLUDGE LIQUORS WITH LEACHATE AFTER 1
AND 50 PVD: SCE MOHAVE LIMESTONE SLUDGE
Constituent
As
Ca
Be
Cd
Cr
Cu
Pb
Hg
Se
Zn
Cl
F
SO3
so4
pH
TDS
Sludge
Liquor,
mg/jf
0.03
26,500
0.02
0.05
0.25
0.56
0.04
CO. 005
0. 12
0.18
28,000
30
1.5
25,000
6.7
92, 500
Aerobic Conditions, mg/4
1st PVD
0.012
0.013
0.011
0.05
0.11
0.15
0.001
0.07
0.060
7700
7.1
0.6
8000
6.0
24,300
50th PVD
<0.004
0.004
<0.001
0.003
0.010
<0.001
<0. 00005
0.004
0.045
130
<0.2
0.3
1300
4.5
2100
Anaerobic Conditions, mg/i
1st PVD
<0.004
--
0.05
0.05
0.07
1.7
<0.0005
0.034
2.7
7500
3.6
0.
8000
7.38
23,000
50th PVD
< 0.004
--
<0.004
<0.015
<0.03
0.08
<0. 00005
0.004
0.13
65
1.4
0.2
1100
7.45
1700
124
-------
TABLE 21. COMPARISON OF CHEMICAL CONSTITUENTS IN
SLUDGE LIQUORS WITH LEACHATE AFTER 1
AND 50 PVD: DLC PHILLIPS LIME SLUDGE
Constituent
As
Be
Ca
Cd
Cr
Cu
Pb
Hg
Se
Zn
Cl
F
so3
so4
pH
TDS
Sludge
Liquor,
mg/1
0.09
0.012
2855
0.023
0.040
0.07
0.18
0.05
0.08
0.09
2700
2.6
27
3150
7.11
9400
Aerobic Conditions, mg/^
1st PVD
0.03
0.004
--
0.003
0.005
0.035
0.12
0.0006
0.07
0.51
2400
1.7
--
2700
4.98
7800
50th PVD
0.004
<0.003
--
<0.001
<0.001
0.008
<0.001
<0. 00005
0.007
0.03
110
0.8
--
1250
4.90
2200
Anaerobic Conditions, mg/Ji
i
1st PVD
0.01
--
--
0.004
0.04
0.04
0.10
0.0004
0.02
0.50
415
2.4
--
2000
7.80
3500
50th PVD
0.003
--
--
0.001
<0.015
<0.01
0.04
<0. 00005
0.001
0.02
65
1.0
0.2
900
9.65
1350
125
-------
TABLE 22. COMPARISON OF CHEMICAL CONSTITUENTS
IN SLUDGE LIQUORS WITH LEACHATE
AFTER 1 AND 50 PVD: APS CHOLLA
LIMESTONE SLUDGE
Constituent
As
Ca
Be
Cd
Cr
Cu
Pb
Hg
Se
Zn
Cl
F
S03
so4
. PH
TDS
Sludge
Liquor,
mg/l
<0.004
2670
0. 18
0.009
0. 21
0. 19
0.01
0. 13
2.5
0.07
1430
0.7
0.9
4400
4.3
9100
Aerobic Conditions, mg/f.
1st PVD
<0.004
--
0.007
0.001
0.019
--
0.016
0.00008
0.05
900
2.4
22
3500
4.6
6500
50th PVD
<0. 004
--
0.004
<0.001
0.002
0.01
<0.001
<0. 00005
0.05
0.04
110
6.1
9.0
1150
5.9
1900
126
-------
TABLE 23. COMPARISON OF CHEMICAL CONSTITUENTS
IN SLUDGE LIQUORS WITH LEACHATE
AFTER 1 AND 40 PVD: GM PARMA
DOUBLE-ALKALI SLUDGE
Constituent
As
Be
Ca
Cd
Cr
Cu
Pb
Hg
Na
Se
Zn
Cl
F
so3
so4
PH
TDS
Sludge
Liquor,
mg/i
<0.004
<0.005
640
<0.02
<0.02
0.06
0.52
0.0005
20, 000
0.075
0. 59
5200
58
140
35,000
12.7
65,000
Aerobic Conditions, mg/i
1st PVD
0.20
0.32
--
0.87
0.95
1.8
1.76
<0.0005
--
0.29
4750
22
48
31, 000
8. 1
58, 000
40th PVD
<0.002
<0 . 004
--
<0.001
<0.001
<0.01
<0.01
<0. 00005
--
0.010
0.04
95
t
0.2
30
1100
6. 1
1650
127
-------
leaching condition than in the aerobic condition, although for the DLC Phillips
lime sludge there was no significant difference between these two. In any
event, when the decrease in concentration is compared with the mother li-
quor in Tables 19 through 23, the concentration of major species in the 50th
pore volume is consistently less under anaerobic leaching conditions than
aerobic conditions. When the same comparison is made with trace elements,
no clear trend is observed except for the case of lead. In every case, lead
concentration in the 50th pore volume was significantly higher when leached
under anaerobic conditions.
For an understanding of this behavior, it is necessary to observe
that the pH of the aerobic leachate was acidic in all cases, whereas the
anaerobic leachate was always basic. The increase in sulfate ion content in
the aerobic leachate could have resulted from acid leaching of the sulfite
ions that subsequently oxidized to sulfate. The increase in chloride ion in
the aerobic leachate is believed to be a consequence of the release of chloride
ion coprecipitated with the sulfite phase (30). The higher lead content in the
anaerobic leachate is consistent with the observations noted previously where
lead increased with increases in the pH of the system. It is believed that the
lead behavior was a consequence of the relative stability of the basic lead
carbonate in the system.
The acidic leachate of the aerobic columns was a consequence of
atmospheric interaction with the leaching water that subsequently reacted with
the sulfite phase to leach bisulfite from the sludge column. The leachate was
then oxidized by exposure to the atmosphere producing sulfate and hydrogen
ions. The pH of aerobic leachate for all sludges ranged between 4.5 and 6.0,
depending on the initial alkalinity of the sludge; apparently, the unbuffered
leaching solutions at pH 7 and 9 could not offset the stronger acid reactions.
Earlier mention was made of the independence of the concentration
of major and trace species on the starting pH of the leaching waters. The
exceptions were that, in the first sampled leachates, the sulfate content was
slightly lower (approximately 10 percent) in the pH 9 column and the pH of the
leachate of this column was always slightly higher than the pH 4 and 7 columns,
which were identical. The other exception was in the concentration of lead,
which was always higher in the initial sample for the pH 9 column. This be-
havior is in complete agreement with other observations which showed that
the concentration of lead was higher in more basic waters.
Except for selenium in the anaerobic leachate of the TVA Shawnee
limestone sludge and copper in the SCE Mohave aerobic leachate, only lead
and zinc were consistently found at high concentrations (>0. Irng/^) in the
initial leachate. In many cases, lead and zinc were found in the first pore
volume at concentrations greater than that of the mother liquor. In the GM
Parma sludge, all trace elements were found in the first pore volume at a
concentration greater than the mother liquor. These elements were quickly
flushed from the column and appeared thereafter at background levels.
128
-------
When leaching behavior is compared in Figures 32 through 36,
it appears that significant differences exist among the leaching results of
major species. However, when this behavior is compared in Tables 19
through 23, certain similarities appear. Specifically, the sulfate concentra-
tion in the aerobic columns after 50 PVD reached a level of stability at values
between 1100 and 1300 Trig/I, completely independent of the concentration of
the mother liquor. Sulfate ion concentration in the 50th pore volume in the
anaerobic columns showed a consistent 20 percent reduction in concentration,
ranging between 900 and 1100 mg/f. In the 50th pore volume, the chloride
concentration was found to have stabilized in the range between 95 and 130
mg/i in the aerobic columns and at about one half of this value in the anaerobic
columns. The concentration of sulfate in the leachate was approximately that
which is predicted from the solubility of gypsum. The concentration of
chloride ion demonstrated a stability value that is not reconcilable with solu-
bility theory and can only be understood if chloride ions are coprecipitated
with both sulfite and sulfate phases.
9.2.5 Pollution by Rainwater Leaching of Chemically
Treated Wastes
Several techniques or processes are available by which sludge may
be treated by chemical additives to reduce its pollution potential and increase
its environmental acceptability as a waste for disposal. The applications and
costs of these processes or techniques are presented in Section X. In most
cases, these processes convert the sludge to a more solidified state by various
means, which may include secondary dewatering and cementitious reactions.
When waste is converted to a solid, disposal management procedures are avail-
able to further reduce or eliminate the pollution potential. However, if these
procedures are not employed, chemically treated sludge may be disposed of in
a manner in which a chemical pollution potential exists. One such disposal
technique is that in which free-standing rainwater is permitted to percolate
through the waste. The resulting leachate will contain the chemical species in
the waste that are free to move. These species can be salts from the mother
liquor, chemical additives, soluble species from the solid waste, or chemical
products resulting from waste reactions with the additives.
Leaching experiments were conducted for determination of the
quality of leachate that could be expected from chemically treated FGD wastes.
Both aerobic and anaerobic leaching conditions were experimentally con-
ducted in the manner described in Section 9.2.4. In all cases, the chemically
treated sludge was dried and pulverized before being packed into a leaching
column, which procedure generated the maximum surface contact area for the
flow of water through the treated sludge. This represents an operational case
in which the cured material is fractured by heavy machinery during final
placement.
The results of the leaching experiments are presented in Figures
40 through 46, in which only major species and trace elements at high concen-
trations are presented. These figures show leaching behavior similar to that
observed in the leaching of untreated sludge. In most cases, the decrease
in concentration of major species in the leachate took place within the first
129
-------
1.0
o
0£
O
Q.
0.1
UJ
OS.
0.01
o
cj
Figure 40
l.o
o
>
0.1
0.01
o
I
z
LkJ
-------
o
>
a:
o
0.1
o 0.01
ee.
t
z
1*J
O
z.
o
o
4 6
LITERS
10
12
10 20 30 40 50 60
PORE VOLUME DISPLACEMENTS
70
Figure 42. Analysis of leachate from TVA Shawnee lime-
stone (IUCS) sludge: anaerobic conditions.
o.i
0.01
8
Cl
I
I
2 3
LITERS
I I I
10 20 30 40 50 60
PORE VOLUME DISPLACEMENTS
70
80
Figure 43. Analysis of leachate from SCE Mohave
(IUCS) sludge: aerobic conditions.
131
-------
a
ce.
o
Q-
0.1
x
o
50.01
o:
3
LITERS
10 20 30 40 50 60
PORE VOLUME DISPLACEMENTS
70
Figure 44. Analysis of leachate from SCE Mohave
(IUCS) sludge: anaerobic conditions.
O
Q-
£ 0.1
^
10.01
z
o
2
LITERS
10 20 30 40 50 60
PORE VOLUME DISPLACEMENTS
so.
70
Figure 45. Analysis of leachate from DLC Phillips
(Calcilox) sludge: aerobic conditions.
132
-------
Sg i
cn
O
a.
0.1
I
0.01
so.
6
LITERS
10
12
i
I
I
10 20 30 40 50 60
PORE VOLUME DISPLACEMENTS
70
80
Figure 46.
Analysis of leachate from DLC Phillips
(Calcilox) sludge: anaerobic conditions.
few PVD. However, contrary to untreated sludge leaching, some major
species did not leach in this way, but rather the decrease in concentration
was more gradual. This was true even when a comparison was made with
anaerobic leaching of untreated sludge. In the chemically treated sludge, no
systematic trend was found in the decrease of the concentrations of major
species when aerobic and anaerobic conditions were compared.
When the absolute-concentrations for untreated materials (Tables
19 through 23) are compared with the values for treated materials (Tables 24
through 27) for the same scrubbing system and leaching conditions, it is
observed that the concentration of major species in the first pore volume of
the treated waste leachate is less than approximately 50 percent of the
untreated sludge leachate. The relative decrease in concentration appears to
depend on the sludge type, chemical treatment, and differences between aero-
bic and anaerobic leaching conditions.
A comparison of the trace element content of untreated sludge
leachate and leachate from chemically treated sludge as given in either the
figures or tables suggests that, in some cases, the origin of the trace elements
was the chemical additive rather than the sludge. Specifically, the high
initial concentration of chromium in the anaerobic leachate of both the TVA
Shawnee limestone and SCE Mohave limestone sludge, chemically treated by
I U Conversion Systems (IUCS), is particularly suggestive of an additive
133
-------
TABLE 24. COMPARISON OF CHEMICAL CONSTITUENTS IN
SLUDGE LIQUORS WITH LEACHATE AFTER
1 AND 50 PVD: TVA SHAWNEE LIMESTONE
SLUDGE (CHEMFIX)
Constituent
As
Be
Cd
Cr
Cu
Pb
Hg
Se
Zn
Cl
F
so3
so4
TDS
pH
Aerobic Conditions, mg/4
1st PVD
0.04
0.008
0.003
0.04
0.05
<0 . 03 5
<0.005
0.01
0.5
1400
0.9
22
3000
7000
4.70
50th PVD
0.006
<0.001
<0.001
<0.001
0.005
<0.001
<0.0005
0.002
0.065
60
0.2
<0. 1
250
500
6.01
Anaerobic Conditions, mg/4
1st PVD
0.03
0.02
0.08
<0.01
--
0.0008
0.008
1.25
1200
1. 1
42
2850
5470
6.93
50th PVD
<0.003
--
0.002
<0.015
<0.01
0.003
0.0004
<0.005
0.015
40
0.6
7
230
500
9.26
134
-------
TABLE 25. COMPARISON OF CHEMICAL CONSTITUENTS
IN SLUDGE LIQUORS WITH LEACHATE
AFTER 1 AND 50 PVD: TVA SHAWNEE
LIMESTONE SLUDGE (IUCS)
Constituent
As
Be
Cd
Cr
Cu
Pb
Hg
Se
Zn
Cl
F
so3
so4
TDS
pH
Anaerobic Conditions, mg/^a
1st PVD
<0.004
0.01
0.60
0.04
0.37
0.0007
0.036
0.35
230
0.8
34
2700
5230
7.76
50th PVD
<0.004
--
0.003
<0.01
<0.01
<0.03
0.0002
0.011
0.033
78
0.2
22
980
1500
10.61
The sample for aerobic conditions was inadequate.
135
-------
TABLE 26. COMPARISON OF CHEMICAL CONSTITUENTS IN
TREATED SLUDGE LEACHATE AFTER 1 AND
50 PVD: SCE MOHAVE LIMESTONE SLUDGE (IUCS)
Constituent
As
Be
Cd
Cr
Cu
Pb
Hg
Se
Zn
Cl
F
so3
so4
TDS
PH
Aerobic Conditions, mg/t.
1st PVD
0.02
0.011
0.02
0.08
0.02
0.07
<0.005
0.02
0.04
330
2.5
1. 5
4000
6800
4.76
50th PVD
<0.004
0.003
<0.001
0.003
0.010
0.033
<0. 00005
0.026
0.032
60
0.2
0.5
200
400
5.63
Anaerobic Conditions, mg/£
1st PVD
0.013
--
0.025
1.8
0.02
0. 13
0.0003
0.23
0.20
300
0.8
34
4300
5660
8.49
50th PVD
<0.004
--
0.0013
<0.015
<0.01
0.03
<0.0005
0.04
0.022
' 75
0.2
17
600
1100
9.77
136
-------
TABLE 27. COMPARISON OF CHEMICAL CONSTITUENTS IN
TREATED SLUDGE LEACHATE AFTER 1 AND
50 PVD: DLC PHILLIPS LIME SLUDGE
(CALCILOX BY DRAVO)
Constituent
As
Be
Cd
Cr
Cu
Pb
Hg
Se
Zn
Cl
F
so3
so4
TDS
pH
Aerobic Conditions, mg/f
1st PVD
<0.01
0.006
0.022
0.04
0.03
0. 18
0.0007
0. 15
0. 16
600
2.6
18
1900
5600
5.03
50th PVD
<0.004
<0.001
0.001
0.001
0.01
0.03
<0. 00005
0.01
0.03
95
0.6
6
650
900
5.51
Anaerobic Conditions, mg/,0
1st PVD
<0.01
--
0.03
0.05
0.02
0. 125
0.001
0. 146
0. 16
600
5.5
22
1925
3750
7.54
50th PVD
<0.004
--
0.003
<0.015
<0.01
0.07
0.0002
0.005
0.037
90
1.7
7
1450
2150
9. 19
137
-------
origin since the chromium content in the untreated sludge was high in either
the sludge or the leachate. In a similar example, selenium appeared in the
initial leachate from the DLC Phillips lime sludge, treated with the Dravo
process, at a concentration of twice that found in the untreated sludge and at
virtually identical concentrations in both aerobic and anaerobic leaching
conditions.
On the other hand, lead and zinc appeared at relatively high
concentrations in the initial leachate from the TVA Shawnee limestone sludge
and again when this sludge was chemically treated, either by the IUCS or
Chemfix process. Lead and zinc were also found in the leachate from the
DLC Phillips lime sludge at relatively high concentrations and appeared
again at approximately the same concentration levels in the leachate after
chemical treatment by the Dravo process.
In the specific case of lead, it was observed that its concentration
was always higher in the anaerobic leachate than the aerobic leachate of
untreated sludges. This behavior was also present in the leachates from
chemically fixed sludges.
It was previously observed that lead and zinc concentrations in
the initial leachate from untreated sludge exceeded the concentration of these
elements in the mother liquor. The same behavior was observed for the
chemically treated sludges; in the chemically treated SCE Mohave limestone
and DLC Phillips lime sludges, the lead concentration continued to rise for
several PVD before receding to lower concentration levels.
In nearly all cases for major species, independent of their initial
concentrations, the chemical species concentrations in the 50th PVD from
chemically treated sludge was less than in the 50th PVD from untreated
sludge. In the case of aerobic leaching conditions, the concentration was
typically reduced to a value of 40 percent or less than that for untreated
sludge. In the case of anaerobic leaching conditions, the concentration
reduction was not as clearly revealed. In one case the concentration in-
creased; in other cases the reduction was small, with the best case repre-
senting only a one-third reduction from the concentration of the untreated
sludge.
The effect of chemical treatment of untreated sludge on the
quality of leachate is generally an improvement in leachate quality dependent
in part on the specific sludge and the specific chemical process. Generally,
the concentration of specific major chemical species-is one half or less in
the leachate from the chemically treated sludge relative to the leachate from
untreated sludge. In the case of aerobic leaching conditions, this relationship
appears to hold for all leaching volumes; however, in anaerobic leaching,
this reduction does not hold true at longer leaching times. In at least one
case, the chemically treated sludge showed no improvement in leachate
quality over untreated sludge.
When these data are examined for trace metals, it is seen that
chemical treatment of sludge does not improve the leachate quality compared
to the leachate of untreated sludge. In those cases in which high initial
138
-------
concentration of trace elements appeared in the leachate of untreated sludge,
the same elements appeared when the sludge was chemically treated and
responded identically under aerobic and anaerobic leaching conditions. In
addition, the data suggest that the chemical additives used in the chemical
processing may contribute trace elements to the leachate. Thus, chemical
treatment by the methods evaluated would not be effective in alleviating the
environmental pollution potential of trace elements. Other techniques, how-
ever, such as (a) reduction of concentration of major species, (b) reduction
of permeability, and (c) superior disposal placement techniques made possi-
ble by chemical treatment can be effective if properly utilized.
9.3 ALTERNATIVE DISPOSAL, TECHNIQUES
9.3.1 Ponding
The method of sludge disposal that represents the least deviation
from state-of-the-art fly ash disposal is direct ponding into a disposal basin.
In this method, the underflow of a clarifier is pumped to the pond where the
solids settle. The supernate is returned to the scrubber because the soluble
salt content is too high for effluent discharge. From the viscosity data and
dewatering experiments, it is seen that in some sludges secondary dewatering
can be conducted before pumping to the pond, while in other sludges the
clarifier underflow has a water content close to the point at which secondary
dewatering would prevent liquid transfer. In most sludges, the settled density
is the lowest density of any dewatering method, and the sludge requires the
largest disposal basin because of the large amount of water retained in the
sludge. At the conclusion of filling the basin, the disposal site can either be
maintained as a pond or allowed to dry, depending primarily upon the pre-
vailing climatic conditions.
The environmental impact of pond disposal is strongly dependent
upon the ability (a) to contain the components of a sludge so as to prevent
environmental pollution and (b) to retire the disposal site in a manner that
does not create a safety hazard or nuisance in subsequent land use. For pond
disposal, the environment can be protected from chemical pollution, princi-
pally from leachate contamination of groundwater, by lining the pond basin
with clay, impermeable soil cements, or synthetic materials. The need
for liner materials is strongly suggested by the high permeability of the
sludge (10 to 100 times greater than most indigenous soils) and the high
concentration of chloride and sulfate ions in the leachate. However, if a
liner material is not used, it is expected that not all trace elements will be
attenuated by the subsoil. Additionally, soils do not significantly attenuate
chloride or sulfate ions.
Pond retirement can be effected either by maintaining the pond as
a lake or by allowing the sludge to dry and covering it with an overburden.
In order to maintain the retired disposal basin as a lake, it is necessary
to provide a balance between water loss and water input. The water loss may
be by evaporation, and, in the case where no liner is used or a breach is
developed in the liner, loss also occurs by percolation through the subsoil.
Precipitation in excess of loss requires a means for eliminating excess water.
139
-------
If precipitation is less than water loss, a means for maintaining water
level must be used or the lake will eventually dry. If specific plans for
overburden coverage are not made at the time of pond retirement, a
dried lake will likely represent an environmental hazard or nuisance.
If a pond is retired by drying the sludge and covering it with an
overburden, certain restrictions may limit reuse of the land. The overburden
can be contoured so as to maximize runoff. If a liner is used, maximizing
rainfall runoff and minimizing percolation of water through the overburden
may be necessary to avoid resaturating the sludge. If no liner is used,
proper contouring is necessary to minimize recharge to the water table and
to minimize the pollution potential of the leachate from the sludge. In areas
where the permeability of the subsoil is low, it is possible to prevent any
contamination of the water table if the sludge leachate does not reach the
water table during the time of active pond filling and, after drying and cover-
ing, if the subsequent disposal site is properly designed so that groundwater
recharge is prevented. Under proper climatical, geological, and hydro-
logical conditions, it is presumed possible to dispose of FGD sludge by
ponding in an environmentally acceptable manner.
9.3.2 Ponding With Under drainage
The method of disposal which includes underdrainage within a
closed basin is a major variation of the ponding method of disposal. This
method retains the advantage of transferring the sludge to the disposal site
by liquid transfer, but, instead of supernate return to the scrubber, the
leachate from the base of the sludge is returned to the scrubber. The advan-
tages of this method are both economical and environmental. By avoiding a
supernate head over the sludge, percolation of sludge leachate into the sub-
soil can be avoided during the active fill period. Upon retirement of the dis-
posal site, the sludge is more easily dried and can be covered without regard
for climatic conditions otherwise required for drying. Subsequent cover con-
touring is necessary for the reasons discussed in ponding, but the under-
drainage system provides a means of sampling and elimination of leachate if
required to prevent groundwater contamination. The primary economic
advantage of this method is the elimination of the requirement for a disposal
basin liner to prevent leachate percolation, except in those cases where the
subsoil is highly permeable, and the relative ease of retirement.
The sludge that is allowed to drain packs to only a slightly higher
density than settled sludge; thus, no savings in the size of the disposal basin
can be gained. Also, limiting restrictions on subsequent land use are
expected. Compaction techniques used with fly ash and other materials to
increase density are shown not to be effective with FGD sludge, but compac-
tion was shown to decrease permeability. Thus, while the underdrained
sludge is physically amenable to compaction, it is not cost-effective for
increasing density or strength, and it is detrimental to the free percolation of
leachate through the sludge. Disposal by ponding with underdrainage offers
added assurance of environmental acceptance over ponding, and in some cases
it can be a more economical method of total disposal.
140
-------
9.3.3 Dry Sludge Disposal
An alternative method of sludge disposal is that which uses a
secondary dewatering method such as filtration or centrifugation, and the
sludge cake is trucked to a disposal basin. From the dewatering experi-
ments, it has been seen that, for most sludges tested, secondary dewatering
does not remove a sufficient amount of water to gain structural stability from
the sludge. Thus, a dry centrifuge or .filter cake is not produced, and the
resulting cake is not easily transported because of its thixotropic nature.
However, the addition of fly ash, if available, can alleviate this problem.
Since trucking is intrinsically more expensive than pumping, and because of
the additional difficulties created by the thixotropic nature of the sludge
(including rewetting), this method of disposal is attractive only in those
cases where alternative methods are not possible.
The structural stability of sludge dewatered by filtration or cen-
trifugation varies among the FGD sludges such that the restrictions that
may be subsequently placed on land use must be determined on a case by
case basis. The environmental acceptability of FGD sludge disposed of by
this method will be highly dependent upon disposal site management
techniques.
9.3.4 Chemically Treated Sludge Disposal
Several commercial processes are available by which FGD sludges
are chemically treated to render them environmentally acceptable by several
alternative disposal methods. An evaluation of three of these processes sug-
gests that the soluble salt content in the leachate from treated sludge is typi-
cally one half or less than that of the untreated sludge. Additionally, the
permeability of the treated sludge appears to be at least one order of magni-
tude less than that of the untreated sludge. Therefore, the expected dissolved
salts that are leached from treated sludge and available to the environment are
considerably less than from untreated sludge.
For every process examined, the structural stability of the treated
sludge was far superior to that of the untreated sludge. The treated sludge
texture ranged from soil-like, clay-like, to concrete-like and developed
strength far in excess of natural soils. Restrictions on subsequent land use
will depend upon local conditions and the long-time stability of the fixed sludge.
Laboratory data have not been developed by any source from which it would be
possible to predict the time-dependent stability of treated sludge.
Chemically treated sludges can be used as landfill in both sub-
merged and above-grade conditions. In the submerged condition, the sludge
serves as a lake bottom, and, if the sludge is sufficiently impermeable only a
minimal leachate is available to the subsoil. In an above-grade condition,
rainwater does not penetrate the "impermeable" surface, and a leachate is not
produced. The environmental impact of treated sludge is less than that of
untreated sludge under most disposal methods although the added assurance
afforded by the chemical process increases the cost of disposal. The cost
effectiveness of chemical treatment must be evaluated in the context of
141
-------
specific land use limitations and disposal site restrictions that exist as
unique conditions at each power plant facility.
In the absence of federal regulations for FGD sludge disposal,
there is no evidence that untreated sludge disposal cannot be environmentally
acceptable with proper placement and management. Conversely, chemical
treatment without regard to proper placement and management is not likely
to assure environmental acceptance. Furthermore, a properly placed and
managed untreated FGD disposal site is more likely to meet environmental
acceptance criteria than a poorly managed disposal site for treated FGD
wastes.
9.4 ASSESSMENT OF THE CHEMICAL POLLUTION
POTENTIAL ON THE ENVIRONMENT BY
ALTERNATIVE DISPOSAL METHODS
Pollution of groundwater by the action of rainwater percolating
through FGD sludge is a genuine concern in the disposal of this material.
Laboratory experimentation has shown that relatively high concentrations of
dissolved chemical species in the liquor from untreated sludges persist in
the leachate until at least five PVD have passed through the sludge. There-
after, the concentration depends on the solubility of the chemical phases in
the sludge solids. The rate at which rainwater passes through FGD wastes
has been measured in the range of 10~4 cm/sec, equivalent to soils of silty
sand. The pollution of groundwater, determined by mass loading (mass/unit
area) is calculated from the amount of pollutant that is carried by the leach-
ing water to the groundwater.
The potential pollution that is possible from sludge when treated
by any of several processors can be reduced by several orders of magnitude.
Such an improvement has resulted from the (a) reduction of available chemi-
cal species (lower solubility) in the treated sludge, (b) reduction of standing
rainwater by promoting runoff, and (c) reduction in the rate of water permea-
tion through chemically treated material (lower permeability). In nearly
every case, the eventual pollutant concentration of the leachate was less than
one-half that of the leachate from untreated sludge under similar conditions of
FGD waste placement and disposal site management. Additionally, the runoff
from treated waste can be more easily maximized, making less than one tenth
of the rainfall available for seepage. Since the permeation coefficient of
treated waste is typically reduced by at least an order of magnitude, the mass
loading per year to the groundwater may be reduced by a factor of 1 0 to 10,000.
Thus, the real pollution measured by mass loading can be reduced by large
amounts.
The consequence of much lower permeability rates of chemically
treated sludge is that many years will pass before five PVD in the sludge will
reduce the leachate concentration to the level of the soluble salts. Therefore,
while the pollution of groundwater by untreated sludge can be severe for a
short period of time and low thereafter, the pollution input caused by chemi-
cally treated sludge, at any given time, is never severe but will be sustained
for much longer periods of time. The latter case is more consonant with
nature, which weathers natural deposits slowly but constantly. However,
since the high concentration of the leachate is not quickly flushed from treated
142
-------
FGD wastes, the potential exists that a later unexpected event could result in
the release of soluble chemical constituents with the possible consequence of
environmental pollution. Such an event might be the result of a subsequent land
use which requires extensive digging into the FGD sludge. Therefore, long
term land-use planning will be necessary.
An evaluation of the amount of chemical pollutants that are avail-
able to the environment by leaching can be determined from the results of the
leaching data summarized in Figure 47 and the permeability data summarized
in Figure 17. An evaluation was made for an FGD waste containing 6000 mg/f
dissolved solids in the mother liquor and disposed of by each of the alterna-
tive techniques discussed in Section 9.3. The evaluation was made on the
assumption that the leachate quality would respond according to the presenta-
tion in Figure 47 and that the rate of passage of the leachate through the
sludge would be 10 cm/sec for untreated sludge and lO'-3 cm/sec for treated
sludge. The waste was placed to a depth of 30 ft during a five-year fill
period. In the cases where the method of disposal was undrained ponding, it
was assumed that a maximum hydraulic head of 6 ft was used during filling
and a 1-ft head, thereafter, when retired as a lake.
The results of this evaluation are presented in Figure 48 for
cases described in Table 28. Cases 1 and 2 represent the ponding technique
of disposal in which the final disposition of the pond is a lake. Case 1 is for
untreated FGD waste, and Case 2 is for treated waste. Cases 3 and 4 repre-
sent pond disposal of untreated and treated waste, respectively, in which
supernate is not allowed on the surface of the pond during filling and the dis-
posal site is retired by grading with the landscape. In these cases, the
li-
o.oiLj
I
I
1 10 20 30 40
PORE VOLUME DISPLACEMENTS
50
60
Figure 47. Leachate TDS from treated and untreated sludges
as a function of PVD.
143
-------
1.0
o
co
o
Zi
O
CO
o
LU
O
CO
CO
0.1
0.01
0.001
CASE 1
A END OF 5th PORE VOLUME
I
I
I
I
I
I
1
20 40
60 80
YEARS
100 120 140 1300
Figure 48. Mass loading of TDS to subsoil for various dispo-
sal modes of treated and untreated FGC wastes.
amount of water recharged to the basin represents a normal recharge for
indigenous soil. Case 5 is the condition in which treated FGD waste is dis-
posed of in a managed landfill where only 10 percent of the normal recharge
is allowed to penetrate the waste because runoff has been maximized. Fig-
ure 48 plots the TDS that are expected to reach the disposal site subsoil per
year as a function of time. Also plotted on each curve is an indication of the
point in time at which the flushing action of the leachate reaches five PVD.
This evaluation shows that, depending on the disposal method
selected, the mass loading of pollutants to the groundwater can be reduced by
as much as two orders of magnitude. Moreover, the significance of disposal
site management as a means of preventing chemical pollution to the environ-
ment is clearly shown. Although chemical treatment reduced the pollutant
load under each condition considered, the larger reduction in pollution load
was a consequence of minimizing rainwater recharge to the subsoil of the dis-
posal basin. This evaluation places in perspective the value of chemical
treatment and site management with respect to the environmental acceptabil-
ity of FGD waste disposal.
144
-------
TABLE 28. INPUT DATA FOR STUDY CASES
Case
1
2
3
4
5
Disposal
Method
Lake
Lake
Pond
Pond
Landfill
Surface
Water
Constant
supernate
Constant
supernate
10. in. /yr
recharge
10. in. /yr
recharge
1 in. /yr
recharge
FGC Waste, 5-Year Fill
Waste
Condition
Untreated
Treated
Untreated
Treated
Treated
Depth,
ft
30
30
30
30
30
Permeability,
cm/sec^
to'4
io-5
io-4
io-5
io-5
Fractional
Pore
Volume
0.67
0.67
0.67
0.67
0.67
Assumed maximum hydraulic head of 6 ft during filling, including depth of
wastes; 1 ft constant water cover thereafter.
For all cases, subsoil permeability = 10" cm/sec.
-------
SECTION X
DISPOSAL APPLICATIONS AND ESTIMATED COSTS
Alternatives available for the disposal of flue gas desulfurization
(FGD) wastes comprise several different approaches to ponding and land-
filling, each having distinct characteristics as to complexity, cost, pollution
potential, land requirements, land reclamation, and degree of technical
development. Additionally, the type of disposal selected for a given site may
be dictated or influenced by local conditions such as distance to disposal site,
topography, weather, soil characteristics, depth of water table, and ground
water characteristics. Considerable discussion of the methods available and
already selected has been presented by The Aerospace Corporation (110),
Haxo and White (31), Crowe (32), PEDCo-Environmental Specialists, Inc.
(33), U.S. Environmental Protection Agency (34), U.S. Army Engineer
Waterways Experiment Station (35), and others.
In this section, a general review of disposal methods is given.
The methods selected by power utilities that have scrubbers which are opera-
tional or were to be operational in 1976 are identified, and a disposal cost
summary is given. The EPA field disposal evaluation project at the TVA
Shawnee station (2) and the status, through 1979, of utility power plant fixa-
tion disposal commitments are reported in separate documents. These
studies are continuing, and a general summary of all FGD waste disposal will
be provided in an Aerospace report to the EPA later in 1977.
10. 1 DISPOSAL METHODS
Environmentally sound disposal methods, regardless of the type
used, must provide the means of preventing or minimizing the seepage of
waste liquors to water supplies. These are grouped into two general cate-
gories: (a) ponding of untreated waste and (b) landfilling with chemically
treated or physically conditioned waste. Each method has unique character-
istics, and one or more may be applicable at a given site, depending on site
conditions and the overall objectives of the user. The approaches that are in
use or that show good potential are listed in Table 29, with significant com-
parative factors indicated for each approach. Methods 1 through 4, which
cover ponding and chemical treatment ("fixation"), have been or are being
used or evaluated in the field; methods 5 and 6, landfilling with untreated
wastes, are being evaluated at the EPA Shawnee field evaluation project.
Also, a tabulation of SO2 scrubber waste disposal systems that are operating
or that were planned to be operating in 1976 is given in Table 30, with much of
147
-------
TABLE 29. SIGNIFICANT FEATURES OF DISPOSAL METHODS
Disposal Method2
1 . Pond
2. Landfill
-Fixation
3. Landfill
-Fixation
4. Pond
-Dewatered Sulfite,
Contains Fly Ash
5. Pond
-Dewatered Sulfite,
Without Fly Ash
6. Landfill
-Sulfate (Gypsum).
With or Without
Fly Ash
Unlined
X
X
X
Impervious
Soil on
Site
Impervious
Soil Added
or
Rigid Liner
Plastic or
Rubberized
Liner
X or X or X
X o
r X o
r X
Dam and
Impervious
Soil
X
Supcrnate
Return
X
X
X
Unde r-Drained
and
Recirculated
X
X
Compaction
X
Technique
Demonstrated
X
X
X
Structural
X
X
X
X
X
Soil Cover
Needed
After
Filling
X
xb
xb
Residual
Value
X
X
X
X
X
00
See Table 7 for total disposal costs.
Soil cap contoured to drain.
-------
TABLE 30. FGD DISPOSAL SYSTEMS OPERATING BY END OF 1976
Utility Company
Power Station
Arizona Public Service,
Cholla, No. 1
General Motors,
Chevrolet Parma,
Nos. 1, 2, 3, and 4
Gulf Power Co. ,
Scholz, No. 1A
Kansas City Power and
Light, Hawthorn,
No. 3
Kansas City Power and
Light, Hawthorn,
No. 4
Kansas City Power and
Light, La Cygne,
No. 1
Kansas Power and
Light, Lawrence,
No. 4
Kansas Power and
Light
Kentucky Utilities,
Green. River, Units
1 and 2
New or
Retrofit
Retrofit
Retrofit
Retrofit
Retrofit
Retrofit
New
Retrofit
New
Retrofit
Size of
FGD Unit,
MW equiv.
115
32
20
140
100
820
125
400
64
Absorbent
Limestone
Double alkali
Double alkali
Limestone injection and
wet scrub
Limestone injection and
wet scrub
Limestone
Limestone injection and
wet scrub
Limestone injection and
wet scrub
Lime
Startup
Date
10/73
3/74
2/75
11/72
8/72
2/73
12/68
11/71
9/75
FGD Waste
Treatment
Process or
Processor
None
None
None
None
None
None
None
None
None
Disposal
Method
Pond
Haul to landfill
Lined pond
Pond
Pond
Pond
Pond
Pond
Pond
£>
vD
-------
TABLE 30. FGD DISPOSAL SYSTEMS OPERATING BY END OF 1976 (Continued)
Utility Company
Power Station
Key West Utility
Board, Stock Island
Plant
Montana Power Co. ,
Colstrip, No. 1
Nevada Power,
Reid Gardner, No. 1
Nevada Power,
Reid Gardner, No. 2
Public Service Co. of
Colorado, Valmont,
No. 5
Montana Power,
Colstrip, No. 2
Nevada Power,
Reid Gardner,
No. 3
Northern States Power,
Sherburne, No. 1
Rickenbacker AFB,
Rickenbacker
Springfield City
Utilities, Southwest,
No. 1
Arizona Public Service,
Four Corners,
No. 5A
New or
Retrofit
New
New
Retrofit
Retrofit
Retrofit
New
New
New
Retrofit
New
Retrofit
Size of
FGD Unit,
MW equiv.
37
360
125
125
50
360
125
680
20
200
160
Absorbent
Limestone
Lime
Sodium carbonate
Sodium carbonate
Limestone
.-
Lime
Sodium carbonate
Limestone
Lime
Limestone
Lime
Startup
Date
10/72
10/75
4/74
3/73
10/74
7/76
6/76
5/76
3/76
7/76
2/76
FGD Waste
Treatment
Process or
Processor
None
None
None
None
None
None
None
None
None
None
None
Disposal
Method
Pond
Pond
Pond
Pond
Pond
Pond
-------
TABLE 30. FGD DISPOSAL SYSTEMS OPERATING BY END OF 1976 (Continued)
Utility Company
Power Station
Central IL. Light Co. ,
Duck Creek, No. 1
Detroit Edison,
St. Clair, No. 6
Commonwealth Edison,
Will County, No. 1
Duquesne Light Co. ,
Phillips
Duquesne Light Co. ,
Elrama
Gulf Power Co. ,
Scholz, Nos. IB
and 2B
Columbus and Southern
Ohio Electric,
Conesville, No. 5
Louisville Gas and
Electric, Cane Run,
No. 4
Pennsylvania
Power Co. ,
Bruce Mansfield,
No. 1
Texas Utilities Co. ,
Martin Lake, No. 1
Louisville Gas and
Electric, Paddys
Run, No. 6
New or
Retorfit
New
Retrofit
Retrofit
Retrofit
Retrofit
Retrofit
New
Retrofit
New
New
Retrofit
Size of
FGD Unit,
MW equiv.
100
180
167
387
210
23
400
178
825
793
65
Absorbent
Limestone
Limestone
Lime
Lime
Lime
Lime
Lime
Limestone
Lime
Startup
Date
6/76
3/76
2/72
7/73
10/75
3/75
8/76
6/76
4/76
9/76
4/73
FGD Waste
Treatment
Process or
Processor
None
None
Lime and fly ash
Amer. admixtures
Decision pending
IU Conversion
Systems
Chiyoda gypsum
IU Conversion
Systems
Yes
Calcilox
(Dravo)
Fixation (Research
Cotrell)
Compacted with
fly ash
Disposal
Method
Pond
Planned landfill
Haul to landfill
Tiered landfill
Lined pond
Tiered landfill
Dam and lake
Landfill
Borrow pit
-------
the data based on information provided by PEDCo (33). These projects are
referred to as "current disposal systems" in this discussion. Of these, 21
units are using ponding, and 8 are using fixation or physical conditioning, with
a total capacity of 4338 and 2983 MW, respectively. Commitments to 1983
(3, 33, 19) identify approximately 80 units committed to nonregenerable scrub-
bers (~ 32,000-MW capacity), of which 15 units (7155-MW capacity) are com-
mited to chemical treatment of FGD wastes.
Each of the methods listed in Table 29 offers a different form of
protection against water pollution and a different approach to land reclama-
tion. Since the long-term effects of ponding and landfilling are still being
evaluated, no single method is considered superior for all cases. Table 31
summarizes the most significant factors associated with each form of
disposal.
10.2 DISPOSAL COST ESTIMATES
Cost estimates for ponding and chemical treatment (fixation) plus
landfilling have been made and reported by Aerospace on several occasions.
In recent studies, performed during the task of coordinating the EPA field dis-
posal evaluation at the TVA Shawnee station, detailed cost estimates were
made of fixation and disposal and were reported (2). The ponding cost esti-
mates presented in (1) and (5) are updated in Section 10. 2. 1 and are compared
with a summary of the fixation disposal costs in Section 10. 2. 2.
10.2.1 Ponding Cost Estimates
Disposal pond costs are dependent on construction and equipment
factors such as dike construction, ancillary equipment, liner materials and
installation, and land costs, as well as factors which are a function of geo-
graphic location such as shipping and labor. An analysis of these factors
was made by Aerospace (1) in 1973 and was subsequently revised (2) in
November 1974 to reflect sharp price increases during that period for flexible
liner materials, equipment, and construction. Since that time, additional
work has been published (3 4), which correlates well with the earlier work.
Since 1974 the price changes have been more moderate, and in this report we
have updated the earlier estimates using economic indices as of 1 January
1976 (5).
The estimates are based on a 1000-MW power plant burning coal
with 3 percent sulfur and 12 percent ash. The service life of the power plant
is assumed to be 30 years, operating at an average load factor of 50 percent,
with an annual average sludge (50 percent solids, including ash) production
of 930,000 metric tons (1,025,000 short tons). The annualized costs include
the costs for labor, maintenance, and capital charges of 18 percent. The
annualized costs include the costs for labor, maintenance, and capital charges
of 18 percent. The capital charges include replacements, insurance, taxes,
cost of capital based on 50/50 debt-equity funding, and the use of straight-
line depreciation.
152
-------
TABLE 31.
SIGNIFICANT ENVIRONMENTAL FACTORS ASSOCIATED
WITH ALTERNATIVE WASTE DISPOSAL METHODS
Disposal
Method
Ponding
Chemical
Treatment
Landfill
Untreated
Landfillb
Land
Reclaimable
No
Yes
Yes
Leachate
Concentrations
Meet Drinking
Water Criteria
No
No
No
Leachable
Constituents
Large
Minimal
Large
Lining
and/or
Underdrain
Yes
Noa
Yes
Long-Term
Pollution
Control
Questionable
Good Potential
Good Potential
For extreme cases, e. g. , highly porous soil, lining or underdrainage may be appropriate.
This method has little or no controlled field evaluation but will be evaluated by EPA (2).
Lining or underdrainage and impermeable cap after completion offer good potential for long-
term control.
-------
Figure 49 presents a summary of the typical installed costs of
various liner materials. The range of costs shown reflect variations in the
cost of materials, transportation, labor, and supplier discount practices.
Two of these materials, 20-mil polyvinyl chloride (PVC) and 30-mil Hypalon,
are considered to be representative of the spread in practical liner costs,
and each was used in estimating 30-year average disposal costs. The results,
shown in Figures 50 and 51 are presented for waste depths varying from 10
to 40 ft. These costs were estimated for a total of three ponds, each having
a 10-year sludge capacity. Also, it was assumed that all of the land needed
for the 30-year output would be purchased at the beginning of the project.
Equipment costs were included for pumps and a two-way pipeline of 1. 6 km
(1 mile), instrumentation to monitor effluent to the pond, and instrumentation
to measure soil resistivity as a means of detecting leaks in the pond linear.
Total labor costs were included at an annual rate of $102,000. It was also
assumed that decreasing quantities of land would be needed over the 30-year
operating period as a result of both the diminishing quantities of sludge being
produced and savings realized in the joint use of dike walls by multiple ponds.
Land costs were estimated to vary between $1000/acre and $5000/acre. The
results show that disposal costs, including land at $1000/acre, for 50 percent
solids, are about $2. 85/ton for PVC and $3.90/fcon for Hypalon at a waste
depth of 30 ft. (These values increase to $3. 50 and $4. 30/fcon, respectively,
when a land cost of $5000/acre is used.) Ponding costs are converted to a dry
FGD waste basis and compared with fixation disposal costs in Table 7. The
total capital investment and annual operating costs for both PVC-20 and
Hypalon-30 are given in Tables 32 through 37.
10.2.2 Cost Comparisons; Ponding Versus Fixation
A comparison of the ponding cost estimates developed in Sec-
tion 10.2. 1 and the fixation disposal cost estimates presented in (2) is shown
in Table 7. Those costs are given in units of dollars per ton of waste (dry),
dollars per ton of coal, and mills per kilowatt hour. This comparison in-
dicates that the cost of ponding with liner installed is approximately 75 percent
of the fixation disposal costs .or 40 percent if the pond has a natural clay liner.
Excluded from these values are additional costs for the reclamation of the site,
which may include the addition of top soil and nutrients on fixed landfills, and
a yet-to-be-defined procedure for closing down a pond for permanent environ-
mental protection. Also excluded is any cost reduction in terms of realizing
the residual value of the land where applicable and any benefit to be realized
from including current ash disposal with FGD total waste disposal where
applicable.
Background information for these cost estimates is given in
Tables 32 through 37 which provide capital costs for ponding at 1, 000 MW
station. Values consider cost variations of liner materials and land, and
are given for three different dike heights. Also given is the conversion of
capital costs to annual operating costs in dollars per ton of wet sludge (50 per-
cent solids) for a 30-year lifetime.
154
-------
POLYETHYLENE - 10 mils
POLYVINYL CHLORIDE - 10 mils
POLYVINYL CHLORIDE - 20 mils
PETROMAT FABRIC - 125 mils
^ 10 1 BUTYL RUBBER - 30 mils
o CHLORINATED
< ' oi POLYETHYLENE - 30 mils
1 mil = 0.0254 mm
1 in. 2.54 cm I o-1 HYPALON - 30 mils
EPDM RUBBER -
30 mils
CLAY - 18 in.
BITUMINOUS CONCRETE -
4 TO 6 in.
2 3 4
INSTALLED COST, $/sq yd
Figure 49. Installed liner costs.
155
-------
c
o
w-
l/l
O
O
CO 7
O *
Q_
50% SOLIDS INCLUDING FLY ASH
30-YEAR AVERAGE
JANUARY 1976 DOLLARS
Note.- To convert to dry
basis, double these
costs
TOTAL COST INCLUDING
LAND AT $5000/acre
INSTALLED
LINER
ONLY
TOTAL COST INCLUDING
LAND AT $1000/acre
TOTAL COST
WITHOUT LAND
DIKE AND PUMPLINE CONSTRUCTION,
PUMPS, POWER, INSTRUMENTATION,
MONITORING EQUIPMENT, AND OPERATING LABOR
10 20 30
SLUDGE DEPTH, ft
40
Figure 50. FGD waste ponding disposal costs: 20-mil PVC.
156
-------
I 5
w-
co
CO
O
0 4
CO
o
Q_
co
50% SOLIDS INCLUDING FLY ASH
30-YEAR AVERAGE
JANUARY 1976 DOLLARS
Note: To convert to dry
basis, double these
costs
TOTAL COST INCLUDING
LAND AT $5000/acre
INSTALLED
LINER
ONLY
TOTAL COST INCLUDING
LAND AT $1000lacre
TOTAL COST
WITHOUT LAND
I
I
DIKE AND PUMPLINE CONSTRUCTION,
PUMPS, POWER, INSTRUMENTATION,
MONITORING EQUIPMENT, AND OPERATING
LABOR
_J I
10 20 30
SLUDGE DEPTH, ft
40
Figure 51. FGD waste ponding disposal costs:
30-mil Hypalon.
157
-------
TABLE 32. TOTAL PONDING COSTS: PVC-20 LINER; 10-FT SLUDGE DEPTH
(JANUARY 1976 DOLLARS; 1000-MW PLANT)a
Total Investment, $> Millions
Installed Liner = 16.5
Construction = 2.4
Liner + Construction + $1000/acre Land = 20.56
Liner + Construction + $5000/acre Land = 27.20
00
Annual Operating Costs
Cost Item
Liner
Construction/Equipment
Operating Labor
Subtotal
$1000/acre Land
$5000/acre Land
Total
Total Cost, $ Millions
2.97
0.43
0. 102
3.502
0.300
1.49
3.802 4.992
Cost/Wet Ton, $C
2.90
0.42
0.10
3.42
0. 29
1.45
3.71 4.87
aBased on disposal of 1.025 X 10 wet tons/year, 50 percent solids,
30-year average.
Construction includes dike construction, pipeline, pumps, operating
power, instrumentation, and monitoring equipment.
cMultiply by 2 to convert to $/ton (dry).
-------
TABLE 33. TOTAL PONDING COSTS: PVC-20 LINER; 20-FT SLUDGE DEPTH
(JANUARY 1976 DOLLARS; 1000-MW PLANT)a
Total Investment, $ Millions
Installed Liner = 10.05
Construction = 5. 36
Liner -t- Construction + $1000/acre Land = 16.346
Liner + Construction + $5000/acre Land = 20.09
Annual Operating Costs
Cost Item
Liner
Construction/Equipment
Operating Labor
Subtotal
$1000/acre Land
$5000/acre Land
Total
Total Cost, $ Millions
1.81
0.97
0. 102
2.882
0. 168
0.842
3.050 3.724
Cost/Wet Ton, $C
1.77
0.94
0. 10
2.81
0. 16
0.82
2.97 3.63
Based on disposal of 1.025 X 10 wet tons/year, 50 percent solids,
30-year average.
Construction includes dike construction, pipeline, pumps, operating
power, instrumentation, and monitoring equipment.
°Multiply by 2 to convert $/ton (dry).
-------
TABLE 34. TOTAL PONDING COSTS: PVC-20 LINER; 40-FT SLUDGE DEPTH
(JANUARY 1976 DOLLARS; 1000-MW PLANT)a
Total Investment, $ Millions
Installed Liner = 5.92
Construction = 12.26
Liner + Construction + $1000/acre Land = 18.788
Liner + Construction + $5000/acre Land = 21.220
Annual Operating Costs
Cost Item
Liner
Construction/Equipment
Operating Labor
Subtotal
$1000/acre Land
$5000/acre Land
Total
Total Cost, $ Millions
1.07
2. 21
0. 102
3.382
0. 109
0.547
3.491 3.929
Cost/Wet Ton, $°
1.04
2. 16
0. 10
3. 30
0.11
0.53
3.41 3.83
Based on disposal of 1.025 X 10 wet tons/year, 50 percent solids,
30-year average.
Construction includes dike construction, pipeline, pumps, operating
power, instrumentation, and monitoring equipment.
Multiply by 2 to convert to $/ton (dry).
-------
TABLE 35. TOTAL, PONDING COSTS: HYPALON-30 LINER; 10-FT SLUDGE DEPTH
(JANUARY 1976 DOLLARS; 1000-MW PLANT)a
Total Investment, $ Millions
Installed Liner = 31.93
Construction = 2.4
Liner + Construction + $1000/acre Land = 35.99
Liner + Construction + $5000/acre Land = 42.63
Annual Operating Costs
Cost Item
Liner
Cons tructi on/Equipment
Operating Labor
Subtotal
$1000/acre Land
$5000/acre Land
Total
Total Cost, $ Millions
5.75
0.432
0. 102
6.284
0.300
1.49
6.584 7.774
Cost/Wet Ton, $°
5.61
0.42
0. 10
6.13
0.29
1.45
6.42 7.58
Based on disposal of 1.025 X 10 wet tons/year, 50 percent solids,
30-year average.
Construction includes dike construction, pipeline, pumps, operating
power, instrumentation, and monitoring equipment.
°Multiply by 2 to convert to ft/ton (dry).
-------
TABLE 36. TOTAL PONDING COSTS: HYPALON-30 LINER; 20-FT SLUDGE DEPTH
(JANUARY 1976 DOLLARS; 1000-MW PLANT)a
Total Investment, $ Millions
Installed Liner = 16.63
Construction =5.36
Liner + Construction + $1000/acre Land = 22.926
Liner + Construction + $5000/acre Land = 26.670
Annual Operating Costs
Cost Item
Liner
Cons truction/Equipment
Operating Labor
Subtotal
$1000/acre Land
$5000/acre Land
Total
Total Cost, $ Millions
2.99
0.965
0. 102
4.057
0. 168
0.842
4.225 4.899
Cost/Wet Ton, $c
2.92
0.94
0. 10
3.96
0. 16
0.82
4.12 4.78
Based on disposal of 1.025 X 10 wet tons/year, 50 percent solids,
30-year average.
Construction includes dike construction, pipeline, pumps, operating
power, instrumentation, and monitoring equipment.
Multiply by 2 to convert to $/ton (dry).
-------
TABLE 37. TOTAL PONDING COSTS: HYPALON-30 LINER; 40-FT SLUDGE DEPTH
(JANUARY 1976 DOLLARS; 1000-MW PLANT)a
Total Investment, $ Millions
Installed Liner = 9.62
Construction = 12.26
Liner + Construction + $lOOO/acre Land = 22.49
Liner + Construction + $5000/acre Land = 24.92
Annual Operating Costs
Cost Item
Liner
Construction/Equipment
Operating Labor
Subtotal
$1000/acre Land
$5000/acre Land
Total
Total Cost, $ Millions
1.732
2. 210
0. 102
4. 044
0. 109
0. 547
4.153 4.591
Cost/Wet Ton, $C
1.69
2. 16
0. 10
3.95
0. 11
0.53
4.06 4.48
Based on disposal of 1.025 X 10 wet tons/year, 50 percent solids,
30-year average.
Construction includes dike construction, pipeline, pumps, operating
power, instrumentation, and monitoring equipment.
"Multiply by 2 to convert to $ /ton (dry).
-------
REFERENCES
1. J. Rossoff and R. C. Rossi, Disposal of By-Products from Non-
Regenerable Flue Gas Desulfurization Systems: Initial Report,
EPA-650/2-74-037-a, U.S. Environmental Protection Agency,
Research Triangle Park, North Carolina (May 1974).
2 R. B. Fling, et al. , Disposal of Flue Gas Cleaning Wastes: EPA
Shawnee Field Evaluation: Initial Report, EPA-600/2-76-070. (NTIS
No. PB Z51-876/AS) U. S. Environmental Protection Agency, Research
Triangle Park, North Carolina (March 1976).
3. P. P. Leo and J. Rosoff, Control of Waste and Water Pollution from
Power Plant Flue Gas Cleaning Systems: First Annual R and D Report,
EPA-600/7-76-018, U.S. Environmental Protection Agency, Research
Triangle Park, North Carolina (October 1976).
4. J. Rossoff, et al. , "Study of Disposal and Utilization of By-Products
from Throwaway Desulfurization Proces.ses," Proceedings of EPA Sym-
posium on Flue Gas Desulfurization, New Orleans (14-17 May 1973).
EPA-650/2-73-038 (NTIS No. PB 250-901 /AS), U.S. Environmental
Protection Agency, Research Triangle Park, North Carolina (1973).
5. J. Rossoff, et al. , "Disposal of By-Products from Non-Regenerable
Flue Gas Desulfurization Systems: A Status Report, " Proceedings of EPA
Symposium on Flue Gas Desulfurization - Atlanta, November 1974,
Vol. I, EPA-650/2-74-126-a (NTIS No. PB 242-572/AS), U.S. Envi-
ronmental Protection Agency, Research Triangle Park, North Carolina
(December 1974).
6. J. Rossoff, et al., "Flue Gas Cleaning Waste Disposal - EPA Shawnee
Field Evaluation, " Proceedings of EPA Flue Gas Desulfurization Sym-
posium - New Orleans. March 1976, Vol. II, EPA-600/Z-76-156b, U.S.
Environmental Protection Agency, Research Triangle Park, North
Carolina (March 1976).
7. J. W. Jones, et al. , "Disposal of By-Products from Non-Regenerable
Flue Gas Desulfurization Systems, "Proceedings of American Society
of Civil Engineers Annual and National Environmental Engineering Con-
vention - Kansas City, Missouri, October 1974, U.S. Environmental
Protection Agency, Research Triangle Park, North Carolina (October
1974).
165
-------
8. J. W. Jones, et al. , "Flue Gas Cleaning Waste Characterization and
Disposal Evaluation," Proceedings of Fourth International Ash Utili-
zation Symposium - St. Louis, March 1976, MERC/SP-76/4, Energy
Research and Development Agency, Washington, D. C. (1976).
9. "Study of Disposal By-Products from Non-Regenerable Flue Gas Desul-
furization Systems," Mid-Term Status, Industry Briefing prepared for
U.S. Environmental Protection Agency, Research Triangle Park,
North Carolina, by The Aerospace Corporation, El Segundo, California
(19 December 1973).
10. "Shawnee FGD Waste Disposal Field Evaluation," Industry Briefing
presented at Raleigh, North Carolina, for U.S. Environmental Pro-
tection Agency, Research Triangle Park, North Carolina, by The
Aerospace Corporation, El Segundo, California (24 September 1975).
11. L. J. Bornstein, R. B. Fling, F. D. Hess, R. C. Rossi, and
J. Rossoff, Reuse of Power Plant Desulfurization Waste Water,
EPA-600/2-76-024, U.S. Environmental Protection Agency,
Research Triangle Park, North Carolina (February 1976).
12. "National Interim Primary Drinking Water Regulations," Water Pro-
grams, Part 141, Federal Register, Vol. 40, No. 248 (24 December 1975).
13. "Resource Conservation and Recovery Act of 1976, " Public Law 94-580,
94th Congress (21 October 1976).
14. "Alternative Waste Management Techniques for Best Practicable Waste
Treatment," Federal Register, Vol. 41, No. 29 (11 February 1976).
15. "United States Public Health Service Drinking Water Standards, 1962,"
Publication No. 956, U.S. Public Health Service, Washington, DC (1962).
16. A. M. Di Gioia, Sr. , and W. L. Nuzzo, "Fly Ash as Structural Fill, "
J. Pwr. Div. , Proceedings of the American Society of Civil Engineers,
pp. 77-92 (June 1972).
17. J. E. McKee and H. W. Wolf, "Water Quality Criteria, " Second Edition,
State Water Resources Control Board, Publication 3-A, Sacramento,
California (April 1971).
18. "Standard Method of Test for Compressive Strength of Molded Concrete
Cylinders," ASTM-C-39-64, American Society for Testing and Materials,
Philadelphia, Pennsylvania (1964).
19. Flue Gas Desulfurization Installations and Operations, U.S. Environ-
mental Protection Agency, Washington, DC (September 1974).
£
(S
166
-------
20. Handbook for Analytical Quality Control in Water and Wastewater
Laboratories, Analytical Quality Control Laboratory, National En-
vironmental Research Center, Cincinnati, Ohio (1972).
21. Methods for Chemical Analysis of Water and Wastes, Analytical Quality
Control Laboratory, National Environmental Research Center,
Cincinnati, Ohio (1971).
22. D. J. von Lehmden, R. H. Jungers, and R. E. Lee, Jr. , "Determina-
tion of Trace Elements on Coal, Fly Ash, Fuel Oil, and Gasoline - A
Preliminary Comparison of Selected Analytical Techniques, " Analytical
Chemistry, Vol. 46, No. 2, p. 239 (1974).
23. F. G. Mesich, K. Schwitzgebel, et al. , Coal Fired Power Plant Trace
Element Study, Vol. 1, Radian Corporation, Austin, Texas, EPA
Contract No. 68-01-2663 (September 1975).
24. R. R. Ruch, H. J. Gluskoter, and N. F. Shimp, Occurrence and
Distribution of Potentially Volatile Trace Elements in Coal,
EPA-650/2-74-054, U.S. Environmental Protection Agency, Research
Triangle Park, North Carolina (July 1974).
25. E. M. Magee, H. J. Hall, and G. M. Varga, Jr., Potential Pollutants
in Fossil Fuels, EPA-R2-73-249, U.S. Environmental Protection
Agency, Research Triangle Park, North Carolina (June 1973).
26. T. Kessler, A. G. Sharkey, Jr., and R. A. Friedel, "Analysis of
Trace Elements in Coal by Spark-Source Mass Spectrometry, " Bureau
of Mines R.I. . p. 7714(1973).
27. V. E. Swanson, "Composition and Trace Element Content of Coal and
Power Plant Ash, " Southeast Energy Study, Part II, Appendix J. U. S.
Geological Survey, Denver, Colorado (January 1972).
28. P. N. Cheremisinoff and R. T. Fellman, Power Engineering, p. 54
(October 1974).
29. P. S. Lowell, D. M. Ottmers, T. I. Strange, K. Schwitzgebel, and
D. W. DeBerry, A Theoretical Description of the Limestone Injection
Wet Scrubbing Process, Vol. 1, Final Report for NAPCA Contract No.
CPA2269138, Radian Corporation, Austin, Texas (9 June 1970).
30. R. H. Borgwardt,"Limestone Scrubbing at EPA Pilot Plant," IERL/RTD
Progress Report No. 14, U. S. Environmental Protection Agency,
Research Triangle Park, North Carolina (June 1975).
31. H. E. Haxo, Jr., and R. M. White, First Interim Report, Evaluation
of Liner Materials Exposed to Leachate, Prepared for U. S. Environ-
mental Protection Agency, Cincinnati, Ohio, by Matrecon, Inc. , Oakland,
California (27 November 1974).
167
-------
32. J. L. Crowe and H. W. Elder, "Status and Plans for Waste Disposal
from Utility Applications of Flue Gas Desulfurization Systems," Pro-
ceedings of EPA Flue Gas Desulfurization Symposium - New Orleans,
March 1976. Vol. II, EPA-600/2-76-156b. U.S. Environmental Protec-
tion Agency, Research Triangle Park, North Carolina (March 1976).
33 Summary Report: Flue Gas Desulfurization Systems, Prepared for U.S.
Environmental Protection Agency, Research Triangle Park, North Caro-
lina, Contract No. 68-02-1321, Task No. 28, by PEDCo-Environmental
Specialists, Inc., Cincinnati, Ohio (January-February 1976).
34. Sulfur Oxide Throwaway Sludge Evaluation Panel (SOSEP), Final Report,
EPA-650/2-75-010-a, (NTIS No. PB 242-618/AS), U.S. Environmental
Protection Agency, Research Triangle Park, North Carolina (April 1975).
35. J. L. Mahloch, et al., Pollutant Potential of Raw and Chemical Fixed Haz-
ardous Industrial Wastes and Flue Gas Desulfurization Sludges: Interim
Report, U.S. Army Engineer Waterways Experiment Station, Vicksburg,
Mississippi, EPA 600/2-76-182, U.S. Environmental Protection Agency,
Research Triangle Park, North Carolina (July 1976).
36. C. L. Parker, "Estimating the Cost of Wastewater Treatment Ponds, "
Pollution Engineering, p. 32 (November 1975).
37. "Economic Indicators, " Chemical Engineering, Vol. 23, No. 5, p. 7
(1 March 1976).
38. L. G. Sillen and A. E. Martell, "Stability Constants of Metal-Ion Com-
plexes," Special Publication No. 17, The Chemical Society, London (1964).
39. J. Johnston and C. Grove, Journal of the American Chemical Society,
Vol. 53, p. 3976 (1931).
40. J. J. Bucy, The Economic Feasibility of Producing Sulfuric Acid From SO2
in Power Plant Stack Gases, University of Nebraska (February 1975).
41. L. L. Boulden, "Moving Solids Like Liquids," Machine Design, p. 87
(24 February 1972).
42. E. J. Wasp, T. L. Thompson, and T. C. Aude, "Slurry Pipeline Eco-
nomics and Applications," Paper presented at the 26th Annual ASME
Petroleum Mechanical Engineering Conference, Houston, Texas (19-23
September 1971).
43. FGD Sludge Fixation and Cost Study, Dravo Corporation, Denver,
Colorado, TVA Contract No. 75F71-59402 (8 August 1975).
44. An Evaluation of the Disposal of Flue Gas Desulfurization Wastes in
Mines and Oceans, Initial Assessment, Arthur D. Little, Inc., Cambridge,
Massachusetts, Contract No. 68-02-2334, U.S. Environmental Protection
Agency, Research Triangle Park, North Carolina (to be published).
168
-------
APPENDIX A
CHARACTERIZATION OF SLUDGE SOLIDS
The characterization of the solids portion of flue gas desulfuri-
zation (FGD) sludges included the identification of the separate crystalline
phases; determination of the amount of each of these phases; and delineaton
of the morphological traits of these phases relative to particle size, shape,
and distribution. The crystalline phase identification was done by x-ray
diffractometry by standard methods and provided a knowledge of the crystal
structures and the relative amount of each structure within each sample
material. The amount of each phase was determined by chemical analysis
of the solid portion of each sludge by the techniques described in
Appendix C. Reconstitution of phases was accomplished not only from the
chemical analyses but also from the determination of the chemistry of the
soluble and hydrous forms of the solids portion and from the evaluation of
the relative quantities indicated by the x-ray analyses.
The physical characteristics of the individual particles were
determined through observation by scanning electron micrography. These
observations revealed the shape and growth habits of the crystalline phases,
the particle sizes and distribution, and the presence (or absence) of phases
indicated by either x-ray or chemical analyses.
A. 1 X-RAY DIFFRACTION CHARACTERIZATION
The objective in the use of x-ray diffraction techniques on power
plant FGD waste was to provide quantitative characterization of the crystal-
line phases in the solid waste. The results of the study indicated that,
although this method is not suitable for quantitative analyses, it is useful
for phase identification that can not be performed by other techniques.
Sludges were oven-dried at temperatures between 70° and 80° C,
and the resultant dry powders were mounted in a conventional powder sample
mount for x-ray diffraction. Copper radiation at1.50 kV,|.a nickel filter, and
a scintillation counter were used. Diffraction peaks between 10- and 50-deg
Bragg angle were recorded, and the resulting patterns were interpreted.
The results of these analyses are presented in Table A-l. The relative in-
tensity values refer only to the intensity among phases in each specific
sludge and indicate no relationship between sludges. Furthermore, since
no standards were used, relative intensities do not necessarily infer relative
concentration of the various phases within a sample because of the differences
in diffraction behavior among differing crystalline phases.
169
-------
TABLE A-l. RELATIVE X-RAY INTENSITIES OF CRYSTALLINE PHASES
IN DRIED SLUDGE SAMPLES
Phase,
Atomic
Formula,
ASTM Index No.
Gypsum,
CaSO4'2H2O,
6-46
Ca Sulfite Hemihydrate,
CaSCy 1/2H2O,
4-588
Ca Sulfate Hemihydrate,
2CaSO4' H2O,
14-453
Anhydrite,
CaSO4,
6-226
Calcite,
CaCO3.
5-586
TVA Shawnee
Limestone,
7/12/73
Strong
Moderate
Moderate
TVA Shawnee
Limestone,
11/27/73
Strong
Moderate
Very Weak
TVA Shawnee
Limestone,
6/15/74
Very Strong
Strong
Weak
TVA Shawnee
Lime,
3/19/74
Weak
Strong
SCE Mohave
Limestone,
3/30/73
Strong
Weak
Very Weak
Moderate
GM Parma
Doable Alkali,
7/18/74
Moderate
Strong
Weak
APS Cholla
Limestone,
4/1/74
Strong
Weak
DLC Phillips
Lime,
6/17/74
Strong
Moderate
UPL Gadsby,
Double Alkali,
3/9/74
Weak
Strong
Weak
Weak
Intensity values relative to the individual phases in each sample.
-------
The results of the x-ray study revealed that in most cases sev-
eral calcium sulfate and sulfite phases were present, in addition to the
expected phases of gypsum and calcium sulfite hemihydrate. In addition,
two other major difficulties developed that prevented consideration of quanti-
tative x-ray diffraction techniques: (a) the line broadening that occurred as
a consequence of the fine particle size of the sulfur phases and (b) the pre-
ferred particle orientation that occurred in sample mounting and resulted,
in some cases, in the complete absence of some of the major diffraction
peaks from the x-ray patterns, particularly with calcium sulfite hemi-
hydrate. The techniques that have been developed to overcome these
difficulties are awkward to use, are time consuming, and create major
uncertainties in the final results. It was decided, therefore, that the x-ray
study would serve solely as a qualitative indication of the phases present.
The Tennesse Valley Authority (TVA) Shawnee limestone sludges
showed strong gypsum peaks (Table A-l). The calcium sulfite hemihydrate,
in all cases, was the second most prominent phase in the x-ray patterns,
appearing as a moderately strong phase. Residual limestone (calcite)
appeared, also, as a weak phase in all samples.
The TVA Shawnee lime sludge displayed an x-ray pattern con-
taining only the calcium sulfite hemihydrate as a strong phase and gypsum
as a weak phase. No residual lime or calcium carbonate phases were de-
tected. Specifically, one sample showed an absence of major diffraction
peaks for the sulfite phase, whereas other samples showed only an irregular
peak intensity distribution for this phase. In nearly every sample, the dif-
fraction peaks were displaced toward greater interatomic spacing.
The Southern California Edison (SCE) Mohave sludge revealed
the presence of gypsum as the major phase. In addition, weak patterns of
calcium sulfate hemihydrate appeared, indicating a thermal decomposition
during drying at 80°C. A moderate calcium carbonate signal was revealed
from residual limestone at a measured concentration of 0. 5 to 0. 3 percent
CaCO3 in the slurry (wet basis. )v
The General Motors (GM) Parma sludge revealed the calcium
sulfate hemihydrate only as a major phase and a weak presence of calcium
carbonate. The strong x-ray diffraction peak of the sulfate hemihydrate in
this sludge probably indicated the presence of gypsum before oven drying.
The line broadening of the x-ray pattern indicated a fine particulate size,
and it is probably because of the fine particles that thermal decomposition
was complete at drying temperatures. (Thermal decomposition of gypsum
to hemihydrate occurs at about 100°C at 1 atm. )
The Arizona Public Service (APS) Company Cholla sludge x-ray
pattern consisted primarily of gypsum and calcium sulfite hemihydrate peaks.
No residual limestone was detected.
Private communication, Dr. Alex Weir, Southern California Edison
Company, Los Angeles, California.
171
-------
The Duquesne Light Company (DLC) Phillips lime sludge revealed,
in its x-ray pattern, the presence of both calcium sulfate dihydrates and cal-
cium sulfite hemihydrate.
The Utah Power and Light (UPL) Company Gadsby sludge x-ray
pattern revealed relatively strong peaks from calcium sulfate hemihydrate
and weak peaks of gypsum and calcium carbonate. In addition, the calcium
anhydrite phase was revealed. The fine particle size indicated by line broad-
ening in the x-ray pattern probably contributed to its thermal decomposition,
as in the GM Parma sludge.
In summary, the x-ray diffraction characterization, while not
providing quantitative data, did provide valuable information in identifying
phases not otherwise expected in the sludge and also provided semiquantita-
tive data relative to the amount of these phases in the sludge. Line broaden-
ing and the absence or irregular distribution of peak intensities inferred the
presence of fine particles and the preferred orientation of these particles in
the sample, which, relative to ASTM standards, are assumed to be randomly
oriented.
A. 2 SOLIDS PHASE COMPOSITION BY CHEMICAL ANALYSIS
The composition of the solids fraction of each of the sludges
sampled was determined by chemical means and is presented in Table A-2;
a description of the analytical techniques used are presented in Appendix C.
The wide range in composition for each of the major solid constituents re-
flects the various design differences that exist among the scrubber systems.
Wastes from systems having high-efficiency fly ash collection facilities up-
stream of the scrubber have compositions that contrast with those from
systems having less efficient collection methods. The calcium sulfate con-
tent of the waste reflects in each case the capability of the calcium sulfite
to oxidize, and this condition usually occurs in the scrubber or reaction
tank; exceptions do take place in particular system designs, e.g., the UPL
double-alkali system. The amount of total sulfur waste (sulfite and sulfate)
does not directly reflect the sulfur content of the coal since other parameters
are more influential. For example, the SCE Mohave sludge consists
primarily of gypsum, but its coal sulfur content, less than 0.5 percent, is
the lowest among the coals sampled. The relative quantity of sulfur waste
products in this sludge is primarily a consequence of the low fly ash input to
the scrubber. At the other extreme, the TVA Shawnee facilities do not col-
lect fly ash upstream of the scrubber, but the high sulfur content of the coal
more than offsets the ratio of fly ash to sulfur waste when compared to other
systems with lower sulfur contents. The high sulfur total for each sludge is
believed to be a consequence of phase dehydration, which cannot be accur-
ately calculated.
A. 2.1 Crystalline Morphology
A portion of sludge solids from each scrubbing facility was se-
lected from the materials prepared for x-ray characterization (Section A. 1)
172
-------
TABLE A-2. PHASE COMPOSITION OF FGD SLUDGES IN WEIGHT PERCENT
Atomic
Formula
CaS04'2H20
CaSOj- 1/2H2O
CaS04- 1/2H20
CaCO3
MgS04'6H20
Na2SO4"7H2O
NaCl
CaS04a
Fly Ash
Total
TVA Shawnee
Limestone,
2/1/73
21.9
18.5
38.7
4.6
20. 1
103.8
TVA Shawnee
Limestone,
7/12/73
15.4
21.4
20.2
3.7
40.9
101.6
TVA Shawnee
Limestone,
6/15/74
31.2
21.8
4.5
1.9
40. 1
99.5
TVA Shawnee
Lime,
3/19/74
6.3
48.8
2. 5
1.9
40.5
100.0
SCE Mohave
Limestone,
3/30/73
84.6
8.0
6.3
1.5
3.0
103.4
GM Parma
Double Alkali,
7/17/74
48.3
12.9
19.2
7.7
6.9
7.4
101.4
APS Cholla
Limestone,
4/1/74
17.3
10.8
2.5
58.7
89.3
DLC Phillips
Lime,
6/17/74
19.0
12.9
0.2
59.7
91. 8
UPL Gads by
Double Alkali,
8/9/74
63.8
0.2
10.8
17.7
8.6
101. 1
Phase not explicitly measured; presence deduced from x-ray study.
-------
for subsequent examination on the scanning electron microscope (SEM). The
dry powders were sprinkled onto an aluminum sample holder and vapor-
coated with a thin layer of conducting carbon before insertion into the SEM
viewing chamber. A series of photomicrographs were taken of each sludge
sample, from which typical examples of the crystalline morphology of the
solids content were selected for presentation here. The photomicrographs
and description of the solid particulate material characteristic of each sludge
are presented in the following text.
A.2.1.1 TV A Shawnee Limestone Sludge
The distribution of particles shown in Figure A-l represents the
solids content of the sludge from the TVA Shawnee turbulent contact absorber
(TCA) system using limestone absorbent. This sludge was characterized by
large platelets of calcium sulfite hemihydrate and fly ash. A broad range of
particle sizes is visible for both phases, with fly ash particles as large as
50-(ojn diameter and fine particles of less than 0. l-jj-m diamter. The calcium
sulfite phase existed as stacks of platelets with edge dimensions as large as
20 to 50 fim for the largest particles and platelet thickness of approximately
0.1 to 0. 5 (J.m. The occasional observation of a fly ash particle incorporated
within a sulfite platelet stack suggests that the agglomeration of sulfite plate-
lets occurred during crystalline formation and was not necessarily a conse-
quence of surface tension agglomeration during drying. In addition to these
two phases, large particles of residual limestone and blocky sulfate particles
were occasionally seen but were seldom greater than 5 nm in their longest
dimension. Other small, unidentifiable particles were seen and could have
been magnesium sulfate or calcium chloride phases that crystallized during
the drying of the solids. Both of these phases crystallize with a large quan-
tity of chemically bound water. The drying procedure and the vacuum ex-
posure necessary for SEM viewing may cause a disruption of crystalline
structure, and the material debris not readily recognizable may be a conse-
quence of this decrepitation.
A.2.1.2 TVA Shawnee Lime Sludge
The photomicrographs shown in Figure A-2 are of sludge particles
from the TVA Shawnee venturi spray tower system using lime absorbent.
This sludge was characterized by agglomerates of calcium sulfite rosettes
and fly ash. The range of fly ash particles was nearly identical to that ob-
served in the TVA Shawnee limestone sludge, but fewer of the large size
fractions were observed. The calcium sulfite hemihydrate was found almost
exclusively in interwoven clusters, which were unmistakenly a consequence
of growth phenomena. Individual platelet size was rarely over 10 (im, with
the vast majority being in the range of 5 (j.m. The platelet thickness was typi-
cally 0. 1 to 0. 5 Him. Fly ash particles were often seen within or associated
with a rosette cluster, but this is not interpreted as an indication that fly ash
serves as a nucleating site. Generally, particles of calcium sulfite were ad-
hered to fly ash particles; this is not interpreted as a growth phenomena but
as a consequence of adherence during drying. In contrast to the limestone
system, no calcium carbonate particles were observed, and no clearly dis-
cernible gypsum particles were observed. It is presumed that all calcium
174
-------
Figure A-l. TVA Shawnee limestone sludge.
175
-------
Figure A-2. TVA Shawnee lime sludge,
176
-------
sulfate in the solids fraction existed as a solid solution in the sulfite crystal.
The mechanism that caused the differentiation between the calcium sulfite
crystal habit in the lime and limestone systems is not understood. Typically,
the lime system operated at a higher pH than the limestone system and the
scrubber design is different. Whether either of these variables affected the
growth habit of calcium sulfite hemihydrate is not fully clear.
A.2.1.3 SCE Mohave Limestone Sludge
The characteristics of the solids content of the SCE Mohave
centrifuge cake are shown in Figure A-3. This material was unique among
the sludges sampled in several respects. From chemical and x-ray analysis,
the major portion of the material was identified as gypsum, but the crystal
habit was not typical of gypsum. Instead of lath-like gypsum crystals, the
shape of the crystalline phase was bulky and equiaxial. Typically, particle
sizes range between 20 and 50 fJ-m, with a few being as large as 100 |J.m and as
small as 5 (J.m. No calcium sulfite hemihydrate crystals were observed, but
some particles appeared to be made up of stacked crystal platelets, possibly
sulfite, that had undergone a transformation, probably, by oxidation to sulfate
Other crystals appeared as agglomerates of fine gypsum particles that had
grown together into larger particles. A conspicuous absence of fly ash was
apparent except for a small quantity of fine particles, typically, less than
1 -|j.m in diameter and found adherent to the surfaces of the gypsum particles.
Also, little unreacted limestone was identified although chemical analysis
indicated greater than 6 percent calcium carbonate in the sludge solids.
(Sampling was done during a limestone depletion run; the residual limestone
may have been reduced in particle size during this run such that the lime-
stone present could not be properly identified in the field of view. )
The apparent absence of limestone is understood in relation to the
sampling condition, and the small quantity and size range of fly ash are under
stood relative to the efficiency of the mechanical and electrostatic separators.
The multiple-fold recirculation of the system liquors that occurred because
of the low sulfur content can explain both the extent of oxidation and the pres-
ence of large particles. However, the change in gypsum particle shape,
from lath-like to equiaxed particles, is not understood. The morphological
changes in crystalline growth reported in the literature are most commonly
associated with the chemical poisoning of growth faces. It is presumed that
such poisoning of pyramidal growth faces by the high concentration of dis-
solved salts, primarily sodium chloride, may possibly have promoted relative
growth on the prismatic faces.
A. 2.1.4 DLC Phillips Lime Sludge
The particles in the photomicrographs presented in Figure A-4
are characteristic of the crystalline phases in the thickner underflow of the
DLC Phillips station scrubbing system using lime as the absorbent. The
large intergrown crystals of gypsum are ideal as structural plaster. A
continual size distribution of gypsum particles was evident, with sizes rang-
ing upward to 100 \im along the length of the gypsum lath and 10 to 20 |j.m in
width. A high concentration of fly ash was present, with the largest particle
177
-------
Figure A-3. SCE Mohave limestone sludge,
178
-------
Figure A-4. DLC Phillips lime sludge.
179
-------
diameters ranging upward to 10 to 15 |J.m and downward to the submicron
range. The x-ray and chemical analyses indicated a significant presence of
calcium sulfite hemihydrate, but the morphology of this phase, observed in
the lime and limestone wastes of the TVA Shawnee system, was not apparent
in the DLC Phillips lime waste. There is some solid solubility of the sulfite
phase in gypsum, but the solubility limit would not account for the quantity
determined by the chemical analysis. An unidentifiable phase, fine in struc-
ture, that appeared to be closely associated with fly ash was visible in nearly
every field of view and could have been the sulfite phase in a state of decom-
position. No other phases were present.
A.2. 1.5 APS Cholla Limestone Sludge
The morphology of particles in the APS Cholla waste is repre-
sented in Figure A-5. A high portion of the solids content in this waste is
fly ash. Particles of fly ash represented a broad range of particle sizes,
with diameters of about 15 p.m maximum, and downward to less than 0. 1 fJLm.
Only one phase of sulfur waste was observed in the APS Cholla waste al-
though several phases were expected from chemical and x-ray diffraction
analysis. This phase did not have the lath-like crystal habit of gypsum but
rather the thin platelet habit of the calcium sulfite hemihydrate phase. These
platelets were dissimilar to the rosette structures of most sulfite crystals
and much more like the sulfite crystals of the TVA Shawnee limestone sludge.
However, they were not observed in stacks and their thickness was several
times greater than those found in the TVA Shawnee limestone sludge. The
absence of more recognizable crystalline morphology is not understood even
though gypsum was expected as the major phase. A better understanding of the
crystal growth phenomena is required in this case.
A.2.1.6 GM Parma Double-Alkali Sludge
The filter cake from the GM system is represented by the photo-
micrographs in Figure A-6. The most prevalent phase in this material was
gypsum that appeared in the form of blocky and lathlike particles, with the
largest being about 20 (im along an edge. A distinct decrepitation was appar-
ent in the gypsum particles as a consequence of thermal dehydration during
drying. The x-ray analysis revealed the presence of an anhydrous phase
which, in this case, was undoubtedly a consequence of decrepitation. A small
amount of fly ash was present, with maximum particle size in the range of
5-(o.m diameter. Calcium sulfite hemihydrate represents about 13 percent of
the solid fraction of the filter cake as determined by chemical analysis. This
phase appeared in fine agglomerates of individual particles in a rosette-like
array. Individual sulfite crystallite sizes were typically from 0. 5 to 1.5 \±m
along an edge, amassed in agglomerates that seldom exceeded 10 (J-m in equi-
valent diameters. Some calcium carbonate and sodium sulfate were expected
in this material but were not identified.
180
-------
Figure A-5. APS Cholla limestone sludge,
181
-------
Figure A-6. GM Parma double-alkali sludge,
182
-------
A.2.1.7 UPL Gadsby Double-Alkali Sludge
The filter cake of the UPL Environtech scrubbing system is
shown in Figure A-7. This material was similar to the GM double-alkali
system in only one feature: the small amount of fly ash in the scrubber
solids, with maximum particle size in the range of 5 |JLm. The phase that
represented the major part of the solids was gypsum, which appeared as
small crystallites, typically, of 1 to 5 p.m in major dimension and agglomer-
ates of more massive particles of nominally 25 nm in diameter. In addition,
calcium sulfite hemihydrate particles were observed in greater excess than
anticipated from either chemical analyses or x-ray analysis. These parti-
cles were morphologically like the calcium sulfite phase observed in the TVA
Shawnee lime process as dense clusters of thin platelets. The clustered
particles were nominally 25 (Jim in diameter although the individual platelets
were typically 2 to 3 urn along an edge and thin, with a typical jthickness of
less than 0. 2 (j.m. The x-ray diffraction analysis indicated the presence of a
significant quantity of anhydrous calcium sulfate. No separately identified
particles of anhydrite were detected, but it is presumed that this phase re-
sulted from the transformation of the smallest size fraction of the gypsum
phase. Thus, it was not a phase that could be easily identified, in contrast
to the gypsum phase.
183
-------
Figure A-7. UPL Gad sby double-alkali sludge.
184
-------
APPENDIX B
DESCRIPTION OF SCRUBBER SYSTEMS
Samples were obtained from the flue gas desulfurization (FGD)
scrubbing facilities of seven power plants (Table B-l), representing lime,
limestone, and double-alkali systems. This appendix describes these
facilities and identifies the sample source locations on the schematic
diagrams of the scrubber systems.
TABLE B-l. FGD SYSTEMS SAMPLED AS DATA BASE
Power Plant
TVA Shawnee
Steam Plant
TVA Shawnee
Steam Plant
Arizona Public
Service Company,
Cholla Power
Plant
Duquesne Light
Company,
Phillips Power
Station
General Motors
Corporation,
Chevrolet- Parma
Power Plant
Southern California
Edison, Mohave
Generating Station
Utah Power ;. >
Light Com pan/.
Gadsby Station
Scrubber
System
Venturi and
spray tower.
prototype
Turbulent
contact
absorber,
prototype
Flooded-disk
scrubber,
wetted film
absorber
Single- and
dual- stage
venturi
Bubble-cap
tower
Turbulent
contact
absorber,
pilot plant
Venturi, and
mobile bed,
pilot plant
Scrubbing
Capacity,
MW (equiv)
10
10
120
410
32
< 1
< 1
Coal
Source
Eastern
Eastern
Western
Eastern
Eastern
Western
Western
Absorbent
Lime
Limestone
Limestone,
fly ash
Lime
Soda ash.
lime
Limestone
Soda ash,
lime
U85
-------
B. 1 TVA SHAWNEE POWER STATION
The Tennessee Valley Authority (TVA) Shawnee steam plant provides
150 MW of power from its unit no. 10 coal-fired boiler. Part of the flue gas
from unit no. 10 is diverted around the electrostatic precipitators into each
of two prototype scrubber systems designed to remove both SO2 and fly ash.
Each scrubbing module can handle 30, 000 ACFM of flue gas, equivalent to
10 MW of power generation.
B. 1. 1 Venturi and Spray Tower
The venturi and spray tower scrubbing system incorporates a dual-
stage lime reactant scrubber made by the Chemical Construction Corporation
(Chemico) and is shown schematically in Figure B-l. A detailed schematic
diagram of the scrubber is shown in Figure B-2.
The first-stage scrubber is a vertical venturi with the flue gas
entering at the top. An adjustable plug in the venturi throat controls the
scrubber pressure, while a slurry steam, entering above the plug, initiates
the removal of particulates. The gas passes into the adjacent after-scrubber
spray tower and rises in countercurrent flow against the descending slurry
spray injected from four nozzle headers. Slaked lime in the slurry reacts
with the SC"2 in the flue gas to form a calcium sulfite precipitate in both the
venturi and spray tower. The sulfur waste drains out with the lime-depleted
slurry from the lower portion of the spray tower into a reaction tank. The
scrubbed flue gas passes through chevron demisters at the top of the tower
and through a steam reheater to the stack.
The effluent slurry from the scrubber is collected in a reaction tank,
where a holding time delay ranging from 4 to 60 min may be introduced to
allow for complete neutralization of the sulfurous acid with the lime slurry
feed (pH adjustment) and to permit the growth of calcium sulfite crystals so
as to facilitate their removal in the clarifier.
The reaction tank effluent slurry is recirculated to both stages of
the scrubber. For the purging of sulfur wastes, a bleedoff from the recircula-
tion line is taken to a 20-ft diameter clarifier. The clarifier overflow goes
to a hold tank, where the process makeup water is added; the discharge from
this tank is pumped back to the reaction tank, making up for the water lost to
the flue gas and that occluded with the solid waste. The underflow from the
clarifier is dewatered in a vacuum filter; the clear filtrate is returned to the
clarifier overflow circuit; and the filter cake is reslurried and pumped to a
disposal pond.
Sample sets were taken on three occasions from the EPA-TVA
venturi and spray tower scrubbing system. The operating conditions which
these samples represent and the locations at which the samples were taken
are presented in Table B-2. The sampling locations are shown on the sche-
matic system diagram, Figure B-l. Samples from location S2, the scrubber
effluent, and from location S3, the clarifier effluent, were collected in dupli-
cate. One sample was immediately filtered to separate the liquid from the
solid fraction of the slurry; the other sample was collected with the liquid
186
-------
COAL
£\
BOILER
I /rr\
FLUE!
GAS |
59)
BOTTOM
ASH
VENTURI
SCRUBBER
CIRCLED NUMBERS INDICATE SAMPLING LOCATIONS
Figure B-l. Schematic process diagram of the EPA-TVA Shawnee
steam plant venturi and spray tower scrubber system
showing sampling locations.
187
-------
CHEVRON
DEMISTER
AFTER-SCRUBBER
INLET
SLURRY
THROAT
ADJUSTABLE
PLUG
VENTURI
SCRUBBER
FLUE
GAS
OUTLET
DEMISTER
WASH
INLET SLURRY
FROM SCRUBBER
EFFLUENT
REACTION TANK
Figure B-2. Schematic of EPA-TVA Shawnee
steam plant venturi and spray
tower scrubber.
188
-------
TABLE B-2. TVA SHAWNEE VENTURI AND SPRAY TOWER SCRUBBER
SYSTEM LIME PROCESS OPERATING CONDITIONS AND
SAMPLE LOCATIONS
Operating Condition
Coal, % sulfur
Gas Flow Rate,
3 -1
m sec
ACFM
Equivalent Power, MW
SO2 Input, ppm
SO2 Removal, %
pH Control
Dust Input,
g (tn)~
gr (scf)~
Scrubber Effluent, solids %
Clarifier Underflow, solids %
Filter Cake, solids %
Liquid- to- Gas Ratio,
gpm (1000 cfm)~
jj
Sampling Location
Process Makeup Water
Scrubber Effluent
Clarifier Effluent
Vacuum Filter Filtrate
Vacuum Filter Cake
Lime Slurry
Coal
Fly Ash
Bottom Ash
19 Mar 1974
3.4
11.8
25, 000
8.3
26, 000 ,
93
8.0
8.0
3.5
8
18
45
54
19 Mar 1974
SI
S2
S3
S4
S5
S6
S7
S8
S9
16 May 1974
3.4
11.8
25, 000
8.3
3,200
78
8.0
8.0
3.5
8
18
45
54
16 May 1974
S2
S3
S4
27 Jun 1974
3.4
11.8
25, 000
8.3
2,700
84
8.0
8.0
3.5
8
18
45
54
27 Jun 1974
S2
S3
S4
All samples were taken by TVA personnel.
189
-------
retained with the solid. In every case, both samples were chemically
characterized.
B.I.2 Turbulent Contact Absorber (TCA)
The TCA limestone scrubbing process is operated in parallel with
the venturi lime process unit described in the previous section and is shown
schematically in Figure B-3. This system is also supplied with 30,000
(maximum) ACFM of flue gas from boiler unit no. 10, diverted around the
electrostatic precipitators, and is designed to scrub both SO2 and flue dust.
The TCA scrubber, manufactured by Universal Oil Products (UOP) uses a
fluidized bed of low density spheres to provide liquid gas contact surfaces and
is shown in detail in Figure B-4.
The flue gas enters laterally near the bottom of the scrubber tower
and rises in a countercurrent flow against the descending scrubbing slurry.
Intimate contact between the two phases is obtained in a three-deck central
section, partially packed with mobile plastic spheres levitated by the gas
flow between lower and upper retaining grids. Crushed limestone in the
slurry reacts with SC>2 in the flue gas to produce a precipitate of calcium
sulfite and releases CO2 to the flue gas. In a secondary reaction, the excess
O2 in the flue gas combines with sulfite ion to produce sulfate ion, which
forms a precipitate with calcium ion.
In the upper portion of the tower, two sets of nozzles are installed:
One is directed downward and sprays the slurry over the rising gas; the other,
a stream sparge, is directed upward against the underside of a Koch tray and
keeps it from clogging. The Koch tray breaks up the flow of gas as it impinges
on a demister placed in the uppermost part of the scrubber. The wet flue gas
passes out at the top of the scrubber and through a steam reheater into the
stack.
The slurry, after reacting with the SO2, drains from the bottom of
the scrubber into a reaction tank, where it may be held to provide a recircu-
lation residence time of 5 to 20 min, allowing the sulfite precipitation to pro-
ceed further toward completion. From the reaction tank, the recirculation
loop carries the slurry back to the scrubber spray nozzles. In addition, fresh
limestone slurry is added to the reaction tank to replace that depleted by the
SOo reaction.
A bleedoff from the reaction tank discharge removes part of the
slurry to the clarifier tank where a separation of liquids and solids takes
place. The clear overflow is returned to the reaction tank via a hold tank
where process makeup water is added to replenish stack and pond loss. The
underflow from the clarifier is pumped as a sludge to a disposal pond.
Sample sets were taken on four occasions from the TCA scrubbing
system. Table B-3 presents the operating conditions for each sampling
occasion and the sample locations shown on the schematic process diagram
in Figure B-3. The scrubber effluent sample T2 and the clarifier effluent
sample T3 were obtained in duplicate. One sample was filtered immediately
at the time of sampling; the other sample was collected with the liquid retained
with the solid. Both samples, in every case, were chemically characterized.
190
-------
CIRCLED NUMBERS INDICATE SAMPLING LOCATIONS
Figure B-3. Schematic process diagram of the EPA-TVA
Shawnee steam plant TCA system showing
sampling locations.
191
-------
FLUE
GAS
OUTLET
CHEVRON DEMISTERS
INLET KOCH TRAY
WASH LIQUOR
KOCH TRAY
STEAM
SPARGE
RETAINING GRIDS.
FLUE ..
GAS HS
IKII FT ^V^
INLET
EFFLUENT KOCH
TRAY WASH LIQUOR
~gj
INLET SLURRY
FROM REACTION
TANK
-MOBILE PLASTIC
SPHERES
EFFLUENT
SLURRY
-H SEE FIGURE B-3 I
I I
Figure B-4. Schematic of EPA-TVA Shawnee
steam plant TCA scrubber.
192
-------
TABLE B-3. TVA SHAWNEE TCA SCRUBBER SYSTEM LIMESTONE
PROCESS OPERATING CONDITIONS AND SAMPLE
LOCATIONS
Operating Condition
Coal, % sulfur
Gas Flow Rate,
3 -1
m sec
ACFM
Equivalent Power, MW
SO 2 Input, ppm
SO2 Removal, %
pH Control
Dust Input,
g (m)'3
gr (scf)'1
Scrubber Effluent
solids %
Liquid-to-Gas Ratio,
gpm (1000 cfm)~
Sampling Location
Process Makeup Water
Scrubber Effluent
Clarifier Effluent
Limestone
Coal
Fly Ash to Scrubber
Fly Ash from Scrubber
Bottom Ash
1 Feb
1973
9.7
20, 500
6.8
2,400
83
5.9
5
7.0
43.6
1 Feb
1973
Tl
T2
T3
T4
T5
T6
T7
T8
12 Jul
1973
9.7
20, 000
6.7
2,400
74
5.7
8
3.5
8.5
40.0
12 Jul
1973
Tl
T2
T3
T4
T5
T6
T8
27 Nov
1973
3.4
9.7
20, 500
6.8
2,700
83
5.9
8
3.5
16.0
54.6
27 Nov
1973
Tl
T2
T3
T4
T5
T6
T8
15 Jun
1974
3.4
9.7
20, 500
6.8
2,400
86
5.5
8
3.5
8.0
54.6
15 Jun
1974
Tl
T2
T3
T5
aAll samples were taken by TVA personnel.
193
-------
B. 2 SOUTHERN CALIFORNIA EDISON (SCE) COMPANY
MOHAVE STATION TCA SCRUBBER SYSTEM
The SCE Mohave station limestone scrubbing system at the time of
sampling was operating on a pilot-plant scale prior to the construction of a
full-scale FGD scrubbing installation. The investigation described here was
limited to the pilot operation of a TCA scrubber manufactured by UOP, and it
is not known whether it is applicable to the full-scale TCA unit subsequently
installed at Mohave. The pilot plant scrubbing system is shown schematically
in Figure B-5.
The scrubber is a three-stage TCA tower, 27 ft high, 19 in. in
diameter, and rated at approximately 0. 5-MW equivalent power generation.
The primary reaction between the recirculating limestone slurry and SC>2 in
the flue gas yields a relatively insoluble calcium sulfite precipitate, which
undergoes subsequent oxidation to the sulfate phase. Slurry from the scrubber
is retained in the recirculation loop between the scrubber and hold tank for an
appreciable time because the low-sulfur coal burned and the low particulate
concentration in the flue gas do not produce a large quantity of solids relative
to the quantity of flue gas scrubbed. Thus, only a small bleed stream from
the recirculation loop is necessary to maintain a constant solids level in the
recirculating slurry.
This bleed is pumped to a centrifuge that separates liquid from a
solid cake. The liquor is returned to the hold tank, at which point process
makeup water is added to compensate for water lost to the stack and centrifuge
cake. The hold tank serves not only as a surge tank but as a reaction tank to
which fresh limestone reactant is added. The centrifuge cake is discharged
to a disposal site.
A single set of samples was taken at the SCE Mohave station for
analysis. Operating conditions at the time of sampling were defined as
limestone depletion tests, and no attempt was made to maintain a controlled
slurry concentration level. The sample locations are identified in Table B-4
and are shown on the schematic system diagram in Figure B-5. All samples
were taken by Bechtol personnel.
TABLE B-4. SCE MOHAVE TCA SCRUBBER SYSTEM
SAMPLE LOCATIONS
Sampling Location 30 Mar 1973
Process Makeup Water Ml
Scrubber Effluent M2
Scrubber Recirculation M3
Centrifuge Centrate M4
Centrifuge Cake M5
Limestone M6
194
-------
PROCESS MAKEUP
WATER
CIRCLED NUMBERS INDICATE SAMPLING LOCATIONS
Figure B-5. Schematic process diagram of the
SCE Mohave station TCA pilot
plant scrubber system showing
sampling locations.
195
-------
B. 3 DUQUESNE LIGHT COMPANY (DLC) PHILLIPS
STATION SINGLE- AND DUAL-STAGE VENTURI
SCRUBBER SYSTEM
The DLC Phillips power station scrubber system was designed to
scrub the total flue gas produced by the full plant capacity of 387-MW gen-
erating power with three of four parallel venturi scrubber modules such that
one module is available as reserve. Three modules are single-stage venturi
scrubbers designed primarily for particulate removal; the fourth module con-
sists of a dual-stage venturi scrubber designed to remove fly ash in the first
stage and SC>2 in the second stage. A schematic of the process diagram of
the single-stage system is presented in Figure B-6, and the dual-stage system
is presented in Figure B-7. Details of the venturi scrubber manufactured
by Chemico and used in both systems are shown in Figure B-8.
The flue gas from the boilers pass through mechanical and electro-
static dust collectors which remove the bulk of the fly ash but allow 10 to 30
percent of the fly ash to enter the first-stage scrubber (or single scrubber).
Flue gas enters the venturi scrubber at the top, passes through a venturi
throat, and emerges laterally at the bottom. Half of the recirculating
scrubber liquor is introduced above the cone, forming the center of the throat;
the other half is sprayed in through tangential nozzles. In the dual-stage sys-
tem, the flue gas is pumped to a second venturi operating in an identidal
manner and then emerges through a reheater to the exhaust stack. In the
single-stage system, the flue gas is passed through mist eliminators and the
reheater before entering the exhaust stack.
In the first-stage scrubber (or single scrubber), intimate contact
between liquor and flue gas occurs in the throat area, where fly ash and about
60 percent of the SC>2 are removed. The second-stage scrubber removes an
additional 30 percent of the SO? in the flue gas. The lower part of the scrubber
separates liquor from the flue gas and serves as a reaction vessel where
sulfurous acid commences to react with lime to precipitate a calcium sulfite
waste. Half of the slaked lime slurry used as the absorbent is introduced
into the system at this point; the other half is introduced into the clarifier,
which receives the bleed from the recirculating scrubber slurry from all
scrubber modules. Overflow from the clarifier is pumped into the recircula-
tion system; the underflow from the clarifier is pumped to an interim disposal
pond. A supernate from the pond is returned to the clarifier. Process make-
up water enters the system in the lime slurry, pump seal water, and fan and
demister sprays.
Sample sets were taken on two occasions from the DLC Phillips
station scrubber system. The first sample set was taken at a time when only
the single-stage system was operating; the second sample set was taken at a
time when the dual-stage system was in operation. The operating conditions
for both sampling occasions and the description of the sample locations are
presented in Table B-5. The sample locations are also shown in Figures B-6
and B-7. All samples were taken by DLC personnel.
196
-------
1
HOT FLUE GASES
FROM DUST
COLLECTORS
BOILER
(vs)
COAL
SCRUBBER
LIQUOR
RECYCLE
STREAM
DEMISTER,
SPRAY /"
MAKEUP
PUMP
1
SINGLE-STAGE
SCRUBBER
[See Figure B-8
for details)
RECYCLE
PUMP
SPRAY
WATER
[
WET GAS
₯
ELIMINATOR
REHEATER
DRAINS
I. D. FAN
STACK
SUMP
SUPERNATE
INTERIM
DISPOSAL
POND
Figure B- 6.
Schematic process diagram of the DLC Phillips
station single-stage scrubber system showing
sampling locations.
197
-------
COAL
(DS)
BOILER
HOT FLUE GASES
FROM DUST
COLLECTORS
COLD FLUE GAS
SCRUBBER
LIQUOR
RECYCLE
STREAM
DEMISTERL.
SPRAY ~
00
MAKE-UP
PUMP
FIRST STAGE
SCRUBBER
(See Figure 4-12
for details)
RECYCLE
PUMP
BLEED
OVERFLOW
SPRAY
WATER
FAN
SCRUBBER
LIQUOR
RECYCLE
STREAM
OVERFLOW
FROM
CLARIFIER
SECOND STAGE
BLEED
SECOND
STAGE
SCRUBBER
DAMPER
PURGE AIR
BLOWER
WET GAS
REHEATER
STACK
LIME SLURRY
FEED
UNDERFLOW U(D4
CALCILOX
MIXING TANK
SUPERNATE
VIBRATOR
WEIGH FEEDER
SLAKER
TRANSFER
TANK
LIME SLURRY
PUMP
INTERIM DISPOSAL PONDS
CIRCLED NUMBERS INDICATE SAMPLING LOCATIONS
Figure B-7.
Schematic process diagram of the DLC Phillips station dual-stage
scrubber system showing sampling locations.
-------
HOT FLUE
GAS INLET
SCRUBBER
LIQUOR
RECYCLE
STREAM
TANGENTIAL
y- NOZZLES
/. -,
ADJUSTABLE THROAT
DAMPERS
FLUE GAS
EXIT
DEMISTER
SPRAY
LOWER CONE
TO RECYCLE
PUMP
Figure B-8. Schematic of DLC Phillips power
station venturi scrubber.
199
-------
TABLE B-5.
DLC PHILLIPS SCRUBBER SYSTEM OPERATING
CONDITIONS AND SAMPLE LOCATIONS
Operating Condition
Coal, % sulfur
Gas Flow Rate,
3 -1
m sec
ACFM
Equivalent Power, MW
SO2 Input, ppm
SO2 Removal, %
pH Control
Dust Input,
g (m)'3
gr (scf)~
Scrubber 'Effluent, solids %
Liquid-to-Gas Ratio,
gpm (1000 cfm)~
Sampling Location
Process Makeup Water
Scrubber Effluent
Clarifier Overflow
Clarifier Effluent
Coal
Lime
Interior Pond Sludge
Interior Pond Sludge, with Calcilox
4 Oct 1973
2.0
274
580,000
145
1,400
60
6.0
0.57
0.25
5
29.9
4 Oct 1973
VI
V3
V4
V5
V6
17 Jun 1974
2.0
236
500,000
120
1,400
70
6.0
0.57
0.25
5
29.9
17 Jun 1974
Dl
D2
D3
D4
D5
D7
D8
200
-------
B.4 ARIZONA PUBLIC SERVICE (APS) COMPANY
CHOLLA STATION FLOODED DISK SCRUBBER
(FDS) AND PACKED ABSORPTION TOWER
SCRUBBER SYSTEM
The APS Cholla scrubber system utilizes limestone scrubbing to
remove SO2 and particulates from flue gas produced by a 120-MW boiler.
Two scrubbers are installed in parallel, each handling 60- MW equivalent
power and sharing a common limestone feed and scrubbing liquor circulation.
One scrubber is filled with a wet film packing for high SO2 absorption effi-
ciency; the other is unpacked and operates as a spray tower. The packed
scrubber unit is illustrated in Figure B-9; details of this scrubber are shown
in Figure B- 10.
Prior to entering the scrubbers, the flue gas has about 80 percent of
its dust removed by a mechanical collector system. A forced-draft booster
fan then forces the flue gas into the first scrubbing stage, a venturi-type FDS
designed to remove the remaining fly ash by injecting the FDS slurry tangen-
tially above the venturi throat and at the variable setting dish. The FDS
slurry passes into the absorber tower, where it is separated from the gas in
a cyclonic demister section, and drains out the bottom to the FDS slurry tank.
The flue gas, containing 70 to 85 percent of the initial SO2> moves
upward through the absorption section of the tower, where it comes in contact
with a wet- film packing and is scrubbed by the counter flowing limestone slurry.
This slurry drains out laterally and is recirculated independently of the FDS
slurry circuit. By this means, the initial scrubbing action in the FDS elim-
inates particulates, while the absorber slurry, high in absorbent content,
effectively removes SO2- The scrubbed gas passes through a 2 percent stage
demister, leaves the absorber tower, and after passage through a steam
reheater enters the stack.
The FDS slurry from the foot of each scrubber tower is pumped to
an agitated FDS slurry tank, from which it is recirculated to the FDS.
Similarly, the absorber slurries are collected in an absorber tower feed tank,
where the limestone slurry required to replenish the reactant is added. Pro-
cess makeup water is added to both the FDS slurry tank and the absorber
tower feed tank, while a crossover line provides a bleedoff from the absorber
tower slurry to the FDS slurry and, at the same time, provides the pH con-
trol for the FDS circuit by introducing limestone from the absorber tower feed.
A bleedoff line from the FDS slurry tank carries the sulfur waste
product, a calcium sulfite-sulfate sludge, and fly ash to the disposal pond.
This bleed is also used to slough mechanically collected fly ash to the dis-
posal pond.
Sample sets were collected from the APS Cholla station on two
occasions. The operating conditions and sampling locations representing
each sampling occasion are presented in Table B-6. Sample locations are
also indicated in Figure B-9. Samples C2 and C3 were collected in duplicate
on the first occasion. One sample was filtered immediately upon collection;
the other sample retained liquor and solids in contact until they were pre-
pared for chemical analysis.
201
-------
FLUE GAS FROM
FDSAND/UNPACKED
ABSORPTION TOWER
COAL
irA\
BOILER
REHEATER
STACK
BOOSTER
FAN
TO ASH
DISPOSAL
POND
PACKED
ABSORPTION
TOWER AND
CYCLONIC
DEMISTER
(See Figure
B-10
for details)
DEMISTER
'SPRAY WASH
FLOODED
DISK
SCRUBBER
(FDS)
LIMESTONE
\
"©
p
LIMESTONE
SLURRY
ABSORPTION
TOWER FEED
TANK
CIRCLED NUMBERS INDICATE SAMPLING LOCATIONS
Figure B-9.
Schematic process diagram of the APS Cholla station
FDS and packed absorption tower showing sampling
locations.
202
-------
SECONDARY MIST
ELIMINATOR
PRIMARY MIST
ELIMINATOR
FLUE GAS
INLET
*
FLOODED
DISK
SCRUBBER
IFDSI
EXIT CAS
TO STACK
ABSORBER
PACKING
PACKED
ABSORPTION
TOWER
t
CYCLONIC
DEMISTER
FROM
i ABSORPTION
TOWER FEED
TANK
<
TO ASH
DISPOSAL
POND
FDS
SLURRY TANK
TO ABSORPTION
TOWER FEED
TANK
Figure B-10.
Schematic of APS Cholla station FDS and
packed absorption tower scrubber.
203
-------
TABLE B-6. APS CHOLLA SCRUBBER SYSTEM OPERATING
CONDITIONS AND SAMPLE LOCATIONS
Operating Condition
Coal, % sulfur
Gas Flow Rate,
. 3 -1
in. sec
ACFM
Equivalent Power, MW
SO Input, ppm
SO2 Removal, %
pH Control, Absorber Tower Circuit
Dust Input,
g (m)~
gr (scf)'1
FDS Slurry Tank Effluent, solids %
Absorber Tower Effluent, solids %
Liquid-to-Gas Ratio,
gpm (1000 cfm)"1
Sample Location
Process Makeup Water
FDS Slurry Tank
Absorption Tower Effluent
Coal
Fly Ash
Limestone
1 Apr 1974
0.5
151
320, 000
115
360
75
6.5
0.41
0. 18
15
15
70.9
1 Apr 1974
Cl
C2
C3
C4
C5
C6
7 Nov 1974
0.5
151
320, 000
115
360
75
6.5
0.41
0. 18
15
15
70.9
7 Nov 1974
Cl
C2
C3
204
-------
B. 5 GENERAL MOTORS (GM) CORPORATION PARMA
POWER STATION DOUBLE-ALKALI SCRUBBER
SYSTEM
The GM double-alkali system is unique among the scrubber systems
investigated in this report in that it removes SC>2 by a regenerable absorbent.
The installation consists of four parallel Koch valve tray scrubber modules
which remove SC>2 from the flue gas on the 32-MW industrial boilers at the
GM Parma plant.- Figure B-ll shows schematically one scrubber module in
the scrubbing system.
Each scrubber consists of a vertical tower fitted with a series of
three bubble-cap trays. Flue gas, cleaned of particulates by a mechanical
fly ash remover, enters the bottom of the scrubber and passes upward through
the tray orifices in a flow that is countercurrent to the descending sodium
hydroxide scrubbing liquor. This liquor passes out through a bottom drain
and is recirculated to the scrubber via a recycle tank.
For the recirculating liquor to be purged of its sulfite content, a
bleedoff from the recycle tank discharge carries a portion of the liquor to
mix tank no. 1 (neutralizer tank), where the weak acid liquor is blended with
the alkaline clarifier no. 2 underflow. The liquor passes to mix tank no. 2
(recausticizer), where it is treated with a lime slurry, regenerating sodium
hydroxide and producing a calcium sulfite precipitate. This is also the point
in the system where a portion of the process makeup water is added, as a
replacement for the water lost in the filter cake and hot stack gas. The under-
flow becomes the feed stream to clarifier no. 1, where the solid and liquor
portions are separated. The underflow from clarifier no. 1, a thickened
slurry high in calcium sulfite, goes to the vacuum filter for dewatering. The
overflow contains calcium hydroxide in solution and is passed to clarifier no. 2
where the calcium ions react with soda ash (sodium carbonate) precipitating
calcium carbonate and preventing the subsequent deposition of a calcium
sulfate scale in the scrubber. The high carbonate concentration in this clari-
fier depresses the calcium content of the liquor, and fresh soda ash replaces
the sodium lost in the filter cake. The clarifier no. 2 overflow is recirculated
to the scrubber; the underflow, a sodium-calcium carbonate slurry, goes to
mix tank no. 1 as an alkalizer to commence the liquor regeneration process.
The vacuum filter receives the slurry underflow from clarifier no. 1
and produces a clear filtrate, which is returned to the clarifiers. The filter
cake is washed and refiltered before being trucked to a landfill area for
disposal.
A single sample set was collected by GM personnel and delivered to
The Aerospace Corporation for analysis. The scrubber system operating
conditions and description of sample locations shown in Figure B-ll are pre-
sented in Table B-7.
,20~5
-------
FLUE GAS
OUTLET
COAL
J©
BOILER
i
t
FLUE GAJ
INLET
t
. le\w\
^L7
\
L
1
SCRUB-
BER
SOLIDS TO
LANDFILL
CIRCLED NUMBERS INDICATE SAMPLING LOCATIONS
Figure B-ll. Schematic process diagram of the GM
scrubber - ; ' ^m showing sampling
locations.
206
-------
TABLE B-7. GM DOUBLE-ALKALI SCRUBBER SYSTEM OPERATING
CONDITIONS AND SAMPLE LOCATIONS
Operating Condition 18 Jul 1974
Coal, % sulfur 2.5
Gas Flow Rate Scrubber,
m3 sec"1 13.3
ACFM 28,000
Equivalent Power, MW 32
SO2 Input, ppm 1,300
SO2 Removal, % 90
pH Control,
Scrubber Recycle Liquor 6.0
Clarifier Return to Scrubber 12
Dust Input,
g(m)"3 0.7
gr(scf)"1 0.3
Filter Cake, solids % 50
Liquid-to-Gas Ratio, gpm (1000 cfm)~ 20
Sampling Location 18 Jul 1974
Process Makeup Water Gl
Mix Tank No. 2 Underflow G2
Clarifier No. 1 Overflow G3
Clarifier No. 1 Underflow G4
Clarifier No. 2 Overflow G5
Clarifier No. 2 Underflow G6
Filter Filtrate G7
Filter Cake G8
Coal G9
Fly Ash G10
Lime Gil
207
-------
B. 6 UTAH POWER AND LIGHT (UPL) COMPANY
GADSBY STATION ENVIROTECH DOUBLE-
ALKALI SCRUBBING SYSTEM
The Envirotech scrubber system is a pilot plant installed in a
100-MW coal-fired boiler that scrubs flue gas from an equivalent of approxi-
mately 1-MW generating capacity. The system is a regenerating system
using soda ash and lime as absorbents and is shown in Figure B-12. The
scrubbing of the flue gas is done in two stages: the first stage is a venturi-
cyclone separator scrubber; it is followed by a two-tray mobile bed scrubber
as the second stage.
Flue gas enters the scrubbers after passing through mechanical and
electrostatic precipitators and passes down through the venturi counter-
current to the scrubber liquor as it passes upward through the mobile bed
scrubber. The flue gas then passes through a mist eliminator and is exhausted.
Scrubber liquor drains from the bottom of the scrubber into an
oxidation tank, which serves also as a hold tank for the scrubber recircula-
tion system. The bleed from the oxidation tank is pumped to a reaction tank
where the alkaline underflow from reactor clarifier no. 2 is added for initial
neutralization. The liquor passes then to a second reaction tank where slaked
lime is added for further pH adjustment. The liquor is then passed to reactor
clarifier no. 1, where the initial separation of solids and liquid takes place.
The overflow passes to reactor clarifier no. 2, where soda ash is added as a
softening step. The underflow from reactor clarifier no. 2 goes to the first
reaction tank; the overflow returns to the oxidation tank. The underflow from
reactor clarifier no. 1 is pumped to a thickener for further solids and liquid
separation. The underflow from the thickener is passed to a vacuum filter;
the filtrate is returned to the system; and the filter cake is discharged as a
waste.
A single sample set was taken by Envirotech personnel for this study.
The scrubber operating conditions and description of sample locations are
shown in Table B-8 and Figure B-12.
208
-------
CAKE WASH
O
vD
I I I (FRESH WATER)
r *~\ SULFATE CAKE
(70-80% suspen solids)
ROTARY
VACUUM FILTER
(belt type)
RECIRC. WASH WATER
FILTER
VACUUM
PUMP
FINAL MIST
ELIMINATOR
FILTRATE
PUMP
SLAKED LIME
MOBILE
BED
SCRUBBER
VENTURI
QUENCHER
REACTOR-
CLARIFIER No. 1
REACTION OCA(-TinK1
TANK REACTION
OXIDATION
TANK
REACTOR
CLARIFIER No.2
REGENERATED
FRESH WATER
OXIDATION
BLOWER
LIQUOR TANK
40% SUSPEN.
SOLIDS
PUMP
PUMP
PUMP
Figure B-12. Schematic process diagram of the UPL Envirotech sulfate-mode
double-alkali scrubber system.
-------
TABLE B-8. UPL ENVIROTECH DOUBLE-ALKALI SCRUBBER SYSTEM
OPERATING CONDITIONS AND SAMPLE LOCATIONS
Operating Condition 9 Aug 1974
Coal, % sulfur 0.4
Gas Flow Rate,
3-1 ,o
m sec 1.2
ACFM 2500
Equivalent Power, MW 0.8
SO2 Input, ppm 800
SO2 Removal, % 90
pH Control 6. 5
Dust Input,
g(m)"3 0.6
gr (scf)"1 0.25
Filter Cake, solids % 75
Liquid-to-Gas Ratio,
gpm (1000 cpm)~ 20
Oxidation Rate, % 90
210
-------
APPENDIX C
CHEMICAL ANALYSIS TECHNIQUES USED BY AEROSPACE
C. 1 INTRODUCTION
This appendix describes the analytical techniques used to determine
the concentration of constituents in the liquid portions of flue gas desulfuriza-
tion (FGD) sludges.* The constituents present in the liquor are divided into
the following: major chemical species (calcium, magnesium, sodium, sul-
fite, sulfate, and chloride), trace metal species, and additional chemical
species. Other water quality tests are also described. Solids analyses dif-
fered from liquor analyses primarily by sample preparation methods.
The precision and accuracy chosen for the measurements were
strongly influenced by the relative cost of the analysis and the relative sig-
nificance of the resultant value. The data obtained,have as high a precisiqn
and accuracy as was considered warranted.* Although the basis for selecting
the proper analytical technique was to minimize any interference from other
species, the presence of chemical species interfering with a particular anal-
ysis was fully acknowledged. Only when the interference was considered sig-
nificant were corrections applied.
C.2 MAJOR CHEMICAL SPECIES
C.2.1 Calcium Determination
The method selected for calcium determination has an accuracy of
4 percent. Calcium oxalate was precipitated and filtered from the solution;
the filter cake was redissolved in HC1; and the solution was titrated against
KMnO4 to the characteristic purple end point. Correction was then made for
excess permanganate.
An alternative technique of atomic absorption spectrophotometry
was used for calcium analyses. Results for solutions analyzed by both
methods demonstrated comparable agreement, to within 6 percent. At con-
centrations greater than 1000 mg/S., the precision of each method was 3 to
*
The methods defined in (20) and (21) were used wherever they were applicable
to the requirements of the analyses.
Precision is defined as the relationship between a measured value and the
statistical mean of measured values, and accuracy is the relationship
between the true value and the measured value.
211
-------
5 percent, but at lower calcium concentrations the data spread in the oxalate
method is greater than that of the spectrophotometric method. Atomic absorp-
tion will be used for all future analyses.
C.2.2 Magnesium Determination
Magnesium was determined by atomic absorption spectrophotometry
to an accuracy and precision ranging between 3 and 5 percent.
C.2.3 Sodium Determination
Atomic absorption spectrophotometry or flame emission photometry
was used to determine sodium ion concentrations, depending on whether the
concentrations were relatively low or high. Total errors were generally
less than 10 percent and typically less than 5 percent.
C.2.4 Sulfite Determination
Total sulfite was determined by use of a specific ion electrode, and
no significant interferences were observed. The oxidation of the sulfite ion
to sulfate in the scrubber liquor was found to be a rapid reaction. Liquor
protected from the atmosphere typically reveals concentrations of several
hundred milligrams per liter of the sulfite ion; however, a brief atmospheric
exposure causes oxidation and reduces these concentrations by one or more
orders of magnitude. The reported sulfite measurements were for samples
analyzed immediately upon arrival in the laboratory. No specific action was
taken to inhibit oxidation other than to ensure that the samples were trans-
ported from the power plant scrubber to the analytical laboratory in sealed
containers. The exposure to air during sampling, filtering, and measuring,
however, resulted in the sulfite values reported. It is presumed that these
concentrations would probably more closely represent the oxidation state of
liquors in the event of their potential discharge.
No value of accuracy or precision can be reported because of the
relative instability of the sulfite ion. At concentrations greater than approxi-
mately 100 mg/S., the iodine titration method is preferred. At concentrations
below this value, the titration method is not applicable, and only the specific
ion electrode can be used. Accuracy of 3 percent and precision of less than
5 percent were determined against sodium sulfite at approximately 50 mg/j? .
For the lower concentrations, the titrimetric technique cannot be validated
against the specific ion method.
C.2.5 Sulfate Determination
Standard nephelometry techniques were used for this analysis. A
barium sulfate precipitate was formed by the reaction of the sulfate ion with
a barium chloranilate reagent. The resulting turbidity was determined by a
spectrophotometer and compared to a curve from standard sulfate solutions.
Although multiple dilutions are necessary to bring the concentration to a
range of optimum reliability, the resulting error is less than 10 percent.
212
-------
C.2.6 Chloride Determination
A specific ion electrode was used to determine the concentration of
chloride ions. Comparisons were made'with results of titrations with silver
nitrate. Two titrimetric methods of chloride analysis were conducted, and
the results have been compared with the results of ion selective electrode
measurements of chloride concentration. Titrations with silver nitrate were
made, with chromate ion as the end-point detection (Mohr method), and these
results were compared with potentiometric titrations. Both titration methods
gave precisions within 1 percent, and the results of both methods agreed to
approximately 1 percent. The precision of the ion selective electrode mea-
surements was approximately ±5 percent, and the results of the titration
methods of the same solutions with AgNO3 were well within the uncertainty
band of the electrode measurements. The chloride concentrations of these
samples were between 2000 and 4000 mg/j? . At much lower chloride concen-
trations, the ion selective electrode method has an advantage because the
precision of the electrode measurement is not appreciably affected by the
concentration level, in contrast to either of the titrimetric methods for which
the precision decreases at lower concentration values.
C.3 TRACE ELEMENT SPECIES
Since most trace element species are highly sensitive to atomic absorp.
tion spectrophotometry, the flameless technique was used for the following
elements: aluminum, beryllium, cadmium, chromium, copper, lead, and
zinc. Results were verified by the National Bureau of Standards (NBS) anal-
ysis standards and by comparative analyses of elements present in relatively
high concentrations by means of gravimetric or volumetric methods. Preci-
sion and accuracy are dependent upon the means of activation, the specific
element, its relative concentration, and the extent of interference by other
elements and matrix effects. The precision and accuracy of the measure-
ments at concentrations of all of these elements between 1.0 and 0.010 mg/j?
ranged between 5 and 50 percent. However, with furnace activation, the
precision of trace elements occurring at even lower concentration levels de-
creases as the detection limit is approached.
Mercury was also determined by this technique; however, the mer-
cury was reduced to the elemental state with stannous chloride, and the ab-
sorption of the resulting mercury vapor was measured. This method has a
precision of about 20 percent and an accuracy of about 20 percent in the
range 0.005 to 0.0005 mg/jC.
Arsenic was determined by reacting arsine with mercurous bromide
to produce Hg3As; the unknown was compared colorimetrically against stan-
dards. For this application, this technique has a precision of about 25 per-
cent and an accuracy of about 25 percent in the range 0. 10 to 0. 01 mg/j? .
A fluorimetric technique that has a sensitivity within 1 mg/jj was
used to determine selenium. It has a precision.of about 10 percent and an
accuracy of 60 percent between 0. 20 to 0. 02 mg/j? .
213
-------
Plans to change the arsenic and selenium analyses to atomic
absorption spectrophotometry are being implemented.
C.4 ADDITIONAL CHEMICAL SPECIES
C.4. 1 Potassium Determination
Potassium was determined by the same techniques as sodium, and
the precision and accuracy obtained were generally between 3 to 5 percent.
C. 4. 2 FLUORIDE DETERMINATION
The fluoride ion was determined by the specific ion electrode using
a Beckman Model 4500 digital pH meter. There were no significant inter-
ferences in the scrubber liquors. This method has a precision of about
5 percent; an accuracy of 20 percent is attainable at the low levels measured.
For future analyses, a Bellack distillation will be performed prior to specific
electrode measurements.
C.5 OTHER WATER QUALITY TESTS
C.5.1 pH Determination
This parameter was measured with a Beckman Model 4500 digital
pH meter to a precision of 0. 002 pH units and an accuracy of 0. 005 pH units.
C.5.2 Total Dissolved Solids (TDS) Determination
The TDS were determined gravimetrically by evaporating a 25-rru?
sample overnight in a tared weighing bottle, under vacuum at 120°F. Since
two of the major constituents (calcium and sodium sulfates) form stable
hydrated salts and are hygroscopic in the anhydrous state, prolonged drying
and minimal exposure of the dried residue were mandatory. This method
has a precision of about 2 percent, and an accuracy of about 5 percent.
C.6 ANALYTICAL METHODS APPLICABLE TO SLUDGE SOLIDS
Sludge solids were analyzed for calcium, sulfate, sulfite, and car-
bonate, as well as total solids and inert material (fly ash).
Calcium was determined by a volumetric method, following an oxa-
late separation. The sample, commonly 0.25 g, was dissolved in hydro-
chloric and nitric acids, diluted, and filtered, and calcium oxalate was pre-
cipitated by ammonium oxalate from a slightly alkaline solution. The pre-
cipitate was filtered off, redissolved in sulfuric acid, and titrated with
standard potassium permanganate.
Sulfate was determined gravimetrically, taking a 0.25-g sample
which was dissolved in hydrochloric acid. The solution was filtered, and
barium chloride added to the hot filtrate to precipitate barium sulfate. This
was filtered off through a tared Gooch crucible with a glass filter pad. It
was then dried and ignited at 800°C, cooled, and weighed.
214
-------
Sulfite was determined volumetric ally. A 2-g sample was placed
in a three-necked flask, fitted with a dropping funnel for adding sulfuric acid,
an entrance tube for nitrogen used to sweep out the evolved gases, and an exit
tube dipping into an absorbent solution of I/10-nitrogen-sodium hydroxide.
After SO2 was evolved and collected as sodium sulfite, the excess sodium
hydroxide was neutralized and the sulfite titrated with standard iodine.
Carbonate was determined by a gravimetric method after evolution
as CO2, along with SO2, by acidifying a 2-g sample in a tared flask. The
flask was warmed gently to expel all gases, cooled, and weighed. The weight
decrease represents CO£ + SC>2 and must be corrected for the SC^ content as
determined volumetrically.
215
-------
APPENDIX D
CHEMICAL CHARACTERIZATION DATA SHEETS
The following tables present the results of the chemical analyses
performed on the liquid and solid portions of flue gas desulfurization (FGD)
sludge samples taken at various locations within the waste streams of FGD
scrubbing systems. The concentration values, as presented, represent
either the mean or median of a minimum of three and a possible maximum
of nine independent measurements. Three independent commercial labora-
tories were used to supplement the in-house analyses. Data acceptance was
made only after the results of two (and sometimes three) independent anal-
yses were evaluated. (Each analysis is the mean of three separate
measurements.)
A description of the scrubbing facilities is given in Appendix B, and
a description of the analytical methods is given in Appendix C. An evalua-
tion of these data is presented in Section 7. 3. 1, and detailed discussions of
the relationship of these data to system parameters are presented in
Section 7. 4.
Tables D-l through D-13 present results from chemical analyses
of the liquid samples from various locations within the scrubbing systems,
as defined in Appendix B, Tables B-l through B-6. Tables D-14 through
D-17 are summation tables for four of the scrubbing systems, giving more
extensive analytical results. Table D-18 through D-27 are analyses of solids
from these scrubber systems, also defined in Tables B-l through B-6.
217
-------
TABLE D-l. ANALYSIS OF SCRUBBER LIQUORS FROM TVA
SHAWNEE STEAM PLANT
Date:
Cone:
1 Feb 1973
mg/4
TCA Scrubber-Limestone System
Scrubber
Liquor
Constituents
Aluminum
Arsenic
Beryllium
Cadmium
Calcium
Chromium
Copper
Lead
Magnesium
Mercury
Potassium
Selenium
Sodium
Zinc
Chloride
Fluoride
Sulfate
Sulfite
TDS
PH
Sampling Locations
Process
Makeup
Water
<0.2
< 0.0005
< 0.0005
< 0.001
0.012
<0. 005
0. 007
-
<0.01
-
< 0.2
_
-
-
-
-
7.4
Scrubber
Effluent
(Separated)
1.7
1.2
0.006
0.009
1800
0.025
0.041
0. 030
100
-
-
0. 35
-
4
_
-
-
460
5800
2. 3
Scrubber
Effluent
(Retained)
0,7
0.8
0.015
0.001
1600
0.006
0.021
0.026
53
-
24
0. 15
12
10
1000
-
2500
160
5400
7.8
Clarifier
Effluent
(Retained)
0.2
1.7
0.012
0.004
860
0.015
0.051
0.039
42
-
28
0.54
10
11
900
3.4
1280
180
3200
7.2
Dash indicates sample not analyzed or insufficient sample.
218
-------
TABLE D-2. ANALYSIS OF SCRUBBER LIQUORS FROM TVA
SHAWNEE STEAM PLANT
Date: 12 Jul 1973
Cone: mg/4
TCA Scrubber-Limestone System
Scrubber
Liquor
Constituents
Aluminum
Arsenic
Beryllium
Cadmium
Calcium
Chromium
Copper
Lead
Magnesium
Mercury
Potassium
Selenium
Sodium
Zinc
Chloride
Fluoride
Sulfate
Sulfite
TDS
pH
Sampling Locations
Process
Makeup
Water
0.04
0. 02
0.0012
0.0005
-
0.001
0.005
0.005
-
0.0001
-
0.01
-
0.4
20
1.2
-
-
7.3
Scrubber
Effluent
(Separated)
1.5
2.0
0.020
0.0072
2200
0. 13
0.060
0. 19
100
0. 11
-
2.0
-
29
4800
0.4
-
-
2.44
Scrubber
Effluent
(Retained)
0.5
2.0
0.028
0.0092
3100
0. 17
0.064
0. 32
50
0. 14
37
3. 1
80
38
6000
0.25
2100
110
11,500
8.4
Clarifier
Effluent
(Retained)
0.3
1.8
0. 026
0. 0089
2600
0.20
0.052
0.28
160
0. 04
43
2.7
87
27
5000
3. 1
1800
90
10,200
9.01
Dash indicates sample not analyzed or insufficient sample.
219
-------
TABLE D-3. ANALYSIS OF SCRUBBER LIQUORS FROM
-------
TABLE D-4. ANALYSIS OF SCRUBBER LIQUORS FROM TVA
SHAWNEE STEAM PLANT
Date:
Cone:
15 Jun 1974
TCA Scrubber-Limestone System
Scrubber
Liquor
Constituents
Aluminum
Arsenic
Beryllium
Cadmium
Calcium
Chromium
Copper
Lead
Magnesium
Mercury
Potassium
Selenium
Sodium
Zinc
Chloride
Fluoride
Sulfate
Sulfite
TDS
pH
Sampling Locations
Process
Makeup
Water
-
0.004
0.004
0.005
-
0.002
0.01
0.006
-
<0. 05
-
-
-
-
_
-
-
-
7. 1
Scrubber
Effluent
(Separated)
-
0.39
0.068
0.006
660
0. 16
0.02
0.35
2800
<0. 05
-
<0. 2
-
0.04
3600
2.3
9000
1550
16,500
4.6
Scrubber
Effluent
(Retained)
2.7
0. 39
0.074
0.004
840
0. 13
0.02
0.30
2800
<0. 05
-
0.08
-
0.03
3300
2.2
10,000
1150
17,800
5.5
Clarifier
Effluent
(Separated)
-
0. 13
0.052
0.004
520
0.09
0.01
0.21
2600
<0. 05
32
0.1
76
0.03
2300
6.5
10,000
55
15,000
8.0
Clarifier
Effluent
(Retained)
0.6
0. 14
0.054
0.003
600
0.09
0.01
0.25
2750
<0. 05
41
<0. 2
79
0.02
2250
6.2
9800
110
15,500
8.3
Dash indicates sample not analyzed or insufficient sample.
221
-------
TABLE D-5. ANALYSIS OF SCRUBBER LIQUOR FROM SCE
MOHAVE GENERATING STATION
Date: 30 Mar 1973
Cone: mg/
-------
TABLE D-6. ANALYSIS OF SCRUBBER LIQUORS FROM TVA
SHAWNEE STEAM PLANT
Date: 19 Mar 1974
Cone: mg/Jt
Venturi-Spray Tower-Lime System
Scrubber
Liquor
Constituents
Aluminum
Arsenic
Beryllium
Cadmium
Calcium
Chromium
Copper
Lead
Magnesium
Mercury
Potassium
Selenium
Sodium
Zinc
Chloride
Fluoride
Sulfate
Sulfite
TDS
pH
Sampling Locations
Process
Makeup
Water
_
< 0.004
0.003
-
-
0.004
0.01
0.03
-
0.001
-
0.003
-
0.03
_
-
-
-
-
7.8
Scrubber
Effluent
(Separated
0.22
0. 15
0.050
0.02
980
0.02
0.08
0.03
53
0. 10
8.4
0.08
71
0.09
1230
<0. 3
1000
450
3500
5.2
Scrubber
Effluent
(Retained)
0.06
0.21
0. 0.023
0.01
900
0.03
0. 10
0.008
56
0.04
8.6
0. 10
33
0.07
1290
<0. 3
800
0.8
3300
5.4
Clarifier
Effluent
(Separated)
0. 12
0. 17
0.020
0.02
840
0.01
0.04
0.04
28
0.09
6.8
0.08
36
0.02
1210
1.4
1350
1.8
3500
9.5
Clarifier
Effluent
(Retained)
0.03
0. 30
0.027
0.03
800
0.02
0.07
0.06
25
0.07
13
0.09
28
0.01
1040
1.4
1000
2.2
3200
9.0
Filter
Effluent
(Filtrate)
0.08
0. 15
0.026
0.03
660
0.03
0.05
0.01
24
0.07
11
0.09
36
0.01
1050
1.4
900
1.7
2800
9.4
Lime
Slurry
0.21
0. 10
0.004
0.01
720
0.01
0.02
0.009
1
0.07
29
0.08
88
0.01
720
40
100
-
1800
12.7
Oash indicates sample not analyzed or insufficient sample
223
-------
TABLE D-7. ANALYSIS OF SCRUBBER LIQUORS FROM TVA
SHAWNEE STEAM PLANT
Date: 16 May 1974
Cone: mg/j?
Venturi-Spray Tower-Lime System
Scrubber
Liquor
Constituents
Aluminum
Arsenic
Beryllium
Cadmium
Calcium
Chromium
Copper
Lead
Magnesium
Mercury
Potassium
Selenium
Sodium
Zinc
Chloride
Fluoride
Sulfate
Sulfite
TDS
PH
Sampling Locations
Scrubber
Effluent
(Separated)
0.12
0.15
0.05
0.04
2360
0.02
0.03
0. 13
220
<0.05
21
1.9
108
0.02
4400
1.5
1500
3.0
8700
5.6
Scrubber
Effluent
(Retained)
0.5
0.04
0.04
0.006
2500
0.008
0.02
0. 11
215
<0. 05
24
1.7
106
0.03
4400
2.2
1400
1.8
8200
7.0
Clarifier
Effluent
(Separated)
0.5
0.06
0.20
0.004
2340
0.03
0.06
0. 14
210
<0.05
30
1.5
108
0.02
4300
2.6
135C
4.6
8000
8.4
Clarifier
Effluent
(Retained)
0.3
0.03
0.07
0.004
2580
0.01
0.07
0. 13
220
<0. 05
29
1.9
104
0.02
4200
4.5
1350
2.3
7800
9.1
Filter
Effluent
(Filtrate)
0.1
0.01
0.05
0.013
2420
0.02
0.04
0. 13
200
<0. 05
27
1.9
109
0.02
4200
3.0
1250
2.7
8400
8.8
Dash indicates sample not analyzed or insufficient sample.
224
-------
TABLE D-8. ANALYSIS OF SCRUBBER LIQUORS FROM TVA
SHAWNEE STEAM PLANT
Date: 27 Jun 1974
Cone: mg/A
Venturi-Spray Tower-Lime System
Scrubber
Liquor
Constituents
Aluminum
Arsenic
Beryllium
Cadmium
Calcium
Chromium
Copper
Lead
Magnesium
Mercury
Potassium
Selenium
Sodium
Zinc
Chloride
Fluoride
Sulfate
Sulfite
TDS
pH
Sampling Locations
Scrubber
Effluent
(Separated)
1.54
0.04
<0. 002
0. 12
3000
0.04
0.01
0.33
420
<0. 001
25
<0. 02
126
0. 18
5800
0.2
800
12
10,800
2.7
Scrubber
Effluent
(Retained)
<0. 01
<0. 002
0. 12
3060
0.04
0.005
0.37
410
< 0.001
27
<0. 02
122
0.44
5200
0.9
700
< 0. 6
10,000
5.4
Clarifier
Effluent
(Retained)
<0. 1
0.02
< 0. 002
0. 11
2820
0.04
< 0.002
0.39
450
< 0.001
32
< 0.02
125
0. 11
5900
4.0
800
0.8
10,400
9.0
Filter
Effluent
(Filtrate)
0.24
0.02
< 0. 002
0. 11
2520
0.03
0.002
0.33
420
< 0.001
28
< 0.02
127
0.08
4900
3.3
800
0.9
9400
8.7
225
-------
TABLE D-9. ANALYSIS OF SCRUBBER LIQUORS FROM APS
CHOLLA POWER PLANT
Date: 1 April 1974
Cone: mg/4
Venturi-Absorber-Limestone System
Scrubber
Liquor
Constituents
Aluminum
Arsenic
Beryllium
Cadmium
Calcium
Chromium
Copper
Lead
Magnesium
Mercury
Potassium
Selenium
Sodium
Zinc
Chloride
Fluoride
Sulfate
Sulfite
TDS
PH
Sampling Locations
Process
Makeup
Water
< 0.004
0.10
0.08
140
0.02
0.04
0.004
5.8
0.04
-
1.2
-
0.05
-
300
<1
-
-
Slurry
Tank
(Separated)
2
< 0.004
0. 14
0.011
680
0. 14
0.20
0.01
3
0.07
14
2.2
2150
O.J1
1700
0.7
4000
0.9
8700
3.04
Slurry
Tank
(Retained)
2
< 0.004
0. 18
0.009
700
0.21
0.19
0.01
6
0. 13
16
2.5
2250
0.07
1430
0.6
4000
<1
9100
4.3
Absorbent
Tower
Effluent
0.6
< 0.004
0.08
0.007
580
0.02
0.03
0.02
7
0.007
-
1.0
800
0.02
620
2.4
2200
1
4300
6.6
Dash indicates sample not analyzed or insufficient sample.
226
-------
TABLE D-10. ANALYSIS OF SCRUBBER LIQUORS FROM APS
CHOLLA POWER PLANT
Date:
Cone:
7 Nov 1974
Venturi-Absorber-Limestone System
Scrubber
Liquor
Constituents
Aluminum
Arsenic
Beryllium
Cadmium
Calcium
Chromium
Copper
Lead
Magnesium
Mercury
Potassium
Selenium
Sodium
Zinc
Chloride
Fluoride
Sulfate
Sulfite
TDS
pH
Sampling Locations
Process
Makeup
Water
_
< 0.004
<0.003
0.0065
27
< 0.004
0.014
0.07
-
0.0015
9-
0.0006
570
0.027
940
1. 1
250
2.5
2438
8.4
Slurry
Tank
(Separated)
_
-
0.038
0.044
770
0.024
0.16
0.37
4
<0.05
28
< 0.0006
1650
0.47
4200
1.5
3750
3500
14,000
3.4
Absorbent
Tower
Effluent
2.1
0.02
< 0.003
0.012
390
0.004
0.010
0. 15
9
<0.5
8
0.033
370
0.036
760
1.0
1360
21
3300
6.8
Dash indicates sample not analyzed or insufficient sample.
227
-------
TABLE D-ll. ANALYSIS OF SCRUBBER LIQUORS FROM DLC
PHILLIPS STATION
Date: 4 Oct 1973
Cone: mg/Z
Venturi-Lime System
Scrubber
Liquor
Constituents
Aluminum
Arsenic
Beryllium
Cadmium
Calcium
Chromium
Copper
Lead
Magnesium
Mercury
Potassium
Selenium
Sodium
Zinc
Chloride
Fluoride
Sulfate
Sulfite
TDS
pH
Sampling Locations
Process
Makeup
Water
_
<0.001
< 0.0005
0.003
50
<0.001
0.003
0.0012
40
0.003
-
<0.01
-
0.013
_
-
-
-
-
Thickener
Overflow
_
0.085
0.012
0.022
1300
0.037
0.06
0.08
220
0.09
20
0.8
1680
0.12
1800
4.8
4500
<1
9400
9.2
Thickener
Underflow
_
0.09
0.012
0.023
1400
0.040
0.07
0. 18
410
0.05
22
0.8
2400
0.09
2700
2.6
6450
27
14,000
7.1
Dash indicates sample not analyzed or insufficient sample.
22%
-------
TABLE D-12. ANALYSIS OF'SCRUBBER LIQUORS FROM DLC
PHILLIPS STATION
Date: 17 Jun 1974
Cone: mg/Ji
Venturi-Lime System
Scrubber
Liquor
Constituents
Aluminum
Arsenic
Beryllium
Cadmium
Calcium
Chromium
Copper
Lead
Magnesium
Mercury
Potassium
Selenium
Sodium
Zinc
Chloride
Fluoride
Sulfate
Sulfite
TDS
PH
Sampling Locations
Process
Makeup
Water
-
< 0.004
< 0.0002
-
-
-
-
< 0.006
-
0.0002
-
0.005
-
-
_
-
-
-
-
Scrubber
Effluent
_
0.06
0.002
0. 1
660
-
-
0.05
-
0.0004
10
0.33
440
-
540
8
2700
1.7
4600
8.9
Thickener
Overflow
_
< 0.004
0.002
-
680
-
-
0.06
-
0.0002
2.6
0.20
380
-
350
2
2800
0.8
4400
4.1
Thickener
Underflow
_
< 0.004
0.003
0. 05
600
-
-
<0. 04
-
0.0002
26
0.028
320
-
470
10
2720
20
4200
10.7
Pond
Sludge
Liquid
_
< 0.004
0.002
-
600
. -
-
0.04
-
0. 0004
22
0.095
340
-
420
7
1000
4.8
4000
10.4
229
-------
TABLE D-13.
ANALYSIS OF SCRUBBER LIQUORS FOR GM
UTILITY BOILER
Date:
Cone:
18 Jul 1974
mg/l
Bubble Cap Tray-Double Alkali System
Scrubber
Liquor
Constituents
Aluminum
Arsenic
Beryllium
Cadmium
Calcium
Chromium
Copper
Lead
Magnesium
Mercury
Potassium
Selenium
Sodium
Zinc
Chloride
Fluoride
Sulfate
Sulfite
TDS
PH
Sampling Locations
Process
Makeup
Water
_
< 0.004
<0.002
< 0.005
-
<0.005
< 0.005
<0. 02
0.0004
-
0.002
-
0.07
_
-
-
-
-
Mix Tank
No. 2
Underflow
_
< 0.004
< 0.005
<0. 02
420
<0. 02
0.06
0.42
0.002
95
0. 15
-
0.06
3500
58
25,000
400
51,000
12.6
Clarifier
No. 1
Overflow
_
< 0.004
< 0.005
<0.02
640
<0. 02
0.02
0.56
< 0.0002
140
0. 14
-
0.09
3700
96
34,000
250
67,000
12,6
Clarifier
No. 1
Underflow
_
-
<0.005
<0. 02
640
<0. 02
0.06
0.55
0.0009
110
0.087
20,000
0.63
4400
82
30,000
160
59,000
12.8
Clarifier
No. 2
Overflow
_
< 0.004
<0.005
<0.02
290
<0.02
0.05
0.55
0.0002
120
0. 19
20,000
Oo05
3100
92
33,000
340
62,000
12.5
Clarifier
No. 2
Underflow
_
< 0.004
< 0.005
<0.02
300
<0.02
0.06
0.53
0.0014
160
0.26
-
0.06
4300
46
30,000
260
68,000
12.5
Filter
Effluent
(Filtrate)
_
< 0.004
<0.005
<0.02
470
<0. 02
0.06
0.52
0.005
-
0.075
20,000
0.59
5200
58
35,000
140
65,000
12.7
Dash indicates sample not analyzed or insufficient sample
230
-------
TABLE D-14. ANALYSES OF SCRUBBER LIQUORS FROM TVA
SHAWNEE STEAM PLANT: TCA SCRUBBER SYSTEM
(Scrubber test conditions at time of sampling shown in Table B-3)
Scrubber
liquor
constituents3
Aluminum (Al)
Antimony (Sb)
Arsenic (As)
Beryllium (Be)
Boron (B)
Cadmium (Cd)
Calcium (Ca)
Chromium (Cr) (total)
Cobalt (Co)
Copper (Cu)
Iron (Fe)
Lead (Pb)
Magnesium (Mg)
Manganese (Mn)
Mercury (Hg)
Molybdenum (Mo)
Nickel (Ni)
Potassium (K)
Selenium (Se)
Silicon (Si)
Silver (Ag)
Sodium (Na)
Tin (Sn)
Vanadium (V)
Zinc (Zn)
Total carbonate
Chloride (Cl)
Fluoride (F)
Sulfite
Sulfate
Phosphate
Total nitrogen
Chemical oxygen demand
Total dissolved solids
Total alkalinity
Conductance, mho/cm
Turbidity, Jackson
units
pH
In-process data
Potential discharge point data
Sample location
Scrubber effluent
Clarifier underflow
Sample date
11/27/73
--
0.2
0.01
--
0.04
1800
0.04
--
0.05
0.06
900
--
--
.-
0.16
6.3
0.2
--
63
--
--
0.84
3400
2.3
--
2700
<0. 1
<0.005
--
12000
--
<3
5.90
6/15/74
2.7
2.0
0.4
0.07
--
0.005
840
0. 16
0. 16
0.02
0.35
0.35
2800
--
<0.5
..
0.44
.-
--
--
0.008
--
--
--
0. 03
3300
2.3
1400
9500
<0. 1
<0.005
--
17800
--
0.027
<3
4. 64
11/27/73
--
0.3
0.004
--
0.004
1600
0.5
--
0.4
.-
0. 12
600
--
0.05
--
0.50
5.9
0.2
--
59
--
--
0.35
3100
2.4
..
2100
<0. 1
<0.005
--
11000
150
<3
9.50
6/15/74
0.6
1.4
0. 1
0.05
--
0.004
520
0.09
0. 10
0.01
0.02
0.23
2750
--
<0.05
--
0.33
41
0. 1
--
0.005
--
--
--
0.02
2300
6.5
80
10000
<0. 1
<0.005
--
15000
--
0.015
< 3
7.96
Concentration in milligrams per liter unless otherwise indicated.
-231
-------
TABLE D-15.
ANALYSES OF SCRUBBER LIQUORS FROM TVA
SHAWNEE STEAM PLANT (VENTURI AND SPRAY
TOWER SCRUBBER SYSTEM)
(Scrubber test conditions at time of sampling shown in Table B-2)
Scrubber
liquor
cons ti tuents^
Aluminum (Al)
Antimony (Sb)
Arsenic (As)
Beryllium (Be)
Boron (B)
Cadmium (Cd)
Calcium (Ca)
Chromium (Cr) (total)
Cobalt (Co)
Copper (Cu)
Iron (Fe)
Lead (Pb)
Magnesium (Mg)
Manganese (Mn)
Mercury (Hg)
Molybdenum (Mo)
Nickel (Ni)
Potassium (K)
Selenium (Se)
Silicon (Si)
Silver (Ag)
Sodium (Ma)
Tin (Sn)
Vanadium (V)
Zinc (Zn)
Total carbonate
Chloride (Cl)
Fluoride (F)
Sulfite
Sulfate
Phosphate
Total nitrogen
Chemical oxygen demand
Total dissolved solids
Total alkalinity
Conductance, mho/cm
Turbidity, Jackson
units
PH
In-process data
Potential discharge point data
Sample location
Scrubber effluent
Clarifier underflow
Drum vacuum filter filtrate
Sample date
3/19/74
0.22
0.39
0. 15
0.05
--
0.02
980
0.02
--
0.08
0.77
0.03
53
--
0. 10
-_
0.5
8.4
0. 08
0.4
0.09
33
--
--
0.09
<10
1230
<0.3
450
1000
<0. 1
<0.001
220
3500
54
0.006
<3
5. 19
5/1 &/74
0. 12
2. 1
0. 15
0.05
--
0.04
2360
0.02
0.6
0.03
0. 14
0. 12
220
0.4
<0.05
..
0.25
21
1.9
1.8
0.01
108
--
--
0.02
<10
4400
1. 5
3.0
1500
<0. 1
<0.001
--
8700
--
0.013
<3
5.67
6/27/74
1.54
1.01
0.04
<0.002
56
0. 12
3000
0.04
0.31
0. 01
1.81
0.33
420
--
<0.001
6. 1
0.29
25
<0. 02
2. 1
0.03
126
3.5
--
0. 18
<10
5400
0.2
12
1800
<0. 1
<0.001
149
10800
63
0.019
<3
5.41
3/19/74
0.03
0.55
0.30
0.027
--
0.03
800
0.02
-.
0.07
0.08
0.06
25
--
0.07
-.
0.08
13
0.09
0.4
0.06
36
-.
<0.001
0.01
<10
1040
1.4
2.2
1000
<0. 1
<0.001
160
3200
49
0. 004
<3
9.02
5/16/74
0.3
2.3
0.03
0.07
--
0.004
2580
0.01
0.6
0.07
0.27
0. 13
220
0.09
<0.05
0.23
29
1.9
1.8
0.01
104
--
--
0. 02
<10
4200
4.5
2.3
1350
<0. 1
<0.001
--
7800
--
0.014
<3
9.12
6/27/74
<0. 1
1.11
0.02
<0.002
46
0. 1 1
2820
0.04
0.32
<0.002
0. 10
0.39
450
0.46
<0.001
6.3
0.24
32
<0.02
1.0
0.03
125
3.5
--
0. 11
<10
5900
4.0
0.8
800
<0. 1
<0.001
98
10400
82
0.014
<3
8.99
3/19/74
0.08
0.46
0. 15
0.026
--
0. 03
660
0.03
--
0.05
0.02
0.01
24
--
0.07
--
0.05
11
0.09
0.2
0.06
36
--
--
0.01
<10
1050
1.4
1.7
900
<0. 1
<0.001
85
2800
57
0.004
<10
9.43
5/16/74
0. 1
1.6
0. 01
0.05
--
0.013
2420
0.02
0.7
0.04
0. 10
0. 13
200
0.2
<0.05
--
0.31
27
1.9
1.6
0.01
109
--
--
0.02
<10
4200
3.0
2.7
1250
<0. 1
<0.00:
--
8400
--
0.012
<10
8.81
6/27/74
0.24
1.01
0.02
<0.002
41
0. 10
2520
0.03
0.35
<0.002
0.06
0.33
420
0.84
<0. 001
5.3
0.21
28
<0. OZ
2.7
0.02
127
3. 1
.-
0.08
<10
4900
3.3
0.9
800
<0. 1
<0.001
89
9400
76
0.013
<10
8.68
Concentration in milligrams per liter unless otherwise indicated.
232
-------
TABLE D-16.
ANALYSES OF SCRUBBER LIQUORS FROM APS CHOLLA
STATION (FDS AND ABSORPTION TOWER SCRUBBER
SYSTEM)
(Scrubber test conditions at time of sampling shown in Table B-6)
Scrubber
liquor
constituents*
Aluminum (Al)
Antimony (Sb)
Arsenic (As)
Beryllium (Be)
Boron (B)
Cadmium (Cd)
Calcium (Ca)
Chromium (Cr) (total)
Cobalt (Co)
Copper (Cu)
Iron (Fe)
Lead (Pb)
Magnesium (Mg)
Manganese (Mn)
Mercury (Hg)
Molybdenum (Mo)
Nickel (Ni)
Potassium (K)
Selenium (Se)
Silicon (Si)
Silver (Ag)
Sodium (Na)
Tin (Sn)
Vanadium (V)
Zinc (Zn)
Total carbonate
Chloride (Cl)
Fluoride (F)
Sulfite
Sulfate
Phosphate
Total nitrogen
Chemical oxygen demand
Total dissolved solids
Total alkalinity
Conductance, mho/cm
Turbidity, Jackson units
pH
In-process data
Potential discharge data point
Sample location
Absorption tower tank
FDS tank
Sample date
4/1/74
0.06
0.03
<0.004
0.08
--
0.007
580
0.02
0.05
0.03
0. 17
0.02
7
0.30
0.007
..
1.0
--
1.0
1 .7
0.01
800
--
--
0.02
-------
TABLE D- 17. ANALYSES OF SCRUBBER LIQUORS FROM DLC PHILLIPS
STATION: SINGLE- AND DUAL-STAGE VENTURI
SCRUBBER SYSTEMS
(Scrubber test conditions at time of sampling shown in Table B-5)
Scrubber
liquor
Aluminum (Al)
Antimony (Sb)
Arsenic (As)
Beryllium (Be)
Boron (B)
Cadmium (Cd)
Calcium (Ca)
Chromium (Cr) (total)
Cobalt (Co)
Copper (Cu)
Iron (Fe)
Lead (Pb)
Magnesium (Mg)
Manganese (Mn)
Mercury (Hg)
Molybdenum (Mo)
Nickel (Ni)
Potassium (K)
Selenium (Se)
Silicon (Si)
Silver (Ag)
Sodium (Na)
Tin (Sn)
Vanadium (V)
Zinc (Zn)
Total carbonate
Chloride (Cl)
Fluoride (F)
Sulfite
Sulfate
Phosphate
Total nitrogen
Chemical oxygen demand
Total dissolved solids
Total alkalinity
Conductance, mho/cm
Turbidity, Jackson units
pH
In-process data
Potential discharge point data
Sample location
Scrubber effluent
Clarifier underflow
Sample date
10/4/73
..
--
0.085
0.012
--
0.02Z
1300
0.037
--
0.06
0.08
220
--
0.09
--
20
0.8
--
0.02
1680
--
--
0. 12
<1
1800
4.8
<1
4500
<0.05
<0.005
--
9400
61
0.0063
<3
9.20
6/17/74
..
--
0.06
0.002
--
0. 1
660
--
--
--
0.5
--
--
0.0004
--
10
0.33
--
-.
440
--
--
--
<1
540
8
1 .7
2700
<0.05
<0.005
65
4600
78
0.0033
<3
8.92
10/4/73
..
--
0.09
0.012
--
0.023
1400
0. 040
0.07
0.026
0. 18
410
--
0.05
--
22
0.8
--
2400
.-
--
0.09
<1
2700
2.6
<1
6450
<0.05
<0.005
--
14000
--
0.01
<3
7.11
6/17/74
..
--
<0.004
0.003
--
0.05
600
--
--
0.04
--
--
0.0002
_-
26
0. 028
--
320
--
--
<1
470
10
20
2720
<0.05
<0.005
60
4200
--
0.0034
<3
10.70
Pond
supernate
6/17/74
--
<0.004
0.002
--
..
600
--
--
..
0.4
--
--
0.0004
22
0.095
--
344
--
<1
420
7
4.8
1000
<0.03
<0.005
-.
4000
41
0.0030
<3
10.44
""Concentration in milligrams per liter unless otherwise indicated.
234
-------
TABLE D-18. TRACE METAL ANALYSIS OF SCRUBBER SOLIDS FROM TVA
SHAWNEE STEAM PLANT
Date: 1 Feb 1973
Cone:
TCA Scrubber-Limestone System
Scrubber
Solids
Constituents
Arsenic
Beryllium
Cadmium
Chromium
Copper
Lead
Mercury
Selenium
Zinc
Sample Description
Coal
4
0.2
1
20
8
30
0.6
3
180
Limestone
6
6
0.12
10
4
0. 1
-
1
-
Fly
Ash
(In)
32
34
0.42
230
25
3
3
8
290
Fly
Ash
(Out)
50
0.2
-
440
110
7
7
2
1600
Bottom
Ash
7
30
0.4
700
220
1
0.4
16
-
Scrubber
Effluent
52
0.2
3
15
9
2
1.2
11
-
Clarifier
Effluent
33
6
1
66
9
1
1
2
110
OJ
-------
TABLE D-19. TRACE METAL ANALYSIS OF SCRUBBER SOLIDS FROM TVA
SHAWNEE STEAM PLANT
Date:
Cone:
12 Jul 1973
ppm
TCA Scrubber-Limestone System
Scrubber
Solids
Constituents
Arsenic
Beryllium
Cadmium
Chromium
Copper
Lead
Mercury
Selenium
Zinc
Sample -Description
Coal
19
0.2
5
10
8
34
0.5
6
220
Limestone
7
3
0.6
9
2
1.5
1
2
-
Fly
Ash
(In)
12
34
4
9
7
4
3
7
600
Bottom
Ash
6
33
0.4
17
9
1
0. 1
7
-
Scrubber
Effluent
7
0.2
2.5
180
20
27
0.4
12
-
Clarifier
Effluent
-
0.3
3.2
250
18
21
1
5
430
-------
TABLE D-20. TRACE METAL ANALYSIS OF SCRUBBER SOLIDS FROM TVA
SHAWNEE STEAM PLANT
Date:
Cone:
27 Nov 1973
ppm
TCA Scrubber-Lime stone System
ro
UJ
Scrubber
Solids
Constituents
Arsenic
Beryllium
Cadmium
Chromium
Copper
Lead
Mercury
Selenium
Zinc
Sample Description
Coal
12
0.6
1
18
15
27
-
4
100
Limestone
1
0.5
0.5
0
4
2
-
3
200
Fly
Ash
(In)
18
3
2
17
11
3
-
2
140
Fly
Ash
(Out)
-
1
6
23
2
6
-
-
230
B ottom
Ash
-
3
0.3
13
17
4
-
2
130
Scrubber
Effluent
3
5
3
140
11
2
-
5
180
Clarifier
Effluent
30
3
0.7
100
8
2
-
7
160
-------
TABLE D-21. TRACE METAL ANALYSIS OF SCRUBBER SOLIDS FROM TVA
SHAWNEE STEAM PLANT
Date: 15 June 1974
TCA Scrubber-Limestone System
00
Scrubber
Solids
Constituents
Arsenic
Beryllium
Cadmium
Chromium
Copper
Lead
Mercury
Selenium
Zinc
Sample Description
Coal
16
0.8
0.5
12
9
17
0.07
3
60
-------
TABLE D-22. TRACE METAL, ANALYSIS OF SCRUBBER SOLIDS FROM SCE
MOHAVE GENERATING STATION
Date: 30 Mar 1973
Cone: ppm
TCA Scrubber-Limestone System
Scrubber
Solids
Constituents
Arsenic
Beryllium
Cadmium
Chromium
Copper
Lead
Mercury
Selenium
Zinc
Sample Description
Coal
3
<0. 1
0.6
45
50
4
0.05
1.5
10
Fly*
Ash
9
2
<0. 5
105
50
25
0.05
10
50
Scrubber
Effluent
0.8
0.06
0.5
9
8
0.25
0.005
5
. 40
Centrifuge
Effluent
(Centrate)
0.6
0.05
0.5
10
9
0.23
0.001
8
45
ro
OJ
sO
*Data from Reference
-------
TABLE D-23. TRACE METAL ANALYSIS OF SCRUBBER SOLIDS FROM TVA
SHAWNEE STEAM PLANT
Date: 19 Mar 1974
Cone: ppm
Venturi-Spray Tower-Lime Process
Scrubber
Solids
Constituents
Arsenic
Beryllium
Cadmium
Chromium
Copper
Lead
Mercury
Selenium
Zinc
Sample Description
Coal
10
2
10
7
12
40
0.5
3.4
280
Fly Ash
(In)
75
12
40
50
80
60
0.4
4
350
Bottom
Ash
5
3
1
6
16
5
0.6
0.7
90
Scrubber
Effluent
13
8
20
25
35
25
0.2
6.5
130
Clarifier
Effluent
18
8
15
20
28
25
0. 1
7.2
280
Filter
Cake
13
11
4
15
30
15
0.2
7.8
200
ro
t^
O
-------
TABLE D-24. TRACE METAL ANALYSIS OF SCRUBBER SOLIDS FROM APS
CHOLLA POWER PLANT
Date: 1 April 1974
Cone: ppm
Venturi-Absorption Tower-Limestone System
Scrubber
Solids
Constituents
Arsenic
Beryllium
Cadmium
Chromium
Copper
Lead
Mercury
Selenium
Zinc
Sample Description
Coal
2.1
0.5
0.1
12
39
60
0.05
5
55
Limestone
<0. 1
-
0.01
11
2
0.5
-
2
-
Fly
Ash
0.4
2
0.03
160
31
165
-
5
150
Slurry
Tank
Effluent
2
1
0.08
52
76
80
4
17
120
Absorption
Tower
Effluent
0.8
0.2
0.06
48
6
14
5
2
60
-------
TABLE D-25. TRACE METAL ANALYSIS OF SCRUBBER SOLIDS FROM DLC
PHILLIPS STATION
Date: 4 Oct 1973
Cone: ppm
Venturi-Lime System
Scrubber
Solids
Constituents
Arsenic
Beryllium
Cadmium
Chromium
Copper
Lead
Mercury
Selenium
Zinc
Sample Description
Coal
16
0.4
4
60
30
45
-
8
80
Lime
<2
0.02
0.03
6
3
0. 1
-
22
300
-------
TABLE D-26. TRACE METAL ANALYSIS OF SCRUBBER SOLIDS FROM DLC
PHILLIPS STATION
ro
Date: 17 June 1974
cone: ppm
Venturi-Lime System
Scrubber
Solids
Constituents
Arsenic
Beryllium
Cadmium
Chromium
Copper
Lead
Mercury
Selenium
Zinc
Sample Description
Coal
6
0.8
0.5
12
11
12
-
3
30
-------
TABLE D-27. TRACE METAL ANALYSIS OF SCRUBBER SOLIDS FROM GE
INDUSTRIAL BOILER
Date:
Cone:
18 Jul 1974
ppm
Bubble Cap Scrubber-Double Alkali System
Scrubber
Solids
Constituents
Arsenic
Beryllium
Cadmium
Chromium
Copper
Lead
Mercury
Selenium
Zinc
Sample Description
Coal
14
1
0.4
7
10
16
0.4
6
27
-------
APPENDIX E
EFFECT OF SCRUBBING PROCESS VARIABLES ON WASTE
SYSTEM CHEMISTRY DATA PLOTS
The following figures show the trace element relationships (or lack
of relationships) that were observed within the liquor portion of scrubber
samples as a function of in-process changes. In each case, major chemical
changes during the wet-end processing of scrubber wastes have been identi-
fied. The following figures show those changes in trace element content as
a consequence of changes in the major chemical species. Relationships were
not observed in every case, and the lack of a function relationship with a
major chemical change is presented since the result may be important for
the proper interpretation of overall results.
Discussions of these data are presented in Section 7. 4 and relate
to functional relationships between trace metal content in scrubber liquor
and process variables.
245
-------
10. Op
EPA- TV
TCA
Al
0. 1
ro
2 O.C
o
o
1.0
AS
0.1
0.01
HT
ill
m
o 2/1/73
o 7/12/73
A 11/27/73
EPA-TVA DLC VENTURI
VENTURI GM DBL ALK
SPRAY TOWER SCE TCA
fr
\
* - -
371
* i
T
--
^ .
^
j=*
n"
i ,
- -t
1
^J TffT
Z" "i-X
- - -T-
1
]Ti
rr! j.
0.1
EPA-TVA
TCA
0.01
Be
0001
3/19/74
5/16/74
6/27/74
SE SCRUBBER EFFLUENT
WO WASTE OUTLET
TCA TURBULENT CONTACT ABSORBER
DBL ALK DOUBLE ALKALI
o 2/1/73
o 7/12/73
A 11/27/73
EPA-TVA
VENTURI
SPRAY TOWER
o 3/19/74
o 5/16/74
A 6/27/74
DLC VENTURI
GM DBL ALK
SCE TCA
Ml' I-J-I-
M
' ~ - - - -
::.-::::
:::. J-
2.
...-t-
..mji..
Ill
j±4|LJ
. . LL
1 1
1 1
j[
SE| Hwo|
A 7/18/74
o 6/17/74
o 3/30/73
GM GENERAL MOTORS PARMA
DLC DUOUESNE PHILLIPS
SCE SOUTHERN CALIFORNIA EDISON
MOHAVE
Figure E-l. Changes in chemical concentration of system liquor as it proceeds
from scrubber effluent to waste outlet.
-------
10, 000 n
1,000 :
Ca
100 :
:'
z"
o in
S 1 Or
0
z
o
o
01-
Cr
0.01 i
-
EPA-TVA
TCA
i
-*t,t_ "
_. -.
ft:
-:-n -
i T
= ::
::=.
4 t
.X
4
1
r
- < i
=^
j^
1 44
^
, rYv
if
pf
EPA-WA OLC VENTURI EPA-TVA DLC VENTURI
VENTURI CM DBL ALK EPA-TVA VENTURI CM DBL ALK
SPRAY TOWER SCE TCA , ^ TCA SPRAY TOWER SCE TCA
l.H
&-. .
_ _ _ i .
. .
j ; j j
4+
:>'i 1
'r* j i
ITT
...::5i
-;:;--
: i 1
; ' ; .':
^
Ui :t-
J C - "
F f-^-5
<> hr
» ^P
^+4-
it
- t -; i-'
IIU ~L
L ffliii
'»....
::::: 1,000,;;;:::
::::!!: Mg :::::::
-...,
100:::::::
in
CL
44414-m 0.1:::: = ;:
--- || -a . . . .
I |
._. -fj<>
r1'
L.. 0.01:::::::
-ffrSI IN Mill
;T 111 " Ho
I t;f "9
-Lxilf
. ; _Jl 0.001
+-, , -j-
4 '*
n nnni
::::::: :;;::::
.
.-"
lt....
j t
,
= - -
( '
'
[ml
-
o 2/1/73 o 3/19/74 a 7/18/74 o 2/1/73 o 3/19/74 a 7/18/74
° 7/12/73 ° 5/16/74 o 6/17/74 ° 7/12/73 o 5/16/74 o 6/17/74
A 11/27/73 * 6/27/74 o 3/30/73 A 11/27/73 A 6/27/74 o 3/30/73
SE SCRUBBER EFFLUENT
WO WASTE OUTLET
TCA TURBULENT CONTACT ABSORBER
DBL ALK DOUBLE ALKALI
DLC DUQUESNE PHILLIPS
SCE SOUTHERN CALIFORNIA EDISON
MOHAVE
Figure E-l. Continued
-------
IS)
*-
00
EPA-TVA DLC VENTURI EPA-TVA DLC VENTURI
EPA-TVA VENTURI GM DBL ALK EPA-TVA VENTURI GM DBL ALK
TCA SPRAY TOWER SCE TCA TCA SPRAY TOWER SCE TCA
loooj., 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 rn Hiiiiiiitirrn L0htiiiiiiiiiiin ttfiiiiiin'iFfi ntiniitiiiiiti
100:::::::
K :?::::
£.-.-
10:::::::
g1
0
5 L0
Of lOmilll
§ mm
o
§ '4--.
o
1.0:::::::
Se
1 1
o.i|^|
1
1 6 g
ii-- - -
::::::: : ^ is ^
r 1 1 1 1 1 1
-11
II
j L
/
-'-'
-I*
::::JF
=::=«t: =»=;;;
5
s
g:E5?E EEE±H
:::->: 0.1:::::::
-$-- .|
tu
'+ -f'"
::: = :: 0.01 :"::::
Onni
IIIHIII 1-OFFrffF
fa.TL: +
.Lf.L,1: . .
lAtl
i. ..
::::::: 0.1 : : S i
. : : + T
Pb tj
T 5 "f
!:;:": 0.01 :-|-T
::s:::: ::::n::
u
^--ir- -I
. - '
:::::::: :^:;:::
: : : qr i : J±: ; - : ±
Kffiffl F^tit
T--T-- Ifftfe
imis M r
^ I|wo| |SE [ -[wpj | SE| {wo] [SE| {wpj
o 2/1/73 o 3/19/74 ^ 7/18/74 o 2/1/73 o 3/19/74 a 7/18/74
a 7/12/73 D 5/16/74 D 6/17/74 ° 7/12/73 a 5/16/74 a 6/17/74
A" 11/27/73 A 6/27/74 ° 3/30/73 A 11/27/73 * 6£7/74 o 3/30/73
SE SCRUBBER EFFLUENT
WO WASTE OUTLET
TCA TURBULENT CONTACT ABSORBER
DBL ALK DOUBLE ALKALI
GM GENERAL MOTORS PARMA
DLC DUQUESNE PHILLIPS
SCE SOUTHERN CALIFORNIA EDISON
MOHAVE
Figure E-l. Continued
-------
100
EPA-TVA
TCA
EPA-TVA
VENTURI
SPRAY TOWER
tv>
gz,
g 1.0
o
o
0.1
0.01
|SE|HWO] |SE|
o 2/1/73
D 7/12/73
A 11/27/73
* 6/15/74
o 3/19/74
o 5/16/74
A 6/27/74
DLC VENTURI
GM DBL ALK
SCE TCA
^
f
/
'..-<>
at | - | »u|
A 7/18/74
a 6/17/74
o 3/30/73
100.000
10.000
Na
1,000
100
10
SE SCRUBBER EFFLUENT
WO WASTE OUTLET
TCA TURBULENT CONTACT ABSORBER
DBL ALK DOUBLE ALKALI
EPA-TVA DLC VENTURI
EPA-TVA VENTURI GM DBL ALK
TCA SPRAY TOWER SCE TCA
-» =
( >
{>
A. - - - -
= £sc; :::::::
_.:;:. ...?"..
= = 'fi
J : = , ..
. J
- " " " T
__.-<>
|SE| ~[woj |SE| -{wpj [SE|
o 2/1/73 o 3/19/74 A 7/H
o 7/12/73 o 5/16/74 o 6/11
A 11/27/73 A 6/27/74 o 3/3D
v 6/15/74
,....-.
[wo
/74
/74
/73
GM GENERAL MOTORS PARMA
DLC DUQUESNE PHILLIPS
SCE SOUTHERN CALIFORNIA EDISON
MOHAVE
Figure E-l. Continued
-------
ui
O
EPA-
100.0Mm=FFFH=
10.000:::: -:
Cl- :-! = : = =
1.000:^:..=
^ -------
0>
E
z
o
i lOOrri 1 1 1 1 1
Z
UJ ----_..
O
10
F- ---,< =
1.0
-----£-
7
/
I?
0 i
EPA-
TVA VENT
A SPRAY
= -11- -1>"::
" "5 1 * ""
-->'-
f*<-
/ , '
^ ,-
... . .^r::::
c r
f r
TVA OLC V
URI GM D
TOWER SCE
i>:::;
=..][.
** 1 1
.Ic=:=M
--=Si5-
j >
ENTURI
BL ALK EPA-
TCA TC
--.-.-.--si.
100,000
miff jj TOS mill
! '{::::
10,000 .!:::::
. . . _ '.,!....
i ooo' .
Illim inn nnn^ | ! 1 1 j 1 1
^i^r 10.000:?i::;.
so4 -nk.=::
:::::-- 1,000 ::
inn
EPA-
TVA VEN1
A SPRAY
::::::: :3t :: : :
M ...... ....
;:::ix: :::::::
" " * * 1 » D - - - .
TVA DLC VE
fURI GM DE
TOWER SCE
::::::: : p ! : :
E = ::B: :::::::
.J 1 :-. : -.
' ii '
i 1 II 1 1 i I I. 1 1 I 1 I 1
5
( t - - -
= " :::::::
NTURI
L ALK
TCA
:::::;(
= = I B 41 .
..1i
"""
- - - - |
[SE] Hwo| ISE | -{wpj |SE| -jwpj |SE| {wpj |SE | -|wo] L§D "JWO'I
o 2/1/73 ° 3/19/74 A 7/18/74 o 2/1/73 o 3/19/74 A 7/18/74
D 7/12/73 ° 5/16/74 D 6/17/74 o 7/12/73 o 5/16/74 a 6/17/74
A 11/27/73 ^ 6/27/74 o 3/30/73 A 11/27/73 A 6/27/74 o 3/30/73
SE SCRUBBER EFFLUENT
WO WASTE OUTLET
TCA TURBULENT CONTACT ABSORBER
DBL ALK DOUBLE ALKALI
DLC DUQUESNE PHILLIPS
SCE SOUTHERN CALIFORNIA EDISON
MOHAVE
Figure E-l. Continued
-------
Sb B Co Si
IQiiiMiiiium-nlOOFFFFmimuiii IQi 1111111111 ITFB 10
1.0
0.1
0.01
10
1.0
0.1
1.0
o.i
0.01
1.0
0.1
0.01
SE WO SE WO SE WO SE WO
Fe Mn Ni Ag
IQi 11 n i M ii 11 i=m IQi 11111111111 im IQi 111111111 niTn 1.0
1.0
0.1
0.01
1.0
0.1
001
1.0
0.1
0.01
0.1
L01
1001
SE -WO SE -WO SE WO SE WO
SE SCRUBBER EFFLUENT O 3/19/74 1
WO WASTE OUTLET D 5/16/74 } EPA-TVA VENTURI SPRAY TOWER
& 6/Z7/74 J
tf 6/5/74 EPA-TVA TCA
Figure E-l. Continued
251
-------
Ui
tv)
oEPA-TVA TCA (clarifier effluent)
DEPA-TVA VENTURI-SPRAY TOWER (filter effluent)
AAPS-CHOLLA [slurry tank (separated)]
vDLC-PHILLIPS (thickener underflow)
o
I
§
E
o
0.001
0.0001
1000
100
1.0
1.0
0.1,
0.01
0.001
As
0123456 7 8 9 10 11 12 13 14 15 16 17 18 19 0 123456 7 8 9 10 11 12 13 14 15 16 17 18 19
MONTHS
Figure E-2. Change in chemical concentration of system liquor as a function of
time of scrubber operations.
-------
ro
01
OJ
oEPA-TVA TCA (clarider effluent)
oEPA-TVA VENTURI-SPRAY TOWER (filter effluent)
AAPS-CHOLLA [slurry tank (separated)]
vDLC-PHILLIPS (thickener underflow)
10.000
1,000,
100
1
a
I
0.1
0.01
0.001
Ca
0.1
0.01
10,000
Cu
1.000
100
10
Na
0123456 7 8 9 10 11 12 13 14 15 16 17 18 19 0123456 7 8 9 10 11 12 13 14 15 16 17 18 19
MONTHS
Figure E-2. Continued
-------
oEPA-TVA TCA (clarifier effluent)
QEPA-TVA VENTURI -SPRAY TOWER (filter effluent)
AAPS-CHOLLA [slurry tank (separated)]
vDLC-PHILLIPS (thickener underflow)
1000
100
10
1.0,
o
I
<
o
z
o
0.1
100,000
10,000
1000
9'-
\
so,
IDS
1.0
o.u
0.01
0.001
0.0001
0.1
o.c
0.001
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
ma
Se
Be
17 18 19 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
MONTHS
Figure E-2. Continued
-------
o EPA-WA TCA (clarifier effluent)
aEPA-TVA VENTURI-SPRAY TOWER (filter effluent)
AAPS-CHOLLA (slurry tank (separated))
ADLC-PHILLIPS DUQUESNE (thickener underflow)
Ul
Ul
10 <
1
*, 0.1,
np
Z
o
S n ni ,
i
o
z
o
o
0.1
0.01;
i ' ' II ' i 1 i i i i ' i
i 1 1 1 1 1 | 1 1 1 |
I!1 !l .--ft 'i
tow IT n T
-1 hH- --r + 4--r\ i ; i ~-r4-t
, i i 1 1 . 1 1 1 | | I M 1
!| ! ! ! ! ! V 1 ! !
iii Nil
^;S^M^Mr;
Ml 11 Ml XP M 1 1 I°J
III i -i"" hi lil l|>.
^3tt HT HIltT *
-tr1 - - *f- T^ -^ tHt ^tr - T^ -
i / .1.1 1 1 1 1 1 1 1 . i. ...
/ i i 1 1 1
'it
|[| | | |||| lljjijll | Ulllll
5 , £.
4-fp;£
""'it''' «'' "--I---
* t
-i-y
i£
--Tt-TT In -T
1 1
""lliniliMMIHiiHM!!!!!!!
|:tf:|::j::::::::|:: ,:::::::|:::::::::::::-:::
1 .7
' ."'
' : S:::: :::::::::::: 100 :::,«:: : ,: S S 1 :=::::::::::::::
+B3f:f;;;f;;;;;;;:;; ';?!;;;;;::!;:;;;;;;;;;;:;;;;;
i i if
J T
sii.^^,....^ ,=......
. . -. i
::::::::::::::::::::::::Pb 10 ::::::::::::::::::::: 1 1 :::::
:::::::::::;;!::::::::::: L^l;:::;:::;::::::::::
J LL JJ L'J LU L'J LU LU LU ' ' o.Uiiiimiiiiiimiiiiiiiiim
444--
..-''
F"
--.;,.<::
1 234567 8 9 10 11 12 13 14 15 16 17 18 19 "*0 12 3456 7 8 9 10 11 12 13 14 15 16 17 18 19
MONTHS
Figure E-2. Continued
-------
oEPA-TVA TCA (clarifier effluent)
aEPA-TVA VENTURI-SPRAY TOWER (filter effluent)
A APS CHOLLA [slurry tank (separated)]
VDLC PHILLIPS (thickener underflow)
10,000
1000
0.1
0.01
E
O
u
8
0.001
10.000
1000'
100
so.
Cd
I 1 I
Cl
0 12 3456 7 8 9 10 11 12 13 14 15 16 17 18 19
MONTHS
Figure E-2. Continued
256
-------
l.U
0.1
Ac
0.01
^j
z*
, . - -
-. - .
r:t£
:::^
i
, &
-4-*_ 1
rt
6k'
- - -
o
>P o
1 | L
3.oi
"TTTr"
tV\A .
W.I
,h,
O
*6
**
*V
......
hi:;x
^-' :
Onn/
.UIW
M. .
n
* r
r~
1.
,*-
2 2 4 6 8 10 12
5 P"
<£l l.U
LLJ
O
O
0.1
p J
LO
n 01
4.
-k . . .
' .
* * * 5>
;:::
-j-'f--
.-,-.
>' '
ba;::
^;
-i.^g.
^^^
i-^.-.
i i i__
.. , .^
a
, t>
* *^
A 7 , K
.
- » i
-*..
..-_
p (I
M' V
-«*T-
w W
. . . ..
i -
.....
^^:
E:
ItILZ-1.
:t
.:.
r 4-
2 4 6 8 10 12
PH
- -
-
O
'"9
*;;
-o - -
cO.bp
. : : .
p
o°-
1 1
1 ! . .
:
d1
> :
cPc
^j. . . .
;*d<
°
'
0.003
3*2---(
D .
*~- '
- 4 * -
fO^- .
>'
»
i * . *
1
- -
- - ,
r
.
l.U
0.1
o m
Cr
2 4 6 8 10 12
PH
Figure E-3. Trace element concentration of system liquor
as a function of system pH.
257
-------
1.0
0.1
Cu
0.01
>
{
e-
1
c
)
e
t
i
c.
1
<
u
t J
y
L
0.04
p
4
(
(
f
i
<
,
I
)
:O.OOJ
>
2 4 6 8 10 12
pH
2
C
>
c
±-
0
f
*
*-
f
Ul_
c
f
c
I"
>f
'*
n i
t
-i
)
44-
--i-
--
1.0
0.1
Pb
0.01
E
0
t
<
1 1-0 B
UJ
'2.
o
0.1 =
Se
0.01 :
n nm .
r
>
0
4 6 8
PH
g
.> o
.- _ -, _ - «
[t<0.01
:== = B|-^^b,
e t
<0.002 1
< 0.002
10 12 't
>
7
! 4
C
1
9 3
5 |
6
~:
l«-
. .
c
P
i
2
g
H
6
c
(
(
,?
>-
10
i
_
i;
)
1.0
0.1
Zn
0.01
246 8 10 12
PH
0.001
Figure E-3. Continued
258
-------
1.0
0.1
0.01
Ol
t
o
<
1 1.0
UJ
o
o
0.1
0.01
:
:
-
-
,
c-
-
r
-
-
- 0
CO °
o>c°
CD
CO
O
1 1 1 1 Mil 1 1 1 1 1 II ll 1 ill 1 III!
!
-
-
~
00°° 0
CWOj CO
o
0 0
8
O CO
of-01
1 (tfvi O <0. 004
' 111 11 11
, innil , , ,,,,,,1 , , ,,,,!
P-
~
r
-
~
~
,
pb
orff °
o
op o
o
* 0
1 1 1 ll III 1 1 1 1 1 1 1 ll 1 1 1 1 1 1 II 1
o Cu
o
°<%
0 0
CO)
o o o
o
<0.002
11
ft
I 1 1 Illlj 1 1 1 1 Mill 1 1 1 ll 1 III
0.01 0.1 1.0
IONIC STRENGTH, mp.
0.01 0.1 1.0
IONIC STRENGTH, m/i
Figure E-4. Trace element concentration of system liquor as a func-
tion of system ionic strength.
259
-------
l.Oi^
-
0.1 r-
-
-
0.01 r
I*
z" i
<
g 1.0 P-
o 5
o :
0
0.1 r
-
-
0.01 r
_
1
8<5
°co
O
o o o
<&
# <0.02 <0.02
1 1 1 1 1 1 1 1 1 1 1 1* 1 1 1 1 1 1 1 1 1 1 1 1 1 1
1
o
0 0 Zn
$°° °
o
O
0
I i I
i i L in ii i i i 1 1 1 iii i i i 1 1 mi
E o Cr
O
00° o
r °
- cPG> o
- S ? °
r ° s
: o
L 1 1 1 1 llll 1 1 1 1 1 llll 1 1 1 1 11111
0.01 0.1 1.0
IONIC STRENGTH, m
0.01 0.1 1.0
IONIC STRENGTH, m/i
Figure E-4. Continued
260
-------
APPENDIX F
MATHEMATICAL DETERMINATION OF PROGRAM SCRUB
A solid of composition MmNn upon dissolution dissociates into m
positive ions Mz+ and n negative ions Nz~ per mole, where z is the ionic
charge. The thermodynamic equilibrium constant K for the dissolution pro
cess is given by
_, m n m m n n
K =aa = C
where a represents activities, c is the molar concentration, and y is the
activity coefficient of the respective ions. (The activity of the solid phase
is unity by definition.) Since activity coefficients of inidividual ions are un-
defined theoretically and impossible to determine experimentally, the mean
activity coefficient of the electrolyte \ is used:
Therefore,
T, m n m+n
/_, ., \
(F'3)
The solubility product constant K , which is defined as a product of molar
concentrations
tf m n TT -(rn+n) /T, .-.
Ksp = CM CN = KY± (F'4)
is dependent on the ionic strength of the solution |JL and becomes equal to the
thermodynamic equilibrium constant K = KSp(|j. -. o), or K = K°n as |JL
approaches zero and, therefore, as \ approaches unity. The ionic strength
is defined as
(F-5)
261
-------
From the modified form of the Debye-Huckel expression,
where A, B, and C are empirical constants, and the mean activity coefficient
can be related to- the ionic strength. The alternative power series form of the
Debye-Huckel equation leads to following empirical expression if only the
first two terms are retained:
-log y± * A'n - C'n ' (F-7)
Taking the logarithm of the equation relating K and K
sp
log Kgp = log K°p - (m + n) log v± (F-8)
and substituting for log \ , one has the final relationship which is used in
Program SCRUB:
log Ksp = log K°p + Xn1/2 - YHL (F-9)
or one may use the relation
pK 5 -log K (F-10)
^ sp & sp v '
and
where X and Y are empirical constants obtained from experimental data
reported in the literature (38, 39). If X and Y and K° = K are known for
each slightly soluble component of the liquor, then ills only necessary to
determine the ionic strength in order to obtain the molar concentration and,
therefore, the solubility of each component. Usually, an iterative procedure
will be necessary to calculate |x, but convergence is rapid so that fewer than
five iterations will suffice to obtain 10 percent accuracy.
262
-------
Experimental data were found in the literature (38,39) from which
the values of pKg , X, and Y were determined:
X Y
CaSO4 4.63 2.77 1.06
CaSO3 6.50 2.76 1.06
CaCO3 8.27 2.21 0.87
Ca(OH)2 5.04 1.59 0.81
It is evident that, for the four calcium compounds listed, there was only a
slight variation in the values of X and Y. Similar results were observed for
compounds of lead and mercury:
X Y_
PbSO4 7.79 2.87 1.28
PbCO3 13.13 3.22 1.12
Hg2SO4 6.17 3.21 1.28
Although data were found to provide the required constants for all the weak
acids involved (exceptions being H2SeO3 and P^AsO^), no additional deter-
minations of pKSp over a sufficiently wide range of ionic strength were found
in the literature. In view of the similarity of X and Y values for the calcium,
lead, and mercury compounds, it was decided to use the values for CaSO4
for all insoluble sulfates; the values of X and Y for CaSO3 were used for all
insoluble sulfites and, similarly, for the carbonates and hydroxides.
With the use of the empirical constants, X and Y for calcium sulfate,
the concentration ratios calculated with computer program SCRUB from the
calcium and sulfate concentrations obtained from analyses of sludge liquors
differed systematically from unity. The mean ratio was approximately 1.2.
r
It should be noted that the pKsp data reported in the literature from which the
values of X and Y have been calculated were obtained by adjusting the ionic
strength of the solutions with NaClC>4. This medium has a different influence
on the activity coefficients of ions than the sludge liquors which are of
interest here.
263
-------
Since analytical data were'available for a large number of sludge liquor
samples which were evidently saturated with gypsum, it was possible to
obtain revised values of X and Y for calcium sulfate from these data. Least
squares fitting of the data for 153 sludge liquor samples was used to obtain
revised values of 4. 13 for X and 2.09 for Y. From these revised constants
the mean concentration ratio obtained with computer program SCRUB for
the 153 samples was 1.00 with a standard deviation of 0. 30. It is apparent
that these revised constants give better agreement between calculated and
experimental data for sludge liquors saturated with gypsum. Therefore,
they have been incorporated in program SCRUB and were used to generate
the calculated results included in this report.
264
-------
2.0
0
&
S
_j
y
<
<_>
|1.5
s
UJ
0-
X
UJ
O
N 5
o ^ 1.0
Ul -,
z
o
5
Q£
UJ
O
z.
0
o
0.5
O
O
0 °
o
0
0 ^0
0
o o ° o 0 o
n 00 MEAN
0 u 0 RATIO
oo 0 o
00°
o °o o
o ° o
o ° o o0
0 °o
o
i i i i i i i i i i i i
1 2 3 4 5 6 7 8 9 10 11 12 13
pH
Figure F-l. Concentration ratio as a function of pH.
-------
r°
O
tv>
o^
O1*
EX PERI MENTAL /CALCULATE
RATION RA
CONCE
o
*»
MEAN
RATIO
- 0
0
oo
°
°
1
0.3
0.6 0.9
IONIC STRENGTH
1.2
1.5
J
1.8
Figure F-2. Concentration ratio as a function of ionic strength.
-------
ts)
O^
^1
TED
CALCU
XPER
RATION RA
CONCE
600
CALCIUM CONCENTRATION, mg//
1200 1800 2400
3000
3600
1*
1
1
0
CALCIUM
0 SULFATE
0
RATIO
-0
CT
ff
1
1
6000
12000 18000 24000
SULFATE CONCENTRATION, mgtf
30000
36000
Figure F-3. Concentration ratio as functions of calcium and sulfate concentrations.
-------
.y o
CT>
O
-5
CALCULATED
O EXPERIMENTAL
o o
o o
10 11
PH
12
Figure F-4. Comparison of experimental and calculated concentra-
tions of arsenic.
ij
E
1*
3
= 0
o>
CO
-5
o
-Q_
CALCULATED
o EXPERIMENTAL
-<0.002--Ld o-o
t
7
PH
10
11
12
Figure F-5. Comparison of experimental and calculated concentra-
tions of beryllium.
268
-------
CALCULATED
o EXPERIMENTAL
11 12
Figure F-6. Comparison of experimental and calculated concentra-
tions of cadmium.
5.-
I 0
.c
o
o-
CALCULATED
o EXPERIMENTAL
o o o
<0.02
-5
I i
I I
I
3456 7 8 9 10 11 12
PH
Figure F-7. Comparison of experimental and calculated concentra-
tions of chromium.
269
-------
5r
O
CJ
cn
o
o o
CALCULATED
O EXPERIMENTAL
O o
o o
cP
I1 o
-5
OD_
-O-
7
pH
8 9 10 11 12
Figure F-9. Comparison of experimental and calculated concentra-
tions of mercury.
270
-------
CALCULATED
o EXPERIMENTAL
ro
o>
-5
12
Figure F-10. Comparison of experimental and calculated concentra-
tions of lead.
5r
O>
CO
-5
^*^.
o o
o o
-Q-
J L
CALCULATED
o EXPERIMENTAL
I I
3456 7 8 9 10 11 12
PH
Figure F-ll. Comparison of experimental and calculated concentra-
tions of selenium.
271
-------
S1
-5
CALCULATED
o EXPERIMENTAL
1111
I I
7
PH
10 11 12
Figure F-12. Comparison of experimental and calculated concentra-
tions of zinc.
272
-------
APPENDIX G
FGC WASTE TRANSPORT COSTS
Cost data have been gathered for five methods of moving flue gas
cleaning (FGC) waste products. The following methods were evaluated:
(a) rail hauling, (b) trucking, (c) barging, (d) pumping, and (e) conveying.
Each of these transport modes could be used separately or in combinations,
depending upon the specific operational situation involved. Data have been
gathered for hauling for distances up to 500 miles; however, in most instances
the disposal site is expected to be within about 5 miles of the power plant. At
such short travel distances, the specific characteristics of the power plant
(e. g. , topography and land availability) would greatly influence the hauling
mode selected. Therefore, the data presented here are considered to be use-
ful in the development of engineering estimates of the relative costs involved.
All estimates were based on the output from a 1000-MW station, producing
between 3700 and 4800 tons per day of sludge (50 percent solids) on a 300 day
per year basis. Also, all estimates are based on one-way distances. In the
case of truck rates, the quotes were made with the understanding that the
trucks would return to their origin. No such specification was made on the
other vehicular transport modes.
These data were used as reference material for a portion of this
study which is referenced in 10. 2. 2 and reported in (2).
G. 1 RAIL HAULING
Inquiries on hauling rates were made with railroad companies in the
west, midwest, and along the eastern seaboard. The quotations received*
were from existing tariff schedules, without benefit of special class rates or
equipment specifically designed for the purpose of hauling sludge. Additional
information was received from the Energy Research and Development Adminis-
tration (ERDA)t on typical rail rates across the nation. The results, as
shown in Figure G-l, indicate that rail hauling is probably not economically
feasible at hauling distances of less than 50 miles at ordinary commercial
rates. At 50 miles, the rates vary between 11 and 14 cents per ton-mile and
decrease to between 3 and 8 cents per ton-mile at 300 miles. The estimates
were made for gondola cars of at least 80 tons capacity.
Private communication: Union Pacific Railroad, 15 May 1975; Penn Central
Transportation Company, 17 July 1975; Chessie System, 8 August 1975.
'Private communication: Murray Chais (ERDA), 18 June 1975.
273
-------
0. It
0.12
0.10
f 0.08
0.06
0.04
0.02
80-TON CAR CAPACITY
-ERDA NATIONAL AVERAGE
UNION PACIFIC. OMAHA. NB
PENN CENTRAL.
PHILADELPHIA. PA
CHESSIE SYSTEM.
BALTIMORE. MD
50
100
150 200
DISTANCE, miles
250
300
350
Figure G-l. Rail rates vs distance.
In view of the large quantities of sludge to be hauled, it was recom-
mended by one of the railroad companies that the power companies should
request a new class rate be established to haul this material in coal cars back
to the mine (44).
G. 2
TRUCK HAULING
The rates for hauling sludge in a bottom dump or trailer truck of at
least 20-ton capacity vary significantly, depending on the locale involved. As
shown in Figure G-2, the rate for hauling a distance of 10 miles varies be-
tween 10 and 12 cents per ton-mile in Wilmington, California, and Pittsburgh,
Pennsylvania, respectively. For distances of 125 to 150 miles, these rates
decrease to between 4 and 5 cents per ton-mile.* For longer hauling dis-
tances, on an interstate rate basis, the cost is approximately 11 cents per
ton-mile at 100 miles and decreases to 9 cents per ton-mile at 350 miles, f
As with the railroad companies, the trucking firms contacted for
rate information recommended that the power companies request a special
hauling rate for the large quantities of sludge involved. One company'
*.
Private communication: W. H. Hutchinson and Sons Service, 1 May 1975;
R. Harrison, Harrison/Nichels, Los Angeles, 12 February 1975.
Private communication: Nationwide Bulk Carrier Conference, Inc., Freight
Tarriff, 11 April 1975.
Private communication: Chancellor and Ogden, Inc., Wilmington, CA.
274
-------
0.12
0.10
f 0.08
s
5 0-06
0.04
0.02
BULK CARRIER CONFERENCE
NATION-WIDE RATE
WILMINGTON, CA
PITTSBURGH. PA
50
100
150 200
DISTANCE, miles
250
300
350
Figure G-2. Truck rates vs distance.
offered to provide new trucks specifically designed to accommodate power
plant sludge disposal, including drivers, equipment, maintenance, and financ-
ing, if a contract for a period of at least five years could be negotiated.
In addition to the hauling costs, a cost must be added for loading and
unloading the trucks. If the sludge is filtered, so as to minimize liquid leak-
age and to improve sludge handling characteristics, then the loading and dis-
charging process would be simpler. It is estimated that a truck can be loaded
with filter cake in approximately 3 to 5 min, and the discharge time would be
approximately the same depending on conditions at the disposal site.
G.3
BARGE HAULING
The typical cost for hauling by a river barge with a 1400-ton capacity
is about 4 to, 5 mills per ton-mile along the tributaries of the Ohio and
Mississippi Rivers and about 1 cent per ton-mile for the coastal-canal route
from New Orleans to Brownsville, Texas.* The variation of barge rates
with distance for the inland waterways is shown in Figure G-3 (.40). These
barges are 195 ft long by 35 ft wide, with a 9-ft draft, and are built with
hoppers having 'sides of approximately 1 5 ft in height. The barges can handle
either clarifier underflow slurry or filter cake. Barge companies have had
experience handling materials of similar consistency. For example, lime
slurry has been barged along the Mississippi River. In this case, the slurry
was pumped onto the barge, and then at the destination a suction hose was
ff
Private communication: John A. Welch, Federal Barge Company, Houston,
Texas, 14 February 1975.
275
-------
0.0200r-
0.0175
0.0150
0.0125
0.0100
0.0075
I
I
100
200 300 400
DISTANCE, miles.
500
600
Figure G-3.
Typical barge hauling costs
for inland waterways.
used to remove the material. Materials of wet-sand consistency are con-
tinuously being loaded by clam shell scoops onto barges and removed in the
same manner. The cost of clam shell operation is about $l/ton handled. One
tug of 5 to 10, 000 hp can handle up to 30 barges in a single tow. One of the
barge companies indicated that materials of the type being barged are exempt
from federal or state control, and, because of the large tonnage, the hauling
costs could probably be negotiated for even less than the amounts noted. In
fact, the barge companies would probably purchase dedicated barges for this
work at $ 100, 000 to $ 150, 000 each. For sludge dewatered to 50 percent
solids, the number of barges required for a 1000-MW scrubber is about seven
every two days (3-1/2 per day). At 65 percent solids, the number of barges
required is reduced to five every two days.
G.4
PUMPING
A number of long-distance pipelines are in operation which pump
materials comparable to FGC waste materials (41). Many of these pipe-
lines are between 100 and 500 miles in length, and the cost associated
with their operation is a function of the tonnage transported, distance,
physical characteristics of the material, terrain conditions, and the annual
capital charges of the pipeline. The two cost factors which are most sig-
nificant are the annual tonnages and the distance transported. For slurries,
similar to FGC wastes, having a specific gravity of 1. 40 and a solids concen-
tration of 50 weight-percent, the transport cost for an annual quantity of
2 million tons is approximately 0.7 cents per ton-mile in 1971 dollars (42).
For a quantity of 1 million tons per year, the cost is estimated to be 1 cent
per ton-mile. These costs do not include the costs associated with prepara-
tion of the material (such as dewatering) or handling the material at its desti-
nation. However, the costs do include operating costs and capital charges.
The capital charges are applied at an annual rate of 15 percent of the total
capital investment.
276
-------
For shorter pipeline distances, i. e. , from 5 to 50 miles, the costs
become site-specific. However, costs have been estimated for an average
throughput of approximately 1 million tons per year (125 tons/hr), using infor-
mation developed as part of the field evaluation at the TVA Shawnee plant (43).
These data show that a 5-mile, 2-way pipeline system, using 500-hp pumps
and a 10-in. diameter pipe on the outgoing line and 30-hp pumps and a 6-in.
diameter return line, would cost approximately 6 cents per ton-mile, includ-
ing capital charges and operating costs.
G.5 CONVEYING
The use of conveyor belts for transporting dewatered FGC sludge
may be a cost-effective option for very short distances or when unusual or
difficult transportation conditions exist. As with the other modes of transport,
the cost for conveyor installations for relatively short distances will vary sig-
nificantly from site to site. However, for the purpose of making an engineer-
ing estimate of the cost of conveyor operations, a quote was obtained from
Bulk Systems of Detroit. v The conveyor was assumed to be 3 ft wide, with
a continuous hood cover and powered by a 400-hp motor. The conveyor por-
tion was priced at $250/ft and the overhead galleries at $ 560/ft. The gallery
spans were assumed to be 60 ft, and the grade was assumed to have a constant
1 percent rise. For a 5-mile installation, the cost, including capital charges,
is 78 cents per ton-mile in 1975 dollars. This price does not include land
costs or special provisions for road crossing. Also, costs of dewatering the
sludge or handling it at the disposal site are not included.
Private communication: J. W. Moll to L. J. Bornstein, for Bulk Systems,
Detroit, Michigan, 7 April 1975.
277
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing}
. REPORT NO.
EPA-600/7-77-052
3. RECIPIENTS ACCESSION NO.
4. TITLE AND SUBTITLE
DISPOSAL OF BY-PRODUCTS FROM
NONREGENERABLE FLUE GAS DESULFURIZATION
SYSTEMS: Second Progress Report
5. REPORT DATE
May 1977
6. PERFORMING ORGANIZATION CODE
Rossoffj R.c.Rossi, R.B. Fling,
W. M. Graven, and P. P. Leo
8. PERFORMING ORGANIZATION REPORT NO.
ATR-77(7297-01)-4
9. PERFORMING ORGANIZATION NAME AND ADDRESS
The Aerospace Corporation
Environment and Energy Conservation Division
P.O. Box 92957
Los Angeles. California 90009
10. PROGRAM ELEMENT NO.
EHE624A
11. CONTRACT/GRANT NO.
68-02-1010
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
Progress; 11/72-11/75
14. SPONSORING AGENCY CODE
EPA/600/13
is. SUPPLEMENTARY NOTES iERL-RTP project officer for this report is J.W.Jones, Mail Drop
61, 919/549-8411 Ext 2915. Initial report in this series is EPA-650/2-74-037a.
16. ABSTRACT,.
The report gives results of the first 3 years of study to determine environ-
mentally sound methods for disposing of wastes from nonregenerable flue gas desul-
furization systems. Untreated and treated wastes from seven different scrubbers at
eastern and western plants, using lime, limestone, or double-alkali absorbents, were
characterized. Concentrations of salts and trace elements are related to potential
environmental pollution for both treated and untreated wastes. Physical properties
(e.g. , bulk density, compression strength, permeability, and viscosity) are given.
Disposal of untreated wastes in impermeable impoundments appears to be environ-
mentally viable; however, the ability to reclaim the land has not been determined.
Chemically treated sludges placed in landfills have been shown to be structurally
adequate; in addition, reduction of leachate intrusion into the subsoil is achieved by
the reduction of solubility, permeability, and surface water. Cost estimates for
chemical treatment/disposal equate to $2 to $3 per ton of eastern coal burned; lined
pond disposal costs are estimated at about 75% of chemical treatment/disposal costs.
Costs for disposal of wastes--dewatered and compacted, or converted to gypsum--
are being determined.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Pollution
Waste Disposal
Flue Gases
Coal
Combustion
Desulfurization
Earth Fills
Scrubbers
Sludge
Calcium Oxides
Limestone
Sorbents
Ponds
Leaching
Pollution Control
Stationary Sources
Nonregeneration
Double Alkali
13B
21B 07B
21D 08G
11G
07A,07D 08H
13C
18. DISTRIBUTION STATEMENT
Unlimited
19. SECURITY CLASS (ThisReport)'
Unclassified
21. NO. OF PAGES
297
20. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
279
------- |