&EPA
United States
Environmental Protection
Agency
Industrial Environmental Research
Laboratory
Research Triangle Park NC 2771 1
EPA-600 7-79-046
February 1979
Disposal of By-products
from Nonregenerable
Flue Gas Desulfurization
Systems:
Final Report
Interagency
Energy/Environment
R&D Program Report
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the INTERAGENCY ENERGY-ENVIRONMENT
RESEARCH AND DEVELOPMENT series. Reports in this series result from the
effort funded under the 17-agency Federal Energy/Environment Research and
Development Program. These studies relate to EPA's mission to protect the public
health and welfare from adverse effects of pollutants associated with energy sys-
tems. The goal of the Program is to assure the rapid development of domestic
energy supplies in an environmentally-compatible manner by providing the nec-
essary environmental data and control technology. Investigations include analy-
ses of the transport of energy-related pollutants and their health and ecological
effects; assessments of, and development of, control technologies for energy
systems; and integrated assessments of a wide range of energy-related environ-
mental issues.
EPA REVIEW NOTICE
This report has been reviewed by the participating Federal Agencies, and approved
for publication. Approval does not signify that the contents necessarily reflect
the views and policies of the Government, nor does mention of trade names or
commercial products constitute endorsement or recommendation for use.
This document is available to the public through the National Technical Informa-
tion Service. Springfield, Virginia 22161.
-------
EPA-600/7-79-046
February 1979
Disposal of By-products from
Nonregenerable Flue Gas
Desulfurization Systems:
Final Report
by
J. Rossoff, B.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, NC 27711
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Washington, DC 20460
-------
ABSTRACT
This report provides results of a four-year study by The
Aerospace Corporation to determine environmentally sound methods for the
disposal of wastes from nonregenerable flue gas desulfurization systems.
The data presented incorporate the results obtained during the fourth year of
the study and the material presented in the second progress report, EPA-
600/7-77-052, May 1977, which covered the first three years of the study.
Untreated and treated wastes from ten different scrubbers at Eastern and
Western plants using lime, limestone, and double alkali processes are
characterized. The report relates concentrations of salts and trace elements
in the wastes to the potential for water pollution. Physical properties, e.g.,
bulk density, load-bearing strength, permeability, and viscosity are given.
Disposal by means of ponding, landfilling of chemically fixed wastes, ponding
with under drainage, and conversion to gypsum are assessed. Disposal cost
estimates for a 1000-MW Eastern plant are 0. 55, 0. 90, and 1. 20 mills/kWh
for ponding on indigenous clay, ponding with liner added, and chemical
treatment/landfill, respectively.
Companion studies within this contract pertaining to field
disposal evaluations and a summary of all U. S. Environmental Protection
Agency projects related to flue gas desulfurization waste and water studies,
have been reported in separate documents.
11
-------
CONTENTS
ABSTRACT
ACKNOWLEDGMENTS
CONVERSION TABLE
I. INTRODUCTION 1
1. 1 Report Coverage 1
1.2 Objectives and Study Approach 2
1. 3 Organization of Report Z
II. CONCLUSIONS AND FINDINGS 3
2. 1 Conclusions 3
2. 1. 1 Characterization 3
2. 1.2 Disposal Methods 3
2. 1. 3 Sludge Utilization 5
2. 1.4 Economics of Disposal 6
2.2 Findings 6
2.2. 1 Leachate Chemistry 6
2. 2.2 Physical Properties 7
2. 2. 3 Chemical Properties 8
2. 2. 4 Chemical Solubility 9
2. 2. 5 Pilot Plant Forced Oxidation of
Sulfite Sludges 9
III. RECOMMENDATIONS 11
3. 1 Field Disposal Evaluation Programs 11
3.2 Chemically Treated Sludges 11
3. 3 Control of Sludge Crystalline Structure 11
ill
-------
CONTENTS (Continued)
IV. SUMMARY 13
4. i General 13
4. 2 Disposal Criteria 13
4. 3 Physical Properties 18
4. 3. 1 Wet Bulk Density 18
4. 3. 2 Coefficient of Permeability 21
4. 3. 3 Viscosity 24
4. 3. 4 Compaction 24
4. 3. 5 Unconfined Compressive Strength 26
4. 3. 6 Bearing Strength 26
4.4 Chemical Characterization 30
4. 4. 1 Constituent Concentrations in Sludge
Liquids and Solids 30
4. 4. 2 Generation of Trace Metals in FGD
Wastes 32
4. 4. 3 Process Variables 32
4. 5 Chemical Solubility Analyses 34
4. 6 Environmental Acceptance of FGD Waste 35
4. 6. 1 Pollutant Access to the Environment 35
4. 6. 2 Alternative Disposal Techniques 36
4. 6. 3 Environmental Benefits of Chemical
Treatment 37
4.6.4 Gypsum 39
4. 6. 5 Alternative Disposal Method for
Untreated Wastes 40
4. 7 Disposal Cost Estimates 41
V. DISPOSAL CRITERIA 43
5. 1 Disposal Alternatives for FGD Sludges 44
5. 2 Environmental Effects of FGD Waste
Disposal Alternatives 48
IV
-------
CONTENTS (Continued)
5. 2. 1 Rainwater Seepage 48
5. 2. 2 Landfill Runoff 50
5. 2. 3 Land Reuse 50
5. 3 Selection of Disposal Criteria 51
VI. PHYSICAL PROPERTIES DETERMINATION 53
6. 1 Background 53
6.2 Solids Characterization 54
6. 3 Wet Bulk Densities 54
6. 3. 1 Experimental Procedure 55
6. 3. 2 Bulk Density Test Results 55
6.4 Viscosity 56
6.4. 1 Experimental Procedure 56
6.4.2 Experimental Results 65
6. 5 Coefficient of Permeability 65
6. 5. 1 Experimental Procedure 66
6. 5. 2 Experimental Results 66
6.6 Bearing Strength 69
6.6. 1 Test Procedure 70
6.6. 2 Test Results 70
VII. CHEMICAL CHARACTERIZATION OF SLUDGE
LIQUORS AND LEACHATES 75
7. 1 Background 75
7. 2 Description of Power Plant Scrubbing
Facilities, Sampling, and Chemical
Analyses 76
7.3 Chemical Analyses 76
7. 3. 1 Experimental Procedure 76
7. 3. 2 Results of Chemical Analyses 78
7.4 Leaching Analysis 83
7.4.1 Experimental Procedures 83
7.4.2 Results of Leaching Experiments 83
-------
CONTENTS (Continued)
7. 4. 3 Assessment of Chemical Data 88
7. 4. 4 Assessment of the Effect of Fly Ash
on Trace Element Contents 88
7. 5 Results of Leaching Fly Ash at Controlled
pH . 92
7.6 Results of Solids Analyses 92
7. 7 Summary and Conclusions 95
VIII. ANALYSIS OF FORCED OXIDATION GYPSUM 97
8. 1 Background of Gypsum from Forced Oxidation
of RTF Pilot Plant Limestone Scrubber
Sludge 97
8. 2 Chemical and Leaching Characteristics 97
8.2. 1 X-Ray Diffraction Measurements 100
8. 2. 2 SEM Measurements 100
8. 2. 3 Wet Chemical Analyses 105
8. 2.4 Leaching Test Results 105
8. 3 Physical Characteristics Ill
8.3.1 Permeability Coefficients Ill
8. 3. 2 Density Ill
8. 3. 3 Unconfined Compressive Strength 114
8. 3.4 Load-Bearing Strength 114
8. 4 Cost Estimates of Gypsum-Producing
Processes 114
8.5 Gypsum Cost Summary 126
IX. EVALUATION OF THE ENVIRONMENTAL
ACCEPTABILITY OF FGD SLUDGE 127
9. 1 Introduction 127
9.2 Background 128
9. 2. i Alternative Routes of Pollution 128
9.2.2 Alternative Disposal Techniques 129
VI
-------
CONTENTS (Continued)
9. 3 The Underdrainage Method for Dewatering
FGD Sludge 130
9.3. 1 Results 131
9. 3. 2 Assessment of Underdrainage
Technology 139
9. 4 Assessment of the Potential Chemical
Pollution of the Environment by
Alternative Disposal Methods 146
X. DISPOSAL COST ESTIMATES 151
10. i Economics of Disposal Processes 151
10.1.1 Ponding of Untreated Wastes 151
10.1. 2 Chemical Treatment and
Disposal 153
10. 1. 3 Economics of Conversion to
Gypsum 153
10.2 Cost Comparison 153
REFERENCES 155
APPENDIX. CRYSTAL MORPHOLOGY 159
-------
FIGURES
1. Permeability of Chemically Treated and Untreated
Sludges 22
2. Viscosity of FGD Sludges 25
3. Dry Bulk Density as a Function of Solids Content
After Removal of Compaction Loading 27
4. Load-Bearing Strength as a Function of Moisture,
Fly Ash Content, and Sludge Origin 29
5. Relationship Between Trace Element Content in
Coal and Sludge Solids 33
6. Relationship Between Trace Element Content in
Coal and Sludge Liquor 33
7. Comparison of Mass Release of Sludge Constituents
for Various Disposal Alternatives Showing
Benefit of Chemical Treatment 38
8. Bulk Density as a Function of Solids Content: TVA
Shawnee Lime Sludge Without Fly Ash 57
9. Bulk Density as a Function of Solids Content: TVA
Shawnee Lime Sludge with 40-Percent Fly Ash 58
10. Bulk Density as a Function of Solids Content: GPC
Plant Scholz Double-Alkali Sludge Without Fly Ash 59
LI. Bulk Density as a Function of Solids Content: GPC
Plant Scholz Double-Alkali Sludge with 30-Percent
Fly Ash 60
12. Bulk Density as a Function of Solids Content: LG&E
Paddy's Run Carbinde Lime Sulfur Sludge with 12-
Percent Fly Ash 61
13. Viscosity of FGD Sludges 64
Vili
-------
FIGURES (Continued)
14. Permeability of Chemically Treated and Untreated
Sludges 68
15. Load-Bearing Strength as a Function of Moisture,
Fly Ash Content, and Sludge Origin 71
16. Concentration of Major Species and TDS in Leachate
Lime Sludge With and Without Fly Ash: TVA
Shawnee, Run F 85
17. Concentration of Major Species and TDS in Leachate
of Sludge With and Without Fly Ash: GPC Plant
Scholz 86
18. Concentration of Major Species and TDS in Filtrate
and Leachate Sludge from LG&E Paddy's Run 87
19. EPA Pilot Plant Forced Oxidation System Tests
with Fly Ash 98
20. EPA Pilot Plant Forced Oxidation System Tests
Without Fly Ash 99
21. RTP First-Stage Slurry Solids Containing
Predominantly Gypsum, Fly Ash, and Small
Quantities of Calcium Sulfite 102
22. RTP First-Stage Slurry Solids Containing
Predominantly Gypsum and Small Quantities
of Calcium Sulfite 102
23. RTP Second-Stage Slurry Solids Containing
Predominantly Calcium Sulfite and Fly Ash 103
24. RTP Second-Stage Slurry Solids Containing
Predominantly Calcium Sulfite 103
25. RTP Filtered Solids Containing Gypsum, Fly Ash,
and Approximately 5-Percent Calcium Sulfite 104
26. RTP Filtered Solids Containing Gypsum and
Approximately 5-Percent Calcium Sulfite 104
27. Concentration of Major Species in Leachate of
First-Stage Slurry Solids of RTP Oxidized
Sludges 109
ix
-------
FIGURES (Continued)
28. Concentration of Major Species in Leachate from
Filtered Solids of RTP Oxidized Sludges 110
29. Concentration of Major Species in Leachate of
Second-Stage Slurry Solids of RTP Oxidized
Sludges H2
30. Load-Bearing Strengths of Dewatered, Filtered
Solids from RTP Oxidized Sludges Containing
5-Percent Sulfite 115
31. Simplified Process Schematic for Wet Limestone
Tail-End Scrubber Forced-Oxidation Gypsum-
Producing System 117
32. Simplified Process Schematic for Wet-Limestone
Integrated Forced-Oxidation Gypsum-Producing
System 118
33. Cost of Producing Saleable Gypsum as a Function of
Sludge Treatment and Disposal Credit, 1977 Dollars 119
34. Gypsum Disposal Cost Allowable to Limit Forced
Oxidation to Cost of Disposal of Comparable Quantity
of Sulfite Sludge, 1977 Dollars 125
35. Load-Bearing Strength as a Function of Moisture,
Fly Ash Content, and Sludge Origin 132
36. Development of Bearing Strength by Dewatering of
TVA Shawnee Lime Filter Cake and Sludge-Fly
Ash Mixtures 133
37. Effect of Rewetting on Development of Bearing
Strength of Underdrained TVA Shawnee Lime
Sludge, Filter Cake, and Sludge-Fly Ash
Mixtures . 135
38. Shawnee TCA Limestone Sludge Solids from
Centrifuge Cake, 45-Percent Solids 136
39. Shawnee TCA Limestone Sludge Solids from
Clarifier Underflow, 29-Percent Solids 136
40. Shawnee TCA Limestone Sludge Solids from
Clarifier Underflow, 35. 7-Percent Solids 137
-------
FIGURES (Continued)
41. Bearing Strength of Underdrained TVA Shawnee
Limestone, Sludge, and Sludge-Fly Ash
Mixtures 138
42. Effect of Rewetting on Development of Bearing
Strengths of Underdrained TVA Limestone Sludge 140
43. Extrapolated Water Balance for Venturi and Spray
Tower Scrubber System 142
44. Base Case: Ponding Without Underdrainage 143
45. Ponding With Underdrainage 145
46. Comparison of Mass Release of Sludge
Constituents for Various Disposal Alternatives 148
Xi
-------
TABLES
1. FGD Systems Sampled as Data Base 14
2. Comparison of Sludge Liquor with Drinking
Water Criteria 16
3. Range of Concentrations of Chemical Constituents
in FGD Sludge Liquors Throughout The Scrubber
Loop 17
4. Sample Sources and Densities 20
5. Permeability Coefficients of Compacted
Untreated Sludge 23
6. Unconfined Compressive Strength Test Results
Summary 28
7. Phase Composition of FGD Waste Solids in Weight
Percent 31
8. Benefits of Chemical Treatment: Case Studies for
Comparison of Mass Release of Sludge Constituents
for Various Disposal Alternatives 38
9. Disposal Cost Comparisons 42
10. Comparison of Sludge Liquors with Drinking
Water Criteria 45
11. Range of Concentrations of Chemical Constituents
in FGD Sludge Liquors Throughout the Scrubber
Loop 46
12. Environmental Effects of Disposal Alternatives 47
13. Chemical Constituents in Leachate from Untreated
and Chemically Treated FGD Waste After 1 and
50 PVD 49
xn
-------
TABLES (Continued)
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
Sample Sources and Densities
Permeability Coefficients of Compacted
Untreated Sludge
FGD Systems Sampled as Data Base
Chemical Analysis: Shawn ee Lime Sludge Liquor
and Leachate
Chemical Analysis: GPC Plant Scholz Double -Alkali
Sludge Liquor and Leachate
Chemical Analysis: Paddy's Run Carbide Lime
Sludge Liquor and Leachate
Range of Concentrations of Chemical Constituents in
FGD Sludge Liquors Throughout the Scrubber Loop
Mass Balance, Charge Balance, and Gypsum Solubility
Ratio of Sludge Liquor and Leachate
Comparative Analytical Results for Shawnee Lime
Trace Elements Leached from Shawnee Fly Ash
at Controlled pH
Composition of Dry Solid Sludge from Three Power
Plants, in Weight Percent
Crystalline Phases in RTP Oxidized Sludges
Identified by X-Ray Diffraction
Composition of RTP Oxidized Sludges by Wet
Chemical Analysis
Analysis of Leachate s and Filtrates of RTP
Oxidation Samples
Physical Characteristics of RTP Oxidized Sludges
Cost of Major Components of Forced Oxidation, Wet
Limestone Scrubbing System for a 500-MWe Plant,
in 1977 Dollars
62
67
77
79
80
81
82
89
91
93
94
101
106
107
113
120
xiii
-------
TABLES (Continued)
30. Base Case Costs for Wet Limestone Scrubbing System
for a New 500-MWe Plant, in 1977 Dollars 121
31. Total Incremental Cost of Forced Oxidation, Gypsum
Producing System for a 500-MWe Plant, in 1977
Dollars 123
32. Estimated Incremental Cost of Producing Gypsum
for Sale or Disposal Relative to the Disposal of
FGC Waste, in 1977 Dollars 124
33. Case Studies for Comparison of Mass Release of
Sludge Constituents for Various Disposal
Alternatives 148
34. Comparison of Conditions for Current Disposal Cost
Estimates with Previous Estimates 152
35. Disposal Cost Comparisons 154
xlv
-------
ACKNOWLEDGMENTS
Appreciation is acknowledged for the guidance provided by the
U.S. Environmental Protection Agency Project Officer, Julian W. Jones,
Industrial Environmental Research Laboratory, Research Triangle Park,
North Carolina. His assistance and constant attention to the program
appreciably contributed to the value of the results obtained.
Valuable contributions were also made by the following
Aerospace personnel: Eugene C. Brouillette, Martha A. Perez, Patricia
A. Riley, Michael A. Rocha, and John R. Shepherd.
xv
-------
CONVERSION TABLE
British
1 inch
1 foot
1 mile
1 square foot
1 acre
1 cubic foot
1 gallon
1 cubic yard
1 pound
1 ton (short)
1 pound per square inch
1 pound per cubic foot
1 ton per square foot
1 part per million
1 British thermal unit
(Btu)
i pound per million Btu
1 Btu per pound
Metric
2. 54 centimeters
0. 3048 meter
1. 609 kilometers
9,290 square centimeters
4,047 square meters
28,316 cubic centimeters
3. 785 liters
0. 7646 cubic meter
0.454 kilogram
0.9072 metric ton
0.0703 kilogram per square
centimeter
0. 01602 gram per cubic
centimeter
9,765 kilograms per square
meter
1 milligram per liter (equivalent)
252 calories
0. 43 grains per million joules;
1. 80 grams per million calories
2. 324 joules per gram; 0. 555
calorie per gram
xvi
-------
SECTION I
INTRODUCTION
1. 1 REPORT COVERAGE
This is the third and final report on a study of the disposal of
the by-products from nonregenerable flue gas desulfurization (FGD) systems.
The total study spans the period of 1973 through mid-1977. This document
contains a summary of the results of the total effort, which consisted of
engineering studies, laboratory experimentation, and data assessments.
This document also includes detailed technical discussions of work per-
formed from January 1976 through March 1977 (with some updating made
during the report writing effort). The most recent efforts consisted of
studies of the physical and chemical characteristics of sludge samples from
three different scrubber facilities: (1) the Tennessee Valley Authority's
(TVA) Shawnee Steam Plant, (2) Louisville Gas and Electric Company's
Paddy's Run Station, and (3) Gulf Power Company's Plant Scholz. Studies
were made, also, of gypsum produced by forced oxidation at the U.S.
Environmental Protection Agency's (EPA) Research Triangle Park pilot
plant, and cost estimates for a commercial-scale forced oxidation system
similar to EPA's were made. The environmental acceptability of FGD sludge
was reviewed, and studies were expanded to include ponding of untreated
sludges in basins with underdrainage systems which return all seepage to
the scrubber. Previous cost estimates were updated to reflect a 1977 dollar
cost basis.
Prior reports on this study (Refs. 1 and 2) have included
results of analyses which were conducted on samples obtained from seven
different scrubber systems. This report serves to expand the data base
reported on previously by examining the properties of (1) lime scrubbing
sludge without fly ash, (2) carbide lime waste as an absorbent scrubbing
sludge, and (3) double alkali sludge with and without fly ash.
Other phases of this program, i. e., the EPA/TVA Shawnee
field disposal evaluation project and a review of all Research and Develop-
ment (R and D) activities concerning the disposal of flue gas cleaning waste
materials were reported on separately (Refs. 3 through 6). A third
annual report on the field evaluation project is in work.
-------
1.2 OBJECTIVES AND STUDY APPROACH
The specific objectives of this study are as follows:
a. To identify environmental problems associated with FGD waste
disposal by determining FGD waste chemical and physical
characteristics.
b. To assess current FGD waste disposal methods, including
feasibility, performance, and costs, by conducting laboratory
studies of wastes under conditions associated with waste
disposal, by providing engineering support and conducting
chemical and physical analyses for the Shawnee field disposal
evaluation, by evaluating other available data, and by conduct-
ing engineering cost studies of disposal methods.
c. To identify alternative disposal methods that provide an accept-
able means of FGD waste disposal by cost-effective methods.
d. To make recommendations regarding alternative disposal
approaches.
The objectives of this study were met by (1) reviewing water
quality standards and waste management regulations and correlating this
study's technical evaluations with potential limitations imposed by these
regulations, as appropriate; (2) performing analyses to determine chemical
and physical characteristics, as appropriate, of FGD wastes, coals, makeup
water, fly ash, and chemically treated wastes from EPA-specified sources;
(3) surveying and analyzing technical and economic data pertaining to FGD
waste disposal; (4) planning, coordinating, evaluating, and reporting the
EPA field disposal evaluation at TVA Shawnee (Ref. 4); and (5) making
recommendations for environmentally sound FGD waste disposal techniques.
1.3 ORGANIZATION OF REPORT
Sections II and in provide conclusions and recommendations,
respectively. Section IV summarizes the total study findings and provides
discussions and appropriate tables and figures to support the findings.
Sections V through X comprise the total technical discussion of the work
performed during this reporting period.
Appendix A provides a description of crystal morphology
pertaining to the sludge samples analyzed during this reporting period.
-------
SECTION II
CONCLUSIONS AND FINDINGS
2.1 CONCLUSIONS
These conclusions are based on results of the total study
discussed herein and include, as appropriate, inputs from two tasks that are
reported on separately, i. e., the EPA/ Aerospace /TVA Shawnee field dis-
posal evaluation project and an Aerospace summary of EPA's research and
development programs for the control of waste and water pollution from
power plant flue gas cleaning (FGC) systems. It should be noted that these
conclusions are made in the absence of federal criteria pertaining to
scrubber sludge disposal.
2.1.1 Characterization
Sludges have been characterized physically and chemically.
These analyses have shown the need to avoid potential water pollution and
land degradation problems.
2.1.2 Disposal Methods
2.1.2.1 Chemical Treatment
It has been shown that sludges which have been chemically
treated by commercially available processes can be disposed of in an
environmentally sound manner and that the disposal site can be reclaimed
as a structural landfill. Chemically treated sludge disposal sites which are
basically "dry disposal" operations must be managed to control runoff
throughout the disposal operation, after which they must be covered with
earth and vegetation. Such site management is necessary to prevent the
discharge of eroded material or dissolved solids into adjacent waterways
or contiguous land surfaces. Chemically treated sites which are classified
as "underwater" disposal must be located in an impoundment area for which
the local hydrological and geological characteristics and history are well
known so that any seepage, overflow, or discharge can be accommodated by
the soil and/or adjacent streams and so that any buildup of sludge constitu-
ents, under worst case conditions, is not appreciable when judged by local
-------
regulations. During the period of preparation of this report, several
full-scale chemically treated sludge disposal sites, both dry and underwater,
have become operational. These are being monitored for environmental
control, and to date no deficiencies have been reported.
2.1.2.2 Ponding
Ponding of untreated sludge on a site containing indigenous or
transported soil considered effective as an impervious base having a coeffi-
cient of permeability of approximately 10'7 or better has been demonstrated
(both in test sites and operationally) to be effective for the control of seepage
and overflow during the period of operation; however, sites such as these
are not considered amenable to land reclamation. Furthermore, serious
doubt exists as to the acceptability of ponding particularly because of the
nonstructural nature of the pond. In this regard, it should be noted that
the EPA has taken one position with respect to sludge disposal in its draft
response to a remand issued on September 10, 1973 by the U. S. Court of
Appeals for the District of Columbia:
EPA considers permanent land disposal of raw
(unfixated) sludge to be environmentally unsound,
because it indefinitely degrades large quantities
of land. Although EPA has no regulatory author-
ity to prevent raw sludge disposal, EPA antici-
pates States and local jurisdictions to require
treatment of sludge . . .
Preventing the degradation of large quantities of land appears to be the
intent,
2.1.2.3 Conversion to Gypsum
The process of forced oxidation of sulfite sludges within the
scrubber loop produces a gypsum by-product which can be filtered to a high
solids content material that is easily handled. Because most of the water
is removed in the dewatering process, a smaller mass of material is pro-
duced for disposal. Advantages exist also within the scrubber loop in that
the size of settling tanks is reduced because of the fast settling property of
gypsum, and the process tends to afford total utilization of the limestone.
As of this writing, sludge is converted to gypsum and disposed of, at one
operational site, in an enclosed basin lined with clay which has been trans-
ported to the disposal site. Further tests are being conducted at the Shawnee
field disposal site to evaluate environmental effects, site maintenance
requirements, and amenability of the material for landfill utilization.
2.1.2.4 Ponding with Underdrainage
A method by which untreated sludge is disposed of in an
enclosed basin that is equipped with an underdrainage system is being evalu-
ated at the EPA/TVA Shawnee field disposal site. This method facilitates
quick settling and densification of the sludge and returns all underdrainage
-------
to the scrubber. No operational site is using this approach; however it
shows promise for producing a structurally sound ponding material which
would be amenable to land reclamation and which may not require a very
low permeability base material. Evaluations are continuing on this method
of disposal.
2.1.2.5 Mine Disposal
Disposal of sludges in mines, particularly coal mines, appears
attractive and is being evaluated separately under EPA sponsorship. Cur-
rently, one disposal site is operational in an active surface mine area
(Milton R. Young Power Station Center, North Dakota, a mine mouth station),
and another site is operational in a manner similar to that appropriate for
mine disposal (Martin Lake, Tatum, Texas). Both of these sites are in
lignite coal areas, and the sludge is dewatered prior to disposal but not
chemically treated.
At the Milton R. Young Station, the fly ash, which is a highly
alkaline material, is collected, slurried, and used as a scrubbing reagent
with lime supplement. All hauling of coal and sludge is by truck. The
sludge is filtered and disposed of both in the pit area and between spoil banks
(V-notch). The EPA is supporting a detailed monitoring effort conducted by
the University of North Dakota and the North Dakota State Geological Survey.
This includes ground water monitoring, characterization of the waste material,
core sampling, and soil sampling.
At the Martin Lake Station, limestone is used as the absorbent
in the scrubber, and the sludge is blended with fly ash for dewatering. This
sludge is transported by train to a landfill which is not an active mine area,
but may be in the future. If so, the sludge may be mixed with overburden.
2.1.2.6 Ocean Disposal
Ocean disposal is being studied under separate EPA sponsor-
ship to better understand the potential effects of the chemical and physical
characteristics of chemically treated and untreated sludges on the ocean
environment. At Stonybrook, Long Island, New York, the New York State
Energy Research and Development Authority and New York State University
are building small artificial reefs of chemically treated sludge blocks to
perform biological tests and to determine the structural integrity of the
treated materials. No detrimental effects have been determined, and
monitoring is continuing. Additionally, the New England Aquarium is con-
ducting biological tests and chemical analyses of chemically treated sludges
immersed in an aquarium.
2.1.3 Sludge Utilization
2.1.3.1 Alumina Extraction
A separate economic study, under EPA sponsorship, of an
alumina extraction process utilizing FGC waste as a source of calcium has
-------
been completed, and the limitations under which this method of utilization
may be competitive with bauxite have been defined. Several key processing
steps have also been identified that require experimental verification. In
general, this process doesn't appear to be economically practical without
an adjoining cement plant which could utilize one of the by-products of the
process.
2. 1. 3. 2 Gypsum
Chemical and physical characteristics of pilot-plant-produced
gypsum by forced oxidation have been determined. Engineering estimates
have indicated that the cost of saleable gypsum produced by this process can
be competitive with domestically mined crude gypsum if appropriate cost
avoidance credits are taken for chemical treatment and disposal. The
marketability of this material requires further study, which is being con-
ducted by TVA under EPA sponsorship.
2.1.4 Economics of Disposal (Sludge Including Fly Ash)
The mid-1977 cost of disposal in the following conditions,
ponding on indigenous clay, ponding with liner added, and chemical treatment/
landfill, were estimated to be 0. 55, 0. 8, and 1. 05 mills/kWh, respectively.
This equates to $4.90, $7. 25, and $9. 70 per ton of sludge on a dry basis,
and $1. 50, $2. 20, and $2. 95 per ton of coal burned.
The cost of gypsum disposal including the forced oxidation
of the sulfite sludge, filtration, and disposal of the gypsum including ash,
in an indigenous clay basin, was estimated to be 1. 05 mills/kWh. On the
basis of an equivalent quantity of nonoxidized sludge, this value is 1. 10
mills/kWh.
2.2 FINDINGS
A brief discussion of findings regarding the chemical and
physical characteristics of sludges is as follows.
2. 2. 1 Leachate Chemistry
In the absence of federal disposal criteria (which are now
being developed under provisions of the Resource Conservation and Recovery
Act of 1976), drinking water criteria specified under the National Interim
Primary Drinking Water Regulations were used as a basis for determining
the necessity for protecting water supplies from intrusion by seepage or
leachate from disposal sites. The trace element and major anion content
(arsenic, cadmium, chromium, lead, mercury, selenium, fluoride,
chloride, and sulfate) of most samples analyzed exceeded the criteria at
least for some of these constituents, as did pH and chemical oxygen demand
(COD).
-------
2.2.2 Physical Properties
2.2.2.1 Bearing Strength
Although no criteria are specified for the strength required
for sludges in a disposal site, tests were made at the field evaluation dis-
posal site to determine the load-bearing capability of various sludges in
differing disposal modes and to verify the capability of these materials to
support construction equipment.
In-situ penetrometer tests at the Shawnee disposal test sites
showed bearing strengths in the range of 100 to 300 psi for the chemically
treated sludges, and the following values for untreated-underdrained sludge
slurries: limestone, 60 to 75 psi; lime, 100 to 240 psi; and gypsum, 300 psi.
All of these sites, except the untreated-underdrained limestone sludge site,
supported moving rubber-tired construction vehicles having tire pressures
of 95 psi. The underdrained limestone site supported a moving vehicle
carrying tire pressures of 22 psi. These are not necessarily the maximum
tire pressure capabilities of these materials, but are further indications of
the variations in strength.
The ability to compact gypsum filter cake and its subsequent
load bearing strength are to be determined at the field site.
2.2.2.2 Wet Bulk Density
Wet bulk densities increase in the following order according to
the dewatering method used: (1) settled, (2) drained, and (3) filtered/
centrifuged. Densities also increase with the type of absorbent used, in the
following order: (1) lime or double alkali and (2) limestone. The presence of
fly ash in lime and limestone sludges causes a minor increase in density over
that of sludges without fly ash. For double alkali systems, the differences
in wet bulk density for sludges with and without ash are smaller and varied
enough that no definite trend was established. In general, wet bulk density
in the settled condition ranged between approximately 1. 3 and 1.45 g/cc except
for gypsum, which was 1. 65; for the drained case, the variation was between
1. 3 and 1. 53 except for gypsum, which was 1. 67; and for the filtered or
centrifuged cases the variation was between approximately 1.45 and 1. 65
except for gypsum, which reached values as high as 1.86.
2.2.2.3 Coefficient of Permeability
The permeability coefficient of untreated sludges is typically
10~4 cm/sec although some untreated samples were found to have coefficients
as low as 2 x 10-5 and as high as 9 X 10'4. When placed in a disposal basin
which is provided with underdrainage, a 40-ft impoundment consolidates such
that a permeability coefficient at the lower level decreases to approximately
10~5 cm/sec. It was found that compaction of untreated dewatered sludges
decreases the permeability coefficient by approximately one order of magni-
tude under a pressure of 30 psi.
-------
Chemically treated sludges typically have a permeability
coefficient of lO'5 or less. During the course of this study, pulverized or
fractured samples were also tested and shown to have coefficients of as high
as 10-4 cm/sec. However, stabilized samples extracted from the Shawnee
disposal site verify the typical value of 10"5, with some samples having
coefficients as low as 10- '.
2.2.2.4 Viscosity
Pumpability (<20 poise), determined by laboratory viscosity
tests, was found for various untreated sludges which had maximum solids
content that ranged between 32 and 70 percent. For all samples analyzed,
each had a critical value of solids content at which the viscosity increased
steeply. However, in general at any given solids content, the double alkali
and carbide lime sludges were the most viscous; lime sludges were some-
what less; and the limestone sludges were the least. Double alkali ash-free
sludges were considered pump able at a solids content of less than about
30 wt%; lime and limestone ash-free, at solids content of less than about
40 wt%; and all sludges with ash, at a maximum solids content in the range
of 40 to 50 wt %.
2.2.3 Chemical Properties
The typical level of total dissolved solids (TDS) in FGD sludge
liquors analyzed in this work was approximately 10,000 mg/f (steady state)
except for double alkali scrubbers, whose liquors have a much higher TDS.
The distribution of trace metals in systems liquors tends to lie between
0. 01 and 1 mg/f or ppm for all elements except mercury, which has a concen-
tration distribution about 1/10 that of other trace elements. In most sludge
liquors, pH ranges between 6. 5 and 9; however, in some samples, pH in
the 10 to 12 range were detected. COD ranged between 40 and 140 mg/£.
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 the steady-state conditions of the major species.
The scrubber pH seems to be responsible for trace metals
leaching from fly ash. The pH of the system downstream of the scrubber
does not significantly affect the concentration of trace elements in the scrub-
ber liquor.
A direct correlation exists between the trace element and
chlorine content of coal and the trace element and chloride content in FGD
sludges.
Fly ash represents the major source of trace elements in all
but the most volatile species (e. g., mercury and selenium) that are scrubbed
from the flue gases.
8
-------
Western coal, having typically lower concentrations of arsenic,
cadmium, mercury, and zinc than Eastern coal, generally produces sludge
having lower concentrations of these elements than does Eastern coal.
2.2.4 Chemical Solubility
For the nine trace elements measured extensively in this study
(arsenic, cadmium, chromium, lead, mercury, selenium, zinc, beryllium,
and copper), there was no evidence that trace element saturation was the
controlling parameter for the trace material concentration in the liquor.
2.2.5 Pilot Plant Forced Oxidation of Sulfite Sludges
Analyses were made of gypsum produced by forced oxidation
of sulfite sludge in an EPA pilot plant to augment limestone and gypsum
settling rate tests being performed by EPA. Results were as follows:
a. Permeability coefficient was approximately 1 x 10"* cm/sec.
b. Pore volume fraction was 0. 55 and 0. 5 for gypsum without
and with fly ash, respectively. Wet bulk densities were 1. 3
and 1.4 g/cm^, respectively, when settled and drained, and
without compaction.
c. Unconfined compressive strength was approximately 60 psi.
The presence of as little as 5 percent sulfite reduced this
value to 15 psi.
d. Gypsum filtered to approximately 85 percent solids.
e. Leachate after 2 to 3 pore volume displacements of water was
saturated with calcium and sulfate ions.
-------
SECTION III
RECOMMENDATIONS
Because of the wide variations in the chemical &nd physical
characteristics of flue gas desulfurization (FGD) sludges as produced and
in their application to various disposal or utilization techniques, certain
parameters, such as scrubber operating conditions as they affect crystal
size, shape, and distribution; freeze-thaw and wet-dry effects; runoff con-
ditions as they affect chemically treated sludges; and alternative disposal
methods, need a broader data base than now available. A broader data base
would enhance the ability to control the characteristics of sludge as well as
predict and control the response of the sludge to environmental, land recla-
mation, or utilization applications. The net effect would be a technological
base which would broaden the applications range of these materials and in
many cases simplify operational procedures which would both improve
reliability and reduce costs. The following recommendations are made in
that regard.
3. 1 FIELD DISPOSAL EVALUATION PROGRAMS
Current programs should be continued, particularly those
that broaden the data base with extended time data incorporating the combined
effects of time, temperature, moisture, and freeze-thaw/wet-dry cycling on
chemically treated and untreated sludges. Evaluations to determine simple,
low-cost methods of gypsum disposal should be continued, as well as low-cost
underdrainage systems for ponds to contain untreated sludges.
3.2 CHEMICALLY TREATED SLUDGES
Further testing should be conducted on chemically treated
sludges for conditions simulating rainfall runoff from landfills on which
uncured treated sludge is continually added. Further testing of slope
strength of chemically treated sludges should also be conducted.
3.3 CONTROL OF SLUDGE CRYSTALLINE STRUCTURE
Scrubber sludges having a high sulfite to sulfate ratio exhibit
undesirable characteristics regarding disposal because, in general, they are
11
-------
highly water-retentive and have low settling rates and settled strength when
compared to these characteristics of sludges having high sulfate to sulfite
ratios. These characteristics of the high sulfite sludges are a result of the
thin platelet structural geometry of the crystals; additionally, their charac-
teristics are difficult to predict because of the wide variations in crystalline
size and shapes that result from the use of different coals, absorbents,
scrubber systems, and scrubber operating parameters. It is recommended
that an experimental program be conducted to identify scrubber operating
procedures which would produce crystalline structure that would enhance
disposal characteristics, with special attention being given to thickening
the platelet structure of high sulfite sludges.
12
-------
SECTION IV
SUMMARY
4. 1 GENERAL
A summary highlighting results of the total effort of this
program 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. Some of these figures and tables are repeated in the
appropriate sections of the body of the report when necessary to complete
the detailed technical description therein.
These summary results consist of an integration of data
obtained during the current reporting period with all data derived previously
and reported in References 1 and 2. Sources of materials analyzed are
identified in Table 1; those analyzed during the current reporting period
were obtained from Plants 8 through 11. It should be noted that the flue gas
desulfurization (FGD) waste disposal studies that were initiated in this
program are now concentrated in an ongoing field disposal evaluation project
at the U. S. Environmental Protection Agency/Tennessee Valley Authority
(EPA/TVA) Shawnee test site in Paducah, Kentucky. Annual reports on that
project will be published in addition to References 3 and 4.
4.2 DISPOSAL CRITERIA
Throughout the study, in the absence of federal or other
criteria relating directly to the disposal of FGD sludges, drinking water
criteria have been used as a basis for establishing requirements for disposal.
Because the trace element and salt content of most samples analyzed ex-
ceeded the drinking water criteria, at least for some of the constituents,
the approach taken in this study has been to consider disposal of the FGD
sludges such that no direct discharge to any water supply would be allowed
and that any seepage or runoff would be minimized or perhaps totally elimi-
nated. Various methods of disposal have been studied and evaluated in
the field (Refs. 3 and 4) and are discussed in Section IX.
13
-------
TABLE 1. FGD SYSTEMS SAMPLED AS DATA BASE
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
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 (CM)
Corporation,
Chevrolet-Parma
Power Plant
Southern California
Edison (SCE) Mohave
Generating Station
Utah Power and
Light (UFL) Company,
Gadsby Station
Gulf Power Company
(GPC). Plant Scholr
Louisville Gas and
Electric (LGbE), Paddy's
Run Station
EPA Pilot Plant. Research
Triangle Park (RTF), NC
TVA Shawnee Steam
Plant Plant (same as item
No. 1, but downstream
of ash collectors)
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
Venturi and
spray tower
Marble bed
absorber
Two- stage
scrubber
Venturi and
spray tower.
prototype
Scrubbing
Capacity, MW
Equivalent
10
10-
120
410
32
< 1
< 1
20
70
0.1
10
Coal
Source
Eastern
Eastern
Western
Eastern
Eastern
Western
Western
Eastern
Eastern
Simulated
Eastern
Eastern
Absorbent
Lime
Limestone
Limestone,
fly ash
Lime
Soda ash.
lime regenerant
(double alkali)
Limestone
Soda ash,
lime regenerant
Soda ash.
lime regenerant
Carbide,
lime (slaked
lime waste)
Limestone
Lime
14
-------
During the period of performance of the final phase of this
study, the Resource Conservation and Recovery Act (RCRA), Public Law
94-580, was enacted (October 1976). Under the provisions of this act, the
EPA has been directed to provide regulations and guidance to control the
disposal of hazardous and nonhazardous wastes including solid waste, which
includes FGD sludges (air pollution control sludges). Under the provisions
of the RCRA, the EPA is to develop a definition of hazardous wastes. Under
various criteria, "toxicity" will undoubtedly be included and could pertain
to FGO sludges. If these sludges are defined to be hazardous, the gener-
ation, transportation, and disposal would be controlled by EPA regulations
as a minimum. If the sludges are determined to be nonhazardous, disposal
would be controlled under the provisions of the RCRA land disposal program.
EPA would produce disposal guidelines but the responsibility for disposal
would be that of the states.
At this time, the hazardous and nonhazardous definitions are
still being formulated, and formal nonhazardous disposal guidelines have
not been promulgated. Therefore, the environmental acceptability of FGD
sludge disposal has been studied in terms of preventing or minimizing the
pollution of groundwaters or surface streams, using the best practical
technology available today, with consideration given to near-term disposal
method improvements.
A comparison of the chemical constituents of sludge liquors
in a discharge stream with the National Interim Primary Drinking Water
Regulation (NIPDWR) is given in Table 2 as a ratio of constituent concen-
tration to water criteria. These ratios are given for the ranges of con-
stituents from all samples analyzed (Table 3), for six independent samples
analyzed in the final phase of this study (Section VII), and for samples
previously analyzed (Ref. 1). The values used in this comparative analysis
represent the initial release of these constituents from the base of an
untreated disposal site; continued leaching of a given sample would result
in a reduction of concentrations.
In Table 2, the ratios for concentrations in the "Range of All
Samples" column show that all elements analyzed plus the total dissolved
solids (TDS) and pH exceed drinking water criteria. However, in observing
ratios for the 10 independent samples shown in the table it can be seen that,
except for selenium in two samples and cadmium in one, no trace element ex-
ceeds the criteria by a factor greater than 10. The TDS are excessive for most
of the samples, and the pH is excessive for two. Although trace elements are
not eliminated as a matter of concern by these data, there are indications
that in many cases the concentrations are quite low and that the item for
concern may generally be the concentration of dissolved solids and, in
some cases, pH. It is difficult at this point, in consideration of poten-
tial experimental data variations, the depletion of the material with leach-
ing time, cation exchange and adsorption in the soil, the dilution between
the disposal site and the consumer tap, to specifically quantify the degree
of pollution potential from the disposal of these sludges. Because of the
comparatively large concentration of dissolved solids and the identification
of random values of high concentrations of trace elements, methods to
-------
TABLE 2. COMPARISON OF SLUDGE LIQUOR WITH DRINKING WATER CRITERIA
NIPDWR
Drinking
Water
Criteria,
mg/l
As 0.05
Cd 0.01
Cr 0.05
Pb 0. 05
Hg 0. 002
Se 0.01
F ~2
TDS 500
pH (actual
values)"
Range of
All
Samples
<0.8 - 2.8
0.4 - 11
0.22 - 5
0.2 - 6.6
0.03 - 2.5
0.28 - 20
<0. 5 - 5
6. 6 - 48. 5
6.7 -12.2
Concentration -f Criteria (Nondimensional)
Sample*
A
0.6
5.0
5.0
0.8
2.5
10.0
__
36
6.7
B
0.4
1.2
0.8
3.0
--
3.3
0.5
6.6.
6.8
C
2.0
0.4
1.8
4.6
--
10.0
3.3
30.0
8.0
D
0.04
--
..
<0.2
<0. 1
4.2
-_
13.4
12.2
E
0.4
11
0.6
6.6
< 0. 5
< 2
1.7
18.8
8.7
F
1.2
1. 3
0.2
0.2
< 0.001
7.8
1
20.5
8.0
G
2.8
_ .
_ _
< 0.2
< 0. 1
20
. _
28
7.8
H
0. 1
_ _
_ _
< 0.2
< 0. 1
14
_ _
18.4
7.3
I
0.8
5
_ _
0.8
0.1
2.8
5
8.4
10.7
J
0.2
2.5
1. 1
< 0. 1
0.03
0.3
< 0.5
48.5
8.9
aSample data are as follows:
Sample Station Absorbent
A Mohave Limestone
B Cholla Limestone
C Shawnee Limestone
D Shawnee Limestone
E Shawnee Lime
F Shawnee Lime
G Shawnee Lime c
H Shawnee Limestone
I Duquesne Phillips Lime
J LG&E Paddy's Run Carbide lime
EPA-proposed secondary regulation is 6. 5 to 8. 5.
cForced-oxidired to gypsum.
% Ash Sampling Date
3 Mar 1973
59 Nov 1974
40 Jun 1974
6 Jan 1977
40 Jun 1974
6 Sep 1976
6 Oct 1976
6 Aug 1977
60 Jun 1974
12 Jul 1976
-------
TABLE 3. RANGE OF CONCENTRATIONS OF
CHEMICAL CONSTITUENTS IN FGD
SLUDGE LIQUORS THROUGHOUT
THE SCRUBBER LOOP - ALL SOURCESa
Scrubber
Constituents
Sludge Liquors,
mg/l
Aluminum
Arsenic
Beryllium
Cadmium
Calcium
Chromium
Copper
Lead
Magnesium
Mercury
Potassium
Selenium
Sodium
Zinc
Chloride
Fluoride
Sulfate
Sulfite
TDS
pH
COD
0.03 to 2.0
0.004 to 0. 14
0.001 to 0. 18
0.004 to 0. 11
10 to 2600
0.011 to 0.25
0.002 to 0. 56
0.005 to 0.33
0. 1 to 2750
0.00005 to 0.005
5. 9 to 760
0. 0006 to 0. 20
10.0 to 55,000
0.001 to 0. 59
420 to 33, 000
0.6 to 10
600 to 84, 000
0.9 to 3500
2800 to 162,700
4. 3 to 12.7
40 to 140
test conditions not necessarily
standard. See Table 2 for repre-
sentative list.
17
-------
dispose of these materials which will prevent their access to public water
supply have been assessed. These include chemical treatment and land-
filling, disposal in impermeable ponds, dry disposal of gypsum (produced by
oxidation of calcium sulfite), and disposal in basins equipped with under-
drainage. These methods are discussed in detail in Sections V and IX,
which compare the degree of pollution control offered by each, and disposal
costs.
4.3 PHYSICAL PROPERTIES
The physical properties measured in this study for the char-
acterization of FGD sludges include wet bulk densities of the sludges as a
function of dewatering processes, coefficient of permeability, viscosity as
a function of water content, compactibility, bearing strength, and unconfined
compressive strength as a function of moisture content. Further charac-
terization included 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
processors were 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 chemical characteristics of each sludge, and an evaluation is made of
these behavioral effects on the requirements for environmentally safe
disposal.
4.3.1 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
economics 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 distribution of particles and the morphology of the particles
determined by the crystalline phase. In this study, four methods of dewater-
ing were investigated in the laboratory: settling, settling with drainage,
vacuum filtration, and centrifugation. These methods are those most often
used or considered because of their relative cost effectiveness.
The results show that all dewatering behavior responds to a
relationship that correlates the wet bulk density pg of a material to the dry
bulk density pD, the weight fraction fa, the true density ps of the solid, and
the density of water pw by the equation:
P P
rsrw
18
-------
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 packaged together. Each dewatering method
and the characteristics of each sludge established a 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 increases relative to the dry bulk
density p_. by the equation:
When sufficient water is added to just fill all void spaces between particles,
the maximum density for the sludge p is reached and is defined as the
coincidence of the two equations.
If the dewatering efficiency of each method is defined as the
percent of the measured wet bulk density relative to the maximum density
attainable, a means of comparing the efficiency of each method is available.
In all cases, the highest dewatering efficiency was obtained by vacuum fil-
tration (av 93. 6 percent); and the lowest efficiency was by centrifugation
(av 81. 9 percent); settling (av 86. 2 percent) and free'drainage (av 87. 2 per-
cent) efficiencies lie in between. However, when actual wet bulk densities
are compared, as in Table 4, little difference is seen between vacuum filtra-
tion and centrifugation, and in some cases centrifugation produces the most
dense sludge among the four methods 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
laboratory equipment, filtration is generally expected to be slightly less
effective with a resultant decrease in dewatering efficiency. On the other
hand, centrifugation is usually less effective in particle packing and more
effective in removing water because of a difference in mechanism for sepa-
rating water from solids. Thus, in field conditions, the efficiency of dewater-
ing by these two methods is more nearly equivalent with a general tendency
for centrifugation to provide 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.
19
-------
TABLE 4. SAMPLE SOURCES AND DENSITIES
Sample Source and Date
TVA Shawnee Limestone,
2/1/73
TVA Shawnee Limestone,
6/15/74
TVA Shawnee Lime,
3/19/74
CM Parma Double Alkali,
7/18/74
UPL Cadsby Double Alkali,
8/9/74
DLC Phillips Lime.
6/17/74
APS Cholla Limestone,
9/1/74
SCE Mohave Limestone,
3/30/73
TVA Shawnee Lime,
9/8/76
TVA Shawnee Lime,
9/8/76
CPC Scholz Double Alkali.
6/20/76
CPC Scholz Double Alkali,
6/27/76
LGfcE Paddy's Run, Carbide
Lime
Fly Ash,
wt %
20
40
40
7
9
60
59
3
6
40
0
30
12a
Bulk Density, g/cc
Settled
1.45
1.46
1.36
1.31
1.33
1.40
1.39
1.65
1.37
1.35
1.35
1.36
1.23
Drained
1.51
1.53
1.34
1.35
1.30
1.48
1.44
1.67
1.42
1.50
1.44
1.38
1.27
Filtered
1.65
1.64
1.51
1.52
1.50
1.52
1.48
1.78
1.49
1.54
1.61
1.55
1.48
Centrifuged
1.56
1.60
1.44
1.43
1.62
1.52
1.58
1.86
1.39
1.44
1.38
1.40
1.31
'includes fly ash and other nonsulfur solids.
20
-------
Although drained sludge usually attains a bulk density in
laboratory tests only slightly higher than settled sludge, the Shawnee lime
sludge without fly ash, to which fly ash was subsequently added, attained a
bulk density in the laboratory close to that of filtered sludge. When fly ash
was admixed to the sludge, it had not developed the intimacy of mixing as
when the fly ash passes through the scrubber. Consequently, the fly ash
separated and settled during draining to the bottom of the column and served
thereafter as a dewatering aid to produce high density.
4. 3. 2 Coefficient of Permeability
The physical parameter that most significantly affects the
pollution potential of a sludge is the permeability of the sludge to rain water
since this parameter governs the amount of leaching water passing through
the sludge. Disposal techniques were developed that can minimize the
recharge of rainwater to subsoils, but in all cases the permeability coeffi-
cient defines the maximum limit to the amount of liquid that can pass through
the sludge.
The coefficient of permeability was measured on untreated
sludges and sludge chemically treated by several processors. Permeation
columns were constructed as described in Section 6. 5 and Refs. 1 and 2,
simulating various methods by which sludge may be disposed of, e. g., as
poured; as cured; and as cured, fractured, and compacted. The rate of
water passage through the columns was measured with time and as a function
of the hydraulic head. In some cases, packed sludge columns were com-
pacted or consolidated to greater particle packing densities, and the
measurements were then repeated. The measured permeability coefficients
are plotted as a function of volume fraction of solids in Figure 1.
The results of these measurements show that the permeability
coefficients of all the untreated sludges in the condition expected in a dis-
posed site are approximately 1 x 10'^ cm/sec (Figure 1). These results
also show that for any specific sludge, there appeared the expected relation-
ship of decreasing permeability with decreasing void volume. Additionally,
there is a predominant correlation with particle size and particle packing
fraction, which determines the permeability behavior of most FGD sludges.
Coarser particles reach higher values of solids fraction and higher perme-
abilities; finer particles tend not to pack as well and have lower perme-
abilities. Fly-ash-containing sludges with bimodal particle distribution tend
to have lower permeabilities than expected from particle size relationships.
When untreated sludge is consolidated to represent the base of a 30- to 40-ft
deep disposal basin, the permeability coefficient decreases to about 10~5 cm/
sec. Table 5 shows permeability coefficients of sludges compacted under
pressures of 30 and 100 psi. These values range from 10'^ and 10~5 cm/sec,
or approximately one order of magnitude lower than noncompacted sludges.
Chemically treated sludges are plotted in Figure 1, and it is
seen that the permeability coefficients of these materials vary considerably,
depending on the treatment process and the condition of the test sample.
21
-------
10
-3
tSJ
o
E
1.6 x U
10
.-7
5.5x10
,-8
SAMPLE SOURCES DATE FLY ASH. *
• TVA SHAWNEE LIMESTONE 6/15/74
o TVA SHAWNEE LIME 3/15/74
o SCE MOHAVE LIMESTONE 3/30/73
• GM PARMA DOUBLE ALKALI 7/18/74
T UPL GADSBY DOUBLE ALKALI 8/9/74
v DLC PHILLIPS LIME 6/17/74
A APS CHOLLA LIMESTONE 4/1/74
X CHEMFIX LAB
0 DRAVO LAB
A IUCS LAB
© POND B TVA SHAWNEE/ DRAVO (Ref. 2)
® POND C TVA SHAWNEE/ IUCS (Ref. 2)
$ POND E TVA SHAWNEE/CHEMFIX (Ref. 2)
* PULVERIZED AND COMPACTED
= FAMILIES OF DATA HAVING SAME SLOPES
40.0
40.5
3.0
7.4
8.6
59.7
58.7
10'
I
a 10
a 20
0.30
a 40 a so
- VOLUME FRACTION OF SOLIDS
0.60
0.70
0.80
0.90
Figure 1. Permeability of chemically treated and untreated sludges
-------
TABLE 5. PERMEABILITY COEFFICIENTS OF
COMPACTED UNTREATED SLUDGE
Sludge
Source
TVA Shawnee Lime
(6% Fly Ash)
TVA Shawnee Lime
LG&E
Paddy's Run
(12% Fly Ash)
GPC Scholz
(Without Fly Ash)
GPC Scholz
(30% Fly Ash)
Sampling
Date
9/08/76
9/08/76
6/20/76
6/27/76
i
Fractional
Solids
Volume
0. 35
0.33
0.47
0.42
0.41
0.25
0.21
0.37
0.34
0.36
0.34
Permeability
Coefficient
cm/sec.
8.9X 10'!?
1.9 X 10'3
7. 3x 10~i?
1.1 X 10'j?
1.4x 10"°
8.4X 10%
3.4 x 10~3
1.4x 10~*
4.4 X 10
4. 8 X 10" e
8.3X 10
Consolidation
Pressure (psi)
100
30
100
30
30
100
30
100
30
100
30
23
-------
For example, in some of these cases, measurements were conducted on
treated sludges that were pulverized in order to pack the permeation column.
In this condition, the columns represented material that had been placed
and cured, then subsequently fractured and relocated. This method of
disposal has been practiced operationally 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 low as 10-' to 10-8 cm/sec. However, permeability
measurements of these sludges, treated under field conditions (Ref. 4).
have shown that the permeabilities can be expected in the range of 10"5 to
10" ' cm/sec.
4. 3. 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 pm and consists of three major phases having markedly differ-
ent morphologies. It is both the particle size distributions and phase
morphologies that are believed to influence the viscosity of the sludges.
Viscosity measurements were performed on sludges at varying water
contents at ambient temperatures, with a commercial viscometer having
a measuring range from 3 to 150 poise. Measurement procedures are
described in Section 6.4. 1.
The results of the viscosity tests, presented in Figure 2
for 13 sludge samples, show 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 show that each viscosity curve can be separated
into two rather distinct slopes.
These data indicate that particle size and particle shape,
particularly the platy sulfite particles, are primarily responsible for the
nature of sludge. Also, these data clearly suggest that fly ash
increases fluidity of a sludge and that high pH may increase its viscosity.
Whereas particle shape, size, and distribution each appear to influence
viscosity behavior, the precise relationship of these parameters is not
clear from these limited data.
4.3.4 Compaction
The compaction capability of FGD sludges was investigated
to determine whether an economic asset might be gained by providing a
means of increasing the mass of waste disposed of within a specified volume.
-------
CURVE
1
2
3
4
5
6
7
8
9
10
11
12
13
SLUDGE FLY
GM PARMA DOUBLE ALKALI
UPL GADSBY DOUBLE ALKALI
TV A SHAWNEE LIME
DLC PHILLIPS LIME
TVA SHAWNEE LIMESTONE
TVA SHAWNEE LIMESTONE
TVA SHAWNEE LIMESTONE
LG&E PADDY'S RUN CARBIDE LIME
TVA SHAWNEE LIME
TVA SHAWNEE LIMESTONE
GPC SCHOLZ SODA ASH DOUBLE ALKALI
GPS SCHOLZ SODA ASH DOUBLf ALKALI
TVA SHAWNEE LIME
ASH. %
7.4
8.6
40.5
59.7
20.1
40.1
40.9
12.4
<1.0
<1.0
<1.0
30.0
40.0
DATE
7/18/74
8/9/74
3/19/74
6/17/74
2/1/73
6/15/74
7/11/73
7/76
9/8/76
9/28/76
6/20/76
6/27/76
9/8/76
120
100
80
o
40
20
30 40 50 60
SOLIDS CONTENT, WEIGHT *
Figure 2. Viscosity of FGD sludges
70
25
-------
Moreover, if compacted sludge behaves like fly ash, significantly reduced
permeabilities and increased strength could be expected which may increase
the environmental acceptability of the sludge.
The compactability of a sludge is dependent upon particles
being rearranged under an applied force by sliding past each other. The
process is enhanced by the presence of some pore water, which acts as a
lubricant, but air voids into which the particles can move must also exist.
The particle size distribution and crystalline morphology are the two most
important parameters influencing the compactability of FGD sludges. The
B^me laboratory compaction test was performed on each of the six FGD
sludges at nominally 100 psi. Samples were loaded into pellet dies for the
tests, and the change in sample heights was noted. There was no excess
water in these samples; therefore, no water escaped during the tests. While
constrained by the die, the samples deflected as much as 15 percent under
the load; however, when removed from the die, the actual permanent com-
paction ranged between 1 and 4 percent, as shown in Figure 3. In contrast,
related studies on fly ash have shown compaction of from 7 to 20 percent
(Ref. 7).
The lack of compaction of laboratory tests of FGD sludges
is 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 to the permanent deformation. The absence of
significant permanent deformation resulted in no measurable increase in
strength, with limited decrease in permeability to water percolation. These
tests suggest that, while some benefit may be gained by mechanical com-
paction, it is not likely to be a cost-effective operation in untreated sludge
disposal. The benefit would result from the compaction of the sludge under
its own weight.
4. 3. 5 Unconfined Compressive Strength
Strength measured in an unconfined condition was compared
with unconfined strength data of soils. These tests provided a means of
estimating the suitability of FGD waste for 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 6 for samples tested in a dry (no free water), a
damp (partial pore water), and a wet (as placed operationally) condition.
The results indicate that, on an overall basis, untreated FGD wastes in a
damp condition (~8 percent water) are stronger than samples that are dried
of all water, as is the case for most soils. Nevertheless, these data indi-
cate that untreated sludges having a solids content of approximately 85 per-
cent or greater have unconfined compressive strengths similar to those
of natural soils.
26
-------
1.4
1.3
n
o*
tl'2
CO
= u
CD
£
O
1.0
0.9
0.8
FLY ASH
APS CHOLLA
TVA SHAWNEE LIMESTONE
TVA SHAWNEE LIME
——•—"••
UPL GADSBY
60
70 80
SOLIDS CONTENT, WEIGHT %
90
100
Figure 3. Dry bulk density as a function of solids content
after removal of compaction loading
27
-------
TABLE 6. UNCONFINED COMPRESSIVE STRENGTH
TEST RESULTS SUMMARY
FGD Waste
Untreated
Chemically Treated
Range of Moisture
Content, %
5.6 to 14.4
0.0
30a
2.4 to 5.6
0.0
35 to 60a
Strength Range,
ton/ft2
2.4 to 0.9
1.5 to 0.6
0.0
48.5 to 25.4
52.3 to 22.3
39 to 1 . 8
Typical in-place conditions
Treated FGD wastes have strengths that are capable of
structural landfill applications. The data do not show an appreciable differ-
ence between the strengths in the dry or wet conditions.
4.3.6
Bearing Strength
Measurements of load-bearing strength (confined) as a func-
tion of solids content were made for 10 different sludges and are plotted in
Figure 4. For these samples, it is shown that each may be dewatered to a
critical value of solids content above which the load-bearing strength
increases steeply. The value of solids content at which this critical value
is reached varies appreciably between samples. This occurs over a range
of approximately 50 to 80 percent solids content, depending on the source
and nature of each sludge. Of particular importance is the effect of fly ash
on bearing strength, as shown for the TVA Shawnee lime and limestone
samples. For both of these samples, the presence of fly ash (40 wt%) at a
given solids content produced a much lower bearing strength than the sample
without fly ash. The same trend holds for the Gulf Sholtz samples with
and without fly ash, except that the difference in bearing strength for the
two samples is not as widely separated as it is for the Shawnee samples.
The results of these tests show that untreated sludges can be
dewatered such that the sludge will exhibit structural qualities and that the
presence of fly ash tends to reduce the strength. Although the degree to
which any sludge must be dewatered to exhibit structural qualities can be
estimated from the curves of Figure 4 by comparing the nature of the sludge
to those shown in the figure, its true load-bearing characteristics should
be determined through experimental analysis. (It should be noted that these
28
-------
250,—
o
<
o
SHAWNEE, 6% FLY ASH - 9/8/76
SHAWNEE, 40% FLY ASH - 9/8/76
SCHOLZ, WITHOUT FLY ASH - 6/20/76
SCHOLZ, 30% FLY ASH - 6/27/76
PADDY'S RUN, 12% FLY ASH -
PHILLIPS, 60% FLY ASH - 6/17/74
CHOLLA, 59% FLY ASH -4/1/74
GADS BY, 9% FLY ASH -8/9/74
SHAWNEE, WITHOUT FLY ASH - 11/30/76
SHAWNEE, 40% FLY ASH - 11/30/76
RTP GYPSUM*, WITHOUT FLY ASH - 12/4/75
RTP GYPSUM*, 40% FLY ASH - 9/30/75
ABSORBENT
L - LIME
DA - DOUBLE ALKALI
CL - CARBIDE LIME
LS - LIMESTONE
« - CONTAINS 5%
SULFITE
60 70
SOLIDS CONTENT, weight %
80
90
Figure 4. Load-bearing strength as a function of moisture, fly ash content,
and sludge origin
-------
materials lose their strength when re wetted and must be dewatered again by
such methods as underdraining or air drying to regain their structural
characteristics.)
4.4 CHEMICAL CHARACTERIZATION
4.4.1 Constituent Concentrations in Sludge Liquids
and Solids
Chemical analyses from all sampling points within lime, lime-
stone, and double alkali scrubber waste streams are summarized in Table 3,
where the range of concentrations for each chemical species is given. These
ranges result from an array of test conditions including some which are not
representative of current technology or standard operating conditions, e. g.,
sample liquor obtained from unwashed filter cake of a double alkali scrubber.
(For a more representative listing, see Table 2.) The distribution of trace
elements in system liquors tends to lie between 0. 01 and 1 mg/£ for all
elements except mercury, which has a concentration distribution about I/10th
that of other trace elements.
Major chemical species concentrations in liquors depend
strongly on the scrubber system parameters and tend to have a TDS content
of about 10,000 mg/£ except during startup and in certain cases, such as with
double alkali. The chloride concentration in the liquor depends primarily
on the chlorine content of the coal. Major chemical species content in the
solids also depends upon system parameters, primarily oxidation conditions
and fly ash collection methods.
The crystalline phases present in the solids portion of each
FGD waste are presented in Table 7. These data show 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 was a consequence of both
unreacted limestone absorbent and carbonate formation by adsorption of
carbon dioxide from the atmosphere.
The source of chemical pollutants in FGD sludges originates
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 samples at various positions along the scrubber circuits. Process
ingredients from which the chemical pollutants originate were also analyzed
in order to provide an understanding of the relationship between the source
and fate of these potential pollutants within the FGD scrubbing processes.
30
-------
TABLE 7. PHASE COMPOSITION OF FGD WASTE SOLIDS IN WEIGHT PERCENT
ATOMIC
FORMULA
CaSO<-2H20
CaSOj-imy)
CtfOj-l/ZHjO
UCO,
MsSOj-oHjO
tto2S04-7H20
Naci
Q«S04b
av ASH
OTHER
TOTAL
TVASHAVmEE
LIMESTONE.
2/1/73
21.9
18.5
38.7
4.6
20.1
103.8
TVA SHAWNEE
LIMESTONE.
7/1203
15.4
2L4
20.2
3.7
40.9
101.6
TVA SHAWNEE
LIMESTONE.
6/15/74
3L2
2U8
4.5
L9
40.1
99.5
TVA SHAWNEE
LIME.
3/19/74
6.3
48.8
2.5
1.9
4U5
ioao
SCEMOHAVE
LIMESTONE.
3/30/73
84.6
8.0
6.3
1.5
3.0
103.4
CM PARMA
DOUBLE ALKALI.
7/17/74
48.3
12.9
19.2
7.7
6.9
7.4
102.4
APS CHOLLA
LIMESTONE,
4/1/74
17.3
10.8
2.5
58.7
10.7°
ioao
DLC PHILLIPS
LIME,
6/17/74
19.0
12.9
a2
59.7
8.2°
ioao
UPL GAOSBY
DOUBLE ALKALI,
wm
a 8
0.2
10.8
17.7
8.6
10L1
TVA
SHAWNEE
LIME.
9/8/76
19.4
69.2
10.3
«LO
99.9
LGtf
PADDY'S
RUN.
LIME 7/76
15.1
37.4
29. S8
7.8
12, 4a
3.5-
105.7
CULF-SCHOLZ
DOUBLE
ALKALI.
6/20/76
94.5
15.3
10.1
«LO
68.1
*The carbide Him used as absorbent Is an acetylene manufacturing plant waste by-product and Is reported to contain 2-2 1/2 percent silica and 3-8 percent CaCf^
Thw not explicitly measured; presence deduced from x-ray study
'Soluble sad phase not determined: quantity by difference
"carbon
-------
4.4.2 Generation of Trace Metals in FGD Wastes
Chemical analyses of FGD sludges identified trace metals in
the liquid fractions of the sludges (Table 3). The specific analyses were
plotted against the analyses for coal (corrected for other process in-
gredients). A relationship for solids is shown in Figure 5 and for liq-
uids, in Figure 6. This relationship shows the direct relationship between
the trace metals in sludge and those in coal. In general, these data indicate
that approximately one percent of the trace metals in the sludge is distrib-
uted in the liquid fraction.
The correlation that exists between the trace metal content
in coal and the trace metal content in fly ash further suggests that fly ash
is the principal source of these trace metals in the sludge. Furthermore,
when the trace metal content of the liquor of a sludge containing no fly ash
is compared with the liquor of sludge with fly ash, it is seen that fly ash
causes the concentration levels to be as much as 10 times greater than
those found in sludge liquors without fly ash. Also, in evaluating the data,
it was found that the data points for sludge with high fly ash content tend to
be on the high side of the bandwidth of Figures 5 and 6, and data points for
low fly ash sludge, on the low side of the data bands.
I "" Since the direct relationship between trace elements in sludge
and in coal is clearly indicated, Western coal, having typically lower con-
centrations of arsenic, cadmium, mercury, and zinc than Eastern coal, is
expected to produce sludges having lower concentrations of these metals.
behavior was found as expected.
4.4.3 Process Variables
An evaluation was made of the trace metal content in the sys-
tem liquor at various positions in the scrubber process. Chemical analyses
indicated that the system liquor pH increased and trace metal content
showed a slight decrease en route from the scrubber to the disposal site.
The results for trace metals 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 takes place en route to the disposal site. The trace metal content in
the liquor may decrease 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 pH changes previously discussed.
An evaluation of all chemical analyses of trace metals 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 primarily variables. However, when chemical data were
compared within a single system, a pH correlation was found. Thus, it is
32
-------
400
100
I
§10
u
0.1
a 01
0.1 i 10 100
AVERAGE TRACE ELEMENT CONTENT OF SLUDGE SOLIDS, ppm
1000
Figure 5. Relationship between trace element content in
coal and sludge solids
100
s
u
a i
UJ
O
I
a ooi
a 01 ai LO 10
AVERAGE TRACE ELEMENT CONTENT OF SLUDGE LIQUOR, mgtf
100
Figure 6. Relationship between trace, element content in
coal and sludge liquor
33
-------
concluded that the primary source of trace metals is the coal and that these
metals 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 increase from startup.
The calcium sulfate saturation level is established quickly in most systems
(although some don't operate saturated with respect to CaSO4 • 2 HzO) 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 chloride ion concentration increases, it effects an
attendant increase in ionic strength that increases the calcium sulfate
equilibrium levels. Trace metal contents do not reflect these changes,
either from a lack of analytical sensitivity at very low concentration, lack
of accuracy of analytical data, or from a lack of response to changes in
major species concentrations.
4.5 CHEMICAL SOLUBILITY ANALYSES
A comparison of experimental scrubber liquor compositions
with compositions calculated from the assumption of equilibrium solubility
was made for each analysis by use of two computer 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 therxnodynamic data, only,
and not from the trace chemical species concentrations.
When the results of the two programs were compared for
major species, it appeared that the two programs were in virtual agreement
with respect to the saturation of calcium sulfate in every analysis and
saturation of calcium sulfite in only a few selected cases. Moreover, the
agreement between the two independent approaches relative to the experi-
mental data provides supportive evidence to the accuracy of the experimental
measurements.
For the nine trace elements measured extensively, there was
no evidence that trace element saturation was the controlling parameter for
the trace element concentration in the liquor. This finding is consistent
with the finding of an absence of trace metal response to the most changes
in major chemical species and that trace elements exceed saturation limits
by several orders of magnitude. Further experimentation is required to
resolve this anomaly. It is postulated, however, that this anomaly is a
consequence of the analysis of ultrafine participates (primarily fly ash)
that pass through filters into the analytical solutions. If this postulation is
correct, the measured concentrations of trace metals in the anomolous
cases would represent condensed phases and not soluble phases, and the
impact on the potential environmental hazard is greatly diminished.
34
-------
4.6 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 environ-
ment and the methods by which the wastes may be disposed. The accessi-
bility of pollutants to the environment can be minimized by (1) decreasing the
permeability of the wastes to reduce the seepage of water, (2) reducing the
leachability of the wastes through a reduction of the solubility of the material,
and (3) managing seepage and runoff to limit the excess of waste constituents
to groundwaters or surface waters. Disposal concepts include (1) chemical
treatment of waste to effect alteration of its permeability and leachability
properties, (2) total impoundment of untreated or treated wastes to isolate
them from the environment, and (3) several methods of impounding untreated,
dewatered wastes such that seepage and runoff are minimized. Some of these
concepts produce a structurally stable material which makes site reclamation
possible. The assessment of these methods shows that although intrinsic
pollution prevention of each method is different from the other, the protec-
tion of the environment is best assured only through the implementation of
appropriate disposal site management techniques.
4. 6. 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 is 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 waste.
During the course of this study, no wind-blown emissions or
effects thereof were noted at air-dried ponds, either untreated or chemi-
cally treated. It is assumed, however, that in arid areas or in humid areas
during drought conditions, fugitive emissions could be a problem and that
such a site would be managed similarly to an earth landfill; i. e., the surface
material could be moistened or the site could be covered and vegetated.
By far the most persistent pollution potential is by water
percolation 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 laboratory experimentation by leaching column tests. Results from these
tests show that the concentration of all chemical species decreases contin-
ually in the leachate, initially at a high rate. After the displacement of
3 to 5 pore volumes, however, the rate of decrease is typically very low,
or the concentration becomes constant. The initial high rate of decrease is
believed to be a consequence of a flushing mechanism; the constant or low-
rate segment appears to be the consequence of the solubility of crystalline
phases.
35
-------
Chemical treatment of a waste by fixation processing produces
a leachate with concentration values of major species nominally one-half
of untreated sludge initially. Highly soluble chemical species like sodium
and chloride ions are removed effectively from the system in the initial pore
flushing, and after approximately five pore volume displacements (PVD), the
TDS consist primarily of calcium and sulfate ions. Leachate of untreated
sludge follows a similar pattern, except the maximum concentration levels
are approximately the same as those of the input liquor (Ref. 4).
Both treated and untreated sludge leachates reach gypsum
saturation concentrations after approximately five PVD.
Chemical treatment did not reveal a discernible difference
from untreated material for trace metals. In most cases, all trace metal
content in the leachate dropped below detection levels after the initial flush-
ing. In the remaining cases, only lead, zinc, and selenium were found for
both treated and untreated sludges to persist at concentration levels sig-
nificantly above the background.
4. 6. 2 Alternative Disposal Techniques
The methods by which FGD waste may be disposed of will be
determined by the environmental acceptability of the waste and the cost of
waste 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 otherwise occur. The physical properties of ponded sludge are such
that structural stability does not develop and subsequent land use is obviated.
Upon retirement of the pond, supernate can be withdrawn, the sludge may
possibly be air dried and an overburden can then be placed on the site.
The additional cost and time required for this type of reclamation even if
possible may increase total costs such that ponding is not cost effective
relative to other methods. An alternative to ponding is the technique whereby
the sludge is sluiced to the disposal basin. Instead of supernate liquor col-
lection, the disposal basin is underdrained, and the excess liquor is returned
to the scrubber. Dewatering by underdrainage enhances structural proper-
ties such that land reclamation can be implemented within a few days.
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 intrinsi-
cally high, the method may offer advantages that offset trucking costs.
Particular attention to disposal site management especially as related to
structural stability may be necessary in this method.
Chemical treatment of FGD wastes provides an additional
assurance of environmental acceptance, but this material must be placed in
36
-------
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 of environmental
protection might otherwise be compromised.
In the case of chemically treated waste, it is possible to dis-
pose 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 potential seepage of leachate caused
by the constant hydraulic head of the impounded water.
4. 6. 3 Environmental Benefits of Chemical Treatment
Chemical treatment has been found to have major benefits
which effectively minimize (and possibly, in some cases, virtually eliminate)
the release of leached sludge constituents to the subsoil through (1) the de-
creased permeability of the treated material, and (2) the amenability of the
treated material to compaction and contouring during placement so that
standing water does not occur on the disposal site. The prevention of stand-
ing water avoids having a hydraulic head on the site; therefore, seepage
through the pores does not occur as a result of hydraulic pressure. This is
accomplished by managing the site so that a major portion of the rainfall
on such a site runs off and is collected in a peripheral ditch which directs
the water to a settling pond, from which decanted liquor is disposed of in
an adjacent stream, if acceptable, or returned to the power plant water
reuse system.
Various examples illustrating the effects of sludge treatment,
effects of different subsoils, and management of the site for different
rainfall recharge rates are presented. The relative amounts of sludge con-
stituents (TDS) released at the sludge base for different modes of disposal
were calculated (Table 8 and Figure 7). The figure illustrates the benefit
of chemical treatment, particularly when disposal is on a highly permeable
soil which otherwise would not be supportive if ponded.
This analysis is based on correlations of laboratory results
and Shawnee field condition data reported in Reference 4. All the cases are
indexed to an untreated slurry pond, Case 1, in which the soil permeability
coefficient is 10~-> cm/sec.
In assessing the effectiveness of chemical treatment, tests to
determine permeability of chemically treated sludges were performed on
cores extracted from the Shawnee field evaluation site. Constant-head per-
meability tests were run on (1) pulverized samples, and (2) samples with
and without visible cracks. Uncracked samples of one material had coeffi-
cients of permeability of about lO'7 cm/sec; the pulverized and the cracked
samples had coefficients of approximately 10~5 cm/sec. Therefore, the
effective coefficient of the treated material could be expected to be between
ID'5 and 10~7 cm/sec. In a conservative case (using a coefficient of 10"5),
37
-------
TABLE 8. BENEFITS OF CHEMICAL TREATMENT: CASE STUDIES FOR
COMPARISON OF MASS RELEASE OF SLUDGE CONSTITUENTS
FOR VARIOUS DISPOSAL ALTERNATIVES
Case
1
2
3
4
5
Disposal
Methoda
Ponded
slurry
Ponded
slurry
Ponded
cake
Dewatered
and ponded
Landfill
Surface
Water
Constant
supernate
Constant
supernate
10-in/yr d
recharge
10-in/yr,
recharge
1-in/yr e
recharge
Sludge ,
Condition
Untreated
Chemically
treated
Untreated
Chemically
treated
Chemically
treated
Sludge
Permeability,
cm/sec
ID'4
io-5
ID'4
io-5
io-5
aFill period = 5 yr; depth = 30 ft.
Porosity (void volumetric fraction) = 0. 67.
cConstant supernate assumes 1-ft depth of surface water.
10-in/yr recharge is unevaporated rainfall.
1-in/yr effective recharge resulting from seepage during runoff of rainfall.
ID
S
a
0.1
0.01
CASE
1. PONDED SLURRY,
UNTREATED
2. PONDED SLURRY.
TREATED
3. PONDED CAKE.
UNTREATED
4. DEWATERED AND
PONDED. TREATED
5. LANDFILL. TREATED
SOIL
PERMEABILITY
COEFFICIENT
10'* cm/sec
Figure 7. • Comparison of mass release of sludge
constituents for various disposal
alternatives showing benefit of
chemical treatment
38
-------
an order of magnitude improvement in impermeability is realized compared
to untreated sludges, which typically have a co .fficient of about 10"4.
The systematic reduction of standing water is illustrated in
Cases 3 through 5, wherein the recharge rate is reduced compared to a
ponded slurry. If only unevaporated rainfall is allowed to recharge, the
mass release into the subsoil is reduced by a factor of about 5 (Case 3
versus 1, and Case 4 versus 2).
The significance of eliminating standing water by runoff is
shown. If it is assumed that 10 percent of the net rainfall is recharged,
then a one order of magnitude reduction is achieved relative to a dewatered
and ponded treated waste (Case 5 versus 4), and trvo to three orders of
improvement depending on soil permeability when comparing the mass
seepage from a chemically treated site to that of a ponded untreated site
(Case 5 versus 1). In addition, compacting the treated materials during the
site filling may reduce crack formation so that an effective coefficient of
permeability better than 10~5 cm/sec may be realized.
Case 2 (Figure 7) considers chemically treated sludge dis-
posal in such a manner that a hydraulic head exists on the site at all times.
The mass release for such a. case is approximately 1/2 to i/iO of that
from an untreated slurry pond, depending on subsoil permeability. Sites
of this type may seep to an adjacent stream or water table, which may
reduce the concentration of constituents because of mixing. Historical data
regarding the stream and water table characteristics and quality, as well
as monitoring, may be mandatory for sites of this type to assure environ-
mental acceptability.
4.6.4 Gypsum
4.6.4.1 Physical Properties
Measurements were made on samples obtained from the EPA
Research Triangle Park (RTP) forced oxidation pilot unit. These measure-
ments were of permeability coefficient, void fraction, water retention,
density, unconfined compressive strength of fly-ash-free first-stage scrubber
(gypsum), second-stage solids (calcium sulfite), and filter solids (gypsum
mixed with 5 to 10 percent calcium sulfite). The effects of fly ash on the
properties of the various materials were also determined:
a. Permeability coefficients for the predominantly gypsum and
predominantly calcium sulfite were virtually the same,
approximately 1 x 10~4 cm/sec.
b. The pore volume fractions of gypsum solids without and with
fly ash range from 0. 55 to 0. 5, respectively. Wet bulk
densities were approximately 1.3 and 1.4 g/cm3, respectively.
39
-------
c. The first-stage solids (gypsum) showed substantially higher
unconfined compressive strength (60 psi) than the second-
stage solids (calcium sulfite). When the gypsum contained
5 to 10 percent calcium sulfite, the unconfined strength was
approximately that of the sulfite (15 psi). The presence of
fly ash did not appreciably affect the unconfined compressive
strength.
d. Gypsum filtered to approximately 85 percent solids.
Field evaluations of gypsum disposal are being conducted at
the Shawnee disposal evaluation site. When completed, results will be
discussed in a separate report on the Shawnee work.
4. 6. 4. 2 Leaching Characteristics
The leachate from the filtered solids and solids from first-
stage slurries after 2 to 3 PVD of water had passed through the samples was
saturated with calcium and sulfate ions. This was typical for all samples
except for leachates from the two second-stage slurry solids at 10 and 22 PVD,
whereby the gypsum supply had been essentially depleted. At that point,
these two were saturated with calcium sulfite. Thereafter, the calcium was
supplied by the dissolution of hydrated calcium sulfite, although the anion
found in the leachate was sulfate because of the rapid oxidation of sulfite ions.
During leaching, the soluble magnesium and calcium chlorides
were washed out of the solids with the residual liquor, leaving the crystalline
gypsum to saturate the leaching water. The ionic strength decreased as the
chlorides were removed, thus reducing the solubility of gypsum, which
leveled out at the same saturation value of about 2200 mg/f.
4. 6. 5 Alternative Disposal Method for Untreated Wastes
The data obtained to date on the concentration of dissolved
solids in the leachates of untreated sludge show that peak levels are reached
which are virtually the same as those of the input sludge liquor. These con-
centration levels have been found to exceed drinking water criteria; there-
fore, some form of control is needed. In addition to the use of impermeable
liners or soils, one method being investigated is the use of an underdrain
system whereby the underdrained water is returned to the scrubber loop for
reuse. An underdrain, vented to the atmosphere, minimizes seepage by
eliminating the hydraulic head of the leachate. The gravity head of any
accumulated surface water is adequate to provide for rapid removal of rain-
fall recharge. This technique has the potential for increasing the bearing
strength of the sludge to levels useful in a landfill. This disposal operation
has been investigated in laboratory tests and is currently being evaluated
in the field for lime and limestone sludges and gypsum (Ref. 4). As of this
writing, each has supported construction equipment, i. e., tractors and
front-end loaders.
40
-------
A water balance analysis (Section IX) has shown that this mode
of closed-loop operation (with underdrainage water return) is possible, in-
cluding the maintenance of tolerable chloride ion levels and minimum impact
of fresh water makeup.
4.7 DISPOSAL COST ESTIMATES
Cost estimates for ponding and chemical treatment/landfilling
have been made and reported by Aerospace on several occasions. During
studies associated with the EPA Shawnee field disposal evaluation project,
Aerospace cost estimates were made of chemical treatment disposal and
were reported in the initial report on that study (Ref. 3). The Aerospace
estimates for lined-pond costs were presented in the initial and second
progress reports on sludge disposal (Refs. 1 and 2) and at the EPA flue gas
desulfurization symposia (Refs. 8, 9f and 10). All estimates were updated
(in a report to EPA on new source performance standards) to a July 1977
basis, and are summarized herein. A summary of these cost estimates is
presented in Table 9, and a discussion of the study is given in Section X.
41
-------
TABLE 9. DISPOSAL COST COMPARISON*
Cost Basis,
Mid- 19 77 $
Mills /kWh
$/ton
sludge (dry)
S/ton
coal
Ponding
Indigenous
Clay
0.55
4.90
1.50
Liner
Added
0.80
7.25
2.20
Landfill
Chemical
Treatment
1.05
9.70
2.95
Gypsum
1.10
10.30m
3.10
Dollar base:
Plant characteristics:
c. Coal burned:
d. Annual average
operating hours:
e. Plant and disposal site
lifetime:
f. SO 2 removal, with
limestone absorbent
g. Limestone utilization
h. Sludge generated:
i. Average annual
capital charges,
30-yr average:
j. Cost of land used for
disposal:
k. Land depreciation:
1. Disposal site:
m. Cost of forced oxidation
and disposal of gypsum
shown as cost/ton of
equivalent quantity of
nonoxidized sludge.
Divide by 1. 08 to con-
vert to gypsum cost.
July 1977
1000 MWe, 8700 Btu/kWh
(0.73-lb coal/kWh)
3. 5% sulfur, 12, 000 Btu/lb, 14% ash
4380 hr/yr (30-yr avg)
30 yr
90%
80% all cases except for gypsum, which
is 100%
4.8 X 105 short tons/yr untreated waste
(dry) including ash
18% of total capital investment
S5000/acre, all land assumed purchased
initially, sludge depth, 30 ft
Total depreciation in 30 yr, straight line
basis
Within one mile of plant
Includes fly ash; disposal is in a clay-
lined pond
42
-------
SECTION V
DISPOSAL CRITERIA
The principal objectives of this study are to evaluate and
determine environmentally acceptable methods for disposal of FGD sludges,
emphasizing landfilling and ponding. Throughout the study, in the absence
of appropriate federal or other criteria relating to the disposal of these
materials, drinking water criteria have been used as a basis for establishing
requirements for disposal. Because the trace element and salt content of
most samples analyzed exceeded the drinking water criteria, at least for
some of the constituents, the approach taken in this study has been to dispose
of the flue gas desulfurization (FGD) sludges such that no direct discharge
to any water supply would be allowed and that any seepage or runoff would be
minimized or, perhaps, totally eliminated. Various methods of disposal
have been studied and evaluated in the field (Refs. 3 and 4), and are discussed
in Section IX.
During the period of performance of the final phase of this
study, the Resource Conservation and Recovery Act (RCRA), Public Law
94-580, was enacted (October 1976). Under the provisions of this act the
U.S. Environmental Protection Agency (EPA) has been directed to provide
regulations and guidance to control the disposal of hazardous and nonhazardous
wastes including solid waste, which includes FGD sludges (air pollution
control sludges). Under the provisions of the RCRA, the EPA is to develop
a definition of hazardous wastes. Under various criteria, toxicity will
undoubtedly be included and could pertain to FGD sludges. If these sludges
are defined to be hazardous, the generation, transportation, and disposal
would be controlled by EPA regulations as a minimum. If die sludges are
determined to be nonhazardous, disposal would be controlled under the
provisions of the RCRA land disposal program. EPA would produce disposal
guidelines, but the responsibility for disposal would be that of the states.
At this time the "hazardous" and "nonhazardous" definitions
are still being formulated, and formal nonhazardous disposal guidelines have
not been promulgated. This section, therefore, discusses the environmental
acceptability of FGD sludge disposal in terms of preventing or minimizing
the pollution of groundwaters or surface streams, using the best practical
technology available today. Criteria discussions given in the second progress
43
-------
report on this study (Ref. 1) are repeated as appropriate and updated to
incorporate progress during the final phase.
A comparison of the chemical constituents of sludge liquors
in a discharge stream with the National Interim Primary Drinking Water
Regulation (NIPDWR) is given in Table 10 as a ratio of constituent concentra-
tion to water criteria. These ratios are given for the ranges of constituents
from all samples analyzed (see Table 11), and for six more recent samples
whose values can be seen in Section 7 for those analyzed in the final phase
of this study, and in Reference 1 for previous analyses. It should be noted
that the values used in this comparative analysis represent the initial
release of these constituents from the base of an untreated disposal site, and
that continued leaching of a given sample would result in a reduction of con-
centrations.
In Table 11 the ratios for concentrations in the "Range of All
Samples" column show that all elements analyzed plus the total dissolved
solids (TDS) and pH exceed drinking water criteria. However, in observing
ratios for the 10 independent samples shown in the table it can be seen that
except for selenium in two samples and cadmium in one, no trace element
exceeds the criteria by a factor greater than 10. (Water criteria for
barium, nitrate, and silver are 1, 10, and 0.05 mg/i , respectively. Limited
field evaluation leachate data show maximum concentrations for these
elements to be about 5, 1, and 0.5 times the criteria, respectively.) The
TDS are high for most of the samples, and the pH is excessive for two of
the samples. Although trace elements are not eliminated as a matter of
concern by these data, there are indications that in many cases the concen-
trations are quite low and that the item for concern may generally be the
concentration of dissolved solids in some cases, pH, and possibly chemical
oxygen demand (COD), as noted in Section 5. 2. 1. It is difficult at this point
in consideration of potential experimental error, the depletion of the
material with leaching time, cation exchange and adsorption in the soil, and
dilution between the disposal site and the consumer tap to specifically
quantify the degree of pollution potential from the disposal of these sludges.
Because of the comparatively large concentration of dissolved solids and
the identification of random values of high concentrations of trace elements,
methods to dispose of these materials while preventing their access to the
public water supply have been assessed. These include chemical treatment
and landfilling, disposal in impermeable ponds, and disposal in basins
equipped with under drainage. These methods are discussed in the following
paragraphs with regard to preventing pollution of public water supplies.
5. 1 DISPOSAL ALTERNATIVES FOR FGD SLUDGES
The general categories of disposal and the considerations
required for environmental control are shown in Table 12. In each case,
seepage of rainwater through the waste and eventual contamination of
groundwater pose an environmental concern for all disposal methods. Runoff
is a potential source of environmental pollution for landfill sites because, by
44
-------
TABLE 10. COMPARISON OF SLUDGE LIQUOR WITH DRINKING WATER CRITERIA
NIPDWR
Drinking
Water
Criteria,
mg/l
As 0.05
Cd 0.01
Cr 0.05
Pb 0. 05
Hg 0.002
Se 0.01
F ~2
TDS 500
pH (actual
values)*3
Concentration -=- Criteria (Nondimensional)
Range of
All
Samples
<0.8 - Z.8
0.4 - 11
0. Z2 - 5
0.2 - 6.6
0.03 - 2.5
0.28 - 20
< 0. 5 - 5
6.6 -48.5
6.7 -12.2
Sample
A
0.6
5.0
5.0
0.8
2.5
10.0
- -
36
6.7
B
0.4
1.2
0.3
3.0
_ _
3. 3
0.5
6.6
6.8
C
2.0
0.4
1.8
4.6
* .
10.0
3.3
30.0
8.0
D
0.04
_ .
_ _
<0.2
<0. 1
4.2
- .
13.4
12.2
E
0.4
11
0.6
6.6
< 0.5
< 2
1.7
18.8
8.7
F
1.2
1.3
0.2
0.2
< 0.001
7.8
1
20.5
8.0
G
2.8
- -
— .
< 0.2
< 0. 1
20
_ _
28
7.8
H
0. 1
__
_ _
< 0.2
< 0. 1
14
_ _
18.4
7. 3
I
0.8
5
_ _
0.8
0. 1
2.8
5
8.4
10.7
J
0.2
2.5
1. 1
< 0. 1
0.03
0. 3
< 0.5
48.5
8.9
in
Sample data are as follows:
Sample Station
A Mohave
B Cholla
C Shawnee
D Shawnee
E Shawnee
F Shawnee
G Shawnee
H Shawnee
I Duquesne Phillips
J LG&E Paddy's Run
EPA-proposed secondary regulation
Forced-oxidized to gypsum.
Absorbent % Ash
Limestone 3
Limestone 59
Limestone 40
Limestone 6
Lime 40
Lime 6
Lime 6
Limestone 6
Lime 60
Carbide lime 12
is 6. 5 to 8. 5.
Sampling Date
Mar 1973
Nov 1974
Jun 1974
Jan 1977
Jun 1974
Sep 1976
Oct 1976
Aug 1977
Jun 1974
Jul 1976
-------
TABLE 11.
RANGE OF CONCENTRATIONS OF
CHEMICAL CONSTITUENTS IN FGD
SLUDGE LIQUORS THROUGHOUT
THE SCRUBBER LOOPa
Scrubber
Constituents
Sludge Liquors,
mg/t
Aluminum
Arsenic
Beryllium
Cadmium
Calcium
Chromium
Copper
Lead
Magnesium
Mercury
Potassium
Selenium
Sodium
Zinc
Chloride
Fluoride
Sulfate
Sulfite
TDS
pH
COD
0.03 to 2.0
0.004 to 1.8
0.001 to 0. 18
0.004 to 0. 11
10 to 2600
0.011 to 0. 5
0.002 to 0. 56
0.005 to 0. 52
0. 1 to 2750
0.00005 to 0.07
5. 9 to 760
0.0006 to 2.7
10.0 to 29,000
0.001 to 0.59
420 to 33, 000
0. 6 to 58
600 to 84, 000
0.9 to 3500
2800 to 162, 700
4.3 to 12.7
40 to 140
All test conditions not necessarily
standard. See Table 10 for repre-
sentative list.
46
-------
TABLE 12. ENVIRONMENTAL EFFECTS OF DISPOSAL ALTERNATIVES
Type of
Disposal
Pond
Basin
Landfill
Condition
of Waste
Untreateda
or
chemically
treatedb
Untreated2*
or
conditioned0
Conditioned0
or
chemically
treatedb
Primary
Drainage
Supe rnate
Supe rnate
Underdrainage
Runoff
Environmental Effect
Seepage
Yes
Yes
Yes
Yes
Runoff
No
No
No
Yes
Land Reu£ e
No
Yes
Yes
Yes
Untreated waste refers to FGD sludges as emitted from primary or secondary
dewatering equipment.
Chemically treated sludges refer to the waste treated by one of several com-
mercial processes that make these wastes suitable for landfill disposal.
°Conditioned waste refers to sludge treated by techniques other than chemical
treatment 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.
-------
definition, these sites are open and do not necessarily return 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.
5.2 ENVIRONMENTAL EFFECTS OF FGD WASTE
DISPOSAL ALTERNATIVES
5.2.1 Rainwater Seepage
One of the difficulties in applying existing water quality
standards to FGD waste waters is that they apply to the quality of the ground-
water or tap water and do not directly define the quality of seepage waters
from a FGD waste disposal site. Nevertheless, possible approaches,
either individually or in combination to control seepage, 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 concentrations of leachate from untreated sludges are given
in Table 11. Reduced concentrations as a result of pore volume displacement
(PVD) can be seen in Table 13, which gives data for untreated and chemically
treated sludges but not necessarily those analyzed for TablelJ. Nonetheless,
reductions as a result of seepage are evident. The conditions represented
in Table 13, however, apply only to cases where fresh sludge is not con-
tinually added to the disposal site.
Mass loading of constituents is determined for site specific
conditions (Section IX), with rainfall quantities, material and soil permeability
coefficients, and method of disposal being the more significant factors.
48
-------
TABLE 13. CHEMICAL CONSTITUENTS IN LEACHATE FROM
UNTREATED AND CHEMICALLY TREATED TVA
SHAWNEE FGD WASTE AFTER 1 AND 50 PVDa
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
1 st PVD
0.02
0.01
0.015
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
1 st 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
Concentrations in mg/i , as appropriate.
49
-------
Therefore, for example, if mixing of seepage with ground waters or
surface waters is considered, the sludge mass and constituent concentrations,
as well as the site specific conditions, would be necessary for determination
of the environmental impact.
In Table 11, it is shown ll-.at values of COD in fresh sludge
ranged between 40 and 140 mg/l . Because of the rapid oxidation characteristic
of sulfite sludge, the COD after 1 PVD was 10 mg/l or less and after 50 PVD
was 6 mg/I or less. All soils are not aerobic; therefore, COD could be
considered a critical parameter. The potential ecological impact of high COD
values was not determined.
5. 2. 2 Landfill Runoff
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, only regulations
pertaining to seepage are applicable. For other cases where runoff is
directed to receiving streams, the most applicable regulations are state
stream standards. Thus, where the receiving stream size is small relative
to water supplied in the stream by landfill runoff, 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 is 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.
d. Divert runoff to flatland where evaporation and seepage
can take place.
e. Dilute landfill runoff with runoff from other sources
before emitting to streams.
Selection of any of these alternatives is dependent upon
site-specific considerations.
5.2.3 Land Reuse
In addition to chemical contamination prevention, environ-
mental concern also includes the attainment of certain structural qualities if
the site is to be reclaimed. In the absence of specific criteria, a conserva-
tive value of 1.8 kg/cm2 (25 psi) minimum unconfined compressive strength
50
-------
was considered because it could be met on the basis of the following
observations (a minimum allowable value was not determined):
a. Heavy equipment on tires carrying 45 psi and driven re-
peatedly onto a saturated, chemically treated pond at
the TVA Shawnee field evaluation site (Refs. 3 and 4)
has maintained traction and left only a mild track in
the material. This material has an unconfined com-
pressive strength of 27 psi in a wet condition.
b. All chemically treated wastes studied attained and usually
exceeded this strength.
c. The standard test for unconfined compressive strength is
simple and inexpensive.
Studies are continuing in the Shawnee field disposal evaluation project
(Refs. 3 and 4) to relate disposal site strength to confined 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 treatment of the material and placement in an impermeable basin
would certainly provide adequate strength and would not allow pollution of
groundwaters. Moreover, a lined siltation pond can assure environmental
protection from any runoff waters. However, this solution may be overly
conservative and incur costs beyond that which may be necessary.
Demonstrations, both small-scale and operational, of treated
and conditioned 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 hydraulic 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 seepage with groundwater, a nearby stream, and a connecting
river. All sites are monitored for water quality.
It is apparent that each disposal site and the material placed
on 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 dis-
posal method chosen for any site will generally be selected on site-specific
conditions. Because of this, the establishment of a single criterion for all
cases may be impractical.
51
-------
At this time, studies are continuing in the Shawnee field
disposal evaluation project, other EPA projects, and in industry to define
and evaluate environmentally sound, least-cost methods of waste disposal,
covering the range of ponding, dewatering and mixing with fly ash, ponding
with underdrainage, conversion to gypsum and dewatering, chemical
treatment, ocean disposal, and mine disposal. Data from these studies
are expected to provide a base from which an appropriate disposal mode
may be selected for any given site.
A detailed discussion of disposal and environmental relation-
ships and the methodology of alternative disposal techniques is given in
Section IX.
52
-------
SECTION VI
PHYSICAL PROPERTIES DETERMINATION
6.1 BACKGROUND
The disposal, handling and transportation techniques applied
to flue gas desulfurization (FGD) wastes are strongly dependent upon the
physical behavior of these wastes and the resultant costs. The predominant
physical properties that limit or restrict the applicability of most of these
techniques include the degree of solids content; the crystalline phase com-
position; and the particle size, shape, and distribution. Experimental tests
were conducted to determine the physical characteristics of the FGD wastes
from power plant scrubbing facilities that supported this program. These
tests were directed toward sludge usage in landfilling and land reclamation
applications.
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 water contents,
permeability as a function of volume fraction solids (converse of void frac-
tion), and bearing capacity as a function of solids content. Additionally,
crystalline structure was observed, using a scanning electron microscope
(SEM), and correlations were made between these observations and the
physical behavior of the respective samples.
The following subsections describe the tests conducted and
results of the tests for each of the FGD wastes investigated. Wherever pos-
sible, these tests results are discussed in terms of the physical characteris-
tics believed to be responsible for the observed behavior.
In addition to FGD waste disposal in its normal state, the option
usually exists to collect or not to collect fly ash simultaneously with the sul-
fur reaction products. Although this option may appear to increase the unit
disposal cost of sludge and fly ash, the combination or recombination of these
materials may improve their environmental acceptability by improving the
physical performance of the waste for a particular end use and, thus may
decrease the overall disposal cost because of the change in physical properties.
Selected tests were conducted on FGD wastes to which fly ash was recombined
to determine both their environmental acceptability and their physical behavior
for structural and landfilling applications.
53
-------
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: hydrated calcium sulfite, cal-
cium sulfate (gypsum), fly ash, and unreacted limestone or lime (usually,
appearing as limestone by reaction with atmospheric CC<2). These solid
phases exist as fine particulates suspended in an aqueous liquor which is
usually saturated with ions of these solids. In addition, an unpredictable
amount of calcium chloride or sodium chloride is also present as totally dis-
solved salts (TDS). A complete chemical characterization of each of the FGD
waste materials studied is given in Section VII.
The relative amounts of each of the solid crystalline phases
are dependent upon system design parameters and include the sulfur content
of the coal and the efficiency of scrubbing, the fly ash in the flue gas passing
through the scrubber and the efficiency of the system to remove fly ash, the
stoichiometric ratio of reactants added relative to sulfur content and the re-
actant utilization efficiency, and the amount of oxidation of the sulfur products
that takes place in the system. The efficiency of scrubbing sulfur 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 that include combustion behavior in the boiler, the nature of min-
eral phases in the coal and its total ash content, the natural reactivity of
lime or limestone, and the particle size and distribution of the sulfite particles.
The number of independent and interdependent variables that affect the com-
position 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.
Furthermore, each crystalline phase with its specific characteristics has
an influence on the behavior of the sludge. Characterization of the solid por-
tion of each sludge relative to phase identification and particle characteristics
is discussed in Section VII and 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.
6.3 WET BULK DENSITIES
The method of sludge dewatering is a critical variable that
affects disposal costs in most cases and determines the selection of the dis-
posal methods in some cases. Measurements of the wet bulk densities of
several sludges dewatered by various alternative methods have been previously
reported (Refs. 1 and 2) and have demonstrated the greater dewatering capa-
bility of filtering or centrifugation relative to settling or free drainage. In
most sludges, there was very little difference in the wet density of sludges
dewatered by settling as compared to free drainage. In almost every sludge,
the particle packing density of these two methods was nearly identical as
determined by the maximum wet density or by dry bulk density. The slight
improvement in wet bulk density of the drained sludge was credited to the
lower water retention made possible by the free draining condition.
54
-------
In a similar manner, it was shown that the difference in density
between filtering and centrifugation waTu^ual]y~quite~smlLfi (Refs. 1 and 2).
It was pointed out that, although filtering was usually slightly superior for de-
watering than centrifugation when using laboratory equipment, this difference
is not always apparent when using industrial equipment. The water retentive
capabilities of all sludges was shown to be so great that in only one sludge
dewatered by filtration was maximum dewatering observed. Generally,
sludges containing coarser particle size distributions dewatered more effec-
tively by all methods relative to sludges with finer particle size distributions.
In the most recent series of experiments, dewatering measure-
ments were made on five samples originating from three scrubbing facilities:
(1) a Tennessee Valley Authority (TVA) Shawnee lime sludge from the venturi-
spray tower scrubber collected without fly ash; (2) a sludge identical to (1)
to which was added fly ash to 40 wt% of the total solids; (3) a double alkali
sludge from the Gulf Power Company (GPC) Plant Scholz collected without
fly ash; (4) a double alkali sludge identical to (3) but collected with 30 wt%
fly ash; and (5) a lime sludge without fly ash from the Louisville Gas and
Electric (LG&E) Paddy's Run station scrubber using a carbide lime sludge
absorbent.
6.3.1 Experimental Procedure
The wet bulk density was determined on FGD sludge samples
after each was dewatered by settling, settling with free drainage, vacuum
filtration, and centrifugation. For each dewatering method, the wet bulk
density was determined by measuring the weight and volume of the dewatered
sample. Subsequently, these samples were dried to constant weight, from
which 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
operates on the principle of Boyle's law in which the volume of air displaced
by the sample is accurately measured.
6.3.2 Bulk Density Test Results
The wet bulk density of the five FGD sludges was calculated as
a function of solids content from the equation
pa pw
pa(l -
where p« is the wet bulk density, ps and pw are the true density of the solids
and water, respectively, and f is the weight fraction solids content. In all
previously reported measurements, the measured wet bulk density precisely
obeyed this relationship. The relationship between wet bulk density and solids
55
-------
content for each sludge and the values of wet bulk density resulting from
dew ate ring by each method are indicated in Figures 8 through 12.
These figures show that the most effective dewatering method
was filtration, and the least effective method was settling for all sludges.
These results are in agreement with previously reported measurements.
However, in this test series, the centrifuged sludge did not dewater as effec-
tively as the sludge previously reported. The most probable reason for this
difference in behavior was the fine particle size of sludge particles in this
test series, which do not respond as well to centrifugal forces. A summary of
wet bulk density as a function of dewatering method for these five samples
and eight samples previously tested is presented in Table 14.
A significant result was observed in the freely drained Shawnee
lime sludge to which fly ash was added. The solids content was adjusted to
approximately 25 percent to simulate clarifier underflow for this sludge.
When poured into the test container, the coarser fly ash particles (typically
50 fun in diameter) settled more rapidly than the sulfur-phase particles typi-
cally 20 fjun in diameter) and formed a fly ash layer on the filter paper used
as a retainer. Supernate water then drained through as in all other cases;
but the presence of a fly ash layer beneath the sludge appeared to aid dewater-
ing. Because water retention in fly ash is relatively low, all the supernate
water passed through this fly ash layer and created air voids However,
since the fly ash is contiguous with the sulfur phase, surface tension forces
between the fly ash particles and water are capable of overcoming the surface
tension between sulfur phase particles and water. Thus, much of this water
is removed from the sludge, and it too is passed through the fly ash layer.
The net consequence of this action is that more effective dewatering takes
place, which is nearly as effective as filtration. Results from other bench-
scale tests and several ponds at the Shawnee Field Evaluation project where
underdrainage is employed were a consequence of this phenomena; these data
are presented in the report on that study (Ref. 4).
6.4 VISCOSITY
The ability to transport sludge either by pumping through pipe-
lines or by truck is dependent upon the ease (or resistance) of the sludge to
flow as a function of retained water. In a previous report (Ref. 1), it was
shown that pumpable mixtures (<20 poise) range from a high solids content of
70 percent to a low solids content of 32 percent. Conversely, sludges having
sufficient resistance to flow for truck transport range as low as 45 percent
solids to a high of 80 percent solids. An evaluation of the previously reported
data indicated that the presence of fly ash contributed strongly to the fluidity
of the sludge. Additionally, at equivalent water contents, fine-particle sludge
had high viscosity whereas coarse sludge had low viscosity. There also ap-
peared to be a subtle effect related to particle morphology, but this effect
could not be separated from the particle size effect in the data.
6.4.1 Experimental Procedure
The viscosity measurements were performed at room tempera-
ture using a S & A VT-02 viscometer having a cylindrical sleeve immersed
56
-------
CJ
CJ
1.8
1.7
1.6
1.5
1.4
CO
1.3
1.2
1.1
1.0
T
T
TVA SHAWNEE LIME
RUN F, 0% FLY ASH
FILTERED
DRAINED
CENTRIFUGED
SETTLED
I
20 40 60 80
SOLIDS CONTENT, WEIGHT %
100
Figure 8. Bulk density as a function of solids content:
TVA Shawnee lime sludge without fly ash
57
-------
u
u
CD
1.8
1.7
1.6
1.5
1.4
1.3
1.2
1.1
1.0
TVA SHAWNEE LIME
40% FLY ASH
FILTERED
DRAINED
CENTRIFUGED
SETTLED
1
I
I
20 40 60 80
SOLIDS CONTENT, WEIGHT %
100
Figure 9. Bulk density as a function of solids content:
TVA Shawnee lime sludge with 40-percent
fly ash
58
-------
LLJ
0
GO
1.8
1.7
1.6
1.5
1.4
1.3
1.2
1.1
1.0
I I
GPC SCHOLZ DOUBLE ALKALI
0% FLY ASH
FILTERED
DRAINED
fCENTRIFUGED
SETTLED
I [
20 40 60
SOLIDS CONTENT, WEIGHT %
80 100
Figure 10. Bulk density as a function of solids content: GPC
Plant Scholz double-alkali sludge without fly ash
59
-------
o
,«
en
CD
1.8
1.7
1.6
1.5
1.4
1.3
1.2
1.1
1.0
I I I
GPC SCHOL2 DOUBLE ALKALI
30% FLY ASH
FILTERED
CENTRIFUGED
DRAINED
SETTLED
I
20 40 60 89
SOLIDS CONTENT, WEIGHT %
100
Figure 11. Bulk density as a function of solids content: GPC
Plant Scholz double-alkali sludge with 30-percent
fly ash
60
-------
u
u
UJ
a
CD
1.8
1.7
1.6
1.5
1.4
1.3
1.2
1.1
1.0
LG&E PADDY'S RUN
12% FLY ASH
FILTERED
CENTRIFUGED
DRAINED
SETTLED
I
20 40 60 80
SOLIDS CONTENT, WEIGHT %
100
Figure 12. Bulk density as a function of solids content:
LG&E Paddy's Run carbide lime sulfur sludge
with 12-percent fly ash
61
-------
TABLE 14. SAMPLE SOURCES AND DENSITIES
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
Duquetne
Lime,
6/17/74
Cholla
Limestone,
9/1/74
Mohave
Limestone,
3/30/73
Shawnee
Lime,
0% Fly Ash
Shawnee
Lime,
40% Fly Ash
Scholz
Double Alkali,
0% Fly Ash
Scholz
Double Alkali,
30% Fly Ash
LGfcE Carbide
Lime,
12.4% Fly Ash
Dewatering 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
1.37
1.35
1.35
1.36
1.23
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
1.42
1. 50
1.44
1.38
1.27
Centrifuge
Percent
Solids
59.8
63.3
49.9
50.9
62.2
57.2
60.9
77.0
-
—
—
—
• *<7 "V -
Density,
g/cc
1.56
1.60
1.44
1.43
1.62
1.52
1.58
1.86
1.39
1.44
1.38
1.40
1.31
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
1.49
1. 54
1.61
1. 55
1.48
62
-------
in the FGD fluid waste and rotating at 64 rpm. The 2.4-cm (15/16-in.)
diameter 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.) At values
of viscosity greater than 120 poise (which is beyond the limit of interest), an
anomalous decrease in measured viscosity of the sludge was observed, which
was interpreted as a separation of liquid and particles at the interface with
the rotating sleeve.
Prior to viscosity measurement, the solids content of the
sludge was determined, and the mixture was homogenized by hand stir-
ring. The cylindrical sleeve was then immersed in the sludge and left un-
disturbed for 15 sec, at which time the stirring motion ceased. The viscom-
eter was then turned on, and the viscosity was noted.
Viscosity measurements were made on six samples originating
from four scrubbing facilities: (1) a Shawnee lime sludge from the venturi-
spray tower scrubber collected without fly ash, (2) the identical sludge to
which was added fly ash to 40 percent of the total solids, (3) a Shawnee lime-
stone sludge from the turbulent contact absorber (TCA) scrubber collected
without fly ash, (4) a double alkali sludge from the Plant Scholz collected
without fly ash, (5) the same double alkali sludge collected with fly ash, and
(6) a lime sludge from the LG&E Paddy's Run station scrubber using carbide
lime sludge absorbent. The viscosity data from these sludges are presented
in Figure 13, superimposed on the data reported previously for ease of
comparison.
The Shawnee lime and Shawnee limestone sludges without fly
ash have nearly identical viscosity relationship, both being at lower solids
content than the sludges with fly ash. When fly ash was added to the Shawnee
lime sludge, the resulting viscosity displaced to a higher solids content as
expected, but the resulting curve was not coincident with the previously mea-
sured Shawnee lime sludge having identical fly ash content. The difference
between the two lime sludges with fly ash was a much larger sulfite rosette
size of the more recently measured sludge. The displacement was exactly
parallel at 3 percent greater solids content in the larger sized sludge.
The Paddy's Run station sludge sample was a fine-grained
material, having approximately 12 percent fly ash, and behaved similarly to
the previously reported double alkali sludges. This sludge had a rapidly
rising viscosity at relatively low solids content; this behavior is believed to
be caused by the small particle size of the sludge, similar to the double al-
kali sludges.
The Plant Scholz sludge was also a double alkali sludge and
also fine-grained, but it did not behave like the other double alkali sludges.
The viscosity curves for both Scholz sludges (one with and the other without
fly ash) have relatively high solids contents similar to the lime /lime stone
sludges. The sludge with fly ash showed no significant difference from the
sludge without fly ash. From an examination of the morphology of the sludge
63
-------
CURVE
1
2
3
4
5
6
7
8
9
10
11
12
13
SLUDGE FLY
GM PARMA DOUBLE ALKALI
UPL GADSBY DOUBLE ALKALI
TV A SHAWNEE LIME
DLC PHILLIPS LIME
TV A SHAWNEE LIMESTONE
TV A SHAWNEE LIMESTONE
TV A SHAWNEE LIMESTONE
LG&E PADDY'S RUN CARBIDE LIME
TV A SHAWNEE LIME
TV A SHAWNEE LIMESTONE
GPC SCHOLZ SODA ASH DOUBLE ALKALI
GPS SCHOLZ SODA ASH DOUBLE ALKALI
TV A SHAWNEE LIME
ASH. *
7.4
8.6
40.5
59.7
20.1
40.1
40.9
12.4
<1.0
<1.0
<1.0
30.0
40.0
DATE
7/18/74
8/9/74
3/19/74
6/17/74
2/1/73
6/15/74
7/11/73
7/76
9/8/76
9/28/76
6/20/76
6/27/76
9/8/76
120
100
2
2
80
8
5 .«
40
20
30
40 50 60
SOLIDS CONTENT, WEIGHT %
Figure 13. Viscosity of FGD Sludges
64
-------
(Appendix A), it is seen that the fine-grained sludge is strongly agglomerated
such that the resulting rheological properties are dependent upon the coarse-
ness of the agglomerates rather than the fineness of the individual particles.
When fly ash is present at the same solids content, there is a substitution of
these agglomerates by fly ash. The fly ash has generally fewer particle sizes
than the agglomerates and are nonspherical, typical of cyclone boiler ash.
Thus, the viscosity curve is not displaced to a higher solids content as might
be expected but appears to be displaced slightly to a lower solids content.
6.4.2 Experimental Results
The results from the viscosity data are consistent with those
previously obtained, but additional emphasis is placed on the role of particle
or agglomerate size in the rheological behavior. In Figure 13, the three
sludges that have relatively high viscosities with a low solids content are all
fine-grained and contain very little fly ash. The sludges that retain fluidity
until a very high solids content is reached are coarse-grained; the sludge with
the highest fluidity has coarse platey grain and a very high fly ash content.
Fine-grained sludges that agglomerate show a response much like that of
coarser sludges. It appears that platey morphology and fly ash both provide
fluidity to a sludge, on the basis of a comparison of the sludges from the
Shawnee Station.
The conclusion from these data indicates that particle size,
particle morphology, and fly ash all make a contribution to the rheological
response of a sludge. Although trends have been determined, it is not now
possible to accurately predict the viscosity of a sludge. However, with these
data, the estimation of a viscosity curve for a sludge for which a single vis-
cosity value has been determined should be satisfactory for estimating initial
operating parameters.
6.5 COEFFICIENT OF PERMEABILITY
The percolation of rainwater through FGD sludge and its ultimate
recharge to the water table constitutes a major method by which these sludges
can pollute the environment. Methods of disposal are available to prevent or
minimize the amount of rainwater that can pass through the sludge. However,
the degree of control and the complexity and cost thereof may depend appre-
ciably upon the self-control afforded by the sludge material itself. The mea-
surement of the coefficient of permeability provides a value from which the
amount of leachate from a disposal basin can be calculated under any given
set of environmental or disposal management conditions.
Permeability measurements of sludges previously reported
(Refs. 1 and 2) have shown that most sludges as they might appear in a dis-
posal basin (either by settling or as filter cake) have coefficients of perme-
ability in the approximate range of 2 X 1Q-4 to 5 X 10~5 cm/sec. A relation-
ship was found for the coefficient of permeability as a function of the volume
fraction of solids (complement of pore volume). The effect of a higher volume
65
-------
fraction of solids, either by compaction or consolidation of the sludge, was to
decrease the coefficient of permeability. The amount of decrease in perme-
ability and the effect of compaction or consolidation appeared to be a function
of the particle size of the sludge and the size and distribution of the fly ash
particles.
6.5.1 Experimental Procedure
In the previously reported work, several methods were used
to prepare a column for permeability measurements. It was found that the
method affected the value of the coefficient of permeability only to the extent
that each method would affect the packing density of particles (volume frac-
tion solids). Thus, the method that most accurately simulates a sludge settled
(or drained) in a pond was the method chosen for these tests. This method
uses a slurry of the sludge poured into a column and allowed to settle or drain
as it might in a disposal basin. Pore volume is calculated from the volume
of settled solids, the known weight of solids, and the true density of the solids.
(Filter cake has a pore volume similar to a settled and drained slurry.)
The permeation measurements were begun when all untreated
water dripped through the column. Deionized water was then added to a
known height above the top of the sludge, and the amount of water percolating
through the column was collected and measured as a function of time.
Additional columns with Shawnee lime, LG&E Paddy's Run,
and Gulf Scholz sludges were prepared in the same manner, but before de-
ionized water was added, the materials were consolidated under 30 or 100 psi
loads. A load of 30 psi is approximately the force experienced at the base of
a 40-ft disposal site, and the permeability of the sludge consolidated under
that force represents a value that may control the movement of leachate
through the waste. At a load of 100 psi, the consolidation of the sludge repre-
sents an assumption of lower limit to the permeability. The permeabil-
ity coefficients were determined by the method previously described.
6.5.2 Experimental Results
The results of the permeability measurements of untreated
sludges analyzed during this phase are shown in Table 15. Because consider-
able data had been generated in previous analyses of untreated sludges (Fig-
ure 14), permeability coefficients were determined for these materials after
compaction under 30- and 100-psi loading. In Figure 14, it can be seen that
most untreated sludges have permeability coefficients in the range of 10-3 to
10-4 cm/sec, but the compacted materials (Table 15) have coefficients of
approximately one order of magnitude lower, i.e., 10'4 to 10~5. Specific
cases are discussed below.
For the case of the TVA Shawnee lime sludge, the extrapola-
tion of the previous data to higher values of solids fraction is a reasonable
representation of the sludge consolidated at high pressures. These data
66
-------
TABLE 15. PERMEABILITY COEFFICIENTS OF
COMPACTED UNTREATED SLUDGE
Sludge
Source
TVA Shawnee Lime
(6% Fly Ash)
TVA Shawnee Lime
LG&E
Paddy's Run
(12% Fly Ash)
GPC Scholz
(Without Fly Ash)
GPC Scholz
(30% Fly Ash)
Sampling
Date
9/08/76
9/08/76
6/20/76
6/27/76
Fractional
Solids
Volume
0. 35
0.33
0.47
0.42
0.41
0.25
0.21
0. 37
0.34
0. 36
0. 34
Permeability
Coefficient
cm/sec.
8.9 X 10'J?
1.9 X 10
7. 3X 10'i?
1. 1 x lO'j?
1.4 x 10
8.4X lO'g
3.4 X 10"
1.4x 10'*
4.4 X 10
4. 8 X IQ'l
8. 3X 10
Consolidation
Pressure (psi)
100
30
100
30
30
100
30
100
30
100
30
67
-------
10
,-3
OT
-1-0*
PONDB
00
10
,-5
PONDE
1.6 x 10
PONDC
3x10
,-7
10
,-7
5.5 x 10
SAMPLE SOURCES DATE FLY ASH. %
• TVA SHAWNEE LIMESTONE 6/15/74
o TVA SHAWNEE LIME 3/15/74
a SCE MOHAVE LIMESTONE 3/30/73
• CM PARMA DOUBLE ALKALI 7/18/74
* UPL GADSBY DOUBLE ALKALI 8/9/74
v DLC PHILLIPS LIME 6/17/74
A APS CHOLLA LIMESTONE 4(1/74
X CHEMFIX LAB
0 DRAVO LAB
A (UCS LAB
© POND B TVA SHAWNEE/ DRAVO (Ref. 2)
$ POND C TVA SHAWNEE/ IUCS (Ref. 2)
$ POND E TVA SHAWNEE/ CHEMFIX (Ref. 2)
* PULVERIZED AND COMPACTED
rrz FAMILIES OF DATA HAVING SAME SLOPES
40.0
40.5
3.0
7.4
8.6
59.7
58.7
10
,-8
I
a 10
a 20
0.30
0.40 a 50
VOLUME FRACTION OF SOLIDS
a 60
0.70
a so
0.90
Figure 14. Permeability of chemically treated and untreated sludges
-------
indicate that an expected lower limit of permeability may be in the range of
10~5 cm/sec. The data for the Shawnee lime sludge to which 40-wt% fly ash
was added show permeabilities similar to the sludge without fly ash but at a
higher solids fraction. This is believed to be a consequence of less than com-
plete mixing of the sludge and fly ash.
The permeability coefficients for the compacted GPC Plant
Scholz sludges was approximately 10"^ cm/sec. The sludge without fly ash
had permeabilities about five times higher than the sludge with fly ash, but
the solids fraction was nearly identical for both materials. The explanation
is believed to be related to the agglomeration of the fine sludge particles,
which respond much like coarser materials; without fly ash, passage of water
through it is relatively rapid. When fly ash was present the solids fraction
did not change, as the fly ash, having a broad particle size distribution, fills
pore passages such that water passage rate through the waste is reduced.
The bimodal particle distribution between coarse sludge and fine fly ash in
the Arizona Public Service (APS) Cholla sludge was previously shown (Ref. 1)
to produce the same effect.
The LG&E sludge was a fine particle sulfite sludge that did not
consolidate well and did not reach a high solids fraction even under high loads.
Moreover, surface tension forces between fine particles and water is high,
and these forces must be overcome for water to pass through. Thus, the low
permeability shown for this sludge is believed to be a consequence of the
particle size of the sulfite phase.
The results from the measurement of permeability coefficients
indicate that sludges that are disposed of by either settling or by placement
of filter cake will have permeabilities generally between 5xlO~5to2X 10~4.
Generally, coarser particles settle to higher fraction solids and have higher
permeability coefficients (such as the Mohave sludge) than fine particle
sludges. When the presence of fly ash gives the waste a bimodal particle
distribution, such as the sludges from Cholla, the permeability tends to be
less than otherwise expected. Fine particle sludges do not settle or con-
solidate well, but because of high surface tension forces, their permeabilities
tend to be lower than those of coarse particle sludge with an equivalent solids
fraction content. Consideration of sludge equivalent to that occurring at the
base of a 40-ft high disposal site will decrease permeabilities to about 10~->
cm/sec.
6.6 BEARING STRENGTH
Preliminary measurements of the load-bearing strengths of
two FGD sludges were reported in Reference 1. Since that time, alternative
dewatering methods have been employed, and the method of testing has been
slightly modified. Recent measurements of load-bearing strengths as func-
tions of solids content have been made for a number of sludges. These results
complement the previous data and serve to reinforce the conclusions drawn
69
-------
from them, namely, that sludges may be dewatered to critical values of solids
content above which the load-bearing strengths increase steeply to values well
above the minimum for safe access of personnel and equipment.
6.6.1 Test Procedure
Load-bearing strengths were measured for sludges from six
power plants, using a method substantively similar to the standard bearing
ratio test for soils (Ref. 11). A 0.95-cm-diameter (3/8-in.) rod was forced
into the sludge at a constant rate of 0. 25 cm/min (0.1 in. /min) with an
Instron loading machine. Penetration loads were recorded continuously to a
maximum penetration depth of 2. 54 cm (1 in.). The cylindrical sample con-
tainer was sufficiently large, 7-cm (2. 7-in.) diameter by 10-cm (4-in.)
depth, that edge and bottom effects were negligible. Tests were also made
with 1. 27-cm (0. 5-in.) and 2.54-cm (1-in.) diameter rods which produced
the same load-penetration curves for all three rods and confirmed that rod
size was not a variable in these measurements. Samples of partially de-
watered sludge, which initially had negligible load strength, were placed in
the containers on a 1. 27-cm (0. 5-in.) deep layer of sand. The containers had
perforated bottoms to allow the sludge to dewater by means of underdrainage.
Load-penetration curves were obtained for each of four solids contents as the
samples dewatered. Solids contents were obtained by vacuum oven drying of
small specimens removed from the top of the sample immediately after each
test.
6.6.2 Test Results
The loads measured at the maximum penetration depth of
2.54 cm (1 in.) were plotted against the corresponding solids contents in Fig-
ure 15 for 11 sludge compositions. The lime process sludge from the TVA
Shawnee Station was tested as received, as well as after intermixing 40-wt%
fly ash (dry basis). The curves of Figure 15 show that similar strengths can
be expected from both samples, but at a proportionally higher solids content
for the sample containing fly ash. The ratio of solids contents of the two
samples required to produce a given load is approximately the ratio of bulk
densities of underdrained samples. Load strengths at 1-in. penetration in
excess of 7 kg/cm2 (100 Ib/in. 2) were obtained for ash-free samples with
solids content greater than 58 percent and for sludge containing 40-wt% fly
ash if the solids content was greater than 65 percent. Negligible strengths
were observed for ash-free samples with solids content of less than 51 percent
and for samples with 40-wt% ash if the solids content was less than 59 percent.
Double alkali process sludge from the Gulf Scholz plant was
essentially ash-free for one sample, while the other sample contained about
31 percent ash. The results of bearing tests of these two samples were quali-
tatively similar to the results for the TVA Shawnee lime samples, although
the separation of the two curves is not as great (Figure 15). However, the
data points for the ash-free samples were scattered so that quantitative com-
parisons are less reliable. It appears that load-bearing strengths in excess
of 7 kg/cm2 (100 Ib/in. 2) can be expected of ash-free sludge with solids
70
-------
250,—
a 150 -
o
to
o
GC.
•<
O
o
SHAWNEE, 6% FLY ASH - 9/8/76
SHAWNEE, 40% FLY ASH - 9/8/76
SCHOLZ, WITHOUT FLY ASH - 6/20/76
SCHOLZ, 30% FLY ASH - 6/27/76
PADDY'S RUN, 12% FLY ASH -
PHILLIPS. 60% FLY ASH - 6/17/74
CHOLLA. 59% FLY ASH -4/1/74
GADS BY, 9% FLY ASH -8/9/74
SHAWNEE, WITHOUT FLY ASH - 11/30/76
SHAWNEE, 40% FLY ASH - 11/30/76
RTP GYPSUM*, WITHOUT FLY ASH - 12/4/75
RTP GYPSUM*, 40% FLY ASH - 9/30/75
ABSORBENT
L - LIME
DA - DOUBLE ALKALI
CL - CARBIDE LIME
LS - LIMESTONE
* - CONTAINS 5%
SULFITE
60 70
SOLIDS CONTENT, weight %
80
90
Figure 15. Load-bearing strength as a function of moisture, fly ash,
and sludge origin
-------
content greater than about 55 percent and for sludge containing 30 percent ash
if the solids content is greater than about 56 percent. Negligible strengths
were observed for both samples if the solids contents were less than 48
percent.
Results of compaction strength tests on a sample of carbide
process sludge produced at the LG&E Paddy's Run station are also shown in
Figure 15. Dewatering of this sample was not continued beyond 52. 5 percent
solids, which corresponds to a load-bearing strength at one-inch penetration
of 4 kg/cm2 (57 Ib/in. 2). The curve for LG&E sludge appears to have the
same shape as the curve for the Scholz ash-free sludge (Figure 14). Negli-
gible strength was observed for samples with solids contents of less than
44 percent.
Sludge from the double alkali process was obtained from the
Utah Power and Light (UPL) Gadsby station. This sludge was tested for load-
bearing strength as it de watered by means of under drainage. Results of the
tests plotted in Figure 14 show that a load-bearing strength in excess of 7 kg/
cm2 (100 Ib/in. 2) was developed when the solids content became greater than
68 percent. Below 60 percent solids, negligible strength was observed.
The compaction strength versus solids content curve for lime
process sludge containing fly ash from the Duquesne Light Company (DLC)
Phillips station was similar in shape to other sludges previously described,
although it was displaced toward a higher solids content, as shown in Fig-
ure 14. Load-bearing strengths in excess of 7 kg/cm2 (100 Ib/in. 2) were
observed for solids contents greater than 70 percent, and negligible strengths
were found at solids contents below 65 percent. Although not plotted in Fig-
ure 14, a load-bearing strength of 33 kg/cm2 (470 Ib/in. 2) was measured
when the sample had dewatered to a solids content of 73 percent.
Limestone sludge from APS Cholla was tested for load-bearing
strength. This sludge had to be dewatered to the highest solids content of all
sludges tested before it developed significant strength. At a solids content
of less than 76 percent, the sludge showed negligible strength. Load-bearing
strengths in excess of 7 kg/cm2 (100 Ib/in.2) were observed with solids con-
tents greater than approximately 80 percent. Although not shown in Figure 15,
a load-bearing strength greater than 44 kg/cm2 (630 Ib/in. 2) was observed
for the Cholla sludge sample when the solids content reached 83 percent.
*
Limestone process sludge (35 wt% solids) , containing less
than 3 percent fly ash, was taken directly from the clarifier underflow at
TVA Shawnee and was tested as received. Also tested was a sample to which
was admixed 40-wt% (dry basis) fly ash, without any preliminary dewatering.
*
This sludge slurry had a high settling rate. Under examination with the SEM,
some very large crystals of CaSOs • 1/2 HgO £>2 |im in length and>0. 2 \ucn
in thickness) were observed.
72
-------
The compaction strength versus solids content curves for the Shawnee limestone
and limestone with fly ash samples (Figure 15) are slightly different in shape
from those typical of the other sludges tested. A greater spread was observed
between the solids contents corresponding to a just measurable load-bearing
strength (5Z percent for the sludge and 55 percent for the sludge with fly ash
mixture) and the solids contents corresponding to a load-bearing strength
greater than 7 kg/cm^ (100 Ib/in. 2), 63 and 67 percent, respectively. Although
this might have been caused by the low initial solids content of the clarifier
underflow, a similarly shaped curve was reported previously (Ref. 1) for the
compaction strength of Shawnee limestone sludge. As expected, the addition
of fly ash displaced the compaction strength curve toward a higher solids con-
tent, without significant alteration of the shape.
It is apparent that each of these 11 sludges can be dewatered by
a simple drainage process to a solids content which corresponds to a load-
bearing strength well in excess of the minimum required for accommodating
powered equipment, and, in most cases, a load-bearing strength suitable for
construction can be obtained.
73
-------
SECTION VII
CHEMICAL CHARACTERIZATION OF SLUDGE
LIQUORS AND LEACHATES
7. 1 BACKGROUND
Previous laboratory characterization and chemical analyses
were performed on sludge from six flue gas desulfurization (FGD) waste
sources, and the results of these experiments have been reported (Ref. 1).
The chemical characterization of sludge from three additional sources is
herein reported. The primary objective of this task was to investigate the
effect of the presence of fly ash in the flue gas being scrubbed on the con-
centrations of trace elements in the sludge liquors and leachates. All of the
sludges previously analyzed contained fly ash which varied from 20 to 60 per-
cent, except fox three which contained 3 to 9 percent. Previously reported
analyses for trace elements in system liquors and leachates were only for
those sludges produced in FGD scrubbing systems that collected some or
all of the fly ash in the scrubber. Therefore, it was not possible to deter-
mine unequivocally whether the trace elements in the liquors and leachates
resulted from the scrubbing of fly ash. It was necessary, therefore, to
analyze sludges from scrubbing operations in which fly ash was removed from
the flue gas ahead of the scrubber.
Additionally, leaching tests were performed with sludge and
fly ash mixtures, prepared by the addition of fly ash to ash-free FGD sludge,
and also with fly ash alone using leaching water controlled at a pH ranging
from 4 to 9. A comparison was made of the concentrations of trace elements
in the liquors and leachates from ash-free sludge and sludge containing fly
ash for two power plants for the purpose of identifying the source of these
species. An attempt was made to correlate the magnitude of the measured
trace element concentrations in sludge liquors and leachates with fly ash,
pH, and conditions of scrubbing or leaching.
A secondary objective of this task was to provide a broader
base for the chemical characterization of FGD sludges. The previous study
(Ref. 1) showed that the waste liquors were saturated with gypsum in their
disposal state. The liquors from the sludges herein reported were analyzed
for major and minor constituents and compared with the existing data base
to determine their relative state of chemical saturation.
75
-------
7.2 DESCRIPTION OF POWER PLANT SCRUBBING
FACILITIES, SAMPLING, AND CHEMICAL ANALYSES
The chemical characterization for this study was generated
from samples taken from the three power plant scrubbing systems listed in
Table 16, using lime or double alkali systems and Eastern coals. For the
purpose of this study, the effect of fly ash collection alternatives was inves-
tigated relative to the scrubber waste chemistry and the potential of these
wastes as an environmental pollutant source. The sludge from the Tennessee
Valley Authority (TVA) Shawnee system was evaluated in the ash-free state
in which it was collected, as well as in a recombined state representing
the phase constituents as they would have been if no fly ash was collected
upstream of the scrubber. The sludge from the Gulf Power Company (GPC)
Plant Scholz was evaluated in the ash-free and combined ash and sludge
states as they were collected.
Essentially ash-free lime process filter cake was obtained
from TVA Shawnee. This sludge was analyzed as received, as well as after
mixing with 40 wt% fly ash (dry basis). The fly ash was 67 percent mechani-
cally separated and 33 percent electrostatically separated ash. Two samples
of sludge from the double alkali process at the Gulf Scholz station were
supplied by A. D. Little, Inc. One sample contained approximately 31 per-
cent fly ash, and the other had no fly ash. A fifth sludge sample tested was
obtained from the carbide lime process at Louisville Gas and Electric
(LG&E) Paddy's Run station and supplied by Combustion Engineering, Inc.
The sludge from this power plant contained approximately 12 percent fly
ash. No sludge from this power plant had previously been tested; therefore,
the test data were obtained to enlarge the data base for FGD sludge
characterization even though no data were available for ash-free sludge to
determine the influence of fly ash.
7. 3 CHEMICAL ANALYSES
7. 3. 1 Experimental Procedure
As a general procedure, samples were taken by the scrubber
system operating personnel. Some samples were filtered at the site to stop
reactions between liquor and solid reagents. All samples were packaged in
plastic 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 of the fil-
trate was determined at once. Liquid samples were stored in sealed polyethy-
lene bottles. Solids samples were stored in sealed polyethylene storage bags.
Samples were stored for a few days, and in some cases for several months be-
fore being analyzed. The time factor had no effect on sludge chemistry except
for insignificant reductions of pH and total alkalinity due to absorption of CO,.
£»
All chemical analyses were performed by standard chemical
analysis techniques to at least minimum criteria identified by the U. S.
Environmental Protection Agency (EPA) Analytical Quality Control Laboratory
76
-------
TABLE 16. FGD SYSTEMS SAMPLED AS DATA BASE2
Power Plant
Tennessee Valley
Authority (TV A),
Shawnee Steam
Plant
Louisville Gas and
Electric Co. ,
Paddy's Run Station,
Unit No. 6
Gulf Power Company
Scholz Plant
Scrubber
Type
Venturi and
spray tower,
prototype
Marble
bed
absorber
Venturi and
spray tower
Scrubbing
Capacity
MW (equiv)
10
70
20
Coal
Source
Eastern
Eastern
Eastern
Absorbent
Lime
Carbide
lime
(slaked lime
waste)
Soda ash
(lime
regenerant)
Other pertinent data are given in Tables 1 and 2.
77
-------
(Refs. 12 and 13). A description of the chemical analyses is presented in
detail in Appendix C of Reference 1 for liquids and solids. In every case,
both the accuracy and precision of the method were determined with respect
to analyses of the scrubber liquors.
Chemical analyses were made for nine trace elements of con-
cern in FGD sludge liquors: arsenic (As), beryllium (Be), cadmium (Cd),
copper (Cu), chromium (Cr), mercury (Hg), lead (Pb), selenium (Se), and
zinc (Zn). In addition, analyses were made for boron (B), and fluorine (F),
which were present at concentration levels much higher than those of the
nine trace elements, and for the major constituents calcium (Ca), chloride
(Cl), sulfate (SOJ, magnesium (Mg), potassium (K), and sodium (Na).
Measurements were made of pH, and total dissolved solids (TDS), and mass
balances, charge balances, and gypsum solubility ratios were calculated for
each sample to check the internal consistency of the data.
The solid phases of each sludge were identified by examination
with the scanning electron microscope (SEM). Analyses were made for each
major component of the sludge solids. Comparative analytical data were
supplied by A. D. Little, Inc., for the sludge samples from Gulf Scholz.
7.3.2 Results of Chemical Analyses
The results of all chemical analyses performed during this
reporting period are presented in tabular form in Tables 17 through 19. The
results of the chemical analyses from sources including those previously
reported and those herein presented are summarized in Tables 20 and 2.
In Tables 17 through 19, the ranges of chemical constituents were obtained
from the analyses of the sludge samples taken after the final stage of
de watering.
From these data, several observations can be made that
identify the chemical constituent concentrations that may be expected
generally from FGD scrubbers. Among the trace elements in sludge liquors,
the ranges of concentrations tend to lie between 0.01 and 1 mg/l with the
exception of mercury, which has a concentration distribution about 1/10
that of other trace elements. Inmost sludge liquors, the pH ranges between
6. 5 and 9; however, in some samples, pH in the 10 to 12 range was detected.
COD ranged between 40 and 140 mg/l .
A correlation exists between the liquor and solids concentra-
tions of beryllium, cadmium, and mercury, in which the values are generally
lower than the other elements, in both phases. This appears to indicate that
these elements exist in the sludges in relatively minor amounts. On the
other hand, solid concentrations are high for Cr and Zn (possibly, also Cu),
but corresponding higher concentrations are not observed in the liquor.
Whereas, in the first case, the correlation strongly suggests a concentration
limitation based on input amounts, in the second case, the lack of correlation
suggests that chemical parameters may be controlling the liquor concentrations
of these elements.
78
-------
TABLE 17. CHEMICAL ANALYSIS: SHAW NEE LIME SLUDGE
LIQUOR AND LEACHATEa
Constituent
PH
TDS
Araenlc
Beryllium
Boron
Cadmium
Calcium
: Chromium
Copper
Lead
Magnesium
Mercury
Potassium
Selenium
Sodium
Zinc
Chloride
Fluoride
Sulfate
Without Fly Ash
9/8/76 Run F
Filtered
Sludge Liquor
8.0
10260
0.058
<0.00i
76
0.013
650
0.011
0.005
0.010
1730
<0. 00006
24
0.078
137
<0.001
1320
1.9
4500
9/8/76 Run F Leachate
First Pore
Volume
8.3
4480
35
....
690
0.025
0.011
<0.002
400
14
75
0.007
550
2.6
2150
Fifth Pore
Volume
7.1
1770
0.120
<0.001
3
0.013
450
0.010
<0.002
<0.002
30
0.00006
2
<0.0004
47
0.01
128
2.5
1100
Run F + 40% Fly Ash Leachate
First Pore
Volume
7.9
4330
65
—
600
0.010
<0.002
<0.005
310
7
72
0.003
400
6
2700
Seventh Pore
Volume
7.4
2430
0.360
<0.001
16
0.010
640
0.004
• <0.002
<0.005
10
0.00024
4
0.051
42
0.02
130
1.2
1450
'Concentrations inmg/l, as appropriate.
79
-------
TABLE 18.
CHEMICAL ANALYSIS: GPC PLANT SCHOLZ DOUBLE-
ALKALI LIQUOR AND LEACHATEa
Constituent
PH
TDS
Arsenic
Beryllium
Boron
Cadmium
Calcium
Chromium
Copper
Lead
Magnesium
Mercury
Potassium
Selenium
Sodium
Zinc
Chloride
Fluoride
Sulfate
6/20/76 Run with No Fly Ash
Leachate
Filtrateb
12.1
155700
43
12
0.1
320
53600
5000
24
80000
First Pore
Volume
12.5
9140
5
1000
—
—
—
<0.01
—
43
2260
1070
48
3700
Sixth Pore
Volume
12.3
3550
<0.004
<0.001
<0. 5
0.016
1100
0.0025
<0.002
0.005
<0.01
0.00030
1
<0. 0004
12
0.007
310
13
1160
6/27/76 Run with 30% Fly Ash
Leachate
Filtrate5
12.0
162700
40
7
0.1
380
55300
4900
4
84000
First Pore
Volume
10.4
17330
<0.5
550
0.2
74
4720
1050
3
10100
Ninth Pore
Volume
9. 1
2140
0.019
<0.001
<0.5
0.010
800
0.024
<0.002
<0.002
6.3
0. 00024
11
<0. 0004
82
0.013
134
0.9
1415
Concentrations in mg/i, as appropriate.
Incomplete analysis for filtrate samples is a consequence of insufficient
sample quantity.
c
Approximately 90 percent of the reduction of concentrations occurs by the
third PVD.
80
-------
TABLE 19. CHEMICAL ANALYSIS: PADDY'S RUN SLUDGE
LIQUOR AND LEACHATEa
Constituent
PH
TDS
Arsenic
Beryllium
Boron
Cadmium
Calcium
Chromium
Copper
Lead
Magnesium
Mercury
Potassium
Selenium
Sodium
Zinc
Chloride
Fluoride
Sulfate
Filtrate
8.9
24230
0.011
<0.001
18
0.025
515
0.054
0.0045
<0.005
3400
0.00006
760
0.0028
260
0.003
5600
<1
15000
Leachate
First Pore
Volume
7.4
5240
2
410
<0.0008
0.004
<0.002
470
125
40
0.015
410
<0.1
2800
Tenth Pore
Volume
8.1
1650
0.023
<0.001
<0.5
0.004
260
<0.0008
<0.002
<0.002
70
0.00006
21
0.006
3
0.005
157
<0.1
920
Concentrations in mg/Jt, as appropriate.
12% ash.
81
-------
TABLE 20. RANGE OF CONCENTRATIONS OF
CHEMICAL CONSTITUENTS IN FGD
SLUDGE LIQUORS THROUGHOUT
THE SCRUBBER LOOP - ALL SOURCES'
Scrubber
Constituents
Aluminum
Arsenic
Beryllium
Cadmium
Calcium
Chromium
Copper
Lead
Magnesium
Mercury
Potassium
Selenium
Sodium
Zinc
Chloride
Fluoride
Sulfate
Sulfite
TDS
PH
COD
Sludge Liquors,
m.g/1
0.03 to 2.0
0.004 to 0. 14
0.001 to 0.18
0.004 to 0. 11
10 to 2600
0.011 to 0.25
0.002 to 0. 56
0.005 to 0. 33
0. 1 to 2750
0.00005 to 0.005
5. 9 to 760
0.0006 to 0.20
10.0 to 55,000
0.001 to 0. 59
420 to 33, 000
0.6 to 10
600 to 84, 000
0.9 to 3500
2800 to 162, 700
4.3 to 12.7
40 to 140
All test conditions not necessarily
standard. See Table 2 for repre-
sentative list.
82
-------
An apparent exception is Se, which has the broadest
concentration range in the liquor (nearly four orders of magnitude) but has
the narrowest concentration range in the solid (factor of eight). In this case,
the sample that contributed the high value in the solid had the lowest pH of
all samples analyzed.
Among the major chemical species, concentrations and range
of concentrations are almost completely dependent upon the chemical para-
meters of the system. Na, as an example, ranges from a low value of
10 mg/jj to nearly 5 percent in the liquid phase in the double alkali system.
When high values of Na and sulfate (or Cl) are present, significant amounts
of sodium salts may be expected in the solids. The major FGD waste,
sulfur phases, can exist as either sulfate or sulfite. As may be expected,
the solids sample having the highest sulfite content also had the lowest
sulfate content.
One of the most important chemical parameters for the
asssessment of the environmental quality of the liquid portion of FGD sludge
is the TDS content. In most cases, TDS concentration is typically
10,000 mg/fc . Higher values were observed for the double alkali samples and
one sample in which a salt brine was used as makeup water. Lower values
were observed in two systems during startup and in a third system in a
partial open-loop operation. From these data, it is reasonable to expect
values of TDS in scrubber liquors to range near 10,000 mg/ft for steady-state
closed systems, except in cases having exceptional circumstances; these
cases can usually be identified from system design parameters.
7.4 LEACHING ANALYSIS
7.4.1 Experimental Procedures
In the previously reported work (Ref. 1), several methods
were used to prepare columns, and it was determined that the results were
independent of the method. For the work herein reported, the following
method was used to prepare a leaching column. Approximately one liter of
sludge, as a slurry of about 50 percent solids, was poured into a 6-cm-
diameter column containing a porous base plate covered with a filter paper,
and the solids were allowed to settle. Excess liquor, including supernate
above the settled solids, was allowed to pass through the solids and filter
until dripping stopped. Deionized water was added to the column, and the
leaching experiment was begun by collecting aliquots of leachate based on
pore volume displacements (PVD). Pore volumes were determined from
the volume and the solids content of the bed. All column tests were conducted
under aerobic conditions.
7.4.2 Results of Leaching Experiments
In previously reported studies (Refs. 1 and 2), leaching tests
were continued until 50 pore volumes had been displaced. In most cases,
83
-------
however, it was observed that 90 percent of the decrease in concentration
of major leachable species had taken place after 3 pore volumes had been
displaced. For the present studies, it appeared unnecessary to continue
leaching after 10 pore volumes had been displaced. The complete leaching
curves were determined by measurement of the TDS in each sample of
leachate. Complete analyses for major and minor constituents were made
only on the first and last leachate sample. In Figures 16 through 18, the
concentrations of major species (804, Cl, Ca, and Mg or Na) and TDS have
been plotted against the average PVD of the leachate for sludges obtained
from three power plants. The curves of Figure 16 show the results of
leaching Shawnee ash-free lime sludge and also sludge mixed with 40-wt%
fly ash. For additional comparison, the concentrations of major species
and TDS of the sludge liquor (filtrate taken from vacuum filter) have also
been plotted. The TDS in the leachates of both sludge samples decreased
from initial values comparable to that of the filtrate and leveled off at
2000 to 2500 mg/f after three pore volumes had been displaced. Similarly,
the sulfate leveled off at about 1000 to 1500 mg/fc . The calcium concentra-
tions decreased to below the 100 mg/! level. These filtrate and leachate
analytical data are also shown in Table 17.
Leachate data were plotted for two sludge samples obtained
from Gulf Scholz in the curves of Figure 17; one was essentially ash-free,
and the other contained 31 percent fly ash. The corresponding analytical data
for the two liquor (filtrate) samples were not plotted because their concentra-
tion exceeded the scale of the figure; however, complete data are shown in
Table 18. For these sludge samples, the extraordinarily high concentra-
tion of dissolved sodium sulfate in the occluded liquor (> 15 percent) was
rapidly reduced so that after five pore volumes had been displaced the TDS
in the leachate stabilized at about 2200 mg/! . The sulfate concentrations
decreased from the 80,000 mg/l level in the liquor and leveled off at 1000
to 1500 mg/| . The calcium concentration increased from about 10 mg/l in
the liquor and leveled off below 1000 mg/l . The sodium and chloride con-
centrations, initially 50,000 and 5000 mg/! , respectively, in the liquor,
decreased to levels below 100 mg/! .
The data of Table 19 and Figure 18 show analogous leachate
and liquor analytical data for a sludge sample containing fly ash from
LOSE Paddy's Run. The leachate from this sludge at the 10th pore volume
differed in the last leachate samples from all the other sludges in that the
TDS and 804 concentrations did not level off at their characteristic values of
2200 and 1300 mg/fc . These characteristic values are the result of saturation
of the leaching water with gypsum. Lower values observed for the 10th PVD
leachate sample from the Paddy's Run sludge represent-under saturation of
the leaching water and indicate that most of the gypsum had been already
dissolved.
84
-------
NO FLY ASH
40* FLY ASH
O
10,000
O — IDS — •
A — S04 — A
D — cr — •
0 — Ca — •
O — Mg — •
5,000
I
o
o
34567
AVERAGE PORE VOLUME DISPLACEMENT
Figure 16. Concentration of major species and TDS in leachate
lime sludge with and without fly ash: TVA Shawnee,
Run F
85
-------
25,000
20,000
5! 15,000
O>
CJ
O
C_3
10,000
5,000
w/o
FLY ASH
IDS
S04
Cl
Ca
Na
2 4 6 8 10 12
AVERAGE PORE VOLUME DISPLACEMENT
Figure 17. Concentration of major species and TDS in leachate
of sludge with and without fly ashtGPC Plant Scholz
86
-------
25,000
20,000
*! 15,000
CD
o
LLJ
CJ
8 10,000
0
O IDS
A
D
0
X
S04
Cl
Ca
Mg
0 2 4 6 8 10 12
AVERAGE PORE VOLUME DISPLACEMENT
Figure 18. Concentration of major species and TOS in filtrate
and leachate of sludge: LG&E Paddy's Run
87
-------
7. 4. 3 Assessment of Chemical Data
#
In Table 21, gypsum solubility ratios calculated for all of the
sludge liquors and leachates are presented. The spread of these ratios is
from 0.48 (for the 10th PVD leachate sample from The Paddy's Run sludge)
to 1.46 (for the first PVD leachate from the Gulf Scholz sludge), and the
mean of the 15 values is 1.04. In view of the spread of the ionic strengths
of these liquors and leachates from 0.04 to 3.0, corresponding to a range
of sulfate concentrations from 900 to 84,000 mg/f , gypsum saturation is
indicated in every case with the possible exception of the Paddy's Run 10th
PVD leachate sample.
Table 21 also contains mass and charge balances calculated
from the analytical data for each of 15 samples of sludge liquors and
leachates. The spread of the calculated mass ratio j, given by the sum of
the ion concentrations divided by the TDS (corrected for the water of hydra-
tion of CaCl2 • 2H2O), is 0.85 to 1. 19, with a mean of 0.97. This indicates
that all of the major species present in the analyzed solutions have been
accounted for within the accuracy of the measurements. The ionic charge
ratio, given by the sum of the equivalent concentrations of cations divided by
the sum of the equivalent concentrations of anions, is an even more severe
test of the internal consistency of the analytical data. The charge ratios for
the 15 samples span a range of from 0.71 to 1. 38, with a mean value of
1.08. By a comparison of the three ratios for a given sample, it is possible
to identify individual analyses that are questionable. Examination of the
analytical results in Tables 17 through 19 for the 15 samples indicates that
several of the analyses for calcium may be in error by 10 to 20 percent.
However, each of the samples involved had high ionic strength, which makes
it difficult to obtain higher accuracy in the analyses for metal ions, and,
particularly, trace elements.
7.4.4 Assessment of the Effect of Fly Ash on
Trace Element Contents
In Tables 17 through 19, complete chemical analyses for
major components and for 10 selected trace elements are reported for the
sludge liquors and leachates from the three power plants. In Table 17,
analytical results are given for Shawnee sludge filtrate liquor and for leachate
samples from leaching tests of both the ash-free sludge and the sludge with
40-wt% fly ash mixture. Results for both ash-free sludge and sludge con-
taining 30 percent fly ash from the Scholz plant are given in Table 18. The
This ratio is defined as the product of the measured calcium and sulfate con-
centrations divided by the solubility product constant (corrected for the
effects of ionic strength.)
#*
These fly ash percentages represent a dry weight percent of the solids
content of the sludge liquor.
88
-------
TABLE 21.
MASS BALANCE, CHARGE BALANCE, AND GYPSUM
SOLUBILITY RATIO OF SLUDGE LIQUORS AND
LEACHATES
Sample Source
TVA Shawnee Sludge
Run E Filtrate
Run F Filtrate
Run F Leachate (No Fly Ash):
1st PVD
6th PVD
Run F Leachate (40% Fly Aah)i
let PVD
7th PVD
Gulf Scholz Sludge
Filtrate (No Fly Aah)
Leachate (No Fly Ash):
1st PVD
6th PVD
Filtrate (30% Fly Ash)
Leachate (30% Fly Ash):
1st PVD
9th PVD
LGfcE Paddy's Run Sludge
Filtrate
Leachate:
1st PVD
10th PVD
Average
Mass Ratio
1.01
0.88
0.93
1.03
1.01
0.97
0.91
1.01
0.86
0.90
0.98
1.17
1.19
0.85
0.91
0.97(10.10)
Charge
Ratio15
1.29
1.38
1.18
1.02
0.87
1.03
1.29
1.08
1.08
1.27
0.98
1.32
0.71
0.92
0.82
1.08 (±0.20)
Ionic
Strength
0. 14
0.27
0. 12
0.05
0. 12
0.07
2.91
0.21
0.09
3.03
0.36
0.07
0.71
0.13
0.04
Solubility
Ratio0
0.97
0.85
1.07
0.82
1.20
1.23
1.19
1.46
1.22
0.91
1.17
1.36
0.86
0.80
0.48
1.04 (±0.26)
Mass ratio
Charge ratio
cSolubility ratio
£ all ion concentrations/TDS (corrected for water in CaCl2' 2H2O not
lost in drying).
£ equivalent concentrations of cations/£ equivalent concentrations
of anions.
Measured calcium concentration X measured sulfate concentration/
K (solubility product constant corrected for ionic strength).
89
-------
results for the Paddy's Run plant are given in Table 19 for filtrate and
leachate of a sludge containing 12 percent nonsulfur solids.
A comparison of analyses are given in Table 22 for Shawnee
lime sludge filtrate liquors from two scrubber test runs in 1976, in which
the fly ash was removed ahead of the scrubber and from three runs made
early in 1974, in which the fly ash was present in the flue gas. The pH of
each of the liquors is in the range of 8 to 9. The analytical results for the
startup run of 3/19/74 differ from the others in that the TDS, Na, K, and
Mg concentrations are much lower, and the As and Hg levels are much
higher than those of any of the other runs. The results of this run have
been retained in Table 22, so as to include all of the previous data from the
lime scrubber at Shawnee. The Mg and 804 concentrations are much higher
in the 1976 filtrate as a result of the deliberate addition of MgO with the
lime in the scrubbing operation.
Comparisons of trace element concentration levels in samples
with and without fly ash can be seen in Table 22. With a single exception,
the results for non-fly-ash filtrates (Runs £ and F) all are approximately
at the level of the lowest of the three samples that contained fly ash or, in
several cases, range between the lowest and the median values for the
1974 runs. The single exception is Zn which is appreciably lower in the
Run E and Run F filtrates.
Examination of the leachate analytical results in Table 17 for
Shawnee Run F ash-free sludge and sludge to which 40-wt% fly ash had been
added shows that five trace element concentrations are greater in the sludge
with fly ash leachate; four are approximately the same in both leachates;
and two trace element concentration levels are higher in the ash-free sludge.
Only for B and Se are the differences greater than by a factor of three.
However, for As, Hg, and Zn, the levels in the leachate of the sludge with
fly ash mixture are higher than those in the Run F ash-free filtrate liquor.
Furthermore, the last PVD of leachate show higher Zn concentrations than
the first PVD. It is possible that the trace element Zn is incorporated in
sulfite or carbonate solid phases, whose solubilities are sufficiently
greater at the slightly lower pH of the last leachate samples to release the
greater amounts of Zn observed. A similar pattern is observed for the Zn
concentrations in the leachates from the Paddy's Run sludge in Table 19.
The behavior of As and Hg may also be a consequence of their release from
a solid phase during leaching.
No leaching tests were made on the previous (1974) samples
of Shawnee lime sludge which contained fly ash; therefore, comparative
data were not generated for the leaching of sludge containing fly ash that
was scrubbed with the flue gas. Such a comparison can be made of the
analytical data for the leachates from the two Scholz sludge samples shown
in Table 18. However, the pH range of the double alkali process used at
the Gulf Scholz plant is too high to expect to see evidence of acid leaching
of fly ash during the scrubbing operation. The high pH of the liquors, ~ 12,
90
-------
TABLE 22. COMPARATIVE ANALYTICAL RESULTS FOR
SHAWNEE LIME SCRUBBER FILTRATE
LIQUORS
Constituent
pH
TDS
Arsenic
Beryllium
Boron
Cadmium
Calcium
Chromium
Copper
Gold
Magnesium
Mercury
Potassium
Selenium
Sodium
Zinc
Chloride
Fluoride
Sulfate
With Fly Asha
3/19/74
9.4
2800
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
5/16/74
8.8
8400
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
6/27/74
8.7
9400
0.02
<0.002
46
0. 11
2520
0.03
0.002
0.33
420
<0.001
28
<0.02
127
0.08
4900
3.3
800
Without Fly Asha
8/23/76b
8.9
5650
<0. 004
<0.001
9
0.025
1450
0.008
<0.003
0.010
220
<0. 00006
22
0.022
130
0.002
1850
1.3
1100
9/P/76c
8.0
10, 260d
0.058
<0.001
76
0.013
650
0.011
0.005
0.010
1730d
<0. 00006
24
0.078
137
<0.001
1320
1.9
4500d
Concentrations in mg/£ , as appropriate.
Designated in Shawnee Scrubber Test Program as Run E.
CDesignated in Shawnee Scrubber Test Program as Run F.
Magnesium added to lime to evaluate absorbent efficiency.
91
-------
is the cause of the virtual absence of Mg, which was precipitated as the
hydroxide and only begins to reappear in the leachates after elution has
lowered the pH to below 9. Examination of the data of Table 18 shows that,
of the trace elements listed, only As and Cr are found in higher concentra-
tions in the leachate from the sample that contained fly ash. The lower pH
of the ninth PVD leachate sample from the Scholz sludge containing fly ash
may also be responsible for the higher concentrations of these two elements.
It is noteworthy that the fluoride concentrations are higher in the liquors
and leachates of the ash-free sludge.
7. 5 RESULTS OF LEACHING FLY ASH AT
CONTROLLED pH
Fly ash obtained from TVA Shawnee was mixed in the pro-
portion of two parts mechanically separated ash to one part electrostatically
separated ash. To this mixture, dilute HCi was added as required to
maintain a constant pH, using a glass electrode as an indicator. Two acid
equilibrations, one at pH 4. 0 and the other at pH 7. 1, were continued until
the pH remained constant and then were allowed to stand with constant
stirring for 24 hr.
The concentrations of trace elements measured in the solutions
at pH 4 and pH 7 are shown in Table 23. Only the concentrations of Cd, Cu,
and Pb appeared to be significantly higher in the sample at pH4 than in the
sample at pH 7. Selenium was found in higher concentrations in the sample
at pH 7 than in the sample at pH 4. Of all the trace element concentrations
measured in the fly ash leaching tests, only the concentrations of Cd, Cu,
and Pb in the sample at pH 4 vere found to be significantly higher than the
values for these elements meat -red in the sludge with fly ash leaching tests.
The concentration levels of all o a.r trace elements in the fly ash leaching
were comparable to those in the s. *ge with fly ash leachate except for the
concentrations of As in the fly ash leachates, which were lower than those
measured in the sludge with fly ash leaching tests.
7.6 RESULTS OF SOLIDS ANALYSES
Dry sludges from three power plants were analyzed for their
major components. The results of these analyses are shown in Table 24.
The samples of Gulf Scholz sludge that was received from A. D. Little, Inc.
were accompanied by the analytical results for the same sample. These data
have been included in Table 24. Comparison shows good agreement between
the analytical results from both laboratories. It may be noted that material
balances are well within 5 percent for all three sludges.
Each of the sludge solids contained greater than 15 percent
gypsum. Neither the Scholz nor the Shawnee sample contained a significant
amount of fly ash, but the Paddy's Run sludge contained about 12 percent
ash and, in addition, about 3.5 percent carbon. The Scholz sludge also
92
-------
TABLE 23. TRACE ELEMENTS LEACHED FROM SHAWNEE
FLY ASH AT CONTROLLED pHa» b
PH
Arsenic
Beryllium
Cadmium
Chromium
Copper
Lead
Mercury
Selenium
Sodium
Zinc
4.0
0.039
<0.001
0.036
0.006
0.036
0.022
0.00012
0.006
10
0.01
7. 1
0.058
<0.001
0.009
0.009
0.003
<0.005
0.00015
0.026
10
0.005
Concentrations in mg/4, as appropriate.
JShawnee fly ash mixture of two parts mechanically separated ash and
one part electrostatically separated ash, September 1976.
93
-------
TABLE 24. COMPOSITION OF DRY SOLID SLUDGE FROM THREE
POWER PLANTS, IN WEIGHT PERCENT
Component
CaSO4a
CaS04b
CaSO3c
CaCO3
Insolublesein HC1
Na,SO/
Z 4
Total
TVA Shawnee,
Lime (Run F)
19.4
—
69.2
10.3
—
—
98.9
LGfcE Paddy's Run,
Carbide Lime
15.1
-
37.4
29.5
15.96
—
97.9
Gulf Scholz,
Double Alkali
15.3
-
68.1
10.1
-
3.7
97.2
Gulf Scholz
(analysis by
A. D. Little, Inc.)
—
13.9
68.1
8.2d
—
3.5
93.7
a Expressed as CaSO4*2 H2O. Calculated from excess Ca.
Expressed as CaSO4' 1/2 H£O. e Contained 3.5 percent carbon (loss on
cExpressed as CaSO3- 1/2 H2O. ignition [LOl]).
Calculated from Na.
vO
-------
contained about 4 percent sodium salts from the double alkali process.
Residual calcium carbonate amounted to about 10 percent for the Shawnee and
Scholz sludges and almost 30 percent for the Paddy's Run sludge.
7.7 SUMMARY AND CONCLUSIONS
The results of the fly ash equilibration and leaching tests
show that low pH scrubbing of fly ash could produce higher levels of several
trace elements than is produced in high pH sludge with fly ash liquors. Of
the nine trace element concentrations measured, three were higher in the
low pH sample; one was higher in the high pH sample; and five were not
significantly different in the two samples. Moreover, the data show that
the concentrations of trace metals in fly ash leachate are essentially
equivalent to the concentration of these metals in sludge with fly ash leachate
having the same pH.
A comparison of trace element concentrations in the filtrate
liquors of Shawnee ash-free lime sludge and sludge obtained by scrubbing
flue gas containing fly ash show that only one element, zinc, had a
significantly lower concentration in the liquor from the ash-free sludge.
For the other eight elements, the concentrations in the ash-free sludge
liquor were comparable either to the lowest levels or to the median levels
observed for the liquors from sludge containing fly ash. Similarly, for the
leachates from the two Scholz plant sludges, one ash-free and the other
containing fly ash, only two trace element concentrations were significantly
higher in the leachate from sludge containing fly ash. Leachates of the
ash-free Shawnee lime sludge and the sludge mixed with 40-wt% fly ash
showed comparable concentration levels for all except one trace element.
From these data, it is concluded that the presence of fly ash in a sludge will
cause significantly increased concentrations of only a few trace metals in
the liquor or leachate by as much as a factor of 10 over concentrations in
the liquor or leachate of a sludge containing no fly ash.
Examination of the solid phases of each of the ash-free sludges
with the SEM showed the presence of some fly ash particles. It is probable
that the finest fly ash particles, i. e., those with the largest relative surface
area and therefore the highest teachability, are carried past the separators
by the flue gas and are collected in the scrubber. The trace elements found
in the liquors of ash-free sludge may be explained by the presence of minor
quantities of fly ash. Since the finest fly ash particles can remain suspended
in the liquors and leachate s and pass through the pores of the filters used,
it is possible that these suspended particles are subsequently analyzed, and
not all the reported concentration levels of trace elements represent dis-
solved species. Additionally, other input ingredients such as lime, limestone,
and magnesium oxide may contribute a low level of trace elements of the
liquor. It is concluded from these data that removal of fly ash ahead of the
scrubber does not eliminate the trace elements from the sludge liquors and
leachate s, but the concentration levels of some trace elements may be
significantly reduced.
95
-------
SECTION VIII
ANALYSIS OF FORCED OXIDATION GYPSUM
8.1 BACKGROUND OF GYPSUM FROM FORCED
OXIDATION OF RTF PILOT PLANT LIME-
STONE SCRUBBER SLUDGE
In the U.S. Environmental Protection Agency (EPA) Emissions/
Effluent Technology Branch, Research Triangle Park (RTP), North Carolina,
R. Borgwardt has been conducting pilot-plant-scale experiments evaluating
the forced oxidation of sulfite sludges from the limestone and lime scrubbing
of SO2 from flue gas. The experiments were designed to determine limestone
utilization, oxidation efficiency, settling rates and bulk densities of the gyp-
sum product. The pilot plant system consisted of a first-stage spray tower
and a second-stage turbulent contact absorber (TCA) loop (Figures 19 and
20). The first-stage loop contained a holding tank equipped for air sparging
to bring about complete oxidation of the sulfite. A portion of the oxidized
slurry was bled off to a vacuum drum filter. Burner combustion products
simulated flue gas with SO2 and HC1 being introduced into the gas stream.
Fly ash was introduced, when desired, with the limestone in the second-
stage loop.
To augment the numerous chemical analyses and physical
property measurements being made at RTP and reported by Borgwardt Ref-
erences 14 and 19, Aerospace conducted additional characterization tests on
filter cake and slurries from the first- and second-stage loops of samples
from two runs utilizing limestone scrubbing, one with and the other without
added fly ash. Scanning electron microscope (SEM) and x-ray examinations
were made of crystalline phases, and wet chemical analyses were made for
major constituents of both the solid and liquid phases. Also, the leaching
characteristics of the solids were determined. Physical properties including
bulk densities, compressive strengths, and permeabilities were also mea-
sured, and cost estimates of gypsum production by forced oxidation were
made.
8.2 CHEMICAL AND LEACHING CHARACTERISTICS
SEM, x-ray, and wet chemical analyses were made of the
filtered solids and solids from the first- and second-stage slurries. All the
97
-------
2590 ppm
00
RUN SEP 29 - OCT 3
ASH ADDITION
SCRUBBER EFRUENT LIQUOR
JTAL S A
S02
CC?
Ca
Cl
pH
S02
5.66
Ib/hr
^
S 503 2660
1250
41
2640
5500
4.3
2910 ppm S02
/^—
•M
T"
,5gp
^^^•H
LIQUOR
TOTAL S AS SO,
S02
C02
Ca
m Cl
Mg
PH
SATURATION
SOLIDS
TOTAL S AS S03
S02
C02
Ca
OXIDATION * 0.93
2300
870
28
2520
5440
770
4.5
1.2
346
15
5
179
35 Ib/hr
TO FILTER
SETTLING RATE- 2.0cm/min
SETTLED DENSITY - 0.91 g/ml1
UTILIZATION • 0.97
16 gph pH 5.4
LIMESTONE + RY ASH *
25.0 Ib/hr
50.8% SOLIDS
BYPASS
FIRST STAGE
I , "
^"
^
V
1,
/
T
>\J~ JL.\J \l\l\\\
15.5 gpm
AP
8 in. H20
/ /
/ /
47 gal 47 gal
. °f> . °6 _
LIQUOR
SATURATION 1.0
SOLIDS
TOTAL S AS S03 306
S02 218
C02 30
Ca 186
OXIDATION - 0.11
UTILIZATION - 0.82
SECOND STAGE
Figure 19. EPA pilot plant forced oxidation system tests with fly ash
-------
S027.71 Ib/hr
HCI 0.15 Ib/hr
6.8% 02
LIQUOR
TOTAL S AS $03 1880
S02 390
Ca 2420
Cl 4180
pH 4.4
SATURATION 1.14
SOLIDS (12%)
TOTAL S AS S03 503
S02 55
C02 7
Ca 271
OXIDATION - 0.86
UTILIZATION • 0.93
TO FILTER
SETTLING RATE - 2 cm/mm
SETTLED DENSITY • 0.88 g/ml
(60% solids)
3000 ppm S02
6.7 gpm'
AIR
I7.9lb/hr)
(2.8x)
PSig
10
in.
H20
H
(IIG 20)
18ft
106
gal
15.8
min
12
tt/sec
AP
2.0
in.
735 ppm
so2
pH3.9
BYPASS
S02 • 720 ppm
(76%)
21 gpm
RUN DEC 1 - 5, 1975,
12 fps, NO FLY ASH
/\
12
ft/sec
AP
7.0
W9B88
pH5.1
(L/G 64)
SOLID SPHERES
3-ln. BED DEPTH
25.4 Ib/hr
-LIMESTONE + FILTRATE
36.2% SOLIDS
47 gal
\
47 gal
V
\
47 gal
OXIDATION FIRST
TOWER STAGE
Grams dry solid per ml final settled volume.
2.2 min 2.2 min 2.2 min
SECOND STAGE
LIQUOR
TOTAL S AS SO, 1270
S02 190
C02 260
Ca 1430
Cl 2260
pH 5.5
SATURATION 1.07
SOLIDS (8%)
TOTAL S AS SO
S02
C02
Ca
3
455
257
78
309
OXIDATION • 0.29
UTILIZATION - 0.74
Figure 20. EPA pilot plant forced oxidation system tests without fly ash
-------
measurements are in substantive agreement witn each other as to the composition
of the solids. The second-stage slurry contained primarily hydrated calcium
sulfite, CaSO3- 1/2H2O, with a small amount of gypsum. Both the first-stage
slurry solids and the filtered solids were primarily gypsum with small amounts
(<1 percent and <6 percent, respectively) of Ca.SO$" 1/2 J^O. The filtered
solids contained more CaSO3' 1/2 f^O than did the first-stage slurry. How-
ever, the latter is not indicative of the basic characteristics of the process.
The presence of CaSOs' 1/2 H2O in the filtered solids, in view
of its virtual absence from the first-stage slurry, was an unprogrammed
plumbing modification whereby a portion of the second-stage slurry was sent
directly to the filter, bypassing the first-stage loop as a means of controlling
the percent solids in the second stage. SEM examination of filtered solids
from RTF tests of March 1-5, 1976, which were definitely known to have the
first-stage by-pass, also showed some CaSO3' 1 /2I^O, as expected, and
comparable tests of March 15-19, 1976, where the bypass condition was cor-
rected, showed only gypsum. As a result of this plumbing configuration, the
effect of calcium sulfite mixed with gypsum was established, as well as the
primary objectives of determining gypsum properties. The characteristics
of the gyp sum-calcium sulfite combination normally would not have been
determined.
8.2.1 X-Ray Diffraction Measurements
The crystalline phase of gypsum, CaSOx* 2H2»O was identified
in each sample of solids from the RTF pilot plant. Although precise quanti-
tative analyses are not possible with x-ray measurements, the major and
minor components of each of the samples were identified and are shown in
Table 25. Gypsum was the major crystalline phase of both first-stage slurry
and filtered solids and was a minor phase of the second-stage slurry. Calcium
sulfite hemihydrate was the major crystalline phase of the second-stage slurry
and was a minor phase of the filtered solids. The extent to which it could
normally be expected was not determined because of the nonstandard plumbing
configuration. Data included in Table 25 for samples removed from the col-
umns at the conclusion of leaching tests are discussed in Section 8.2.4.
As shown in Table 25, x-ray diffraction measurements verified
that each of the samples dried in ambient air contained gypsum. For those
samples dried at 90°C at atmospheric pressure or at 50° C under vacuum,
some or all of the gypsum was converted to the hemihydrate CaSO4* 1/2H£O.
In one case, there was an indication that some hemihydrate was formed during
ambient air drying. In contrast, the hemihydrate of calcium sulfite appears
to be unaffected by prolonged vacuum drying at 50° C.
8.2.2 SEM Measurements
SEM photography confirmed the presence of gypsum in the
filtered solids and the first-stage solids (Figures 21 through 26). The fiist-
stage slurry solids contained only gypsum crystals, while the filtered solids
100
-------
TABLE 25. CRYSTALLINE PHASES IN RTF OXIDIZED SLUDGES
IDENTIFIED BY X-RAY DIFFRACTION
Source of Sample
Sample Treatment
Major Constituent
Sampled 9/30/75, Contained Fly Ash:
Filtered Solids
Filtered Solids
1st Stage Slurry Solids
2nd Stage Slurry Solids
Filtered Solids
1 at Stage Slurry Solids
2nd Stage Slurry Solids
Dried in air @ 90°F
Ambient air dried
Ambient air dried
Ambient air dried
From leachate column
Dried in ambient air
From leachate column
Dried in ambient air
From leachate column
Dried in ambient air
CaSO.-1/2H.O
4 2
CaSO4-2H2Oa
CaSO4-2H2O
CaSO3-l/2H2O
CaSO4-2H2Oa
CaSO -2H,O
4 2
CaS03-l/2H20
Sampled 12/4/75, Contained No Fly Ash:
Filtered Solids
Filtered Solids
1st Stage Slurry Solids
1 st Stage Slurry Solids
2nd Stage Slurry
Ambient air dried
Vacuum dried at 50°F
Ambient air dried
Vacuum dried at 50°F
Vacuum dried at 50°F
CaS04-l/2H20
CaSO/1.2H0Oa
4 L
CaSO4-l/2H2O
CaSO3-l/2H2O
lSmall quantities of CaSO,- 1/2H O in sample.
-------
Figure 21. RTF first-stage slurry solids
containing predominantly gyp-
sum, fly ash, and small quan-
tities of calcium sulfite (sam-
pled 1/30/75), X1000
Figure 22. RTP first-stage slurry solids
containing predominantly gyp-
sum and small quantities of
calcium sulfite (sampled
12/4/75), X1000
-------
Figure 23. RTP second-stage slurry solids
containing predominantly cal-
cium sulfite and fly ash (sam-
pled 9/30/75), X1000
Figure 24. RTP second-stage slurry solids
containing predominantly cal-
cium sulfite (sampled 12/4/75),
X1000
-------
Figure 25. RTF filtered solids containing
gypsum, fly ash, and approxi-
mately 5-percent calcium sul-
fite (sampled 9/30/75), X1000
Figure 26. RTF filtered solids containing
gypsum and approximately
5-percent calcium sulfite,
X1000
-------
contained mostly gypsum with a small amount of calcium sulfite hemihydrate.
The second-stage slurry solids contained mostly calcium sulfite hemihydrate
with only a small amount of either gypsum or calcium carbonate. Fly ash
was evident as spherical particles in those samples (9/30/75) where fly ash
was included in the process.
8.2.3 Wet Chemical Analyses
The chemical analyses of the dried sludge solids are shown in
Table 26. As a result of the sludge being oven-dried prior to analysis, the
gypsum was converted to the hemihydrate; therefore, the calcium sulfate
analyses are reported as CaSO^ 1/2H2O.
As observed in the SEM photographs, the filtered solids and
the first-stage solids were predominantly sulfate and the second stage pri-
marily calcium sulfite. Although the filtered solids contained primarily the
gypsum phase of the sulfate, approximately 5 percent calcium sulfite was ob-
served in each of the two samples analyzed. The presence of sulfite was not
due to incomplete oxidation in the first-stage loop but was determined to be
the result of an unscheduled modification of the plumbing whereby a portion
of the second-stage slurry bypassed the first-stage loop.
The quantitative results of the wet chemical analyses verified
conclusively that complete oxidation of calcium sulfite to gypsum was achieved
in the first-stage loop.
Calcium, magnesium, chloride, sulfate and total dissolved
solids (TDS) were reported for the filtrates from both samplings. The re-
sults are shown in Table 27. Mass and charge balance calculations were
also included and the results expressed as ratios in Table 27. The close
proximity to unity of these ratios corroborates the results of the chemical
analyses. Since the analytical data appear to indicate saturation of the filtrates
with gypsum, a solubility ratio was also determined. It was defined as the
product of the measured calcium and sulfate ion concentrations divided by the
theoretical solubility product constant corrected for the effect of ionic strength.
These ratios also are approximately unity, thus verifying gypsum saturation.
8.2.4 Leaching Test Results
Leaching test results of RTF limestone-scrubbed FGC gypsum
wastes from sludge samples of September 30, 1975 (with fly ash) and Decem-
ber 4, 1975 (without fly ash) are outlined in Table 27.
On the basis of the drying conditions used in the TDS determi-
nation, the CaCl2' 2H2O in the solids was not dehydrated when the sample was
brought to constant weight; therefore, a correction was applied to reduce the
TDS by the moles of H2O in the hydrated CaCl?-
The results for the filtered solids and first-stage slurry solids
show that, after two to three pore volumes of water have passed through the
105
-------
TABLE 26. COMPOSITION OF RTF OXIDIZED SLUDGE
BY WET CHEMICAL ANALYSIS
Sample Designation
Sampled 9/30/75.
Contained Fly Ash
First-Stage Slurry
Second-Stage Slurry
Filtered Solids
Sampled 12/4/75.
Contained No Fly Ash
First-Stage Slurry
Second-Stage Slurry
Filtered Solids
Total Solids
(% of Slurry)
11.3
12.4
64.9
b
Composition of Solids, Percent by Weight, Dry
Fly Ash
40.0
38.4
38.6
CaSO4' 1/2 H2Oa
62.5
7.8
52.7
94.6
11.8
86.4
CaSO3- 1/2 H2O
<0. 5
41. 4
5.7
0.7
51.8
9.5
CaCO3
<0. 5
11.4
4.9
8.0
37.0
aDrying process converted gypsum to CaSO4' 1/2 HLO.
b
Not determined.
-------
TABLE 27. ANALYSIS OF LEACHATES AND FILTRATES OF
RTF FORCED OXIDATION SAMPLES
Sample and Date
Firat-Stage
Slurry Solids,
9/30/75
Firat-Stage
Slurry Solids.
12/4/75
Second-Stage
Slurry Solids,
9/30/75
Second-Stage
Slurry Solids,
12/4/75
f iltered
Solids,
9/30/75
Filtered
Solids,
12/4/75
Filtrate,
9/30/75
Filtratt.
12/4/75
Contained
Fly Aah
Yes
No
Yes
No
Yes
No
Yes
No
No. of PVD
0.7
26.4
1.0
10.3
0. 3
9.6
0.9
22. 1
0.6
11. 1
0. 5
1-1. 1
_ _
_ _
Concentration, mg/4
TDS
6820
2180
5860
2160
6770
300
5070
330
7100
2190
10100
2310
9230
10490
TDS
Corrected
5840
3140
4775
2120
5855
270
42VO
300
6010
2130
8100
J2oO
7595
HM5
£b
5140
2130
5110
1620
5270
310
4290
380
s -I 70
2065
7940
2100
7290
7750
T)C
0.88
0. 99
1.07
0. 76
0.90
1.14
1. 00
1. 27
0. 91
0.97
o. 98
u.93
0. 'ttj
0. VI
Ca
1300
550
1280
540
9bO
120
1080
100
1220
560
2170
500
1 -I2U
Z200
so,
1540
1500
1450
1000
ivio
150
1500
220
1600
I-1UO
1250
1500
1800
1130
Cl
2000
80
2160
80
1870
40
1530
60
2220
'.'0
•1040
100
3330
wo
Mg
300
0. 3
220
0.04
S40
0. 3
180
0. 3
•130
15
480
0.2
74 U
•140
Ionic
Strength
0. 15
0.08
0. 14
0.05
0. 16
0.01
0. U
0.01
0. 16
0. On
0.23
0. Ot>
0.22
0. 23
SolubUity
Product
d
Ratio
1.16
1. 19
1.12
0. 92
0. 97
0. 08
1. 17
0. 10
1. 03
1. 1 *
0. ''4
1. U
0. 90
0. 89
o
-J
aM-asured TDS corrected for 2H,O in CaCl -2H O.
i
L - sum of Ca, SO,, Cl, and Mg ions.
cMass ratio: L/TDS (corrected.
(Ca) X (SO.)/K_ __ with K_ __. corrected for ionic strength.
4 c*ad(-? . ^*ao(_/.
KCaC03 w" uaed-
-------
samples, gypsum, the principal ingredient of both materials, was being
dissolved and produced a leachate that was saturated with calcium and sulfate
ions.
Each one of the samples of Table 27 was saturated with gypsum,
except for two samples of solids from the second-stage slurry at 10 and 22
pore volume displacements (PVD) whereby the gypsum supply had been essen-
tially depleted. These two were saturated with Ca.SO$. From this point on,
the calcium was supplied by the dissolution of hydrated calcium sulfite,
although the anions found in the leachate were sulfate because of the rapid
oxidation of sulfite ions.
The ratios of the products of the calcium and sulfate molar con-
centrations to the estimated solubility product constants (corrected for the
ionic strength) have been tabulated in the last column of Table 27. It is shown
that these ratios were virtually constant and ranged between 0.9 and 1. 2.
For the two second-stage slurry samples, the solubility product constant of
CaSO3 was used. In these two cases, the low values are attributed to very
small amounts of residual gypsum in the beds which were still contributing
sulfate to the leachates.
In assessing the overall material balance at a PVD of less than
1.0, i.e., TDS versus the summation of the Ca, SO^, Cl, and Mg ions, it is
apparent that considerably better agreement was obtained with those samples
without fly ash than those that had fly ash in the scrubber slurry. This indi-
cates that the major species were covered by Ca, 804, Cl, and Mg for the
no-fly-ash case, whereas some significant additional constituents were prob-
ably leached when fly ash was present. It is also apparent that, after many
PVD (>10), the concentration of the major species was virtually the same
whether or not fly ash was initially present.
The concentrations of TDS and major constituents in leachate
from the first-stage slurry solids are plotted against pore volume displace-
ment in Figure 27.
In both samples, the TDS in the leachates reached the same
minimum gypsum (saturation) value after about three PVD. The value was
unchanged even when the leaching was continued until more than 30 pore vol-
umes had been displaced. The concentrations of magnesium and chloride
were reduced as leaching progressed, but sulfate remained relatively con-
stant and calcium was only partially removed, as shown in Table 27. The
computed solubility ratios also supported the fact that each of the leachate
samples was saturated with gypsum. During leaching, the soluble magnesium
and calcium chlorides were washed out of the bed with the residual liquor
leaving the crystalline gypsum to saturate the leaching water. The ionic
strength decreased as the chlorides were removed, thus reducing the solu-
bility of gypsum.
The results of leaching tests for the filtered solids were simi-
lar to those for the first-stage slurries (Figure 28). The TDS leveled out at
the same saturation value of about 2200 mg/J .
108
-------
80001-
6000
o
4000
o
o
2000
CONTAINED
NO FLY ASH
O
D
CONTAINED
FLY ASH
IDS
Cl
Mg
10 20
AVERAGE PORE VOLUME DISPLACEMENT
Figure 27. Concentration of major species in leachate of first-
stage slurry solids of RTF oxidized sludges
109
-------
8000 -
6000 -
4000
o
o
2000
CONTAINED CONTAINED
NO FLYASH FLY ASH
5 10
AVERAGE PORE VOLUME DISPLACEMENT
Figure 28. Concentration of major species in leachate from
filtered solids of RTF oxidized sludges
110
-------
It is apparent from the plots of Figure 29 and from the data of
Table 27 that the leachate characteristics of the second-stage slurry solids
illustrate the different composition of the solids. The same initial pattern
of changing TDS with PVD that was observed for the first-stage and filtered
solids samples only persists to about five pore volumes. The TDS concen-
trations (Figure 29), particularly the one for the sample that contained no
fly ash, began to level off at the gypsum saturation content of about 2200 mg/t
and then decreased to about 300 mg/t when 20 pore volumes were displaced.
The sulfate concentrations in the final leachate samples from the second-
stage slurries were about 1/10 of the amount in the final leachates of the
filtered solids or the first-stage slurries. Therefore, the solubility ratios
for these leachate samples are approximately 0. 1, as shown in Table 27 and
appear to correspond to saturation of the leachate with calcium sulfite
hemihydrate.
The inflection points on the TDS curves for the solids from the
second-stage slurry solids correspond to the gypsum being depleted. Initially,
the available gypsum in the leachate was 8 and 12 percent, with and without
fly ash, respectively (Table 26), and was ultimately depleted, leaving only
calcium sulfite hemihydrate. This was confirmed, as shown in Table 25, by
x-ray diffraction measurements on the leachate bed solids made at the con-
clusion of the leaching tests.
8.3 PHYSICAL CHARACTERISTICS
Measurements were made of permeability coefficients, void
fraction, water retention, unconfined compressive strength, and load-bearing
strength of fly-ash-free gypsum, calcium sulfite, and gypsum mixed with
5 to 10 percent calcium sulfite. The effects of fly ash on the properties of
the various materials were also determined.
8.3.1 Permeability Coefficients
-4
Permeability coefficients of about 1x10 cm/sec were re-
ported for all the samples (Table 28). These values were determined for
beds that were constructed by loading columns with slurries of the solids and
allowing the solids to settle and drain. The lowest value of 5 X 10"5 cm/sec
was obtained for a bed that was prepared by "rodding" the slurry during the
initial drainage. These results indicate that the second-stage slurry solids,
which are primarily calcium sulfite hemihydrates, showed slightly lower
permeability coefficients than the samples which were predominantly gypsum.
8.3.2 Density
Samples were prepared by casting slurries in cylindrical con-
tainers with perforated bottoms and allowing the settled and drained samples
to dry for several days at ambient temperature until the samples were capable
of maintaining structural integrity. After weighing, the samples were oven-
dried at 50°C to obtain moisture contents. From these measurements and
:n
-------
80001—
7000 r
CONTAINED
NO FLY ASH
O
D
A
0
CONTAINED
FLY ASH
TDS
Cl
r°4
Ca
Mg
10 20
AVERAGE PORE VOLUME DISPLACEMENT
Figure 29. Concentration of major species in leachate of
second-stage slurry solids of RTF oxidized
sludges
112
-------
TABLE 28. PHYSICAL CHARACTERISTICS OF RTF OXIDIZED SLUDGES
Sample Designation
Moisture
Content,
Density.
g/cm3
(wet)
Sampled 9/30/75. Contained Fly Ash
First- Stage Slurryc
Second-Stage Slurry*1
Filtered Solids*
Sampled 1Z/4/75. Co
First-Stage Slurry
Second- Stage Slurry
Filtered Solids
10
14
18
ntained No
15
14
18
1.4 '
1.2
1.5
Fly Ash
1.3
1.0
1.35
Unconfined
Compressive ,
Strength, kg /cm
(wet)
4.4
1.1
1.4
4.2
1.8
(dry)b
r
0.7
0.5
0.9
3.0
1.9
1.6
Density,
g/cmj
(dry)
1.3
1.0
1.2
1. 1
0.8
1.1
Pore
Volume
Fraction
0.50
0.60
0.50
0.55
0.65
0.55
Permeability
Coefficient,
cm/seci
1.6X 10"4
8. IX 10"5
_a
4. 5X 10"s
1. IX 10'4
9.6X 10"5
1. IX 10'4
prepared by "redding" slurry during initial drainage
'Test specimen was
50*C, vacuum.
cGypsum sludge.
dSulfite sludge.
eGypsum sludge with 5 percent sulfite "impurity.'
-------
corresponding dimensional measurements, the wet and dry densities shown
in Table 28 were determined. Densities of the dried samples were generally
about 20 percent lower than the densities measured before drying. Pore vol-
ume fractions computed from the dry densities ranged from 0.50 to 0.65, of
which from 30 to 50 percent remained filled with moisture when the cast sam-
ples were allowed to drain and dry in the ambient air until firm enough for
compressive strength measurements. Both, for samples with and for those
without fly ash, the second-stage solids showed the largest pore volume frac-
tions and the lowest densities, but the filtered solids retained the largest per-
centage of water. The pore volume fractions of the samples that contained no
fly ash were slightly higher and their densities slightly lower than for the cor-
responding samples that had no fly ash.
8.3.3 Unconfined Compressive Strength
The unconfined compressive strengths of wet and dry samples
are shown in Table 28. A comparison of the results for the two sets of sam-
ples shows that when wet, the corresponding members of both sets had com-
parable strengths, but when totally dry the samples that contained no fly ash
had higher compressive strengths. Furthermore, for both sets of samples
the first-stage solids when wet showed substantially higher compressive
strengths than did the second-stage slurry solids or filtered solids.
The analyses of the filtered solids samples showed approxi-
mately 5 percent calcium sulfite hemihydrate (Table 26) with the first-stage
solids containing only negligible amounts. It is believed that the presence
of this hemihydrate in the filtered solids is responsible for the greater re-
tention of water and for the lower compressive strengths relative to the first-
stage solids.
8.3.4 Load-Bearing Strength
Load-bearing strengths were measured on both samples of
filtered solids, with and without fly ash, using an adaptation of the standard
bearing ratio test for soils, ASTM D-1833-73 (Ref. 11). These gypsum sam-
ples, which contained 5 percent sulfite, exhibited strength characteristics of
sludge containing typically sulfite-sulfate ratios of 3 to 4.
Measurements were repeated after further de water ing of the
solids. Results for the filtered solids with and without fly ash are plotted
in Figure 30. Although the measurements were terminated prematurely for
the sample containing fly ash, the curves are similar for the two samples
except that the rapid increase in strength with increasing solids content oc-
curs at about 10 percent higher solids content for the sample containing fly
ash.
8.4 . COST ESTIMATES OF GYPSUM-PRODUCING PROCESSES
Engineering cost estimates were made for producing wallboard
grade gypsum from forced oxidation of limestone-scrubbed flue gas desulfuri-
aation (FGD) sulfite-rich wastes. Two basic processes were considered:
114
-------
175
150
125
100
CD
LU
OC
co
CD
§ 75
op
s
o
50
25
o NO FLY ASH
A WITH FLY ASH
60
65 70
SOLIDS CONTENT, WEIGHT %
75
Figure 30. Load-bearing strengths of dewatered, filtered solids
from RTF oxidized sludges containing 5-percent sulfite
115
-------
a tail-end system incorporating an oxidation tower and an integrated system
which represents the pilot plant illustration of Borgwardt at RTF, which was
tested at the TVA Shawnee Plant with a 10-MWe venturi- spray tower scrubber
(Figures 31 and 32). The estimates are meant to illustrate the range of
costs for saleable-grade gypsum produced from SO 2 abatement processes.
The economics were also evaluated for converting sulfite-rich flue gas clean-
ing (FGC) wastes to a high gypsum content for environmentally sound disposal
above ground or in a landfill.
I Results of the calculations indicated that for a 500-MWe tail
/ end forced oxidation system, total capital costs are approximately 20 percent
/ greater than for a base case system that includes particulate removal by an
1—^electrostatic precipitator (ESP) and a limestone scrubber system producing
sulfite-rich sludge. A schematic of the basic particulate removal and scrub-
ber system considered as a baseline and the major additional equipment to
oxidize the sulfite-rich slurry are depicted in Figure 31. It illustrates a two-
stage (venturi-spray tower) limestone wet-scrubber system with a tail-end
oxidation tower incorporated to oxidize the sulfite-rich slurry to gypsum. In
order to account for the production of saleable material with a sufficiently
low moisture content, centrifugation was included. Capital equipment costs
for the integrated .system (Figure 33) were estimated to be about 13 percent
more than the base case scrubber. This configuration is approximated by
the pilot plant system tested by Borgwardt of EPA (Ref. 14), which is cur-
rently being evaluated at the TVA Shawnee Power Plant, Paducah, Kentucky,
by EPA, TVA, and Bechtel at the 10-MWe level.
The base case scrubber capital costs for a 500-MWe scrubber
system, in 1977 dollars, were estimated at $94/kW. The figure excludes
disposal costs and was the average of five units in the 450- to 550-MW range
reported in Reference 15 and corrected to 1977 costs based on methods;used
in Reference 3.
The major pieces of additional equipment required for oxidizing
the sludge with the tail-end system includes oxidation towers, pumps, air
compressors, and centrifuges. Individual equipment costs were taken from
Reference 16 and corrected to 1977 based on factors obtained from Refer-
ence 17. The total increase in costs for these items, for the tail-end system
which is shown in Table 29, was $10.14 X 10°. A significant savings in
equipment costs can be realized with the integrated system; the correspond-
ing increase in capital equipment costs above the baseline is $6.64 X 106,
(Table 29). -
Annual charges for the oxidation;-related equipment based on a
30-year life for the equipment were taken as 18 percent of the capital costs
(Ref. 3). Annual operating costs of $2.87 X 10& were based on data from Ref-
erence 18 and adjusted to 1977. Limestone costs were estimated at $6/ton,
and for a plant operating load factor of 50 percent, totalled $0. 75 X iO^/year.
The base case capital and operating costs are summarized in Table 30.
Gypsum-production costs discussed in subsequent paragraphs are assumed
116
-------
OXIDATION
TOWER
FLUE
GAS
AIR
COMPRESSOR
ELECTROSTATIC
PRECIPITATOR
(ESP)
MAJOR ADDITIONAL
EQUIPMENT REQUIRED
IS SHOWN WITHIN
DOTTED LINES
Figure 31. Simplified process schematic for wet-limestone tail-end scrubber
forced-oxidation gypsum-producing system
-------
NOTE:
MAJOR ADDITIONAL
EQUIPMENT REQUIRED
IS SHOWN WITHIN
DOTTED LINES
DEMISTER
WASH
FLUE
GAS
ELECTROSTATIC
PRECIPITATOR
(ESP)
FLY ASH
r
00
±
jCENTRIFUGEH-
•^
GYPSUM
ICKENER
MAKE-UP
WATER
VENTURI
SCRUBBER
I J
I
I AIR
(COMPRESSOR
I
I
OXIDA-
TION
TANK
HOLDING
TANK
•*•
LIMESTONE
SLURRY
Figure 32. Simplified process schematic for wet-limestone integrated
forced-oxidation gypsum-producing system
-------
10
9
8
7
' 6
5
4
3
500-MWe PLANT, 30-yr LIFE
50% AVERAGE OPERATING LOAD FACTOR
WET LIMESTONE SCRUBBING, FORCED OXIDATION
-1
-2
-3
LIMESTONE
UTILIZATION
U.S. MINED CRUDE
GYPSUM, FEB 1975
7
8
9
10
11
12
FGC WASTE TREATMENT AND DISPOSAL COST AVOIDANCE,
$/ton dry sludge (including ash)
Figure 33. Coat.of producing saleable gypsum as a function
of sludge treatment and disposal credit, 1977
dollars
119
-------
TABLE 29. COST OF MAJOR COMPONENTS OF FORCED OXIDATION.
WET LIMESTONE SCRUBBING SYSTEM FOR A 500-MWe
PLANT, 1977 DOLLARS*'13
Equipment
Oxidation Towers (15 X 50 ft, 2 each)
Air Compressors (30% spare
capacity)0
Pumps (50% spare capacity)0
Scroll-type Centrifuge (25%
spare capacity)0
Piping (estimate, 10%)
Oxidation Tank (20, 000 gal, 2 each)
Equipment
Life, yr
15
15
10
15
15
Total
Tail -End
System
Total Cost
$ 4. 20 X 106
3.75 X 106
0.52 X 106
0.75 X 106
0.92 X 106
Not req'd
$10. 14 X 106
Integrated
System
Total Cost
Not req'd
3.75 X 106
0. 52 X 106
0.75 X 106
0.92 X 106
0.70 X 106
6. 64 X 106
Ref.
16
16
16
16
16
a30-year plant life; estimate includes installation.
Dry gypsum produced: 1.37 X 105 tons/yr; slurry pumped: 500 gpm.
cSpare capacity based on a minimum of 7000 hr/yr operation.
-------
TABLE 30. BASE CASE COSTS FOR WET LIMESTONE
SCRUBBING SYSTEM FOR A NEW 500-MWe
PLANT IN 1977 DOLLARS*
Capital Costs at $94/kW
Annual Charges at 18%
Operating Costs
Limestone at $6 /ton
Total
$47.00 x 106
--
--
--
Totalb
$/Ton Coal
Mills /kWh
Annual
$ 8.46 X 106
$ 2. 87 X 106
$ 0. 75 x 106
$12.08 X 106C
$12.57
$ 5.53
Ref.
15
3
18C
18
a30 year life; 50% average operating load factor (4380 hr/yr); tons
coal burned, 9. 6 X 105 tons/yr (3. 5% S, 12% ash, 0. 88 Ib/kWh;
kWh generated, 2. 18 X 109/yr; dry sludge, 1. 93 X 10* tons/yr
(includes 0. 87 X 105 tons/yr fly ash); 90% SO2 removal; 0.65
limestone utilization factor; limestone consumption, 1.25 X 10s
tons/yr.
bEscalated at 6% annually from 1973 estimate for 500-MWe plant.
°Does not include FGC waste disposal costs.
121
-------
to be independent of the base case capital and annual operating costs, inasmuch
as gypsum modifications are estimated and presented on an incremental basis.
On this basis, the effect of forced oxidation can be compared directly to the
waste disposal costs by chemical treatment and landfill or by a number of
other disposal options.
The estimate of $10.14 X 10 results in a cost of $17. 51 and
$15.62 per ton of dry gypsum for a limestone utilization factor of 0. 65 and
of 1.0, respectively (Table 31). It does not consider the cost of disposing of
the waste which would be incurred if the sludge were not oxidized. Based on
a midrange value of $9. 00/ton, dry, for FGD sludge chemical treatment and
disposal (Ref. 3), which would offset the gypsum processing costs, and
assuming a $3. 00/ton fly ash disposal cost, the net cost of producing gypsum
in a tail-end system is $6. 73 and $4.84/ton dry gypsum for the 0.65 and
1.0 limestone utilization factors, respectively (Table 32). The effect of im-
proved limestone utilization is highly significant in reducing the net gypsum
cost. A larger effect in reducing the production cost is realized by using an
integrated system. With the economies resulting with that mode of operation,
the estimates are $2. 22 and $0. 31 per dry ton, respectively, for utilization
of 65 and 100 percent. Tests have shown (Ref. 19) that a utilization of virtually
100 percent can be achieved with the process and that goal should be con-
sidered in the oxidation process. The 65 percent factor is provided for both
methods to place the costs in perspective. To be more realistic, the gypsum
may contain as much as 8 percent moisture and therefore would cost approxi-
mately 8 percent less per ton of moist gypsum.
Operating tests have also indicated that the settling properties
of gypsum are significantly superior to those of calcium sulfite. Therefore,
smaller thickeners can be used, which reduces plant capital costs. This re-
finement has not been included in the estimate. In addition, better, trouble-
free scrubber operation has been reported with a forced oxidation system,
providing further potential areas for reducing the operating cost attributable
to forced oxidation.
A comparison of mined gypsum cost with forced oxidation
saleable gypsum costs that include credit for sludge disposal is shown in Fig-
ure 33. For the tail end oxidation systems and 100 percent limestone utiliza-
tion, a $9/ton chemical treatment and disposal avoidance cost results in a
gypsum cost comparable to the 1975 fob price of $4.80/ton for domestic gyp-
sum (Figure 33). At lower disposal costs, forced oxidation economics of tail-
end systems are not as favorable, and at greater than $9/ton they become
increasingly attractive. Over the entire range of sludge disposal cost avoid-
ance of $7 to $11/dry ton of sludge, gypsum from the integrated system is
competitive with the mined gypsum.
An assessment was made of these costs to determine gypsum
disposal costs which when added to the cost of forced oxidation would equal
the cost for disposal of a comparable quantity of sulfite sludge. The results
are plotted in Figure 34 which portrays the values for tail-end and integrated
122
-------
TABLE 31. TOTAL INCREMENTAL COST OF GYPSUM-PRODUCING
SYSTEM FOR A 500-MWe PLANT IN 1977 DOLLARS*
(AS COMPARED TO SCRUBBING)
Forced Oxidation ~
Cost Increment
Capital Coat
Annual Charge at 18%
b
Annual Labor
Limestone
Total
Tail End
10.14 X 106
—
_ _
--
Integrated
6. 64 X 106
--
_ _
--
Total
$/Ton Coal
Mills /kWh
$/Ton Dry Gypsum
Annual
Tail End
Integrated
Utilization Factor
0.65
—
1.82 X 106
0.58 X 106
d
2.40 X 106
2.50
1.09
17.51
1.0
—
1.82 x 106
0. 58 X 106
(0.26 X 106)
2. 14 X106
2.22
0.98
15.62
0.65
--
1.20 x 106
0.58 x 106
d
1. 78 X 106
1.85
0.82
13.00
1.0
--
1.20 X 106
0. 58 x 106
(0.26 X 106)
1.52 x 106
1.58
0.70
11.09
t\>
CO
* Relative to wet limestone scrubbing without waste disposal, 30-year life plant, 50% average load factor.
Prorated on the basis of scrubber system operation.
CLimestone used at 1.0 utilization factor = 0. 81 X 10 tons/yr. ^
Included in scrubber costs.
-------
TABLE 32.
ESTIMATED INCREMENTAL COST OF PRODUCING GYPSUM
FOR SALE OR DISPOSAL RELATIVE TO THE DISPOSAL OF
CHEMICALLY TREATED FGC WASTE, IN 1977 DOLLARS*
Item
Limestone Utilization Factor
Gypsumb
Ash Disposal6
Sludge Treatment and
Disposal Cost Avoidance at
$9.00/Dry Ton Sludge with
Ash
Cost Avoidance; Filtering
Total Incremental Cost of
Gypsum /Ton (Dry):
$/Ton Coal
Mills /kWh
Saleable (Ash -Free)
Tail
0.65
$17.51
1.90
(12.68)
N/A
$ 6.73
0.96
0.42
End
1.0
$15.62
1.90
(12.68)
N/A
$ 4.84
0.68
0.30
Integrated
0.65
$13.00
1.90
(12.68)
N/A
$ 2.22
0.32
0.14
1.0
$11.09
1.90
(12.68)
N/A
$ 0.31
0.04
0.02
Disp
(Indue
Tail
End
1.0
$ 9.55
N/A
(12.68)
(0.55)
$(3.68)
(0.52)
(0.23)
os able
pig Ash)
Integrated
1.0
$ 6.79
N/A
(12.68)
(0.55)
$(6.44)
(0.90)
(0.40)
aAll costs are converted to $/ton dry gypsum except as noted; 500 MWe plant, 30-yr life,
50% average load factor, wet limestone scrubbing, forced oxidation.
Gypsum-sulfite sludge quantities, dry ton/yr:
Dry Ton/Year
Ash-Free With Ash
Gypsum 1.37x10* 2.24X10?
Sulfite Sludge 1. 06 X 105 1.93 X 105
cEstlmated at $3/ton dry ash and converted to $/ton dry gypsum.
-------
to
O
O
12
11
10
9
8
7
6
5
4
3
S* 2
00 C i
.3 1
~ o
CO 1=
o 2
0- o
to c
500-MWe PLANT, 30-yr
LIFE,50% AVERAGE
LOAD FACTOR,
WET LIMESTONE SCRUBBING,
FORCED OXIDATION, 100%
LIMESTONE UTILIZATION
INTEGRATED
TAIL-END
NOTE: COST OF FORCED
OXIDATION PLUS GYPSUM
DISPOSAL BREAK-EVEN
EQUALS SULFITE SLUDGE
DISPOSAL COST
78 9 10 11
'SULFITE SLUDGE DISPOSAL COST,
$/ton dry sludge (including ash fly)
12
Figure 34. Gypsum disposal cost allowable to limit forced oxidation
to cost of disposal of comparable quantity of sulfite sludge,
1977 dollars
125
-------
oxidation systems. Data consistent with the approach shown in Table 32 were
used to determine Figure 34. In that figure, it can be seen, for example, that
if sulfite sludge disposal costs $9/ton of sludge (dry, including ash), the break-
even cost for disposal of gypsum (dry, including ash) from an integrated sys-
tem would be approximately $6/ton. From the standpoint of disposal only,
any cost of disposal for gypsum in excess of $6/ton would make forced oxida-
tion more expensive than sulfite production and disposal. For a tail-end sys-
tem, the gypsum disposal break-even cost would be about $3. 50 if sulfite-
rich sludge disposal is $9. 00 (Figure 34). These comparisons can be repeated
for any sludge disposal cost between $6 and $l2/ton of sludge, using Figure 34.
8.5 GYPSUM COST SUMMARY
In summary, the cost of producing saleable gypsum in integrated
systems results in generally favorable economic trends relative to chemical
treatment and disposal costs of $7 to $11/ton of sludge (dry). Significant to
this fact is a $2/ton reduction in the cost of producing the gypsum, wt.jh can
be achieved by improving limestone utilization from 65 to 100 percent. Land-
fill disposal of gypsum-fly ash mixtures appears to be marginally competitive
with disposal costs at $8/dry ton; gypsum economics improve with increasing
disposal costs.
•
For tail-end systems operating with 100 percent limestone
utilization, the cost of producing saleable gypsum is marginally competitive
if disposal costs are in the $8 to $9/ton range. Limestone utilization of less
than 90 percent tends to make this method of utilization noncompetitive with
chemical treatment. The economics of disposal of gypsum-fly ash mixtures
from a tail-end system do not appear to be competitive with chemical
treatment.
126
-------
SECTION IX
EVALUATION OF THE ENVIRONMENTAL
ACCEPTABILITY OF FGD SLUDGE
9. 1 INTRODUCTION
The data presented in this study pertain to the phenomenological
behavior of flue gas desulfurization (FGD) sludges that is expected in field
disposal. The behavior has been separated into chemical and physical pro-
perties, and experiments have been conducted in a manner so as to simulate
as closely as practicable the environmental conditions expected in actual
field disposal. Thus, the range of experiments has included the disposal
conditions considered pertinent from environmental and economic points of
view.
This section of the report discusses the expected impact of
FGD sludges on the environment with respect to the range of properties
observed in this study. It is presumed in these discussions that the range of
properties observed in the sludge samples are representative of the sludges
that are being and will be produced. In the sampling of this study, an
attempt was made to provide representation 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 require disposal.
In order to evaluate the environmental acceptability 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 disposal 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 impact that is inherent in each
of the various disposal techniques. The assessment of the potential pollution
impacts includes considerations of die chemical state of the pollutant and the
route by which it may enter the environment.
127
-------
9.2 BACKGROUND
In previous reports (Refs. 1 and 2), the various potential
routes of chemical pollution from disposal sites and some of the alternative
methods by which FGD may be disposed were discussed. The following sub-
sections review the highlights from these reports and provide an assessment
of their technical and economic viability for FGD waste disposal.
9.2.1 Alternative Routes of Pollution
Several potential routes of environmental pollution have been
identified, but generally they are reduced to either pollution of air or
pollution of water. Experimental investigation showed that pollution of air
by vapor species, either volatilization or dissociation of sulfite or sulfate,
cannot occur even in extreme climatic conditions. A form of airborne
pollution can exist by wind-whipped spray from the surface of a disposal
pond, but this form of pollution is generally localized and limited in effect.
Pollution of water is by far a more likely source of potential
environmental pollution either from surface runoff or from leaching of the
disposal basin with rainwater. The assessment of data on runoff suggests
that a siltation pond be used during placement of treated FGD waste and that
overburden be used to prevent runoff pollution from erosion attack by rain,
freeze-thaw, or wet-dry cycles.
The most serious form of pollution potential exists as a con-
sequence of the action of rainwater leaching through a disposal basin.
Leaching data from laboratory experiments show that the concentration of
major species in the leachate is rapidly reduced in both chemically treated
and untreated sludges, but the absolute concentration of the treated material
is about one-half that of untreated material initially, after which they tend
to converge. A steady-state concentration of total dissolved solids (TDS)
in the leachate is maintained after about five pore volumes displacement
(PVD) at 2000 to 2200 mg/i and .primarily is representative of the equili-
brium dissolutions of gypsum. Trace elements generally were reduced to
values below detection limits during the initial pore flushing of the column,
and no appreciable effect of chemical treatment could be discovered. Only
lead, selenium, and zinc were ever found in detectable quantities after
steady state was established. The leachate quality of columns controlled in
aerobic conditions differed in a predictable manner from an anaerobic
leachate, but neither created a leachate markedly different from the other.
The primary advantage of chemical treatment is the conver-
sion of the kludge to a load-bearing material and a reduction of the coefficient
of permeability by at least one order of magnitude. When placed in a land-
fill it can be contoured to prevent the collection of rainwater so that the
potential for seepage is eliminated or greatly minimized. If the placement
of the material is such that rainwater and sludge liquor are collected on its
128
-------
surface or if some seepage occurs during a runoff condition, the reduction
of both the permeability and concentration of major species reduces the
mass release of sludge constituents to the subsoil when compared to
untreated sludges.
9.2.2 Alternative Disposal Techniques
9.2.2.1 Ponding
The method that represents the least deviation from state-of-
the-art fly ash disposal is direct ponding into a disposal basin. The environ-
mental impact of pond disposal is strongly dependent upon the ability (1) to
contain the components of a sludge so as to prevent environmental pollution
and (2) 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 environ-
ment can be protected from chemical pollution principally from leachate
contamination of groundwater, by lining the pond basin with clay, imper-
meable soil cements, or synthetic sheeting material. Also, some natural
clay deposits have sufficiently low permeabilities that sludge disposal
can be safely contained in a natural basin.
While chemical pollution may be avoided through proper pond
management, ultimate pond retirement may not be effected in an environ-
mentally sound manner. Although a pond may be evaporated and possibly
air-dried and covered with overburden, an acceptable procedure of this sort
has not been demonstrated. Moreover, the question of land reuse or
eventual pollution by a breach in a pond liner cannot now be answered.
9.2.2.2 Chemical Treatment
The method by which FGD sludge is treated chemically to
increase its structural stability and to reduce leachable chemical com-
ponents or reduce its permeability is offered by several commercial pro-
cessors. Chemically treated sludges can be used as landfill in both sub-
merged and above-grade conditions. For any chosen disposal condition,
chemical treatment provides improvements in the environmental accept-
ability of FGD wastes. While the environmental acceptability of treated
sludges appears to be favorable under most disposal methods, the added
environmental assurance afforded by the chemical process increases the
cost of disposal.
9.2.2.3 Other Methods
An alternative method of sludge disposal is one which
oxidizes sulfite sludges to form gypsum, which can then be dewatered to a
high solids content (greater than 75 percent). This material is easy to
handle after dewatering and when stored or disposed of above ground, it
tends to shed water, generally, with a minimum amount of erosion of the
129
-------
gypsum. Control of runoff may be required. An evaluation of this method
is being conducted at the U.S. Environmental Protection Agency/Tennessee
Valley Authority (EPA/TVA) Shawnee Test Site in Paducah, Kentucky
(Ref. 4).
Another alternative method of FGD waste disposal is by
pond under drainage. This method requires a closed basin in which the
waste at any water content may be placed, and excess water is returned to
the scrubbing system from an underdrainage system. A detailed descrip-
tion of this method is presented in the following section and describes the
work in this report period. A field evaluation of this process is discussed
in Ref. 4.
Also, ocean and mine disposal are being investigated in other
EPA studies. References 5 and 6. Under appropriate conditions, it appears
that these methods can be practical.
9.3 THE UNDERDRAINAGE METHOD FOR
DEWATERING FGD SLUDGE
A simple, economical, and widely applicable method may
exist for changing a scrubber sludge impoundment from what would other-
wise be a structural and environmental liability to a condition acceptable
for landfill and other ultimate uses. The method requires that, prior to
introduction of the sludge, the location for the impoundment be provided with
unde rdrainage facilities for removal of the sludge seepage and rainfall as
it collects subsequently and drains. Because no accumulation of either
liquor or leachate is provided to a hydraulic head, no soluble sludge con-
stituents are carried into the soil and ultimately to the underground water
table. Therefore, pond liners or impervious clay are not necessary pre-
requisites of a sludge impoundment except in rare cases of high soil per-
meability or high water tables. When the impoundment is filled, it can be
retired, equipped with a covering of soil, and the land can then be reclaimed
for other uses. So far as it is now known, this method is applicable to all
the sludges now being produced by current scrubbing processes. This
method is now being evaluated at TVA Shawnee as part of the EPA-
sponsored sludge disposal field evaluation project. Two ponds provided
with facilities for underdrainage have been filled, one with lime process
sludge and the other with limestone process sludge. These two ponds are
now being monitored, along with five others previously filled with either
untreated or chemically fixed sludge. The data presented herein represent
the results of limited laboratory tests conducted immediately prior to the
filling of the ponds at Shawnee. The purpose of the tests was to identify
optimum compositions and conditions for producing a structurally suitable
ponded material. Tests were not designed to study the effects of all
parameters, and not all of those parameters which were investigated were
systematically varied so as to obtain the most definitive conclusions.
130
-------
9.3.1 Results
The load-bearing strength curves of Figure 35 show that
similar relationships between load-bearing strengths and water content
characterize all of the 10 sludge compositions tested. At a certain threshold
solids content that is characteristic of each sludge tested, the load-bearing
strength increases steeply over a small incremental increase in solids con-
tent from a negligible value to well above the minimum for safe access of
personnel and equipment. To obtain the data shown in Figure 35, each of
10 sludge samples was dewatered over a period of a few days only by means
of underdrainage, and it was shown that each of the sludges tested can be
readily dewatered by means of underdrainage to load-bearing strengths
that are acceptable for structural purposes.
The lime and limestone ash-free sludges from TVA Shawnee
were tested more extensively in the laboratory in support of the field dis-
posal evaluation program (Refs. 3 and 4), and were used to determine
the rates of development of load-bearing strength as they were dewatered by
underdrainage. Laboratory tests were conducted to determine the loading
strengths of the ash-free sludges and of sludge and fly ash mixtures as the
samples settled and partially dewatered by means of underdrainage. The
fly ash used was obtained from TVA Shawnee and consisted of a mixture of
two parts mechanically separated ash and one part electrostatically separated
ash. Two samples of filter cake (45 and 48 wt% solids) were obtained from
the venturi lime scrubber at Shawnee. The latter sample was obtained
during a run in which MgO was added to the scrubber slurry. Centrifuge
cake (46 wt% solids) and clarifier underflow slurry (29 wt% solids) were also
obtained from the turbulent contact absorber (TCA) limestone scrubber at
Shawnee. A second sample of clarifier underflow from the TCA limestone
scrubber taken several weeks later was also tested.
In Figure 36, data have been plotted to show the rates of
development of compaction strength of wet, ash-free, lime process filter
cake and filter cake and fly ash mixtures as they dewater by means of under-
drainage. For comparison, Figure 36 also includes plots of the strengths
developed by samples containing similar compositions of sludge and fly ash
which were allowed to settle without removing any of the water that
accumulated at the surface. In the latter case, even after settline for two
weeks, the observed strengths were less than 1 kg/cm2 (14 Ib/in.*2). In
contrast, after one week the underdrained samples had developed strengths
ranging from 5 kg/cm2 (70 lb/in.2) for the sample of ash-free sludge to
approximately 20 to 25 kg/cm2 for the samples of sludge containing 40 per-
cent fly ash. The most rapid development of strength was shown by the
sample that had half of the fly ash mixed with the sludge and the other half
underlying the sludge and fly ash layer.
Tests were made to determine the effect of compacting the
partially dewatered samples. Although repetitive application of compaction
loads of 3 psi resulted in increased load-bearing strengths by as much as a
131
-------
a 150-
u>
t\>
o
s
Q
o
SHAWNEE, 6% FLY ASH - 9/8/76
SHAWNEE, 40% FLY ASH - 9/8/76
SCHOLZ, WITHOUT FLY ASH - 6/20/76
SCHOLZ, 30% FLY ASH - 6/Z7/76
PADDY'S RUN. 12% FLY ASH -
PHILLIPS, 60% FLY ASH - 6/17/74
CHOLLA, 59% FLY ASH -4/1/74
GADS BY, 9% FLY ASH -8/9/74
SHAWNE, WITHOUT FLY ASH - 11/30/76
SHAWNEE, 40% FLY ASH - 11/30/76
RTP GYPSUM*, WITHOUT FLY ASH - 12/4/75
RTP GYPSUM*, 40% FLY ASH - 9/30/75
ABSORBENT
L - LIME
DA - DOUBLE ALKALI
CL - CARBIDE LIME
LS - LIMESTONE
* - CONTAINS 5%
SULFITE
40
50
60 70
SOLIDS CONTENT, weight %
90
Figure 35. Load-bearing strength as a function of moisture, fly ash content,
and sludge origin
-------
350
300
250
CD
ULJ
cc
GO
CD
200
150
CD
I
100
50
SAMPLE
O
•
0
A
A
•
a
UNDER-
DRAINED
NO
NO
YES
YES
YES
YES
YES
FLY ASH.
wt%
NONE
40
NONE
NONE
40
40
20
PERCENT SOLIDS
INITIAL
45
58
45
35
FINAL
50
60
FLY ASH
UNDER-
LAYER.
wt%
NONE
NONE
NONE
40
NONE
NONE
20
- Not determined
0
5 10
SETTLING TIME, days
15
Figure 36. Development of bearing strength by de water ing of
TVA Shawnee lime filter cake and sludge-fly ash
mixtures
133
-------
factor of two, this incremental increase was about equivalent to that which
would have been obtained by several additional days of dewatering by
underdrainage.
Moderate losses in strength resulted from the addition of
water to the surface of samples to simulate the effect of a one .half inch
rainfall. The results of these rewetting tests are shown in Figure 37.
Re wetting caused an initial reduction in strength although the strength was
regained after additional time was allowed for dewatering.
Tests were also made with centrifuge cake and clarifier
underflow from the limestone TCA scrubber at Shawnee. Samples were
prepared by dilution of centrifuge cake with centrate and by removal of
supernate from the clarifier underflow to give the 35 wt% solids content in
both cases. A second sample of the clarifier underflow was taken two weeks
later, which had a 35 wt% solids content and much larger crystals. The
solids of this sample settled more rapidly and scanning electron microscope
(SEM) photos showed much larger crystals of calcium sulfite hemihydrate,
CaSO-j • 1/2 I^O, than those of the earlier clarifier underflow sample.
For comparison, SEM photos of the three TCA limestone solid samples are
shown in Figures 38 through 40. A discussion of crystal morphology is
given in Appendix A. '' Load-bearing strength measurements on these
materials as they dewatered by means of underdrainage are shown in the
plots of Figure 41. It was observed that the limestone sludge, if left
undisturbed, did not dewater as readily as the lime sludge. A skin-like
film that formed on the surface sealed off the sample and inhibited further
drainage. Puncturing of this surface film caused die samples to resume
drainage, with consequent increase in load-bearing strengths as the solids
contents increased. Vibration or agitation also accelerated the drainage
rates after the formation of the surface film. (This condition has not been
evident at any time during two years of monitoring at the field evaluation
sites.) Therefore, rates of development of bearing strengths.of the limestone
sludge and sludge-fly ash mixtures could not be determined with a high degree
of reproducibility in the laboratory.
The first clarifier underflow, which was initially 29 percent
solids, developed a strength of 5 kg/cm2 (70 Ib/in. 2) in one week. When
the initial solids content was adjusted to 35 percent, the ash-free clarifier
underflow had developed a strength of about 9 kg/cm2 (128 Ib/in. 2) at the
end of a week. Slightly greater strengths (10 kg/cm2) were shown by the
centrifuge cake. The most rapid gain in strength was shown by the ash-
free centrifuge cake with an underlayer of fly ash, which had a loadbearing
strength of 18 kg/cm2 (260 Ib/in. 2) after aix days. Samples with the thick-
nesses ranging from 4 to 10 in. were tested. All dewatered to comparable
strengths although the surfaces of the samples with the greater thickness
more readily sealed off and interrupted the drainage.
i
The solids of the second clarifier underflow sample settled
more rapidly, and consequently the rates of development of load-bearing
.134
-------
300
250
8.
I 200
CO
CD
150
<:
o
V
SAMPLE
O
•
O
UNDER-
DRAINED
YES
YES
YES
FLY ASH,
W1%
NONE
40
NONE
PERCENT SOLIDS
INITIAL
55
45
45
FLY ASH
UNDER-
LAYER,
wt%
40
NONE
NONE
"Samples rewet with O.Bin. of surface water
5 10
SETTLING TIME, days
15
Figure 37. Effect of re wetting on development of bearing strength
of underdrained TVA Shawnee lime sludge, filter
cake, and sludge-fly ash mixtures
135
-------
Figure 38. Shawnee TCA limestone sludge
solids from centrifuge cake,
45-percent solids (sampled
11/30/76), X3000
Figure 39- Shawnee TCA limestone sludge
solids from clarifier underflow,
29-percent solids (sampled
11/30/76), X3000
-------
Figure 40. Shawnee TCA limestone sludge
solids from clarifier underflow,
35. 7-percent solids (sampled
12/1/76), X3000
137
-------
300
250
8.
200
CD
CO
£150
CD
100
50
0
SAMPLE
}
UNDER-
DRAINED
YES
YES
YES
YES
YES
YES
FLY ASH,
wt%
NONE
NONE
NONE
NONE
40
NONE
PERCENT SOLIDS,
INITIAL
29
35
35
35
46
-, 46
FLY ASH
UNDER-
LAYER,
wt%
11
NONE
40
NONE
NONE
40
0
5 10
SETTLING TIME, days
Figure 41. Bearing strength of underdrained TVA Shawnee
limestone sludge and sludge-fly ash mixtures
138
-------
strengths were accelerated. Although several sludge-fly ash combinations
were tested, all developed ultimate strengths in excess of 28 kg/cm2
(400 Ib/in.2), approximately double that of the first sample (Figure 41).
Rewetting of the limestone sludge samples after they had
developed significant load-bearing strengths resulted in immediate loss of
strength as the surface was partially reslurried. Much of the former
strength was regained after subsequent draining, as shown in Figure 42.
In summary, all sludge samples that were provided with a
means of dewatering by underdrainage developed load-bearing strengths
greater than 5 kg/cm2 (70 Ib/in.2) within one week. Rewetting results
only in a temporary loss of strength until the sludge again dewatered. The
inclusion of fly ash was beneficial to the dewatering process in all cases.
In some cases, it was preferable to mix the ash with the sludge; in other
cases there was greater advantage in having alternate layers of sludge
(or sludge-fly ash mixtures) and fly ash, but generally, a drainage layer
at the base (in these cases, fly ash) appreciably shortened the time for
drainage and strength development. One of the most important conclusions
is that preliminary partial dewatering by filtration or centrifugation is
unnecessary. The clarifier underflow which is readily pumpable can be
dewatered by underdrainage until load-bearing strengths have developed
that are adequate to support heavy equipment.
9.3.2 Assessment of Underdrainage Technology
The use of underdrainage has demonstrated in laboratory
tests the potential ability to provide environmentally sound disposal of
untreated sludge. Inmost cases, underdrained sludge develops sufficient
bearing strength (> 70 psi in a day or two after placement) to support con-
struction equipment. In addition, the presence of an underdrainage system
prevents the development of a hydraulic head over the subsoil, and only a
minimal amount of leachate water penetrates the soil. This occurs only
during the few years of filling. Thus, chemical pollution of groundwater is
avoided, and, since supernate is not allowed to exist on the surface (except
as noted in the following paragraph), no pollution by runoff is expected.
These factors are discussed in the following example.
In this example of disposal procedure, it is assumed that
FGD waste is placed in a closed basin to a depth of 30 to 40 ft, at the base
of which is an underdrainage system which is buried under 1 ft or more
of bottom ash, fly ash, sand, and gravel. The area being developed repre-
sents approximately one year of waste fill (about 30 acre/yr/1000-MW
capacity during the first year of operation) and the waste may be placed by
sluicing from a clarifier or trucked from a filter. If fly ash is removed
upstream of the scrubber, it is added to the disposal site by alternate or
simultaneous unloading with sludge; no mechanical mixing is necessary.
Water that enters the basin, either with the solids during filling or by
rainfall, is returned to the scrubber for reuse. The method is amenable
139
-------
SAMPLE
O
A
D
UNDER-
DRAINED
YES
YES
YES
FLY ASH.
W1%
NONE
NONE
40
PERCENT SOLIDS,
INITIAL
49
33
33
FLY ASH
UNDER-
LAYER,
wt%
40
40
NONE
'Samples rewet with O.Sin. of surface water
5 10
SETTLING TIME, days
15
Figure 42. Effect of rewetting on development of bearing
strengths of underdrained TVA Shawnee lime-
stone sludge
140
-------
to the continuous expansion of the disposal basin and simultaneous
continuous land reclamation by covering with overburden. Since water
evaporates more easily from damp solids than from a pond, advantages of
this phenomena can be made by closing the underdrain valves and using the
surface of the waste dried during the summer months to evaporate blow-
down from a scrubber (if necessary). Normal disposal operations continue
in an adjacent basin.
9.3.2.1 Water Balance in an Underdrained Disposal Site
The effect of water recirculated from an underdrained disposal
site on the water and chloride ion balance of an FGD scrubber was calculated.
The specific scrubber system chosen for the calculation was based on the
TVA Shawnee 10-MWe venturi-spray tower scrubber using lime absorbent,
and extrapolated to a 1000 MWe plant (Figure 43, Ref. 20). This basic
system, without underdrainage, is representative of a tight, closed-loop
water system which requires a total of 1410 gal/min (gpm) fresh makeup
water to replace 900 gpm lost up the stack and 510 gpm lost by occlusion in
the ponded waste containing 45 percent solids. In addition, supernate
(1390 gpm) is returned to the scrubber loop from the pond; this results
when the ponded clarifier underflow containing 18 percent solids settles to
45 percent solids. The chloride concentrations in the slurry from the
scrubber and the supernate return are the same if the clarifier is not
included; the amount of water to the disposal site and returned as supernate
increases by 2900 gpm (the amount returned from a clarifier if used). This
condition illustrated in Figure 43 was then modified to form a base case by
including rainfall and returning une vapor a ted rainwater in the supernate
from the pond to the plant.
9.3.2.1.1 Base Case
The base case assumed a disposal site of 30 acres, with no
underdrainage, and the unevaporated rainfall returned in the supernate
to the scrubber. A 30-acre operating site, with a depth of 30 ft, was chosen
as a reasonable size in terms of operating efficiency and physical accessi-
bility, with the sludge output of about one year's operation at 7000 hr per
year being representative. The total amount of water falling onto a 30-acre
area (at a rate of 48 in./year) is 41 x 106 gal/year or an average of 78 gpm.
Based on a 30-acre surface area, a dry-bulb temperature of 68° F, and
relative humidity of 50 percent, 85 gpm are evaporated (Figure 44). Al-
though a 50 percent relative humidity and 78 gpm rainfall do not occur
simultaneously, they are considered as concurrent average values for
purposes of this preliminary analysis. Surge capacity to handle heavy rain-
fall conditions may be provided by the pond in service by limiting the under-
flow return during a portion of the fill period, or by using an adjacent pond
not yet used or filled.
Data from the Shawnee evaluation (Ref. 3) indicate that a
steady-state operating condition of 5000 ppm Cl- may be expected in the
pond input liquor and in the supernate.
141
-------
900 gpm
0.55%
VENTURI
AND
SPRAY
TOWER
SCRUBBER
163,000 gpm
100%
STACK
FLUE
GAS
162,100 gpm
99.5*
4800 gpm
2.95%
MAKE-UP
WATER
• PERCENTAGE VALUES ARE
PERCENT OP MAXIMUM
WATER FLOW INTO SCRUBBER
• WATER FLOW RATE IN
GALLONS PER MINUTE (gpm).
TO CONVERT TO cu m/hr
MULTIPLY BY 0.227
POND
EXTRAPOLATED OPERATING CONDITIONS
LIQUID-TO-GAS RATIO: 54.0 gal/1000 cu ft
SCRUBBER EFFLUENT SOLIDS: 8 PERCENT
CLARIFIER UNDERFLOW SOLIDS: 18 PERCENT
WATER RETAINED IN SETTLED SLUDGE: 46 PERCENT SOLIDS
COAL 3.4 PERCENT SULFUR
SO, REMOVAL EFFICIENCY: 90 PERCENT
Figure 43. Extrapolated water balance for venturi and spray tower
scrubber system
142
-------
STACK GAS
(900 gpm at 0 ppm Cf
FLUE GAS
(20.1 Ib/min CH*
IN COAL
RAIN
(78 gpm at 0 ppm CI"
WATER-LIME SLURRY
(458 gpm at 0 ppm CD
500 gpm at
5000 ppm CI
EVAPORATION
52 gpm at 0 ppm CI"
-#*
SUPERNATE RETURN
(26 gpm at v
5000 ppm CD*
_
FRESH WATER MAKEUP
(900 gpm at OppmCI"
26 gpm x 8.5 x 0.005= 1.1
20.1
21.2 Ib/min
DISPOSAL
-SITE
OUT**
500 gpm x 8.5 x 0.005=
21.2 Ib/min
Figure 44. Base case : ponding without under drainage
143
-------
9.3.2.1.2 Underdrained Case
Based on the characteristics of lime sludge remixed with
fly ash, now being evaluated at Shawnee (Ref. 4), an additional 140 gpm
is returned to the scrubber as underdrainage, in addition to the 1383 gpm
returned in the base case (Figure 45). The 140 gpm represents additional
water freed during the settling process (occurring because of under-
drainage) when the solids content in the pond dewater from 45 to 60 percent
solids. Again, assuming steady-state conditions with 22.0-lb/min Cl~
entering from the flue gas, a steady-state Cl- concentration of 6900 ppm
was calculated in the 1523 gpm underdrainage returning to the scrubber.
9.3.2.2 Summary - Underdrained Disposal
The disposal method just described is currently being
evaluated on a small scale in the EPA/TVA Shawnee Field Disposal Evalua-
tion Project (Ref. 4). In addition, operating strategies for a 30-year site
using typical climatic conditions will be formulated. This technique, though
not fully developed, appears to offer several desirable advantages regarding
environmental pollution control as well as potential cost savings under
specific conditions. Also, certain limiting conditions are associated with
this technique. The advantages and limits to which underdrainage can be
applied, as the method is now understood, are as follows.
9.3.2.2.1 Advantages
The advantages of the underdrained disposal method include
the following:
a. Dewater ing equipment such as filters or clarifiers
are not required; i.e., the sludge may be sluiced
to the disposal site, and the underdrainage system
is used to accomplish dewatering.
b. Underdrainage process is workable with sludges
containing fly ash or without fly ash.
c. Load-bearing strengths in excess of 70 psi develop in
a few days. The material loses some strength when
rewetted by rainfall but regains its strength again in
a few days.
d. Water balance of the scrubber loop can be maintained
without detrimental effects, with proper design and
management.
e. Underdrained basins can be used periodically as
evaporation ponds (weather permitting)to evaporate
scrubber blowdown if necessary.
144
-------
STACK GAS
(900 gpm at OppmCf
FLUE GAS
(22.0 Ib/min Cl'
RAIN
(78 gpm at 0 ppm CD
SCRUBBER
SYSTEM
WATER-LIME SLURRY
(377 gpm at 0 ppm CD
EVAPORATION
(85gpm at 0 ppm CI"
11900 gpm at 6900 ppm CI"
UNDERDRAIN V-
RETURN V
1 DISPOSAL SITE (30 acres)
(1513 gpm at 6900 ppm CD**
(370 gpm RETAINED)
FRESH WATER MAKEUP
(900 gpm at 0 ppm CD
CHLORIDE ION BALANCE IN SCRUBBER SYSTEM
OUT**
]N*
UNDERDRAIN RETURN: 1523 gpm 8.5 x 0.069=89.3
FLUE GAS:
22.0
111.3 Ib/min
1900 x8.5 x 0.0069 =
111.4 Ib/min
* Clarifier Included ** Concentration unchanged
without clanfier
Figure 45. Ponding with underdrainage
145
-------
f. Under most conditions, a pond lining is not required
because a hydraulic head of any consequence does not
exist. The material is continuously saturated, however;
therefore, limitations on soil permeability and separation
distance to the water table would need to be determined.
9.3.2.2.2 Limitations
The limitations of the underdrain disposal method are as
follows:
a. An underdrained basin is limited in size because most
of the rainfall is collected and routed through the
underdrained system to the scrubber. Therefore, a
modular approach wherein the disposal site is com-
partmented into basins, each containing, for example,
approximately one year's output from the scrubber.
For a 1000-MW system, operating at 7000 hr per year,
30 acres with a depth of 30 ft are required.
b. The disposal site must be reasonably close to the power
plant to allow the installation of a closed-loop pipeline
limited by the economics of efficient disposal.
c. An underdrainage system, with the potential of low cost
by using a drainage blanket of bottom ash and/or gravel,
sand, and fly ash has not been demonstrated. Such a
development might result in significant savings as
compared to a system of underdrainage pipes.
9.4 ASSESSMENT OF THE POTENTIAL CHEMICAL
POLLUTION OF THE ENVIRONMENT BY ALTERNATIVE
DISPOSAL METHODS
Pollution of groundwater by the action of rainwater percolating
through FGD sludge is a major concern in the disposal of sulfur sludges.
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.
Thereafter, 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 can be expressed as mass
loading of pollutant that is carried by leaching water to groundwater.
The potential pollution that is possible from sludge when
chemically treated by any of several processes can be reduced by several
orders of magnitude. Improvement is accomplished by (1) the reduction of
146
-------
available chemical species in the treated sludge, (2) the reduction of
rainwater available because of runoff, and (3) the elimination or reduction
of the rate of water permeation through the chemically treated material.
In nearly every case, the pollutant concentration of the leachate from treated
sludges is less than one-half that of the leachate from untreated sludge
when fresh material is placed in the disposal site. Additionally, the run-
off from treated waste can be maximized and makes available much less
than one-tenth of the rainfall for seepage, since the permeation coefficient
of treated waste is typically reduced by at least an order of magnitude. Thus,
the real pollutant migration measured by mass loading can be reduced by
very large amounts.
The consequences of much lower permeability rates of chemi-
cally treated sludge are that many years will pass before five PVD in the
sludge can 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 potential of chemi-
cally treated sludge is not as severe but will be sustained for much longer
periods of time. The latter case is more in consonant with nature, which
weathers natural deposits slowly but constantly.
An evaluation of the amount of chemical pollutants that are
available to the environment by leaching was determined from the results
of the leaching data summarized in Section VII and the permeability data
summarized in Section VI. Various examples illustrating the effects of
sludge treatment, the effects of different subsoils, and the management of
the site for different rainfall recharge rates are presented. The relative
amounts of sludge constituents (TDS) released at the sludge base, for
different modes of disposal were calculated (Table 33 and Figure 45). This
analysis is based on correlations of laboratory results and Shawnee field
condition data reported herein and in References 3 and 4. All the cases
are indexed to an untreated slurry pond, namely, Case 1, in which the
soil permeability coefficient is 10-5 cm/sec.
In assessing the effectiveness of chemical treatment, tests
to determine permeability of chemically treated sludges were performed on
cores extracted from the Shawnee field evaluation site. Constant-head
permeability tests were run on (1) pulverized samples and (2) samples with
and without visible cracks. Uncracked samples of the material had coeffi-
cients of permeability of about 10-7 cm/sec, and the pulverized and cracked
samples had coefficients of approximately 10'5 cm/sec. Therefore, the
effective coefficient of the treated material could be expected to be between
10-5 and 10-7 cm/sec. Assuming a conservative case (using a coefficient
of 10-5), an order of magnitude improvement in impermeability is realized
compared to untreated sludges, which typically have a coefficient of about
10-<
The systematic reduction of standing water is illustrated in
Cases 3 through 5, wherein the recharge rate is reduced compared to a
147
-------
TABLE 33. BENEFITS OF CHEMICAL, TREATMENT: CASE STUDIES FOR
COMPARISON OF MASS RELEASE OF SLUDGE CONSTITUENTS
FOR VARIOUS DISPOSAL ALTERNATIVES
Case
1
2
3
4
5
Disposal
Methoda
Ponded
slurry
Ponded
slurry
Ponded
cake
Dewatered
and ponded
Landfill
Surface
Water
Constant
supernate
Constant
supernate
10-in/yr,
recharge
10-in/yr,
recharge
1-in/yr g
recharge
Sludge ,
Condition
Untreated
Chemically
treated
Untreated
Chemically
treated
Chemically
treated
I udge
Permeability,
cm/sec
io-4
ID'5
1C'4
io-5
ID'5
Fill period = 5 yr; depth = 30 ft.
Porosity (void volumetric fraction) = 0. 67.
Constant supernate assumes 1-ft depth of surface water.
10-in/yr recharge is unevaporated rainfall.
1-in/yr effective recharge resulting from seepage during runoff of rainfall.
1.0
0.1
0.01
CASE
1. PONDED SLURRY,
UNTREATED
2. PONDED SLURRY.
TREATED
3. PONDED CAKE.
UNTREATED
4. DEWATERED AND
PONDED. TREATED
5 LANDFILL. TREATED
SOIL
PERMEABILITY
COEFFICIENT
10'' cm/sec
Figure 46. Comparison of mass release of sludge
constituents for various disposal
alternatives showing benefit of
chemical treatment
148
-------
ponded slurry. If only unevaporated rainfall is allowed to recharge, the
mass release into the subsoil is reduced by a factor of about 5 (Case 3
versus 1, and Case 4 versus 2).
The significance of eliminating standing water by runoff is
also shown. If it is assumed that 10 percent of the net rainfall is recharged,
then a one order of magnitude reduction is achieved relative to a dewatered
and ponded treated waste (Case 5 versus 4), and two to three orders of
improvement are seen, depending on soil permeability, when comparing the
mass seepage from a chemically treated site to that of a ponded untreated
site (Case 5 versus 1). In addition, compacting the treated materials during
the site filling may reduce crack formation so that an effective coefficient
of permeability better than 10-5 cm/sec may be realized.
Case 2 (Figure 46) considers chemically treated sludges
disposed of in such a manner that a hydraulic head exists on the site at all
times. The mass release for such a case is approximately 1/2 to 1/20 of
that of an untreated slurry pond depending on subsoil permeability. Sites
of this type may seep to an adjacent stream or water table, which may
reduce the concentration of constituents by mixing. Historical data regard-
ing the stream and water table characteristics and quality, as well as
monitoring, may be mandatory for sites of this type to assure environmental
acceptability.
149
-------
SECTION X
DISPOSAL COST ESTIMATES
Cost estimates for ponding, chemical treatment, and landfilling
have been made and reported by Aerospace on several occasions. During
recent studies associated with the U.S. Environmental Protection Agency
(EPA) Shawnee field disposal evaluation project, Aerospace cost estimates
were made of chemical treatment disposal and were reported in the initial
report on that study (Ref. 3). The Aerospace cost estimates for lined-pond
disposal and chemical treatment disposal were presented in the initial and
second progress reports on the sludge disposal study (Refs. 1 and 2) and at
the EPA 1974, 1976, and 1977 flue gas desulfurization symposiums (Refs. 8
through 10). These estimates were updated to July 1977 costs in a more re-
cent report prepared for EPA on new source performance standards (Ref. 21).
In that update refinements were made to reflect more current data on coal
content, plant efficiency, operating conditions, and land costs. A comparison
of the conditions for the current estimates with the previous baseline assump-
tions are shown in Table 34.
10.1 ECONOMICS OF DISPOSAL PROCESSES
The four flue gas desulfurization (FGD) waste disposal methods
selected for economic evaluation were as follows:
a. Ponding of untreated wastes using a flexible elastomeric
liner.
b. Ponding of untreated sludges using indigenous clay.
c. Chemical treatment and landfill disposal.
d. Forced oxidation of sulfite sludge to gypsum and
landfill disposal.
10.1.1 Ponding of Untreated Wastes
Several waste pond sites currently in operation employ flexible
elastomeric liners and sites lined with indigenous clay (impervious) soil;
however, only one elastomeric lined pond is reported to contain FGD sludge.
151
-------
TABLE 34. COMPARISON OF CONDITIONS FOR CURRENT
DISPOSAL COST ESTIMATES WITH PREVIOUS
ESTIMATES
Condition
Previous
Current
Dollar Base
Plant Characteristics
Coal Burned
Annual Operating Hours
Plant and Disposal
Site Lifetime
SO2 Removal, with Lime-
stone Absorbent
Sludge Generated
Limestone Utilization
Average Annual Capital
Charges, 30-yr av
Cost of Land Used for
Disposal
Land Depreciation
Disposal Site
1975
1000 MWe,
0.88-lb coal/kWh
3. 0% sulfur,
10, 560 Btu/lb,
12% ash
4380 (30-yr av)
30 yr
85%
5.475 x 105 short
tons/yr (dry)
65%
18%
$1000/acre; all land
assumed purchased
initially; sludge depth,
30 ft
Total depreciation in
30 yr; straight line
basis
Within one mile of
plant
July 1977
1000 MWe,
0. 75-lb coal kWh
3. 5% sulfur,
12,000 Btu/lb,
14% ash
4380 (30-yr av)
30 yr
90%
4.798 x 105 short
tons/yr (dry)
80%
18%
$5000/acre; all land
assumed purchased
initially; sludge depth
30ft
Total depreciation in
30 yr; straight line
basis
Within one mile of
plant
152
-------
Since previous work (Ref. 2) has shown that the optimum pond
depth for this type of disposal is 30 ft, which is the depth at which pond con-
struction and land costs are optimum with respect to the cost of liner mate-
rial, the results are reported on the basis of a 30-ft depth of sludge.
Commercial materials which are considered typical of the price
range of installed liners, i.e., PVC (ZO-mil thickness) and Hypalon (30-mil
thickness) were used in this study. The least expensive of these two mate-
rials, i.e., PVC-20, was selected for this analysis, which is summarized in
Table 35. The indigenous soil is assumed to be impervious, with a permea-
bility coefficient of 10~6 cm/sec or better.
10.1.2 Chemical Treatment and Disposal
The cost of chemical treatment and disposal of FGD sludge in
a landfill was estimated in March 1976 (Ref. 3). At that time, estimates of
total disposal costs were made for three chemical treatment processes, i.e.,
Dravo, IU Conversion Systems, and Chemfix, for a 1000-MW plant based on
1975 dollars. With this work as a basis, cost estimates have been updated
for the current conditions referenced in Table 34 and summarized in Table 35.
10.1.3 Economics of Conversion to Gypsum
The cost of producing gypsum as a by-product from line or
limestone 1000-MWe scrubbing processes has been reported for mid-1977 by
Aerospace (References 1 and 11). The estimate included the costs required
to incorporate the forced oxidation processing into a basic lime or limestone
scrubber system. The fly ash was assumed to be separated in electrostatic
precipitators. However, the cost of ESP's was not charged to the cost of
producing the gypsum. Alternatively, the fly ash could be scrubbed in a two-
stage venturi-absorber, forced oxidation system. On the basis of the power
plant and scrubbing conditions shown in Table 34, 2. 65 X 10-* tons of ash-
free gypsum (dry basis) would be produced annually.
The capital equipment costs to produce gypsum in a new 1000-
MWe installation, scrubbing flue gas from 3. 5 percent sulfur coal, and re-
moving 90 percent 80%, is $10.09 X 10° or $10.00/kW. The estimates,
which include the cost of oxidizing the waste to gypsum and the disposal of the
gypsum combined with 222, 000 short tons of ash, are summarized in Table 35.
10.2 COST COMPARISON
A comparison of costs for the various forms of disposal is
given in Table 35. In addition to presenting the costs in mills/kWh, costs
are also presented in terms of cost per ton dry sludge and cost per ton of
coal burned. The disposal cost for gypsum includes the additional cost of
forced oxidation of the sulfite slurry, and, for ease of comparison, the gyp-
sum cost is converted to the cost of an equivalent amount of sludge.
15:
-------
TABLE 35. DISPOSAL, COST COMPARISONS*
Cost Basis,
Mid- 1977 $
Mills /kWh
$/Ton
Sludge (Dry)
$/Ton Coal
Ponding
Liner
Added
0.80
7.25
2.20
Indigenous
Clay
0.55
4.90
1. 50
Landfill-
Chemical
Treatment
1.05
9.70
2.95
Gypsum^
1. 10
10.30C
3. 10
See Table 34 for all referenced conditions.
5100% limestone utilization for gypsum case; all others, 80%.
*
'Cost of forced oxidation and disposal of gypsum sludge (including
fly ash) in an indigenous clay-lined pond is converted to cost/ton of
equivalent FGD nonoxidized sludge including fly ash. To convert to
cost/ton of gypsum including fly ash, divide by 1. 08.
154
-------
REFERENCES
1. J. Rossoff, et al., Disposal of By-Products from Nonregenerable
Flue Gas Desulfurization Systems; Second Progress Report,
EPA-600-7-77-052, U.S. Environmental Protection Agency, Research
Triangle Park, NC (May 1977).
2. J. Rossoff and R. C. Rossi, Disposal of By-Products from Non-
regenerable Flue Gas Desulfurization Systems; Initial Report,
EPA-650/2-74-037a, U.S. Environmental Protection Agency, Research
Triangle Park, NC (May 1974).
3. R. B. Fling, et al., Disposal of Flue Gas Cleaning Wastes: EPA
Shawnee Field Evaluation. Initial Report, EPA-600/2-76-070, U.S.
Environmental Protection Agency, Research Triangle Park, NC
(March 1976).
4. R. B. Fling, et al., Disposal of Flue Gas Cleaning Wastes; EPA
Shawnee Field Evaluation; Second Annual Report, EPA-600/7-78/024,
U.S. Environmental Protection Agency, Research Triangle Park,
NC (February 1978).
5. P. P. Leo and J. Rossoff, 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, NC (October 1976).
6. P. P. Leo and J. Rossoff, Control of Waste and Water Pollution from
Power Plant Flue Gas Cleaning Systems; Second R and D Report,
EPA-600/7-78-224, U.S. Environmental Protection Agency, Research
Triangle Park, NC (November 1978).
7. 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. 72-92 (June 1972).
8. J. Rossoff and R. C. Rossi, "Flue Gas Cleaning Waste Disposal, EPA
Shawnee Field Evaluation, " presented at the EPA Flue Gas Desulfuri-
zation Symposium, New Orleans, LA (March 1976).
155
-------
9. J. Rossoff, et al., "Disposal of By-Products from Non-Regenerable
Flue Gas Desulfurization Systems: A Status Report, " presented at
the EPA Flue Gas Desulfurization Symposium, Atlanta, GA (November
4-7, 1974).
10. P. P. Leo, R. B. Fling, and J. Rossoff, "Flue Gas Desulfurization
Waste Disposal Field Study at the Shawnee Power Station, " EPA,
Symposium on Flue Gas Desulfurization, Hollywood, FL (November
8-11, 1977).
I4. "Standard Method of Test for Bearing Ratio of Laboratory Compacted
Solids," ASTM Designation D1833-73 (July 1973).
12. Handbook for Analytical Quality Control in Water and Wastev/ater
Laboratories, Analytical Quality Control Laboratory, National
Environmental Research Center, Cincinnati, OH (1972).
13. Methods for Chemical Analysis of Water and Wastes, Analytical
Quality Control Laboratory, National Environmental Research Center,
Cincinnati, OH (1971).
14. B. H. Borgwardt, "IERL-RTP Scrubber Studies Related to Forced
Oxidation, " Proceedings; Symposium on Flue Gas Desulfurization,
New Orleans, LA (March 1976); EPA-600/2-76-136a, U.S. Environ-
mental Protection Agency, Research Triangle Park, NC (May 1976).
15. Summary Report - Flue Gas Desulfurization Systems, prepared by
PEDCO-Environmental Specialists, Inc., for the U. S. Environmental
Protection Agency, Research Triangle Park, NC, Contract No.
68-02-1321 (September-October 1976).
16. H. G. Blecker and T. M. Nichols, Capital and Operating Costs of
Pollution Control Equipment Modules, Vol. II, Data Manual,
EPA-R5-73-023b, U.S. Environmental Protection Agency, Washington,
DC (July 1973).
17. "Economic Indicators, " Chemical Engineering, Vol. 84, No. 1, p. 7,
McGraw-Hill Inc., New York, NY (January 3/ 1977).
18. G. G. McGlamery, et al., Conceptual Design and Cost Study, Sulfur
Oxide Removal from Power Plant Stack Gas. EPA-R2-73-244, U.S.
Environmental Protection Agency, Research Triangle Park, NC
(May 1973).
19. R. H. Borgwardt, Sludge Oxidation in Limestone FGD Scrubbers,
EPA-600/7-77-061, U.S. Environmental Protection Agency, Research
Triangle Park, NC (June 1977).
156
-------
20. L. J. Bornstein, et al., Reuse of Power Plant Desulfurization Waste
Water, EPA-600/2-76-024, U. S. Environmental Protection Agency,
Research Triangle Park, NC (February 1976).
21. P. P. Leo and J. Rossoff, Controlling SO? Emissions from Coal-
Fired Steam Electric Generators; Solid Waste Impact,
EPA-600/7-78-0446, Vol. II, U.S. Environmental Protection Agency,
Research Triangle Park, NC (March 1978).
157
-------
APPENDIX
CRYSTAL MORPHOLOGY
For each of the sludges evaluated in this report period, a
portion was selected, dried, and prepared for scanning electron micrographic
(SEM) examination. 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. The photomicrographs and
description of the solid particulate material characteristic of each sludge are
presented in the following text.
A. 1 TVA SHAWNEE LIME SLUDGE. RUN F.
WITHOUT FLY ASH
The characteristics of the solids content of the Tennessee
Valley Authority (TVA) Shawnee venturi spray tower system using lime absor-
bent and containing no fly ash is shown in Figure A-l. The sulfur phase is
represented by calcium sulfite rosettes that range between about 15 and 30-^m
diam with the median cluster size about 20-jxm diam. Individual platelet size
was typically 5 pm in its longest dimension and rarely exceeded 10 ^m. The
platelet thickness was typically 0. 1 to 0. 5 |xm. These particle characteristics
are nearly identical to the characteristics of particles previously observed
(Ref. 1) from this scrubber system. The primary difference between this
sludge and those observed previously is the near absence of fly ash, and it
may be presumed that differences in physical behavior may be ascribed to
fly ash. In the present sludge sample, very fine fly ash particles were occa-
sionally observed, but they constituted less than one percent of the sludge
sample.
A. 2 GPS SCHOLZ DOUBLE ALKALI SLUDGE,
WITHOUT FLY ASH
The solids content of the sludge from the Gulf Power Service
(GPS) Scholz double alkali scrubbing facility are shown in Figure A-2. This
sludge was comprised of very fine participates agglomerated in clusters. The
fine participates were calcium sulfite hemihydrate, with dimensions in the
range of 1 |xm along an edge and thickness of 0. 05 to 0. 1 jam. Cluster sizes
159
-------
Figure A-l. Shawnee.venturi and spray
tower, lime sludge solids,
no fly ash, X2000
Figure A-2.
Scholz double-alkali sludge
solids, no fly ash, X2000
-------
ranged from individual particles up to agglorr erates as large as 50 (Jim in
diameter. Most agglomerates, however, were about 10 to 20 \j.m in diam.
In contrast to the rosette clusters observed in the TVA Shawnee lime sludge,
the particles of the Scholz sludge did not form well-structured agglomerates
but were more randomly formed and showed more clear evidence of inter-
growth, as observed frequently with gypsum sludges. From the nature of
the intergrowth integrity, it is presumed that the physical behavior of this
sludge resembles a material with a particle size and broad particle range
of the agglomerates rather than that of the fine particle size of the
individual particulate.
A. 3 GPS SCHOLZ DOUBLE ALKALI SLUDGE.
WITH FLY ASH
The GPS Scholz double alkali sludge solids which contained
fly ash are shown in Figure A-3. This sample had sulfur phase agglomerates
identical to those described in the previous subsection. In addition, there
were fly ash particles that ranged from submicron to about 25-jim diam, with
the majority of the fly ash particles in the range of 10- to 20-fj.m diam. Many
of the smaller particles were shaped as prolate spheroids, typical of cyclone
boilers.
A. 4 LG&E PADDY'S RUN SLUDGE, CARBIDE LIME
The sludge solids from the Louisville Gas and Electric (LG&E)
Paddy's Run station are shown in Figure A-4. The characteristics of this
sludge differ from others in that few agglomerates were formed. Neverthe-
less, individual particle sizes were small, typically ranging from 0. 1 to
5 jim along an edge. Thickness dimensions of 0. 05 fim and less were ob-
served. Some agglomerates were observed, but few were larger than 5 ^im
in a major dimension. Among all sludges characterized, this sludge contained
the finest, dispersed particle distribution thus far observed. Fly ash particles
were also observed in the sludge, and their particle size range was typically
0. 1 to 5 fim, also, with a mean particle size of about 1 \im. The observed
frequency of fly ash in the SEM micrographs was less than that indicated by
chemical analyses and may be caused by the obscuration of the fly ash by the
fine, dispersed sulfur phase. The physical behavior of this sludge is not ex-
pected to resemble any other among those tested in this report but resemble
more the behavior of the double alkali sludges with fine particle sizes pre-
viously reported (Ref. 1).
A. 5 TVA SHAWNEE LIMESTONE SLUDGE. WITHOUT
FLY ASH. BLOCKY CRYSTALS
During a series of ash-free test runs of the turbulent contact
absorber (TCA) at the TVA Shawnee station, limestone absorbent was used
and scrubber operating parameters, such as slurry solids, hold tank size,
and reaction tank residence time, were varied. A SEM of a sample analyzed
is shown in Figure A-5. This is not considered a typical sample. Nonethe-
less, it is of interest because of the "blocky" shape of the sulfite crystals (10 to
161
-------
ro
Figure A-3. Scholz double-alkali sludge
solids with fly ash, X3000
Figure A-4. Paddy's Run carbide-lime
sludge solids with fly ash,
X3000
-------
Figure A-5.
Shawnee TCA limestone
sludge solids, no fly ash
(not considered typical)
X3000
163
-------
20 (Jjn and 2 to 4 fim thick), which displayed improved draining and load-
bearing strength properties compared to limestone sludge samples from the
same scrubber, which typically produces platelets considerably smaller than
the crystals in Figure A-5.
164
-------
TECHNICAL REPORT DATA
(rlease read Inaructiont on the reverse before completing)
1, REPORT NO.
EPA-600/7-79-046
2.
3. RECIPIENT'S ACCESSION-NO.
4. TITUS AND SUBTITLE Disposal of By-products from Non-
regenerable Flue Gas Desulfurlzation Systems: Final
Report
S. REPORT DATE
February 1979
6. PERFORMING ORGANIZATION CODE
7.AUTHORfSJ' ——
J.Rossoff, R.C.Rossi, R.B. Fling, W. M. Graven,
and P. P. Leo
8. PERFORMING ORGANIZATION REPORT NO.
0. PERFORMING ORGANIZATION NAME AND ADDRESS
The Aerospace Corporation
Environment and Energy Conservation Di vis ion
P.O. Box 92957
Los Angeles, California 90009
10. PROGRAM ELEMENT NO.
EHE624A
11. CONTRACT/GRANT NO.
68-02-1010
12. SPONSOR!NO AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Industrial Environmental Research Laboratory
Research Triangle Park, NC 27711
13. TYPE OF REPORT ANQ PERIOD COVERED
Final; 12/72 - 3/78
14. SPONSORING AGENCY CODE
EPA/600/13
IB. SUPPLEMENTARY NOTES IERL-RTP project off icer is Julian
2489. EPA-650/2-74-037a and EPA-600/7-77-052 are
W. Jones, MD-61, 919-541-
earlier related reports.
18. ABSTRACT
The report gives results of a 4-year study to determine environmentally
sound methods for disposing of wastes from nonregenerable flue gas desulfurization
(FGD) systems. Data presented incorporates results obtained during the fourth year
with material from report EPA-600/7-77-052, covering the first 3 years of the
study. Untreated and treated wastes from 10 different scrubbers at eastern and wes-
tern plants using lime, limestone, and double-alkali processes were characterized.
The report relates concentrations of salts and trace elements in the wastes to the
potential for water pollution. Physical properties (e.g., bulk density, load bearing
strength, permeability, and viscosity) are given. Disposal by ponding, landfilling
of chemically fixed wastes, ponding with underdrainage, and conversion to gypsum
was assessed. Disposal cost estimates for a 1000-MW eastern plant are 0. 55, 0.90,
and 1.20 millsAWh for ponding on indigenous clay, ponding with liner added, and
chemical treatment/landfill, respectively. Companion studies, pertaining to field
disposal evaluations and a summary of all EPA projects related to FGD waste and
water studies, have been reported separately.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b. IDENTIFIERS/OPEN ENDED TERMS
COSATI Field/Group
Pollution
Flue Gases
Desulfurization
Waste Disposal
Scrubbers
Calcium Oxides
Limestone
Ponds
Earth Fills
Gypsum
Pollution Control
Stationary Sources
Nonregenerable Systems
Double Alkali Process
13B 08G
21B 08H
07A,07D 13C
131
07B
18. DISTRIBUTION STATEMENT
Unlimited
19. SECURITY CLASS (Thu Report)
Unclassified
21. NO. OF PAGES
181
20. SECURITY CLASS (This page)
Unclassified
22. PRICE
Porm 232iws (9- TV
165
3.S.
------- |