EPA-600/2-76-024
February 1976
Environmental Protection Technology Series
REUSE OF POWER PLANT
DESULFURIZATION WASTE WATER
Industrial Environmental Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into five series. These five broad
categories were established to facilitate further development and application of
environmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The five series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
This report has been assigned to the ENVIRONMENTAL PROTECTION
TECHNOLOGY series. This series describes research performed to develop and
demonstrate instrumentation, equipment, and methodology to repair or prevent
environmental degradation from point and non-point sources of pollution. This
work provides the new or improved technology required for the control and
treatment of pollution sources to meet environmental quality standards.
EPA REVIEW NOTICE
This report has been reviewed by the U.S. Environmental
Protection Agency, and approved for publication. Approval
does not signify that the contents necessarily reflect the
views and policy of the Agency, 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-600/2-76-024
February 1976
REUSE OF POWER PLANT
DESULFURIZATION WASTE WATER
by
L. J. Bornstein, R. B. Fling, F. D. Hess
R. C. Rossi, and J. Rossoff
The Aerospace Corporation
2350 East El Segundo Boulevard
El Segundo, California 90245
Grant R-802853-01-0
ROAP No. R21AZU-025
Program Element No. 1BB392
EPA Task Officer: Dennis Cannon
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
-------
CONTENTS
CONCLUSIONS 1
RECOMMENDATIONS 3
1. INTRODUCTION 5
1. 1 Introduction 5
1.2 Program Objectives 7
1.3 Study Approach 7
1.4 Organization of This Report 8
2. SUMMARY 11
2. 1 Water Criteria 11
2. 2 Power Plant Scrubbing Processes 14
2. 3 Scrubber Liquor Characterization and
Assessment . 20
2.4 Water Treatment Technology and Costs 21
3. WATER QUALITY CRITERIA 27
4. SCRUBBER LIQUOR CHARACTERIZATION 29
4. 1 Power Plant Scrubber Liquor Sampling 29
4.2 Potential Scrubber Purge Conditions 29
4. 3 Extrapolation of Process Flows 56
4.4 Chemical Conditions Affecting Com-
position of Scrubber Liquors and Water
Imbalance 70
4.5 Assessment 73
5. WATER TREATMENT STUDIES 79
5. 1 Literature and Industry Review 79
5.2 Selected Water Treatment Processes 90
5.3 Estimated Water Treatment Costs 97
iii
-------
CONTENTS (Continued)
REFERENCES 107
APPENDIX A: Description of Chemical Analysis
Techniques 113
APPENDIX B: Costing Background Data 121
-------
FIGURES
2-1. EPA/TBA Shawnee Steam Plant Typical Flow
Diagram for Venturi and Spray Tower
Scrubber System: 8.34 MW Equivalent
Operation 16
2-2. Schematic of EPA/TVA Shawnee Steam Plant
Venturi and Spray Tower Scrubber 17
2-3. Water Balance for EPA/TVA Shawnee Steam
Plant Venturi and Spray Tower Scrubber
System: 8.34 MW Equivalent Operation 18
2-4. Water Balance for Venturi and Spray Tower
Scrubber System 19
2-5. Estimated Costs of Lime-Soda Softening and
Filtration 25
2-6. Estimated Costs of Lime-Soda and Filtration
Pretreatment Plus Reverse Osmosis 26
4-1. Schematic of EPA/TVA Shawnee Steam Plant
Venturi and Spray Tower Scrubber 34
4-2. Typical Flow Diagram for EPA/TVA Shawnee
Venturi and Spray Tower Scrubber System:
8.34 MW Equivalent Operation 35
4-3. Water Balance for EPA/TVA Shawnee Venturi
and Spray Tower Scrubber System: 8. 34 MW
Equivalent Operation 39
4-4. Schematic of EPA/TVA TCA Scrubber 40
4-5. Typical Flow Diagram for EPA/TVA Shawnee
TCA Scrubber System: 6.8 MW Equivalent
Operation 41
4-6. Water Balance for EPA/TVA Shawnee TCA
Scrubber System: 6.8 MW Equivalent
Operation 45
-------
FIGURES (Continued)
4-7. Schematic of Arizona Cholla Station FDS
and Packed Absorption Tower Scrubber 46
4-8. Typical Flow Diagram for Arizona Public
Service Cholla FDS and Packed Absorption
Tower Scrubber System: 115 MW
Equivalent Operation 48
4-9. Water Balance for Arizona Cholla Public
Service FDS and Packed Absorption Tower
Scrubber System: 115 MW Equivalent
Operation 51
4-10. Typical Flow Diagram for Duquesne Phillips
Single-Stage Scrubber System 53
4-11. Typical Flow Diagram for Duquesne Phillips
Station Dual-Stage Scrubber System 54
4-12. Schematic of Duquesne Phillips Station
Single-Stage Scrubber 55
4-13. Water Balance Duquesne Phillips Station
Single- and Dual-Stage Venturi Scrubber
Systems: 120 MW Equivalent Operation 59
4-14. Water Balance for Venturi and Spray Tower
Scrubber System (Extrapolation Based on
EPA/TVA Shawnee Steam Plant Type of
Process): 1000 MW Equivalent Operation 61
4-15. Water Balance for TCA Scrubber System
(Extrapolation Based on EPA/TVA Shawnee
Steam Plant Type of Process): 1000 MW
Equivalent Operation 62
4-16. Water Balance for FDS and Absorption Tower
Scrubber System (Extrapolation Based on
Arizona Cholla Public Service Process):
1000 MW Equivalent Operation 63
vi
-------
FIGURES (Continued)
4-17. Water Balance for Dual-Stage Venturi Scrubber
System (Extrapolation Based on Duquesne
Phillips Station Type of Process): 1000 MW
5-1
5-2.
5-3.
Estimated Water Treatment Costs per
Estimated Water Treatment Costs per
Kilowatt Hour for Power Output of 1000 MW
65
. . 98
99
100
5-4. Estimated Water Treatment Costs per
Million Btu Heat Input for Power Output of
1000 MW 101
5-5. Estimated Water Treatment Costs per Ton
of Coal Burned for Power Output of 1000 MW 10Z
vii
-------
TABLES
1-1. Power Plants Scrubbers Sampled
2-1. Water Quality Criteria and Range of
Concentration of Constituents in Scrubber
Liquors Studied 13
2-2. Summary of Generic Types of Water
Treatment Processes Capable of Meeting
Water Quality Criteria 22
4-1. Power Plants Sampled 29
4-2. Test Conditions for EPA/TVA Shawnee
Venturi and Spray Tower Scrubber 36
4-3. Analyses of Scrubber Liquors from EPA/TVA
Shawnee Venturi and Spray Tower Scrubber
System 37
4-4. Test Conditions for EPA/TVA Shawnee
TCA Scrubber 43
4-5. Analyses of Scrubber Liquors from EPA/TVA
Shawnee TCA Scrubber System 44
4-6. Operating Conditions for Arizona Public
Service Cholla FDS and Absorption Tower
Scrubber System 49
4-7. Analyses of Scrubber Liquors from Arizona
Public Service Cholla FDS and Absorption
Tower Scrubber System 50
4-8. Operating Conditions for Duquesne Phillips
Station: Single- and Dual-Stage Venturi
Scrubber Systems 57
4-9. Analyses of Scrubber Liquors from Duquesne
Phillips Station: Single- and Dual-Stage
Venturi Scrubber Systems 58
viii
-------
TABLES (Continued)
4-10. Summary and Comparison of Water Balance
Projections for 1000 MW Equivalent Scrubbers 66
4-11. Summary of Scrubber Bleed Stream Flow
Projections for 1000 MW Equivalent Scrubbers 68
4-12. Potential Water Uses and Constituents of
Concern 75
5-1. Summary of Generic Types of Water Treatment
Processes Capable of Meeting Water Quality
Criteria 80
5-2. Capability of Water Treatment Processes to
Meet Requirements 91
ix
-------
ACKNOWLEDGMENTS
Sincere appreciation is acknowledged for the guidance
and assistance provided by Dennis Cannon, the present EPA Project
Officer, and by Alden Christiansen and Guy R. Nelson, the prior
Project Officers of this study program.
The authors are especially grateful to the late
Lawrence J. Bornstein, who was the Project Manager for The
Aerospace Corporation during most of this program. His tireless
efforts in producing most of the survey data and evaluations and the
major portion of this report contributed greatly to the accomplishment
of this study.
-------
CONCLUSIONS
The following conclusions were made as a result of
an Aerospace Corporation study to determine the treatments and costs
applicable to the reuse of liquor from nonregenerable flue gas desul-
furization systems if there should be a necessity for a system purge
to provide the capacity for the required fresh make-up water. (It should
be noted that the determination of whether scrubbers should be purged
was not made and was not an objective of this study.) These conclu-
sions are based on the analysis and assessment of four different
scrubber liquors, which were characterized in this study and which are
believed to be representative of most but not all scrubber liquors.
With these data as background and using existing criteria and guide-
lines, some postulated criteria, and an estimation of purge flow rates,
the following conclusions are made regarding the potential reuse of
purged scrubber liquors:
a. Purged scrubber liquor must be treated if it is
to be reused as power plant service water or if
it is to be discharged.
b. Treatment for service water usage in most cases
can be accomplished by chemical precipitation
(e.g., lime-soda softening, filtration, and pH
control when necessary). This includes usages
such as scrubber make-up, cooling tower make-
up, and plant housekeeping. For an expected mi-
nority of cases, excessive chloride content will
require additional treatment such as reverse
osmosis.
-------
For discharge to surface waters, treatment
would consist of a pretreatment and an additional
treatment (e.g., lime-soda softening, filtration,
and pH control plus reverse osmosis). This
would result in costs approximately twice the
cost of treatment for service water in most cases.
Quantities of scrubber liquor in a 1000 Mw
equivalent unit presumed for treatment for
purposes of this study are 100 to 300 gpm
nominally and 700 gpm maximum. Total treat-
ment costs (capital and operating), based on a
30-year equipment life and 18 percent annual-
ized capital charges, are as follows:
Flow rate of
liquor to be
treated
200 gpm
Lime-soda softening
and filtration
$5.00/1000 gal
or
0. 05 mill/kWh
Lime-soda softening
and filtration
pretreatment plus
reverse osmosis
$9.00/1000 gal
or
0. 11 mill/kWh
700 gpm
$2.50/1000 gal
or
0. 09 mill/kWh
$5.00/1000 gal
or
0.22 mill/kWh
To convert:
gpm to cu m/hr, multiply by 0. 227.
dollar/1000 gal to dollar/cu m, multiply by 0. 264.
-------
RECOMMENDATIONS
Analyses determined in this study of waste stream
liquors and a review of related data given in Refs. 1 through 3 show
that most lime and limestone scrubber liquors, including those
produced during steady-state operating conditions, must be treated
if they are to be reused after a scrubber system blowdown. If it is
shown that a blowdown is required, it is recommended that the purged
liquor be treated for reuse as power plant service water.
The potential liquor treatment requirements that were
determined in this study were based on the chemical characterization
of liquors from a limited number of power plants and postulated
purge conditions. Therefore, it is recommended that additional
studies be made to broaden the data base. These should include the
characterization of liquors from additional power plant scrubbing
systems, development of more comprehensive usage criteria, de-
tailed analyses of potential purge conditions, and treatment demon-
strations for those purge conditions, if any, that are shown to be
necessary. Specific recommendations are as follows:
a. Analyses
1. An expanded data base regarding scrubber
liquor chemical characterization should be
determined or assembled. Any additional
factors which were not part of this study,
such as radioactivity, should be considered
in this characterization.
-------
2. Detailed analyses should be conducted on
conditions that may require scrubber purge,
including purge flow rates and purge cycles.
These results should be confirmed by oper-
ational tests.
3. Investigations should be made of the tech-
niques available that may prevent scrubber
purge conditions from arising. These tech-
niques should be tested on operational
systems.
4. If scrubber purge conditions are determined
to be unavoidable, the net effect of waste
disposal should be determined, including the
brine produced in water treatment systems.
b. Usage Criteria
1. Criteria should be established, consistent
with results of the analyses recommended
above, to define power plant service
water quality.
2. Power plant effluent guidelines and standards
should be expanded, consistent with the
analyses recommended above, to relate to
the potential discharge of treated liquors.
c. Demonstration
If it is shown that scrubber purge is unavoidable,
a demonstration of purged liquor treatment is
recommended. At this time, the following five
treatment processes are recommended for con-
sideration in the selection of hardware for
demonstration:
1. Chemical precipitation (e. g. , lime-soda
softening)
2. Chemical precipitation plus reverse osmosis
3. Chemical precipitation plus ion exchange
4. Multistage flash evaporation
5. Brine concentration
-------
SECTION 1
INTRODUCTION
1. 1 INTRODUCTION
The installation of nonregenerable flue gas
desulfurization systems in coal-burning power plants using lime or
limestone as the scrubber absorbent requires large quantities of
water (Refs. 1 and 2). This water circulates through the scrubber
loop as an alkaline slurry. The sulfur waste products that form are
bled from the scrubber, and, in most systems where a settling tank
or clarifier is used, the supernatant liquor is recirculated to the
scrubber. Wastes produced by these systems are generally removed
by disposing of the clarifier underflow, a filter cake, or a centri-
fuge cake. They are typically ponded or placed in a landfill after
possible intermediate treatment for environmental protection. All
supernate liquors resulting from the waste disposal processes are
returned to the scrubber loop; hence, the current scrubber systems
can be described as closed-loop operations. The only water lost is
that which is evaporated with flue gases or occluded with the solid
waste products. Fresh water is added to make up for these losses
and is adequate to prevent the formation of scale that results from
precipitation of solids at critical points within the system.
This study evaluates the potential reuse of scrubber
liquor within water systems of the power plant, in the event that a
purge is required to maintain adequate capacity for fresh make-up
-------
water to the scrubber. The power plant water systems to be evaluated
are those that may accept nonpotable water. These include cooling towers,
housekeeping service, and scrubber make-up. Alternatively, these
waters can be treated for discharge. In addition, those conditions
that exist or may develop in the operation of flue gas scrubbing
systems that may require a system purge are postulated. These
conditions include variable sulfur content of the coal, variable boiler
load factor, and sulfite sludge oxidation. In an evaluation of the reuse
of scrubber liquor in other water systems, it is necessary to deter-
mine the relationship of the water quality of these liquors to the limit-
ing requirements of the intended applications.
Currently, The Aerospace Corporation is conducting
a study, "Disposal of By-Products From Nonregenerable Flue Gas
Desulfurization Systems," under Environmental Protection Agency
(EPA) Contract Number 68-02-1010 (Refs. 2 and 3). Data obtained
in that study have shown that scrubber liquors are unacceptable for
direct discharge and may not be acceptable for most other water
systems within the power plant without prior water treatment.
This program was initiated because of the potential
unacceptability of direct discharge of scrubber liquors and the need to
determine the potential reuse of purged liquors as power plant service
water. Secondary reasons are the national goal of zero discharge by
1985 in accordance with the Federal Water Pollution Control Act
Amendments of 1972, and the requirements for the 1977 and 1983
discharge limitations based on the best practical control technology
available and the best available technology economically achievable
(Ref. 4). Program objectives and the study approach to characterize
the problem and assess the technology applicable for water reuse are
given in Sections 1.2 and 1.3, respectively.
-------
1.2 PROGRAM OBJECTIVES
The principal objectives of this study were to determine
the requirement, if any, for treatment of nonregenerable flue gas
desulfurization waters for reuse and to determine the water treatment
processes and costs applicable for reuse of these waters.
1.3 STUDY APPROACH
The study objectives were accomplished by performance
of the following specific tasks:
a. Determination of the characteristics of scrubber
system liquors from four coal-burning power
plants at the potential discharge points and
quantification of the constituents affecting water
quality.
b. Identification of the scrubber systems with flow
diagrams and corresponding liquor flow rates
under varying scrubber operating conditions
and estimation where possible of those conditions
that may require system liquor blowdown.
c. Assessment of the quality of liquor waste waters
against appropriate water quality criteria.
d. Survey of water treatment processes available
and in development and assessment of the
applicability of those processes deemed capable
of reducing the concentration of constituents
of concern in the waste waters to allow reuse
within the power plant or to allow direct
discharge.
e. Recommendation of appropriate water treatment
processes based upon technical and economic
considerations.
Samples of flue gas desulfurization scrubber system
liquors were obtained from four different scrubber systems
(Table 1-1) using lime or limestone as the sulfur dioxide absorbents.
The concentration of the trace elements and major species in the
liquors of the potential discharge flow streams were determined by
-------
conventional analytical techniques, and the results of these analyses
were assessed against water quality criteria for the nonpotable reuse
of the water within the power plant operations or for its discharge
to surface waters.
Flow-diagrams and rates were determined for the
flow streams of the four scrubber systems. An extrapolation was
made of the flow rates from each of these systems to an identical
system of 1000 MW equivalent capacity.
A technical and economic review was conducted of the
literature and industry to identify potentially acceptable water treat-
ment processes. The economic assessment included initial invest-
ment, annual capital charges, and operating expenses for these
processes.
1.4 ORGANIZATION OF THIS REPORT
Following this introduction, a summary of the total
study is given in Section 2. Appropriate figures and tables are
included to support the findings.
The technical discussion is given in Sections 3, 4,
and 5. Significant features of each of these are as follows:
Section 3 discusses the water quality criteria selected
for water reuse in the power plant as nonpotable service water or
for discharge to surface waters.
Section 4 contains the flow diagrams, flow rates in
the potential discharge flow streams, and water balances for the
four scrubber systems. These data were extrapolated to power plant
scrubbing systems using the same design and operating characteris-
tics, but operating at 1000 MW capacity equivalent. This section also
contains complete tabulated data and an assessment of the chemical
characteristics of the scrubber liquors. An assessment is also made
of the concentration of constituents at the potential discharge points
8
-------
Table 1-1. POWER PLANT SCRUBBERS SAMPLED
Description of
. scrubber and coal
Scrubber
Type
Equivalent capacity,
MW
Manufacturer
Absorbent
Coal
Source
Sulfur content,
percent
Power plant
Tennessee Valley Authority
(TVA) Shawnee Steam Plant,
Paducah, Kentucky
Venturi and
spray tower
10
(EPA/ TVA
prototype)
Chemical
Construction
Corporation
Lime
Eastern
3.4 (avg)
Turbulent
contact
absorber (TCA)
10
(EPA/TVA
prototype)
Universal Oil
Products
Company
Limestone
Eastern
3.4 (avg)
Arizona Public
Service Company
Cholla Station,
Joseph City,
Arizona
Flooded disk
scrubber (FDS)
and absorption
tower
120
(full-scale)
Research
Cottrell, Inc.
Limestone
and flyash
Western
0.5
Duquesne
Light Company
Phillips Station,
South Heights,
Pennsylvania
Single- and
dual- stage
venturi
125 each
(four full-
scale units)
Chemical
Construction
Corporation
Lime
Eastern
2
-------
against water quality criteria for reuse as service water in other
water systems in the power plant, and discharge to lakes and streams
if the need should arise.
Section 5 summarizes the applicable water treatment
processes and equipment described in the literature or in data
available from industrial contractors. A technical assessment is
made of the potentially applicable processes, and engineering cost
estimates are made of the initial investment, annual capital charges,
and operating costs.
Supporting data in the appendixes include a description
of the analytical techniques used to determine the concentration levels
of the constituents in the scrubber liquors, an evaluation of the accu-
racy and precision of the analyses, and the background data used in
preparing the water treatment cost estimates.
10
-------
SECTION 2
SUMMARY
2.1 WATER CRITERIA
In the nonregenerable flue gas desulfurization scrubber
operation, the scrubbing systems water is continuously recycled
through a closed-loop system. Normally, there is no discharge of
water from these systems except for the moisture discharged with the
sludge and that which evaporates from the scrubber. Under off-
design conditions or any other condition not generally employed in
scrubbers at this time (Section 4.2), a certain amount of scrubber
liquor may have to be purged and treated prior to its reuse. This
assessment is based on a review of the chemical analyses of system
liquors sampled from four different scrubbing systems (Section 2.3)
and a comparison of these data with selected water use criteria.
The four systems from which liquors were analyzed do
not necessarily represent a typical cross section of scrubbers installed
or planned in the U.S. However, on the basis of other studies (Refs.
1 through 3), it appears that most system liquors will have charac-
teristics within the ranges of those analyzed in this study. Some ex-
ceptions, such as the Mohave Station scrubber in Nevada and the
General Motors scrubber in Parma, Ohio (both of which have high
11
-------
concentrations of chloride), would require more extensive water
treatment than the four stations represented in this study. The vari-
ations in scrubber liquor characteristics on a national overview basis
should be the subject of a separate study; however, the conclusions
reached in this report pertain to only those stations studied, which are
believed to be representative of most nonregenerable scrubbing systems
now operating or to be installed.
The potential reuse considered for treated scrubber
liquors are power plant service water, scrubber reuse, and direct
discharge to lakes and streams. Reuse as power plant service water
or scrubber make-up was considered as a single category, and direct
discharge as another. No specific criteria were found for power plant
service water usage; therefore, the criteria postulated for this study
was based on an informal survey of water requirements now practiced
within the power industry. These criteria, listed in Table 2-1 as
"nonpotable service water criteria" were used also for scrubber make-
up on the basis of results obtained in The Aerospace Corporation study
of scrubber waste disposal requirements (Ref. 3). For direct dis-
charge, several criteria were available, e.g. , the U.S. Public Health
Service drinking water standards of 1962 (Ref. 5), federal and state
water quality criteria documents (Refs. 6 and 7), EPA proposed cri-
teria for public water supply intake dated October 1973 (Ref. 8), and
the EPA effluent guidelines (Ref. 9). Of the public service and drinking
water criteria available, the EPA proposed critera of October 1973 was
selected as being the most appropriate for this assessment and is given
in Table 2-1. The EPA effluent guidelines at this time are not broad
enough to cover most of the major constituents of scrubber liquors, and
therefore were not listed. They were used, however, in a comparative
sense (Section 2.3).
An examination of Table 2-1 shows that the criteria
for nonpotable service usage is much less stringent than the public
service water criteria. Service water usage is concerned mostly
12
-------
Table 2- 1. WATER QUALITY CRITERIA AND RANGE OF
CONCENTRATION OF CONSTITUENTS IN
SCRUBBER LIQUORS STUDIED
Constituents
Aluminum
Antimony
Arsenic
Beryllium
Boron
Cadmium
Calcium
Chromium (total)
Cobalt
Copper
Iron
Load
Magnesium
Manganese
Mercury
Molybdenum
Nickel
Potassium
Selenium
Silicon
Silver
Sodium
Tin
Vanadium
Zinc
Carbonate
Chloride
Fluoride
Sulfite
Sulfate
Phosphate
Nitrogen (total)
Chemical oxygen demand
Total dissolved solids
Total alkalinity (as CaCOj)
Conductance mho/cm
Turbidity, Jackson units
Pi
Concentration, mg/f (unless otherwise indicated)
Range of
constituent
concentrations
at potential
discharge pointsa
0.03 - 0.3
0.09 - 2.3
<0.004 - 0.3
<0.002 - 0.14
8.0 - 46.
0.004 - 0.11
520. - 3000.
0.01 - 0.5
0.10 - 0.7
<0.002 - 0.2
0.02 - 8.1
0.01 - 0.4
3.0 - 2750.
0.09 - 2.5
0.0004 - 0.07
0.91 - 6.3
0.05 - 1.5
5.9 - 32.
<0.001 . 2.2
0.2 . 3.3
0.005 - 0.6
14.0 - 2400.
3.1 - 3.5
<0.001 - 0.67
0.01 - 0.35
-------
with nonscaling and noncorrosive properties, whereas the public
service water criteria are concerned with health factors. Also,
the demand for nonpotable service water in a power plant is always
high, and the water treatment requirements for public service water
usage are more extensive and expensive than that required for
service water (Section 5.3). Throughout this report assessments are
made for both types of uses. However, it is assumed that the
treatment of scrubber liquors for reuse as public service water
(i.e. , drinking) would be employed only in rare cases. Therefore,
the major emphasis in this study is placed on treatment of scrubber
effluent for reuse as service water.
2.2 POWER PLANT SCRUBBING PROCESSES
As part of this study, a four-step analysis was made
of the scrubbing process at each of the four scrubbers (Table 1-1).
This analysis included:
a. Examining the operating conditions occurring at
the time the scrubber liquor samples were taken
b. Identifying the water flow rates in the various
flow streams of that scrubber system
c. Projecting those flow rates to a scrubber of a
size equivalent to that required for a 1000 MW
power plant
d. Estimating potential purge rates for purposes of
this study.
Liquor samples were taken from the four scrubber
systems of the coal-burning power plants identified in Table 1-1.
These scrubbers ranged in equivalent capacity of 10 to 400 MW. The
plants burned either eastern or western coals, and the flue gases
were scrubbed with a lime or limestone slurry water that reacted
with the sulfur dioxide. The process flow diagram for the EPA/TVA
venturi and spray tower scrubber, which is designed for a 10 MW
14
-------
equivalent capacity, using lime as the absorent, is shown in
Figures 2-1 and 2-2 as an example case. The water balance for this
system is shown in Figure 2-3 (Ref. 10). Folio-wing the scrubber is
a clarifier and a vacuum filter. The separated liquor is recycled
within the scrubbing system. At the times of sampling, an average
of 3.4 percent sulfur coal was burned, 78 to 93 percent removal of
the sulfur dioxide was obtained, and 306 cu m/hr (1350 gpm) of water
was recycled into the scrubber.
The flow diagram illustrates the amount of water flow-
ing in each major flow stream and its percentage as related to the
total feed water to the scrubber. In this system, less than 1 percent
of the feed water to the scrubber was lost with the sludge to the dis-
posal pond and to the stack with the flue gases. Fresh water was
introduced into the system with the lime slurry mix and by direct
addition of fresh make-up water into the hold tank. The extrapolated
values are shown in Figure 2-4 for a water balance for a 1000 MW
scrubber system based on the same process used in the venturi and
spray tower system. The water flow rates were scaled up from the
test conditions that existed during sampling at the EPA/TVA facility.
The calculations were based upon operating at 90 percent sulfur
dioxide removal efficiency, using the identical scrubber design, having
the same conditions in the scrubber and burning the same coal, and
retaining the same scrubber liquid-to-flue gas ratio. For the 1000 MW
capacity scrubber plant, it was estimated that 320 cu m/hr (1410 gpm)
of water would be lost, of which 204 cu m/hr (900 gpm) would escape
from the stack with the flue gases and 115 cu m/hr (510 gpm) would
be occluded with the solids in the filter cake that was sent to the
sludge pond for disposal. The rates of occluded water in the dis-
charge of all four scrubbers were used to estimate potential purge
rates (Section 2.4).
15
-------
VENTURI
SCRUBBER
See Figure
2-2
for details)
CLARIFIER
^
DISCHARGE
TEST CONDITIONS
LIQUID-TO-GAS RATIO: 54.0 gal/1000 CU ft
SCRUBBER EFFLUENT SOLIDS: 8 PERCENT
CLARIFIER UNDERFLOW SOLIDS: 18 PERCENT
FILTER CAKE SOLIDS: 45 PERCENT
COAL: 3.4 PERCENT SULFUR
Figure 2-1.
EPA/TVA Shawnee Steam Plant typical flow diagram
for venturi and spray tower scrubber system:
8.34 MW equivalent operation
16
-------
FLUE
GAS
OUTLET
CHEVRON -.
DEMISTER >*
AFTER-SCRUBBER-
FLUE
GAS
INLET
INLET
IINLt I V. , *
SLURRY ^
-------
7.5 gpm
0.55%
STACK
PERCENTAGE VALUES ARE
PERCENT OF MAXIMUM
WATER FLOW INTO SCRUBBER
WATER FLOW RATE IN GALLONS
PER MINUTE (gpm), TO CONVERT
TO cu m/hr MULTIPLY BY 0.227
43.4 gpm
3.2T
POND
TEST CONDITIONS
LIQUID-TO-GAS RATIO: 54.0 gal/1000 cu ft
SCRUBBER EFFLUENT SOLIDS: 8 PERCENT
CLARIFIER UNDERFLOW SOLIDS: 18 PERCENT
FILTER CAKE SOLIDS: 45 PERCENT
COAL: 3.4 PERCENT SULFUR
Figure 2-3.
Water balance for EPA/TVA Shawnee Steam Plant
venturi and spray tower scrubber system:
8.34 MW equivalent operation
18
-------
900 <
0.5S
am
VENTURI
AND
SPRAY
TOWER
SCRUBBER
163,000 gpm
100%
STACK
4800 gpm
!.95T
PERCENTAGE VALUES ARE
PERCENT OF MAXIMUM
WATER FLOW INTO SCRUBBER
WATER FLOW RATE IN
GALLONS PER MINUTE (gpm).
TO CONVERT TO cu m/hr
MULTIPLY BY 0.227
5200^
POND
EXTRAPOLATED OPERATING CONDITIONS
LIQUID-TO-GAS RATIO: 54.0 gal/1000 cu ft
SCRUBBER EFFLUENT SOLIDS: 8 PERCENT
CLARIFIER UNDERFLOW SOLIDS: 18 PERCENT
FILTER CAKE SOLIDS: 45 PERCENT
COAL: 3.4 PERCENT SULFUR
SO2 REMOVAL EFFICIENCY: 90 PERCENT
Figure 2-4.
Water balance for venturi and spray tower scrubber
system (extrapolation based on EPA/TVA Shawnee
Steam Plant process): 1000 MW equivalent
19
-------
Because each process is configured differently and
has its own individual operating characteristics, the data obtained for
these four scrubbers cannot be extrapolated directly to other systems.
2.3 SCRUBBER LIQUOR CHARACTERIZATION AND
ASSESSMENT
Table 2-1 lists the constituent concentrations found in
the scrubber liquor samples taken from the potential discharge points
of the four scrubbers.
Variations in the concentration of the constituents are
due in part to the different operating conditions that existed among
plants. For example, the Arizona Public Service Cholla Station
operates in a tightly closed scrubber loop situation resulting in high
concentrations of chlorides and total dissolved solids in the bleed
stream from the flooded disk scrubber slurry tank that is presently
the discharge point for this system (Ref.10). Conversely, the
Duquesne-Phillips Station system has experienced operating conditions
requiring high water use within the scrubbing system which diluted
the concentration of the dissolved ingredients (Ref. 11). Variations
in scrubber liquor composition at the potential discharge points were
also caused by differences in the coal composition, in the absorbents,
and in the scrubber process design and operating characteristics.
For the scrubber liquor to be reused in other systems
within the power plant, it will be necessary to reduce the concentra-
tion of boron, magnesium, calcium, or sulfate, from one or more
of the scrubber flow streams studied. An adjustment of the pH of the
water would also be necessary for some of the flow streams. These
reductions and adjustments of the characteristics of purged liquor
would make the liquor applicable for reuse within the scrubber also.
The concentration of lead, mercury, selenium,
chlorides, and sulfates and the pH in the scrubber liquor streams
20
-------
from all four power plants exceed the EPA-proposed criteria for
public water supplies. In addition, in one or more of the streams
the concentration of arsenic, boron, cadmium, chromium (total),
iron, manganese, and silver also exceed the criteria, but not
simultaneously. Because these ingredients exceed the criteria, the
scrubber liquor must be treated before it can be discharged.
The EPA effluent guidelines and standards (Ref. 9),
though not as restrictive or as encompassing as public water supply
criteria, require treatment of some of these liquors because of limi-
tations on iron and pH level. However, this approach for direct
discharge should be used with caution since the major constituents
of the liquor are not specified.
The following section summarizes treatment tech-
nologies available for these purposes and discusses the relative com-
plexities of treatment for reuse as service water and discharge to
surface waters.
2.4 WATER TREATMENT TECHNOLOGY
AND COSTS
A review was made of the literature, and technical
meetings and discussions were held with water treatment processors
and equipment manufacturers to determine the water treatment tech-
niques applicable to the potential reuse or discharge of scrubber
liquor. The results of this review are summarized in Table 2-2,
which identifies the generic types of water treatment processes, the
development status of each, and indicates whether or not the water
would meet the water quality criteria.
It was found that a treatment process such as lime-
soda softening which reduces the concentration of all constituents,
except soluble sodium and chloride salts, would be sufficiently effec-
tive that the water of the four liquors studied would be acceptable
21
-------
Table 2-2. SUMMARY OF GENERIC TYPES OF WATER
TREATMENT PROCESSES CAPABLE OF
MEETING WATER QUALITY CRITERIA
Water treatment
process
Complete system
Multistage evaporation
Tower distillation
Brine concentration
Spray drying
Vacuum freezing
Solar distillation
Rotating bipolar electrodes
Partial operation
Filtration or centrifugation
Ultrafiltration
Reverse osmosis
Ion exchange
Chemical precipitation
(including lime -soda
softening)
Electrodialysis
Selective absorbent
Electrochemical; fluidized
bed
Foam separation
EPA-proposed
public water
supply intake,
October 1973
Yes
Yes
Yes ,
Yes
Yes
Yes
No
No
No
Yes
Yes
No
Yes
No
No
No
Nonpotable
service
water
Yes
Yes
Yes
Yes
Yes
Yes
No
No
No
Yes
Yes
Yes
Yes
No
No
No
Status
Operational
Operational
Operational
Operational
Development
Development
Laboratory
Operational
Operational
Operational
Operational
Operational
Development
Laboratory
Laboratory
Laboratory
Used with other operations to form a total water treatment system
22
-------
(after pH adjustment, if necessary) for use as power plant service
water (e.g., cooling tower makeup, heat exchangers, scrubber
makeup, and housekeeping). In plants -where a chloride concentration
greater than 5000 mg/i is unacceptable, lime-soda softening would
have to be followed by an additional process (e.g. , reverse osmosis).
This value was selected on the basis of an informal survey of the
industry. Variations can be expected depending on the characteris-
tics of any particular plant.
The use of scrubber liquor in a public water supply
would require a complete treatment, e.g., lime-soda softening
followed by reverse osmosis in all cases and, in some cases, pH
adjustment. Therefore, treating the liquor for reuse as drinking
water can require a treatment process and cost appreciably in excess
of that required for reuse as service water.
Other existing treatment processes such as multistage
flash evaporation, brine concentration, and softening-ion exchange
could be used instead of softening and reverse osmosis, but would
generally be less cost effective. Processes in development have not
yet shown that they are cost effective.
A preliminary estimate was made of the amount of
water that may have to be removed from the scrubber loop to provide
the adequate volume for fresh water make-up requirements
(Section 4. 2). Two scrubber systems were considered: a venturi and
spray tower system and a turbulent contact absorber system. The
greatest amount of excess liquor that may have to be removed for
maintenance of the water balance may occur in the case where gypsum
is formed and the excess liquor is subsequently removed from the
scrubber loop. The least amount may be in the case of a reduced
load factor on the system. The quantity of water to be removed from
a 1000 MW equivalent system appears to be in the range of 22.7 to
68. 1 cu m/hr (100 to 300 gpm) but can possibly be as high as
23
-------
160 cu m/hr (700 gpm). A detailed analysis of potential scrubber
purge flow rates was not a subject of this study; however, an estimate
was considered necessary because of the wide variations in treatment
costs as a function of treatment system size. For this reason, costs
were determined and plotted for flow rates up to 1000 gpm. Specific
case values are presented for the lower nominal and the upper end
of the estimated flow rate range, i.e. , 200 and 700 gpm.
The total cost, including capital charges and operating
expenses, for treating 45.4 and 160 cu m/hr (200 and 700 gpm) of
scrubber liquor using a lime-soda process varies from 0. 51 to
0.25 cents per gallon. This converts to a reverse order of approxi-
mately 0.05 to 0.09 mills per kWh at an average annual production
of 4560 million kWh over 30 years. Likewise, when lime-soda pre-
treatment is combined with reverse osmosis, these values are
approximately double, i.e., 0. 9 to 0. 5 cents per gallon and 0. 1 to
0.2 mills per kWh, respectively. These costs are shown on
Figures 2-5 and 2-6 where values are given for flow rates from 200
to 1000 gpm.
24
-------
1974 DOLLARS
1000 MW POWER PLANT SCRUBBER
COSTS BASED ON 30-YEAR ANNUALIZED COSTS, 4560 OPERATING
HOURS PER YEAR, AND CAPITAL CHARGES OF 18%
LAND COSTS NOT INCLUDED
"8 2.0
WATER TREATMENT COST, CENTS PER GAL TREATED
n in nil o 20 0 25 0.30 0.35 0.40
0.45
0.50
O
t;
I
I
1.5 -
1.0 -
0.5
WATER TREATMENT PLANT
CAPACITY, gpm
0.025 0.050 0.075 0.100
WATER TREATMENT COST. MILLS PER KWh PRODUCED
200 gpm
Figure 2-5. Estimated costs of lime-soda softening
and filtration
25
-------
1974 DOLLARS
1000 MW POWER PLANT SCRUBBER
COSTS BASED ON 30-YEAR ANNUALIZED COSTS. 4560 OPERATING
HOURS PER YEAR, AND CAPITAL CHARGES OF 18%
LAND COSTS NOT INCLUDED
8
D
|
IO
g
3
j
2
S 2.0
0.10
WATER TREATMENT COST, CENTS PER GAL TREATED
0.20 0.30 0.40 0.50 0.60 0.70 0.80
0.90
3.0
UJ
i/i
tu
1.0
'TO I
CONVERT
FROM
I
I
I
I
I
TO
MULTIPLY
BY
gpm
gal
cu m/hr
liter
I
WATER TREATMENT
PLANT CAPACITY, gpm
1.00
0.05 0.10 0.15 0.20 0.25 0.30 0.35
WATER TREATMENT COST, MILLS PER kWh PRODUCED
0.40
Figure 2-6. Estimated costs of lime-soda and filtration
pretreatment plus reverse osmosis
26
-------
SECTION 3
WATER QUALITY CRITERIA
The scrubber liquor is continuously recycled in the
flue gas desulfurization scrubbing system. Currently, a bleed stream
of scrubber slurry is withdrawn from the scrubber loop and is
dewatered; the resultant solids are sent to a waste disposal pond.
The supernate liquid (system liquors) and make-up water, which is
added to replace the water lost with the waste product and by evapora-
tion with the flue gases, are returned to the scrubber. If the scrubber
liquors are to be reused within the power plant or discharged, they
must meet the accepted water quality criteria for their intended use.
Large quantities of water are needed in the power
plant operations for nonpotable uses (e.g., cooling water, scrubber
make-up, housekeeping, and irrigation of the power plant vegetation).
These uses can be divided into nonpotable water for use throughout
the plant that will not cause scaling or corrosion of the equipment
and irrigation water. After the identification of the specific water
use, it is necessary to establish criteria to ensure that the water
will be of sufficient quality for that use.
The October 1973 EPA-proposed criteria for public
water supply intake (Ref. 8) is part of the data base for the 1985 goal
of zero discharge of pollution into the nation's waters as defined by
Public Law 92-500. These criteria were used to judge the quality of
the scrubber liquor that might be discharged. A comparison of the
27
-------
EPA criteria with the U.S. Public Health Service (USPHS) 1962
criteria (Ref. 5) shows a similarity except that the USPHS criteria
have more stringent limitations on the concentration of total dissolved
solids (TDS) (500 mg/j? versus no limit) and arsenic (0.05 versus
0. 10 mg/i). Conversely, the EPA criterion is tighter on the con-
centration of nitrates (10 versus 45 mg/jf).
The service water and discharge water criteria are
shown in Table 2-1. Because there are no published or approved
power plant service water criteria, the service water values used
were based upon the literature and discussing requirements with
water treatment processors and power plant operating personnel.
These values are goals assumed acceptable for service water use
within most power plants; they are not absolute requirements. A
specific power plant may have special requirements depending upon
its operation and local restrictions. The EPA-proposed criteria
for public water supply were used as the discharge criteria.
28
-------
4.1
SECTION 4
SCRUBBER LIQUOR CHARACTERIZATION
POWER PLANT SCRUBBER LIQUOR SAMPLING
The data base for this study was prepared from samples
taken from the four scrubbers listed in Table 4-1.
Table 4 -1. POWER PLANTS SAMPLED
Power plant
TVA Shawnee
Steam Plant
TVA Shawnee
Steam Plant
Arizona Public Service
Company Cholla
Station
Duquesne Light
Company Phillips
Station
Scrubber
systems
Venturi and
spray tower
Turbulent con-
tact absorber
Flooded disk and
absorption tower
Single- and dual-
stage venturi
Coal
Eastern
Eastern
Western
Eastern
Absorbent
Lime
Limestone
Limestone and
fly ash
Lime
4.2
POTENTIAL SCRUBBER PURGE CONDITIONS
Power plant scrubbing systems are fundamentally iden-
tical although each installation is unique in detail design features.
Each scrubber vendor has developed hardware having basic system
design differences; cither differences also exist as a consequence of
29
-------
coal type, fly ash collection facilities, and scrubber size. The
similarities among scrubbing facilities are based primarily on the
chemistry of scrubbing. For the lime or limestone systems, an
alkaline slurry is recirculated between the scrubber and a reaction
tank where the effluent from the scrubber reacts with the absorbent
material. A bleed stream from the reaction tank carries off solid
reaction products (and fly ash) and passes through a primary and often
secondary dewatering system. A supernate liquor is returned to the
system and, although clear, it is saturated with dissolved calcium
sulfite and/or calcium sulfate salts.
Under normal conditions, no system bleed has been
found necessary. Normal operating conditions are here considered as
those conditions defined by design criteria and include characteristics
of the boiler, coal, fly ash, existing hardware for retrofit systems,
absorbent, disposal facilities, and any other consideration peculiar to
that facility or design. The only water loss from these systems occurs
by evaporation in the stack and from water lost to the disposal site by
occlusion with the solids. In the operation of scrubber systems studied
for this report and other systems now known to be operating, no system
bleed is considered necessary for continuous operation. The fresh
water used by the pump seals and demister spray provides enough
make-up water to equal the loss of water by evaporation and solids
disposal. The solids disposal includes enough dissolved solids in its
occluded water to normally meet the bleed requirements necessary
to limit dissolved solids build-up.
Several circumstances not now considered normal
operating conditions may exist in the future that could affect the water
balance of the system and create a need for a system purge. One such
circumstance can arise when a scrubber is operated at a flue gas mass
load level below some critical value created by low boiler loading, or
alternatively under circumstances whereby lower sulfur coal is used
than that for which the system was designed. Under these circumstances,
30
-------
reduced solid quantities are formed in the reaction tank and a
reduced bleed stream flow must be established so that a constant
solids content can be maintained in the recirculation system of the
scrubber. (The solids content of recirculating liquor is maintained
so as to provide suitable nuclei in the scrubber onto which newly
precipitated material may crystallize). The reduced bleed decreases
the loss of water from the system. However, the need for fresh
make-up water, which is dictated primarily by the requirement for
demister water for scale prevention, is not reduced. Thus, when
the fresh water requirement of the system is satisfied, an excess in
system water would occur and a system purge would be required to
maintain proper water balance. Of the power plants sampled in this
study, one has experienced this condition while operating the scrubber
at 33 percent of design load.
Another condition that could force a system bleed may
arise under circumstances in which operational changes are made
from that of the system design. For example, the oxidation of sulfite
sludge to a sulfate sludge may evolve into a normal operating prac-
tice as a consequence of the greater efficiency in dewatering gypsum
sludge and improved disposal practices. Under this circumstance,
more sludge liquor would return to the scrubber because of the
improved separation of liquid from the solids. The effect would be a
reduction of water loss from the system. If this reduction reduces
the make-up requirement below the fresh water requirement, the
condition of excess water would result as described in the previous
paragraph and a system bleed would be required to maintain water
balance. This same condition may also exist if any of several im-
proved dewatering methods are used.
As yet another circumstance, major repairs or hard-
ware replacement are unavoidable, although not necessarily consid-
ered part of normal operation. If major maintenance can be
31
-------
prescheduled, it is often possible to divert the liquor from one system
to another over a period of time. However, major repairs are often
the result of more catastrophic circumstances, and the need for a
quick blowdown may become an immediate requirement. Acceptable
plans for such a contingency should be part of every plant's operation,
and the reuse of these waters in other parts of the power plant is
recognized as an alternative to be considered.
In each of the power plants sampled, reported values
for total water balance are those for circumstances that are consid-
ered as normal and typical operating conditions, i.e., conditions
within the range of design criteria and typical for the operation of
>'<
that plant. Normal design criteria include maximum power output
and a level of reduced power output that is usually undefined and
depends on the flexibility of the system. Thus, the point at which
fresh water requirements exceed make-up water requirements such
that normal operation is affected must be determined empirically.
From the power stations sampled, it can be estimated that this point
is reached in the vicinity of about 50 percent of the maximum design
parameter, i.e., 50 percent load factor or 50 percent reduction in
sulfur content. The reduced flue gas mass loading that may accom-
pany startup or shutdown of either boiler or scrubber causes a tem-
porary excess liquor problem over such a short period of time that
the capacity of a disposal pond is usually more than adequate to
accept this excess water and, under normal conditions, to return it
in the recirculation system.
In the one case in which the plant was operating at 33 percent of de-
sign load, a system purge was necessary. Reported values were.
corrected for this condition when extrapolating to full scale.
32
-------
4.2.1 Shawnee Steam Plant Venturi and Spray Tower
Scrubber System
This system is one of three parallel scrubber modules
installed on a 150 MW coal-fire boiler at the TVA Shawnee Steam
Plant, Paducah, Kentucky (Ref. 10). These modules were installed to
provide a large, versatile prototype system as a test bed for wet
scrubbing sulfur dioxide (SCO and particulates from boiler flue gas.
Each module has a 10 MW equivalent capacity; this is achieved by
taking approximately 14. 15 cu m/sec (30,000 acfm) of flue gas from
the ductwork of Boiler No. 10 upstream of the electrostatic precipi-
tators. This dual-stage scrubber, which is shown in Figure 4-1,
consists of a venturi with an adjustable plug that permits control of
the pressure drop in the first stage and a spray tower after-scrubber
in the second stage. It was manufactured by the Chemical Con-
struction Corporation (Chemico).
Flue gas entering the scrubber passes through the
venturi and then upward through the spray tower countercurrent to
the scrubbing liquors that are recirculated from the reaction tank.
The effluent liquors return to the reaction tank where pH is adjusted
with a slaked lime slurry. A bleed stream from the reaction tank is
routed to a clarifier. The clarifier overflow is returned to the
reaction tank; the underflow passes through a vacuum filter; and the
filtrate is returned to the reaction tank. The filter cake is discharged
to a disposal pond. A flow diagram is shown in Figure 4-2.
On three occasions, slurry samples were taken from
three stations in the system: (a) the scrubber effluent to the reaction
tank, (b) the clarifier underflow to the vacuum filter, and (c) the
filtrate from the vacuum filter to the hold tank. The scrubber test
conditions are presented in Table 4-2, and the analysis of the liquors
from these sample points is presented in Table 4-3. Although the'
33
-------
CHEVRON -
DEMISTER
AFTER-SCRUBBER
FLUE
GAS
INLET
THROAT
ADJUSTABLE X
PLUG
VENTURI
SCRUBBER
FLUE
GAS
OUTLET
INLET v. _JL_
SLURRY ^
-------
VENTURI
SCRUBBER
(See Figure
4-1
for details)
CLARIFIER
^>
DISCHARGE
TEST CONDITIONS
LIQUID-TO-GAS RATIO: 54.0 gal/1000 cu ft
SCRUBBER EFFLUENT SOLIDS: 8 PERCENT
CLARIFIER UNDERFLOW SOLIDS: 18 PERCENT
FILTER CAKE SOLIDS: 45 PERCENT
COAL: 3.4 PERCENT SULFUR
ADDITIONAL CONDITIONS AT TIME OF SAMPLING
SHOWN IN TABLE 4-2
Figure 4-2.
Typical flow diagram for EPA/TVA Shawnee
venturi and spray tower scrubber system:
8. 34 MW equivalent operation
35
-------
Table 4-2. TEST CONDITIONS FOR EPA/TVA SHAWNEE
VENTURI AND SPRAY TOWER SCRUBBER
Parameters
Gas rate,
cu m/sec
acfm
Equivalent megawatts
SO- input, ppm
SO? removal, percent
pH control
Dust input,
gm/cu m
gr/scf
Scrubber, effluent
solids, percent
Liquid-to-gas ratio,
gal/1000 cu ft
Sample datea
3/19/74
11.8
25,000
8.3
2,600
93
8.0
8.0
3.5
8
54.0
5/17/74
11.8
25,000
8.3
3,280
78
8.0
8.0
3.5
8
54.0
6/27/74
11.8
25,000
8.3
2,700
84
8.0
8.0
3.5
8
54.0
o
Data base used in extrapolation of scrubber system to 1000 MW
equivalent size scrubber and for determining water balance.
Water balance at time of testing is shown in Figure 4-3, and
values extrapolated to 1000 MW equivalent are shown in
Figure 4-14.
36
-------
Table 4-3. ANALYSES OF SCRUBBER LIQUORS FROM EPA/
TVA SHAWNEE VENTURI AND SPRAY TOWER
SCRUBBER SYSTEM
Scrubber test conditions at time of sampling shown in Table 4.2.
Concentration in milligrams per liter of filtered liquor (Whatman 40).
Scrubber
liquor
constituents
Aluminum (Al)
Antimony (Sb)
Arsenic (As)
Beryllium (Be)
Boron (B)
Cadmium (Cd)
Calcium (Ca)
Chromium (Cr) (total)
Cobalt (Co)
Copper (Cu)
Iron (Fe)
Lead (Pb)
Magnesium (Mg)
Manganese (Mn)
Mercury (Hg)
Molybdenum (Mo)
Nickel (Ni)
Potassium (K)
Selenium (Se)
Silicon (Si)
Silver (Ag)
Sodium (Na)
Tin (Sn)
Vanadium (V)
7.inc (/n)
Total carbonate
Chloride (Cl)
Fluoride (F)
SuUite
Sulfate
Phosphate
Total nitrogen
Chemical oxygen demand
Total dissolved solids
Total alkalinity
Conductance, mho/cm
Turbidity, Jackson
units
PH
In-process data
Potential discharge point data
Sample location
Scrubber effluent
Clarifier underflow
Drum vacuum filter filtrate
Sample date
3/19/74
0.22
0.39
0. 15
0.05
--
0.02
980
0.02
..
0.08 '
0.77
0.03
53
..
0. 10
_ _
0.5
8.4
0.08
0.4
0.09
33
-.
..
0.09
<10
1230
<0.3
450
1000
<0. 1
<0.001
220
3500
54
0.006
<3
5.19
5/16/74
0.12
2. 1
0. 15
0.05
--
0.04
2360
0.02
0.6
0.03
0. 14
0.12
220
0.4
<0.05
0.25
21
1.9
1.8
0.01
108
..
..
0.02
<10
4400
1.5
3.0
1500
<0.1
<0.001
--
8700
-
0.013
<3
5.67
6/27/74
1.54
1.01
0.04
-------
scrubber effluent is not a potential discharge point, it was included to
identify the scrubber chemistry. Analysis of samples taken down-
stream of the reaction tank has shown that they are essentially equiva-
lent in all cases and that they represent the water quality of the liquor
at any potential discharge point.
The water balance for this system is presented in the
simplified flow diagram in Figure 4-3. The actual water flow in gal-
lons per minute is an integrated average for the equivalent power
capacity specified on the three days for the nominal coal sulfur con-
tent of 3.4 percent and for a 78 to 93 percent efficiency in SO2
removal from the flue gas. In each case, flow rates are given as
water flow; slurry flows are greater because of the incorporation of
solids in liquors in varying amounts at different points in the system.
The percentages shown are based on the maximum water flow in the
system, which in this case are represented by the recirculation
liquor from the reaction tank to the scrubber.
4.2.2 EPA/TVA Shawnee Turbulent Contact Absorber
(TCA) Scrubber System
This system is also one of the three parallel scrubber
modules installed on Boiler No. 10 at the Shawnee Steam Plant. This
module has the same design capacity as that of the system described
in Section 4. 2. 1 (Ref. 10). The TCA scrubber (manufactured by
Universal Oil Products) uses a fluidized bed of low density plastic
spheres that are free to move between retaining grids (Figure 4-4).
Three stages of grids are used to obtain sufficient surface for
adequate scrubbing efficiency.
A system flow diagram is shown in Figure 4-5. Flue
gas enters the scrubber from a side port, passes upward through the
grids, contacts the levitated plastic spheres, and exits through
chevron-shaped demisters. Scrubbing liquor from the reaction tank
38
-------
PERCENTAGE VALUES ARE
PERCENT OF MAXIMUM
WATER FLOW INTO
SCRUBBER
WATER FLOW RATE IN
GALLONS PER MINUTE (gpm)
TO CONVERT TO cu m/hr
MULTIPLY BY 0.227
4.2 gpm
0.31%
POND
TEST CONDITIONS
LIQUID-TO-GAS RATIO: 54.0 gal/1000 cu ft
SCRUBBER EFFLUENT SOLIDS: 8 PERCENT
CLARIFIER UNDERFLOW SOLIDS: 18 PERCENT
FILTER CAKE SOLIDS: 45 PERCENT
COAL: 3.4 PERCENT SULFUR
ADDITIONAL CONDITIONS AT TIME OF
SAMPLING SHOWN IN TABLE 4-2
Figure 4-3.
Water balance for EPA/TVA Shawnee
venturi and spray tower scrubber
system: 8. 34 MW equivalent operation
39
-------
FLUE
CAS
OUTLET
CHEVRON DEMISTERS
INLET KOCH TRAY
WASH LIQUOR
KOCH TRAY
STEAM
SPARGE
RETAINING GRIDS
<
FLUE ,.
GAS H>
^^
INLET
EFFLUENT KOCH
TRAY WASH LIQUOR
A A A
o oi,
00 o
<%>-.*
INLET SLURRY
FROM REACTION
TANK
.^MOBILE PLASTIC
* SPHERES
EFFLUENT
SLURRY
REACTION
TANK
Figure 4-4. Schematic of EPA/TVA TCA scrubber
40
-------
REHEATER
STACK
TCA
SCRUBBER
(See Figure
4-4
for details)
TEST CONDITIONS
LIQUID-TO-GAS RATIO: 54.6 gal/1000 cu ft
SCRUBBER EFFLUENT SOLIDS: 8 PERCENT
CLARIFIER UNDERFLOW SOLIDS: 40 PERCENT
COAL: 3.4 PERCENT SULFUR
ADDITIONAL CONDITIONS AT TIME OF SAMPLING
SHOWN IN TABLE 4-4
Figure 4-5. Typical flow diagram for EPA/TVA Shawnee
TCA scrubber system: 6.8 MW equivalent
operation
41
-------
is sprayed against the rising flow of flue gas, percolates downward
through the scrubbing stages, and exits to the reaction tank. The pH
is adjusted with a limestone slurry that is also fed to the reaction
tank. A bleed stream from the reaction tank is passed through a
clarifier. The overflow returns to the reaction tank, and the under-
flow is pumped to a disposal pond. No secondary dewatering is done
in this system.
On two occasions slurry samples were taken from the
scrubber effluent and clarifier underflow. Only the clarifier under-
flow represents the potential discharge points. The test conditions
of the scrubber on these days are presented in Table 4-4, and the
results of the liquor analysis from these sample points is presented
in Table 4-5.
The water balance for this system is presented in the
simplified flow diagram in Figure 4-6. The water flow is given in
gallons per minute, and the percentages are based on maximum flow
to the scrubber. The reported flow rates are for the equivalent power
capacity treated as specified, for the nominal coal sulfur content of
3.4 percent, and for about an 86 percent SO? removal efficiency.
4.2.3 Arizona Public Service Company Cholla Power
Station -- Flooded Disk Scrubber (FDS) and
Packed Absorption Tower Scrubber System
This system consists of two parallel dual-stage scrub-
bers installed on a 115 MW coal-fired boiler at the Cholla Station,
Joseph City, Arizona (Ref. 11). Each scrubber is designed for 60
MW equivalent power generation capacity. The flooded disk and
absorption tower scrubbers (Figure 4-7), (manufactured by Research
Cottrell, Inc.) were designed to handle fly ash and remove SO- from
the flue gas. The FDS is a variable throat, flooded disk, venturi-
type scrubber used principally to remove the fly ash entrained in the
42
-------
Table 4-4. TEST CONDITIONS FOR EPA/TVA
SHAWNEE TCA SCRUBBER
Parameters
Gas rate,
cu m/sec
acfm
Equivalent megawatts
SO- input, ppm
SO_ removal, percent
pH control
Dust input
gm/cu m
gr/scf
Scrubber effluent solids,
percent
Liquid-to-gas ratio,
gal/1000 cu ft
Sample date
11/27/73
9.7
20,500
6.8
2,700
83
5.9
8.0
3.5
16.0
54.6
6/15/74a
9.7
20,500
6.8
2,400
86
5.5
8.0
3.5
8.0
54.6
Data base used in extrapolation of scrubber system to 1000 MW
equivalent size scrubber and for determining water balance.
Represents stable test conditions. Water balance at time of
testing is shown in Figure 4-6, and values extrapolated to
1000 MW equivalent are shown in Figure 4-15.
43
-------
Table 4-5. ANALYSES OF SCRUBBER LIQUORS FROM EPA/
TVA SHAWNEE TCA SCRUBBER SYSTEM
Scrubber test conditions at time of sampling shown in Table 4-4.
Concentration in milligrams per liter of filtered liquor (Whatman 40).
Scrubber
liquor
constituents
Aluminum (Al)
Antimony (Sb)
Arsenic (As)
Beryllium (Be)
Boron (B)
Cadmium (Cd)
Calcium (Ca)
Chromium (Cr) (total)
Cobalt (Co)
Copper (Cu)
Iron (Fe)
Lead (Pb)
Magnesium (Mg)
Manganese (Mn)
Mercury (Hg)
Molybdenum (Mo)
Nickel (Ni)
Potassium (K)
Selenium (Se)
Silicon (Si)
Silver (Ag)
Sodium (Na)
Tin (Sn)
Vanadium (V)
Zinc (Zn)
Total carbonate
Chloride (Cl)
Fluoride (F)
Sulfite
Sulfate
Phosphate
Total nitrogen
Chemical oxygen demand
Total dissolved solids
Total alkalinity
Conductance, mho/cm
Turbidity, Jackson
units
pH
In -process data
Potential discharge point data
Sample location
Scrubber effluent
Clarifier underflow
Sample date
11/27/73
..
--
0.2
0.01
--
0.04
1800
0.04
_-
0.05
0.06
900
--
--
..
0. 16
6.3
0.2
--
--
63
--
--
0.84
3400
2.3
--
2000
<0.1
<0.005
--
12000
--
<3
5.90
6/15/74
2.7
2.0
0.4
0.07
--
0.005
840
0. 16
0. 16
0.02
0.35
0.35
2800
--
<0.5
..
0.44
-.
--
--
0.008
--
--
--
0.03
3300
2.3
1400
9500
<0.1
<0.005
--
17800
--
0.027
<3
4.64
11/27/73
--
0.3
0.004
0.004
..
0.5
--
--
0. 12
600
--
0.05
--
0. 50
5.9
0.2
--
._
--
--
--
0.35
_ _
1900
--
..
--
<0. 1
<0.005
--
11000
150
_
<3
9.50
6/15/74
0.6
1.4
0. 1
0.05
~ ~
0.004
520
0.09
0. 10
0.01
0.02
0.23
2750
--
<0.05
--
0.33
--
--
--
0.005
--
--
--
0.02
2300
6.5
80
10000
<0. 1
<0.005
--
15000
--
0.015
< 3
7.96
44
-------
fT 15 gpm
100%
STACK
28 gpm
2.52
10.1 gpm
.9%
32.6 gpm
2.9%
PERCENTAGE VALUES ARE
PERCENT OF MAXIMUM
WATER FLOW INTO SCRUBBER
WATER FLOW RATE IN
GALLONS PER MINUTE (gpm).
TO CONVERT TO cu m/hr
MULTIPLY BY 0.227
POND
TEST CONDITIONS
LIQUID-TO-GAS RATIO: 54.6 gal/1000 cu ft
SCRUBBER EFFLUENT SOLIDS: 8 PERCENT
CLARIFIER UNDERFLOW SOLIDS: 40 PERCENT
COAL: 3.4 PERCENT SULFUR
ADDITIONAL CONDITIONS AT TIME OF
SAMPLING SHOWN IN TABLE 4-4
Figure 4-6.
Water balance for EPA/TVA Shawnee
TCA scrubber system: 6.8 MW
equivalent operation
45
-------
REHEATER
EXIT GAS
TO STACK
SECONDARY MIST
ELIMINATOR
PRIMARY MIST
ELIMINATOR
FLUE GAS
INLET
FLOODED1
DISK
SCRUBBER
(FDS)
'ABSORBER'
PACKING
PACKED
ABSORPTION
TOWER
CYCLONIC
DEMISTER
DEMISTER
SPRAY WATER
FROM
ABSORPTION
TOWER FEED
TANK
TO ASH
DISPOSAL
POND
FDS
SLURRY TANK
TO ABSORPTION
TOWER FEED
TANK
Figure 4-7.
Schematic of Arizona Cholla Station FDS and
packed absorption tower scrubber
46
-------
flue gas after passing through the upstream mechanical collectors
that remove 80 percent of the fly ash.
A flow diagram is shown in Figure 4-8. A forced draft
booster fan forces flue gas into the FDS and then into the base of the
absorption tower. Recirculating limestone slurry is pumped from the
FDS slurry tank into the FDS scrubber tangentially above the throat
and at the disk. The scrubber slurry is carried with the flue gas
through the cyclonic demister and then returns to the FDS slurry
tank after having removed the fly ash and 15 to 30 percent of the SO-.
The flue gas rises through the packed absorption tower countercurrent
to limestone slurry recirculated from the absorption tower feed tank.
Of the two parallel absorption towers, one is operated as a packed
tower and the other as a spray tower. The slurry returns to the
absorption tower feed tank after scrubbing approximately 40 to 60
percent of the remaining SO- from the flue gas. Adjustment of pH is
made by the addition of a limestone slurry to the absorption tower
feed tank. A crossfeed between the absorption tower feed tank to the
FDS slurry tank provides a bleed for the absorption tower scrubbing
liquors and a supply of absorbent for the FDS liquors. The scrubber
system bleed is made from the FDS slurry tank and is used without
dewatering to slough the mechanically collected fly ash to a disposal
pond.
On two occasions, slurry samples were taken from the
FDS slurry tank and the absorption tower feed tank. Only the output
of the FDS slurry tank can be considered a potential water discharge
point. The operating conditions existing in the scrubber at the time
of sampling are presented in Table 4-6, and the results of the chemi-
cal analyses are presented in Table 4-7.
The water balance for this system is presented in the
simplified flbw diagram in Figure 4-9. The water flow is given in.
47
-------
FLUE GAS FROM
FDS/UNPACKED
ABSORPTION TOWER
INLET
FLUE
GAS
BOOSTER
FAN
REHEATER
STACK
PACKED
ABSORPTION
TOWER/
CYCLONIC
DEMISTER
(See Figure
4-f
for details)
FLOODED
DISK
SCRUBBER
(FDS)
DEMISTER
SPRAY WASH
MAKE-UP
WATER
a
FDS SLURRY
TANK
1
0
TO ASH
DISPOSAL
POND
ABSORPTION
TOWER FEED
TANK
TEST CONDITIONS
LIQUID-TO-GAS RATIO: 70.9 gal/1000 cu ft
FDS SLURRY TANK EFFLUENT SOLIDS: 15 PERCENT
COAL: 0.5 PERCENT SULFUR
ADDITIONAL CONDITIONS AT TIME OF SAMPLING
SHOWN IN TABLE 4-6
Figure 4-8.
Typical flow diagram for Arizona Public Service Cholla
FDS and packed absorption tower scrubber system:
115 MW equivalent operation
48
-------
Table 4-6. OPERATING CONDITIONS FOR ARIZONA
PUBLIC SERVICE CHOLLA FDS AND
ABSORPTION TOWER SCRUBBER SYSTEM
Parameters
Sample datea
4/1/74
11/7/74
Gas rate,
cu m/sec
acfm
Equivalent megawatts
SO- input, ppm
SO- removal, percent
pH control
Dust input
gm/cu m
gr/scf
Absorption tower effluent
solids, percent
Liquid-to-gas ratio,
gal/1000 cu ft
151
320,000
115
360
75
6.5
0.41
0.18
15
70.9
151
320,000
115
360
75
6.5
0.41
0.18
15
70.9
Data base used in extrapolation of scrubber system to 1000 MW
equivalent size scrubber and for determining water balance.
Water balance at time of testing is shown in Figure 4-9, and
values extrapolated to 1000 MW equivalent are shown in Figure
4-16.
49
-------
Table 4-7. ANALYSES OF SCRUBBER LIQUORS FROM
ARIZONA PUBLIC SERVICE CHOLLA FDS
AND ABSORPTION TOWER SCRUBBER
SYSTEM
Scrubber test conditions at time of sampling shown in Table 4-6.
Concentration in milligrams per liter of filtered liquor (Whatman 40).
Scrubber
liquor
constituents
Aluminum (Al)
Antimony (Sb)
Arsenic (As)
Beryllium (Be)
Boron (B)
Cadmium (Cd)
Calcium (Ca)
Chromium (Cr) (total)
Cobalt (Co)
Copper (Cu)
Iron (Fe)
Lead (Pb)
Magnesium (Mg)
Manganese (Mn)
Mercury (Hg)
Molybdenum (Mo)
Nickel (Ni)
Potassium (K)
Selenium (Se)
Silicon (Si)
Silver (Ag)
Sodium (Na)
Tin (Sn)
Vanadium (V)
Zinc (Zn)
Total carbonate
Chloride (Cl)
Fluoride (F)
Sulfite
Sulfate
Phosphate
Total nitrogen
Chemical oxygen demand
Total dissolved solids
Total alkalinity
Conductance, mho/cm
Turbidity, Jackson units
pH
In-process data
Potential discharge data point
Sample location
Absorption tower tank
FDS tank
Sample date
4/1/74
0.06
0.03
<0.004
0.08
--
0.007
580
0.02
0.05
0.03
0. 17
0.02
T
0.30
0.007
_.
1.0
--
1.0
1.7
0.01
800
--
--
o.oz
<1
6ZO
2.4
1.0
ZZOO
<0.1
<0.005
105
4300
52
0.0053
<5
6.59
11/7/74
2.1
0.16
0.02
<0.003
3.8
0.012
390
0.004
<0.01
0.01
0. 13
0. 15
9
0.48
<0.5
0.09
0.06
7.5
<0.033
--
<0.007
370
..
0.07
0.04
<1
760
1.0
21
1360
<0.1
<0.005
90
3300
130
0.00299
<5
6.80
4/1/74
2
0.09
<0.004
0. 14
--
0. Oil
680
0. 14
0.17
0.20
0.42
0.01
3
0.34
0.07
1.5
16
2.2
3.3
0.03
2250
..
_-
0. 11
<1
1700
0.7
0.9
4000
0.41
0.002
340
8700
--
0.0112
<5
3.04
11/7/74
0.22
--
0.04
8.0
0.044
770
0.024
0.1
0.16
8.1
0.37
4
2.5
<0.05
0.91
0.30
28
<0.001
--
0.05
1650
.-
0.67
0.47
<1
4200
1.5
3500
3750
<0.5
<0.005
390
14000
--
0.014
<5 '
3.38
50
-------
49 gpm
STACK
MAKE-UP
WATER
19 gpm
0. 08%
10 gpm
O.S4%
t
i
32 gi
0. P
4500 gpm
19.8%
FLOODE
DISK
SCRUBB
(FDS)
AND PA
ABSORP
TOWER
pm
\%
} i
FDS S
:D
ER
CKED
TION
4493 <
19. 6
r
LURRY
NK
4 18'c^ฐno9pm 4 32 gpm
OU.U/fe I Q \Aฎ/
18,087 gpm
79.9%
xi_ FLUE
^~ GAS
LIMESTOM
ปP,m SLURRY
155 gpm
0.24%
MAKE-UP
WATER
E 0. 05%
"1
1
r
ABSORPTION
_ TOWFR ^fc-
TANK
TO DISPOSAL
AREA
TEST CONDITIONS
LIOUID-TO-GAS RATIO: 70.9 gal/1000 cu ft
FDS SLURRY TANK EFFLUENT SOLIDS:
15 PERCENT
COAL: 0.5 PERCENT SULFUR
ADDITIONAL CONDITIONS AT TIME OF
SAMPLING SHOWN IN TABLE 4-6
PERCENTAGE VALUES ARE PERCENT OF
MAXIMUM WATER FLOW INTO SCRUBBER
WATER FLOW RATE IN GALLONS
PER MINUTE (gpm). TO CONVERT
MULTIF
TO cu m/hr
IPLY BY 0.227
Figure 4-9. Water balance for Arizona Public Service
Cholla FDS and packed absorption tower
scrubber system: 115 MW equivalent
operation
51
-------
gallons per minute, and the percentages are based on the maximum
combined water flow to the two stages of scrubbing.
4.2.4 Duquesne Light Company Phillips Station -- Single -
and Dual-Stage Venturi Scrubber System
This system, which is installed on the coal-fired
boilers at the Phillips Station, South Heights, Pennsylvania, is a full-
scale system designed to treat 500 MW of power generating capacity
in four parallel modules (Ref. 12). Since each module is designed to
handle approximately 125 MW, the full plant capacity of 387 MW is
expected to be handled with three modules; one would be a spare.
Three modules (Figure 4-10) are single-stage venturi scrubbers
designed primarily for particulate removal; the fourth (Figure 4-11)
consists of a dual-stage venturi scrubber designed to remove fly ash
in the first stage and SO2 in the second stage. All four modules were
manufactured by Chemico. An SO- removal efficiency of about 60
percent is obtained in the single-stage venturi scrubbers when a lime
slurry is used as the absorbent; the efficiency is about 90 percent for
the dual-stage unit. A schematic of the design of the single-stage
venturi scrubber is shown in Figure 4-12.
Flue gases pass through mechanical and electrostatic
dust collectors located in series with each boiler, then the gases are
diverted into a common manifold with lines leading to each of the three
single-stage scrubbers and the first stage of the dual-stage scrubber.
The hot flue gas enters the first stage of the dual-stage scrubber or
the single-stage scrubber and impinges upon the upper cone where
half of the scrubber liquor is introduced (Figure 4-12). The other
half of the scrubber liquor enters through the tangential nozzles at a
point above the adjustable throat dampers. The flue gas and scrubbing
liquor make contact in the throat section of the scrubber where the
particulates and some SO- are removed. The gas and liquor continue
52
-------
HOT FLUE GASES
FROM DUST
COLLECTORS
SCRUBBER
LIQUOR
RECYCLE
STREAM
DEMISTER/
SPRAY r~
MAKE-UP
PUMP
I
SINGLE-STAGE
SCRUBBER
(See Figure 4-12
for details)
-WET OAS
STACK
RECYCLE
PUMP
LIME SLURRY
FEED
SUPERNATE
INTERIM
DISPOSAL
POND
Figure 4-10.
Typical flow diagram for Duquesne Phillips
single-stage scrubber system
53
-------
HOT FLUE GASES
FROM OUST
COLLECTORS
COLD FLUE GAS
SCRUBBER
LIQUOR
RECYCLE
STREAM
Ul
DAMPER
PURGE AIR
BLOWER
SCRUBBER
LIQUOR
RECYCLE
STREAM
SPRAY
WATER
FIRST STAGE
SCRUBBER
(See Figure 4-12
for details)
SECOND
STAGE
SCRUBBER
DEMISTER
SPRAY
DEMISTER
SPRAY
MAKE-UP
PUMP
OVERFLOW
FROM
CLARIFIER
RECYCLE
PUMP
SECOND STAGE
BLEED
LIME SLURRY
FEED
RECYCLE
PUMP
SERVICE
WATER
LIME
STORAGE
SILO
LIME
SLURRY
TANK
VIBRATOR
WEIGH FEEDER
INTERIM
DISPOSAL
POND
LIME
SLAKER
SLAKER
TRANSFER
TANK
LIME SLURRY
PUMP
Figure 4-11. Typical flow diagram for Duquesne Phillips Station
dual-stage scrubber system
-------
HOT FLUE
GAS INLET
SCRUBBER
LIQUOR
RECYCLE
STREAM
TANGENTIAL
j- NOZZLES
--- -ป
ADJUSTABLE THROAT
DAMPERS
FLUE GAS
EXIT
DEMISTER
SPRAY
LOWER CONE
TO RECYCLE
PUMP
Figure 4-12. Schematic of Duquesne Phillips Station single-
stage scrubber
55
-------
downward to the separator section where the flue gas enters a
demister, leaves the scrubber through a wet induced draft (ID) fan,
and enters the second stage of the dual-stage train, or through mist
eliminators in the three single-stage scrubbing trains.
Scrubbing liquor leaves the scrubber to be either re-
circulated or pumped to the clarifier. The overflow from the clarifier
is recycled to the scrubber, and the underflow is sent to one of three
interim disposal ponds. Lime slurried with fresh water is added in
each scrubber and in the clarifier. The pump and fan spray waters
also provide make-up water.
On two occasions slurry samples were taken from the
clarifier and scrubber and once from the pond. The conditions that
existed in the scrubber at the time of sampling are given in Table 4-8.
The results of the liquor analysis from these sample points are pre-
sented in Table 4-9. Clarifier underflow and pond supernate repre-
sent potential discharge points in this system.
The water balance for this system is presented in the
simplified flow diagram in Figure 4-13 representing water flow under
operating conditions corrected for water imbalance caused by variances
from system design. The water flow is given in gallons per minute,
and the percentages are based on the maximum water flow in the
system represented by the water flow to the scrubber.
4.3 EXTRAPOLATION OF PROCESS FLOWS
4.3.1 Scale-Up of Four Systems Studied
The actual flows of process liquors as presented in
Figures 4-3, 4-6, 4-9, and 4-13 were used to calculate the process
flow expected for 1000 MW facilities operating at 90 percent SO-
removal efficiency, using the identical scrubber design, operating
conditions, and coal as those on which the data base was established.
In all cases, scale-up of flow rates was presumed to be proportional
56
-------
Table 4-8. OPERATING CONDITIONS FOR DUQUESNE
PHILLIPS STATION: SINGLE- AND DUAL-
STAGE VENTURI SCRUBBER SYSTEMS
Parameters
Sample date'
10/4/73b
6/17/74c
Gas rate,
cu m/sec
acfm
Equivalent megawatts
SO_ input, ppm
SO_ removal, percent
pH control
Dust input
gm/cu m
gr/scฃ
Venturi effluent
solids, percent
Liquid-to-gas ratio,
gal/1000 cu ft
274
580,000
145
1,400
60
6.0
0.57
0.25
29.9
236
500,000
120
1,400
70
6.0
0.57
0.25
29.9
Data base used in extrapolation of scrubber system to 1000 MW
equivalent size scrubber and for determining water balance.
Water balance at time of testing is shown in Figure 4-13, and
values extrapolated to 1000 MW equivalent are shown in Figure
4-17.
Two single-stage scrubbers.
^ ^^
'Two single-stage and one dual-stage scrubbers.
57
-------
Table 4-9. ANALYSES OF SCRUBBER LIQUORS FROM
DUQUESNE PHILLIPS STATION: SINGLE-
AND DUAL-STAGE VENTURI SCRUBBER
SYSTEMS
Scrubber test conditions at time of sampling shown in Table 4-8.
Concentration in milligrams per liter of filtered liquor (Whatman 40).
Scrubber
liquor
constituents
Aluminum (Al)
Antimony (Sb)
Arsenic (As)
Beryllium (Be)
Boron (B)
Cadmium (Cd)
Calcium (Ca)
Chromium (Cr) (total)
Cobalt (Co)
Copper (Cu)
Iron (Fe)
Lead (Pb)
Magnesium (Mg)
Manganese (Mn)
Mercury (Hg)
Molybdenum (Mo)
Nickel (Ni)
Potassium (K)
Selenium (Se)
Silicon (Si)
Silver (Ag)
Sodium (Na)
Tin (Sn)
Vanadium (V)
Zinc (ฃn)
Total carbonate
Chloride (Cl)
Fluoride (F)
Sulfite
Sulfate
Phosphate
Total nitrogen
Chemical oxygen demand
Total dissolved solids
Total alkalinity
Conductance, mho/cm
Turbidity, Jackson units
pH
In-process data
Potential discharge point data
Sample location
Scrubber effluent
Clarifier underflow
Sample date
10/4/73
..
--
0.085
0. 012
--
0.022
1300
0.037
0.06
_ _
0.08
220
--
0.09
..
--
20
0.8
--
0.02
1680
.-
.-
0.1Z
<1
1800
4.8
<1
4500
<0.05
<0.005
..
9400
61
0.0063
<3
9.20
6/17/74
--
0.06
0.002
--
660
--
_ _
0.5
--
--
0.0004
..
--
10
0.33
--
_ _
440
._
.-
--
<1
540
8
1.7
Z700
<0.05
<0. 005
65
4600
78
0.0033
<3
8.92
10/4/73
..
--
--
0.012
--
0.023
1400
0.040
0.07
0.026
0. 18
410
--
0.05
..
--
22
0.8
--
- -
2400
0.09
<1
2700
2.6
<1
6450
<0.05
<0.005
14000
--
0.01
<3
7.11
6/17/74
..
--
<0.004
0.003
--
_ _
600
--
..
0.4
--
-.
<0.002
..
26
0.028
--
. .
320
--
<1
470
10
20
720
<0.05
<0.005
60
4200
--
0. 0034
<3
10.70
Pnnrf
* UI1U
supernate
6/17/74
..
--
<0.004
0.002
--
_ _.
600
--
..
0.4
_-
._
0.0004
_ _
22
0.095
--
..
344
_ _
--
<1
420
7
4.8
1000
<0.03
<0.005
4000
41
0.0030
<3
10.44
58
-------
POND
TEST CONDITIONS
LIQUID-TO-GAS RATIO: 29.9 gal/1000 cu ft
SCRUBBER EFFLUENT SOLIDS: 5 PERCENT
CLARIFIER UNDERFLOW SOLIDS: 38 PERCENT
COAL: 2 PERCENT SULFUR
ADDITIONAL CONDITIONS AT TIME OF
SAMPLING SHOWN IN TABLE 4-8
PERCENTAGE VALUES ARE PERCENT OF
MAXIMUM WATER FLOW INTO SCRUBBER
WATER FLOW RATE IN GALLONS
PER MINUTE (gpm). TO CONVERT
TO cu m/hr MULTIP
I PLY BY 0.227
Figure 4-13. Water balance for Duquesne Phillips Station single'-
and dual-stage venturi scrubber systems: 120 MW
equivalent operation
59
-------
to the flue gas flow rates; therefore, the power generated would be a
direct scale-up when identical coal was used.
The scale-up of the scrubbing system based on the two
EPA/TVA Shawnee Steam Plant prototype facilities represents the
greatest size scale-up of the four facilities examined. However, it
was assumed that individual scrubber modules of several hundred
megawatt capacity equivalence would be constructed for a large
power generating plant of this design rather than a single 1000 MW
unit. Thus, the scale-up factor would probably be within the range of
10 to 30 times rather than a factor of two orders of magnitude.
Accordingly, it was assumed that the relative flow rates would be
unaffected by the number of parallel units required to handle the full
plant capacity. The scale-up of a scrubber system based on an EPA/
TVA Shawnee venturi and spray tower or TCA type scrubber system
was determined on a linear proportional basis whereby the scrubber
liquid-to-gas ratio was held unchanged from those of the sampling
conditions. The results of these calculations are presented in the
simplified schematic flow diagrams in Figures 4-14 and 4-15.
For the Arizona Public Service Cholla Station scrubbing
system, full-scale scrubbers exist; thus, the scale-up to 1000 MW
equivalent power generating capacity was done on the presumption that
additional like-units would be added in parallel. During sampling
periods, the SG>2 removal efficiency of this system was 75 percent
although a scrubber module with a packed absorption tower removes
90 percent of the SO? (the absorption tower performing as a spray
tower is 60 percent efficient in SO2 removal). In performing the
scale-up of this design to 90 percent removal efficiency, it was
assumed that all scrubbers would operate at the flow rates required
for a packed absorption tower. The results of these calculations are
presented in the flow diagram in Figure 4-16.
60
-------
VENTURI
AND
SPRAY
TOWER
SCRUBBER
PERCENTAGE VALUES ARE
PERCENT OF MAXIMUM
WATER FLOW INTO SCRUBBER
WATER FLOW RATE IN
GALLONS PER MINUTE (gpm).
TO CONVERT TO cu m/hr
MULTIPLY BY 0.227
FILTER
1390 gpm
0.86%
>m
POND
EXTRAPOLATED OPERATING CONDITIONS
LIQUID-TO-GAS RATIO: 54.0 gal/1000 cu ft
SCRUBBER EFFLUENT SOLIDS: 8 PERCENT
CLARIFIER UNDERFLOW SOLIDS: 18 PERCENT
FILTER CAKE SOLIDS: 45 PERCENT
COAL: 3.4 PERCENT SULFUR
S02 REMOVAL EFFICIENCY: 90 PERCENT
TEST CONDITIONS USED AS BASIS FOR
EXTRAPOLATION SHOWN IN TABLE 4-2
Figure 4-14.
Water balance for venturi and spray tower
scrubber system (extrapolation based on
EPA/TVA Shawnee type of process): 1000
MW equivalent operation
61
-------
164,000 gpm
100%
TCA
SCRUBBER
163,150 gpm
99.5%
MAKE-UP
WATER
LIMESTONE
SLURRY
REACTION
TANK
4800 gpm
2.9%
PERCENTAGE VALUES ARE
PERCENT OF MAXIMUM
WATER FLOW INTO SCRUBBER
WATER FLOW RATE IN
GALLONS PER MINUTE (gpm).
TO CONVERT TO cu m/hr
MULTIPLY BY 0.227
1490 gpm
0.9T
EXTRAPOLATED OPERATING CONDITIONS
LIQUID-TO-GAS RATIO: 54.6 go I/1000 cu ft
SCRUBBER EFFLUENT SOLIDS: 8 PERCENT
CLARIFIER UNDERFLOW SOLIDS: 40 PERCENT
COAL: 3.4 PERCENT SULFUR
S02 REMOVAL EFFICIENCY: 90 PERCENT
TEST CONDITIONS USED AS BASIS FOR
EXTRAPOLATION SHOWN IN TABLE 4-4
Figure 4-15.
Water balance for TCA scrubber system
(extrapolation based on EPA/TVA
Shawnee type of process): 1000 MW
equivalent operation
62
-------
430 gpm
0.22%
190 gpm
0.
MAKE-UP
WATER
FLOODED
DISK
SCRUBBER
(FDS)
AND PACKED
ABSORPTION
TOWER
39,150 gpm
19,9%
295 gpm
0.1?%
157,500 gpm
80.0%
STACK
MAKE-UP
157,410 gpm I WATER
79.9r
290
0.
FLUE
GAS
39,000 gpm
19.8%
LIMESTONE
SLURRY
FDS SLURRY
TANK
310 gpm
0,16%
0 gpm
0,06%
ABSORPTION
TOWER
TANK
520 gpm
0.26%
TO DISPOSAL
AREA
EXTRAPOLATED OPERATING CONDITIONS
LIOUID-TO-GAS RATIO: 70.9 gal/1000 cu ft
COAL: 0.5 PERCENT SULFUR
S02 REMOVAL EFFICIENCY: 90 PERCENT
FDS SLURRY TANK EFFLUENT: 15 PERCENT
OPERATING CONDITIONS USED AS BASIS FOR
EXTRAPOLATION SHOWN IN TABLE 4-6
PERCENTAGE VALUES ARE PERCENT
OF MAXIMUM WATER FLOW INTO SCRUBBER
WATER FLOW RATE IN GALLONS
PER MINUTE (gpm). TO CONVERT
TO cu m/hr MULTIPLY BY 0.227
Figure 4-16. Water balance for FDS and absorption tower
scrubber system (extrapolation based on
Arizona Public Service Cholla Station
process): 1000 MW equivalent operation
63
-------
The Duquesne Phillips Station scrubbing system is
designed for about 400 MW operating capacity; thus, scaling up this
system to 1000 MW equivalent would require only about a 2. 5 increase
in scrubbing capacity. However, at the time of sampling, flue gases
from only 120 MW generations capacity boilers were being treated,
but the full scrubber design capacity of three 125 MW equivalent
scrubbers was being used. Also at that time, the flue gas velocity
through the collectors was such that fly ash carry-over was four
times that normally experienced. This higher solids content resulted
in an increase in particulate loading which saturated the system's
design solids capacity. To scale up this sytem to 1000 MW at 90
percent SCX removal efficiency and to compensate for the higher
solids loading, it was presumed that dual-stage venturi scrubbers
would be required and deficiencies would be corrected. As in the
other cases, the scrubber liquid-to-gas ratio was assumed to be
unchanged from those of the test conditions. The results of these
calculations are presented in the simplified flow diagram in Figure
4-17.
4.3.2 Summary
Table 4-10 summarizes and compares the water re-
quirements for the four scrubber systems studied and extrapolated
in size to an equivalent capacity of 1000 MW. Several observations
can be made from these comparisons relative to water use in SO,
L*
scrubbing systems. The maximum water flow into the scrubber
varies among the different types of scrubber systems and appears to
be unrelated to the sulfur content of the coal, fly ash content, or the
percentage of the scrubber's design capacity being used. Since the
scrubber liquor requirement appears to be solely a function of
scrubber system design, no extrapolation of water use from a given
system to other systems or other sulfur content coals is possible
64
-------
790 gpm
0.6%
121,400 gpm
95.6%
950
0.
LIME
SLURRY
STACK
DUAL -STAGE
VENTURI
SCRUBBER
200 gpm
5400 gpm
4.3%
5560 gpm
4.4%
1
CLARIFIER
130 gpm
O.f%
POND
EXTRAPOLATED OPERATING CONDITIONS
LIQUID-TO-GAS RATIO: 29.9 gal/1000 cu ft
SCRUBBER EFFLUENT SOLIDS: 5 PERCENT
CLARIFIER UNDERFLOW SOLIDS: 38 PERCENT
COAL: 2 PERCENT SULFUR
S02 REMOVAL EFFICIENCY: 90 PERCENT
OPERATING CONDITIONS USED AS BASIS FOR
EXTRAPOLATION SHOWN IN TABLE 4-8
PERCENTAGE VALUES ARE PERCENT OF
MAXIMUM WATER FLOW INTO SCRUBBER
WATER FLOW RATE IN GALLONS
PER MINUTE (gpm). TO COVERT
TO cu m/hr MULTIPLY BY 0.227
Figure 4-17,
Water balance for dual-stage venturi
scrubber system (extrapolation
based on Duquesne Phillips Station
type of process): 1000 MW equivalent
operation
65
-------
Table 4-10. SUMMARY AND COMPARISON OF WATER BALANCE PROJECTIONS
FOR 1000 MW EQUIVALENT SCRUBBERS
Values based on 90 percent SO-, removal
Scrubber location,
type, and
absorbent
Water balance
shown in
Extrapolation
base
Sulfur in coal,
percent
Flow points
Scrubber
Input
Output
Bleed
Clarifier
Overflow
Underflow
Filter
Filtrate
Filter cake
Water losses
Stack
Pond
Water input
With absorbent
Make-up
TVA Shawnee,
venturi and
spray, lime
Figure 4-14
8.3 MW equiv. ,
83% loading,
8 5% SOฃ removal
3.4
Water
flow,3
gpm
163,000
162, 100
4,800
2,900
1,900
1,390
510
900
510
500
910
Percent
of total
flow
100
99.5 *
Z.95
1.78
1.17
0.86
0.31
0. 55
0.31
0.30
0.56
TVA Shawnee,
TCA, limestone
Figure 4-15
6.8 MW equiv.,
68% loading,
86% SO2 removal
3.4
Water
flow,a
gpm
164,000
163, 150
4,120
3,310
810
Percent
of total
flow
100
99.5
2.5
2.0
0.5
Not
Applicable
850
810
170
1,490
0.5
0.5
0. 1
0.9
Arizona Cholla,
FDS and absorption
tower, limestone
and fly ash
Figure 4-16
115 MW equiv.,
100% loading,
75% SO2 removal
0.5
Water
flow,a
gpm
196,840
196,410
520
Percent
of total
flow
100
99.8
0.26
Not
Applicable
Not
Applicable
430
520
110
840
0.22
0.26
0.06
0.43
Duquesne Phillips,
dual-stage venturi,
lime
Figure 4-17
120 MW equiv.,
33% loading,
65% SO2 removal
2.0
Water
flow,a
gpm
127,750
126. 960
5,560
5,400
490
Percent
of total
flow
100
99.4
4.4
4.3
0.4
Not
Applicable
790
360
1, 150
--
O.fe
0.3
0.9
--
aTo convert gpm to cu m/hr multiply by 0.227.
Values of Arizona Cholla are for FDS slurry tank rather than scrubber.
-------
from these data. Moreover, based on the operating procedures at the
time of sampling, it was presumed that the liquid-to-gas ratios within
a scrubber would be maintained constant so as to obtain the desired
sulfur removal efficiency. Thus, an extrapolation to power plant load
factors other than those observed during sampling would be expected
to be a reasonably linear function. No other variables or conditions
appear to affect the scrubber water requirements.
Table 4-11 summarizes the flows of the liquor bleed
streams. In the bleed stream from the scrubber, the larger streams
come from the higher sulfur bearing coals; however, a direct
correlation is impossible because of the fly ash content variation in
the solids and the total solids content in the slurry. Comparison of
water loss to the pond is also impossible because of the different
sludge disposal designs among the plants.
An analysis was made to establish a scaling factor that
could be used to correlate the ratio of water loss to the pond (based on
a 1000 MW equivalent scrubber capacity) as a function of the sulfur
content of the coal being burned. However, to establish this factor
two normalizing assumptions were made:
a. The sludge from the scrubber would be dewatered
to a solids content of 45 percent solids by weight.
b. All available supernate from the dewatering pro-
cess would be returned to the scrubber system.
On the basis of these two assumptions, it was found that for the four
scrubber systems studied the ratio of F/S is between 150 to Z24 with
an average of 187, where F is the water loss to the disposal site in
gallons per minute (to convert to cu m/hr multiply by 0. 227) and S is
the percent sulfur in the coal.
In a given system, a reduction in sulfur content with
all other system parameters held constant will have the effect of .
reducing the quantity of solid waste and the waste water associated
67
-------
Table 4-11. SUMMARY OF SCRUBBER BLEED STREAM
FLOW PROJECTIONS FOR 1000 MW
EQUIVALENT SCRUBBERS
Values based on 90 percent SO2 removal
Scrubber location.
type, and
absorbent
Water balance
shown in
Coal type and sulfur
content, percent
Scrubber capacity,
MW
Extrapolation base,
MW
Sulfite and sulfate
in total solids,
percent
Scrubber
Bleed stream,
gpmc
Solids, percent
Clarifier
Underflow, gpm
Solids, percent
Filter
Water loss in
cake, gpm
Solids in cake,
percent
Filtrate, gpm
Solids, percent
Pond
Supernate,
gpm
Solids, percent
TVA Shawnee,
venturi and spray
tower, lime
Figure 4-14
Eastern 3.4
10
8.3
50
4800
8
1900
18d
510
4 5
1390
45
Not
Applicable
TVA Shawnee,
TCA, limestone
Figure 4-15
Eastern 3.4
10
6.8
50
4120
8
810
40
Not
Applicable
Not
Applicable
Arizona Cholla,
FDS and absorption
tower, limestone
and fly ash
Figure 4-16
Western 0.5
120
115
40*
520
15
Not
Applicable
Not
Applicable
Not
Applicable
Duquesne Phillips,
dual- stage venturi,
lime
Figure 4-17
Eastern 2. 0
Four units of
125 each
120
40a
5560
5
490
38
Not
Applicable
130
<1
Adjusted to 90 percent SO, removal efficiency.
Values for Arizona Cholla are for FDS slurry tank rather than scrubber.
Water flow rate in gallons per minute; to convert gpm to cu m/hr multiply by 0.227.
Clarifier was undersized resulting in shorter residence time and lower solids content.
68
-------
with the solids. The reduced loss of waste water to the disposal site
will reduce the volume of make-up water required to balance the
system. However, if this reduction is sufficiently large so as to be
inadequate to satisfy minimum fresh water requirements, a system
imbalance may occur. In this case, when minimum fresh water
requirements are met, make-up water will then exceed water losses
by evaporation or to the pond and excess system water will result.
Two potential environmentally acceptable solutions exist: (a)
increase the relative quantity of waste water being disposed with
solids, or (b) treat a portion of the return water from the dewatering
treatment circuit to satisfy the fresh water requirement. In many
cases, the need for fresh water can be reduced by blending untreated
return water with fresh water, but there is a limit to the extent that
blended water can satisfy the demister or pump seal fresh water
requirement. This limit is a system design parameter for each
plant and must be found by experience.
The make-up water requirement for each system is depend-
ent primarily upon the water balance from pond losses and stack
losses. However, in watertight systems, the water balance require-
ment for make-up water may be exceeded by the fresh water require-
ments of the demister and pump seals. In these cases, the limit to
the degree of dewatering may be dictated by this fresh water require-
ment, and there may exist only a slight accommodation for reducing
the sulfur content without major system design changes. Since stack
losses are dependent upon atmospheric conditions and design
variables, adjustments in this value are not easily made; therefore,
it is not an adjustable parameter.
In addition to a change in sulfur content as it affects
the water balance, an .identical effect is experienced with a decrease
in flue gas mass loading to the scrubber. In a case of reduced load
factor, as before, a reduction in sulfur waste products causes a
69
-------
reduction in waste water and in make-up water to meet water balance
requirements. As before, if this reduction exceeds the fresh water
requirements of the system, an identical water balance problem
arises.
4.4 CHEMICAL CONDITIONS AFFECTING COMPOSITION
OF SCRUBBER LIQUORS AND WATER IMBALANCE
In each case of water imbalance previously identified,
excess system water occurs only when fresh water requirements
exceed make-up water requirements and can take place whenever a
critical reduction in water losses occurs. Other system variations
beside those identified here are possible, but the point at which
excess system water is produced can not be determined without
detailed study of the system design or by experimental determination.
It could not be determined from these data whether sulfur removal
efficiency or other specific design parameters could produce inde-
pendently the conditions necessary for excess system water.
The following sections (4.4. 1 and 4.4.2) provide the
results of an overview of fundamental scrubber chemistry and the
potential for water imbalance.
4.4.1 Saturation Concentrations and Solubility of
Ingredients
An evaluation of the scrubber liquor chemical analysis
(shown in Tables 4-3, 4-5, 4-7, and 4-9) indicates that the scrubber
systems operate at saturation concentrations of calcium and sulfate
ions amounting to several thousand milligrams per liter. In addition,
sodium and chloride ions were found to be present in high concen-
trations, but they are not controlled by saturation conditions.
The analytical data for the liquor samples are con-
sistent with the assumption of saturation with gypsum (CaSO^ 2H_O).
70
-------
However, it should not be inferred that the concentrations of cal-
cium or sulfate will not vary among the liquor samples, nor will
the solubility of calcium sulfate (as expressed by the product of the
sulfate and calcium concentrations) remain constant. In addition to the
effect of temperature, the solubility of calcium sulfate is affected by
the presence of other ions (e.g., sodium, magnesium, or chloride).
Although the quantitative relationship is somewhat complicated, in
general the solubility of calcium sulfate will increase by the summation
of all molar ionic concentrations after each has been multiplied by the
square of its ionic charge. Therefore, magnesium ions have a four-
fold greater effect than equimolar concentrations of either sodium or
chloride ions. An approximate correlation may be observed between
the solubility of calcium sulfate and the total dissolved solids (TDS) of
the sample. For example, each of the three data sets for the EPA/TVA
Shawnee-venturi system shows a fivefold increase in the product of the
calcium and sulfate concentrations which corresponds to a threefold
increase in TDS. Similar relationships exist in each of the other
scrubbing systems.
The chemistry of other chemical constituents including
the trace elements in the system are controlled by their saturation
concentrations in the presence of calcium, sulfate, and hydroxyl ions.
Trace elements, including toxic species, tend to be limited by their
equilibrium concentration with the calcium and sulfate ions at low values
of pH and with the hydroxyl ions at high values of pH. Most of these ele-
ments are controlled in the scrubber liquor at concentration levels be-
low 1 mg/t , and many are controlled at about 0.05 mg/1 .
Control of highly soluble species is dependent upon
maintaining steady-state conditions whereby the system's rate of loss
(as soluble species in occluded water associated with solid waste)
balances the rate of intake (either by absorption from flue gas or f-rom
71
-------
process ingredients). The tighter the closed loop, the higher the
specie concentrations become. The most direct effect of these
species is their contribution to the TDS. In each case, sodium,
calcium, sulfate, magnesium, and chloride ions constitute the mea-
surable dissolved solids content in the liquors. High TDS values
increase the ionic strength of the liquor which in turn increases the
saturation concentrations of chemically controlled soluble species in
the system.
The concentration of soluble species in the scrubber
liquors is dependent upon two system variables: the pH of the system
liquors, and the tightness of the water balance within the system. The
latter will affect the TDS, and the calcium sulfate chemistry control
will be affected by both.
In addition to these system variables, a reduction in
the coal's sulfur content will increase the soluble species in the scrub-
ber. In this case, the rate of production of sulfur by-products is re-
duced. As a consequence, the scrubbing of other salts and acids from
the flue gas is proportionately greater and will result in higher soluble
salt contents. The resulting increase in ionic strength will also in-
crease the soluble saturation concentration of calcium sulfate. In
some systems, the increase of soluble calcium sulfate will increase
the demister scaling problem. In all cases, an independent reduction
in sulfur content will increase the level of dissolved solids in the liquor
relative to higher sulfur operation.
4.4.2 Water Imbalance
A reduction in boiler load factor decreases the flue gas
mass flow but does not affect the water quality of the liquor. In this
case, absorption of all species from the flue gas (or from fly ash
carried in the flue gas) is reduced in direct proportion to the reduc-
tion in flue gas mass flow. The rate of production of sulfur by-products
72
-------
is reduced proportionately to all other system rates; thus, no change is
effected in the concentration of soluble species in the liquor. In both
the cases of reduced load factor and reduced sulfur content, it has been
assumed that the reduced need for make-up water to satisfy water
balance remains sufficient to satisfy the minimum requirement for
fresh water make-up.
The chemical analysis of system liquors reveals that
the chemical constituents within these liquors exist at essentially
identical concentrations at all potential system purge points. Minor
variations exist as a consequence of the design choice of make-up
water introduction or absorbent introduction into the system, but these
variations do not affect the general water quality of the liquor. The
exception to this generalization exists between scrubber effluent and
points past the introduction of the absorbent. The scrubber effluent
has a slightly lower pH than the system downstream of the reaction
tank and concentrations of species controlled by the hydroxyl ion tend
to be somewhat higher. Since scrubber effluent is not a potential
point of system purge, it need not be considered for water quality
acceptance.
4.5 ASSESSMENT
Table 2-1 shows the range of concentration of the con-
stituents found in the scrubber liquors at the potential discharge
points, the criteria assumed for the quality of nonpotable service
water for usage in the power plant, and a tabulation of the October
1973 EPA proposed criteria for public water supply intake. It pre-
sents the total range of values of the ingredients found in the liquors
at the potential purge points among the four systems studied. This
range of constituent values depends upon the composition of the coal
and absorbents used, the scrubber design, and system operating
characteristics. The values in the table are given in milligrams
73
-------
per liter; however, the potential discharge rate in pounds per day can
be calculated from the following equation:
DR = 0.012 cf
where
DR = discharge rate of the ingredient in Ib/day (to
convert Ib/day to kg/day multiply by 0.4536)
c = concentration of the constituent in mg/i
/ = flow rate of the water in the scrubber system
flow streams in gpm
For example, a given constituent at 0.004 mg/i at a
flow rate of 400 gpm would mean about 0.009 kg/day (0.019 Ib/day)
of material potentially would be discharged.
An assessment was made of the potential water uses and
the constituents of concern for each use; the results are summarized in
Table 4-12. There is a need for large quantities of nonpotable water
for general operations within the power plant (listed under "Service
water" in Table 4-12) and for other applications that require water free
of constituents that would corrode, scale, or foul the equipment. Al-
though the concentration of many constituents does not present a problem
for scrubber liquors to satisfy these needs, it is necessary to reduce
the concentration levels of the chloride, sulfate, magnesium, and
calcium present from one or more of the liquor streams studied.
A treatment process such as lime-soda softening will
reduce the sulfate, magnesium, and calcium concentrations to pre-
clude scaling of the equipment. In addition, the concentrations of
arsenic (As), boron (B), cadmium (Cd), iron (Fe), lead (Pb),
selenium (Se), mercury (Hg), copper (Cu), zinc (Zn), nickel (Ni),
74
-------
Table 4-12. POTENTIAL WATER USES AND
CONSTITUENTS OF CONCERN
Water use
Constituents of concern
Service water
Scrubber make-up
Air conditioning
equipment
Pump seals
Cooling tower make-up
Evaporative coolers
Housekeeping
Discharge
Calcium, magnesium, and sulfate
Calcium, chloride, magnesium,
sulfate, and total dissolved
solids
Arsenic, boron, cadmium, chlo-
ride, chromium (total), fluoride,
iron, lead, manganese, mercury,
selenium, silver, and sulfate
75
-------
cobalt (Co), and manganese (Mn) will also be reduced by this process.
In existing or new plants where brackish water is acceptable for
service water use this process would be adequate. In cases where
brackish water is unacceptable, the lime-soda softening process
would have to be followed by an additional process such as reverse
osmosis. Furthermore, a high chloride content water requires that
the equipment in contact with this water be constructed from, or
coated with, salt water resistant materials. Although it is within the
current state of the art to build chemical process equipment to with-
stand the corrosive effects of high chloride content water, the equip-
ment costs will be increased. Chlorides can be removed from the
scrubber liquor by the use of reverse osmosis, ion exchange, or
evaporation and condensation techniques. These techniques, in
addition to other concepts such as electrodialysis and vacuum
freezing vapor compression, are being tested in many parts of the
world as methods for converting sea water to potable water.
If the scrubber liquor is to be discharged to surface or
ground waters and is to meet the October 1973 EPA-proposed criteria
for public water supplies, the concentration of the following con-
stituents will have to be reduced for some of the cases studied: As,
B, Cd, chromium (Cr) (total), fluoride, iron, Mn, and silver (Ag).
In addition, it was found that Pb, Hg, Se, chlorides, and sulfates in
the scrubber liquor from all four power plants exceeded the criteria.
The amount of reduction of the constituents will depend upon the spe-
cific composition of the scrubber liquor. For example, analyses of
the individual power plant scrubber liquors showed that the Arizona
Public Service Cholla Station scrubber at the time of sampling was
operating in a tightly closed scrubber loop condition resulting in higher
than acceptable concentration levels in the discharge stream from the
76
-------
FDS tank for all of the aforementioned ingredients, except As. The
EPA/TVA Shawnee TCA scrubber system exceeded the limits for As,
Cr (total), Pb, Hg, Se, Ag, chlorides, and sulfates. The EPA/TVA
Shawnee venturi and spray tower scrubber system exceeded the limits
for As, B, Cd, Cr (total), Pb, Mn, Hg, Se, Ag, chlorides, and sul-
fates in the discharge from the clarifier underflow or from the drum
vacuum filter filtrate on one or more of the sampling dates. Con-
versely, the Duquesne Phillips Station system was experiencing oper-
ating conditions requiring high water flow rates through the system
which diluted the concentration of the dissolved ingredients. The re-
sult was that the liquor from the clarifier underflow exceeded the
proposed EPA criteria only for Cd, Pb, Hg, Se, chlorides, and sul-
fates; these results may not represent the normal concentrations to
be expected in this scrubber liquor.
77
-------
SECTION 5
WATER TREATMENT STUDIES
5.1 LITERATURE AND INDUSTRY REVIEW
The domestic and foreign literature was reviewed, and
technical meetings and discussions were held with representatives
from water treatment processors, consulting firms, equipment manu-
facturers, research laboratories, and governmental agencies conduct-
ing research and development testing in the field of water treatment.
This review was conducted to determine what data are available on
processes and types of equipment for use in reducing the concentra-
tion levels of the sludge liquor constituents (Refs. 13 through 50). The
results indicated that there are many pieces of equipment available
designed to accomplish a specific water treatment and that numerous
laboratories, agencies, and companies are currently engaged in ex-
panding the field of water chemistry and treatment operations. How-
ever, most of this work is being directed toward the treatment of brine
or fresh water for public use or a specific industrial process, and for
sanitary waste disposal; only a limited amount of work has been done
in industrial water reuse.
Generic types of water treatment processes and an in-
dication of their current status is given in Table 5- 1. Also included is
an assessment, which is based upon the data gathered in the technical
review, about the possibility of the treated water meeting the established
water quality criteria for reuse or discharge. A selection of the systems
considered and applicable for scrubber liquor treatment is given in
Section 5. 2.
79
-------
Table 5-1. SUMMARY OF GENERIC TYPES OF WATER
TREATMENT PROCESSES CAPABLE OF
MEETING WATER QUALITY CRITERIA
Water treatment
process
Complete system
Multistage evaporation
Tower distillation
Brine concentration
Spray drying
Vacuum freezing
Solar distillation
Rotating bipolar electrodes
Partial operation
Filtration or centrifugation
Ultrafiltration
Reverse osmosis
Ion exchange
Chemical precipitation
(including lime -soda
softening)
Electrodialysis
Selective absorbent
Electrochemical; fluidized
bed
Foam separation
EPA-proposed
public water
supply intake,
October 1973
Yes
Yes
Yes
Yes
Yes
Yes
No
No
No
Yes
Yes
No
Yes
No
No
No
Nonpo table
service
water
Yes
Yes
Yes
Yes
Yes
Yes
No
No
No
Yes
Yes
Yes
Yes
No
No
No
Status
Operational
Operational
Operational
Operational
Development
Development
Laboratory
Operational
Operational
Operational
Operational
Operational
Development
Laboratory
Laboratory
Laboratory
Used with other operations to form a total water treatment system
80
-------
5.1.1 Multistage Evaporation
Many concepts and designs of single- and multistage
evaporators are available and have been used in the chemical process
industry for many years to concentrate the solids in a liquid mixture
by evaporation of the liquid material (Ref. 13). Multistage evap-
oration uses the concept of a series of evaporators with decreas-
ing pressure from one stage to the next; the vapor from one stage
is fed into the heating chamber for the following stage. This evapora-
tion and condensation concept can produce a condensate of high purity
water that meets the requirements for discharge to bodies of water,
for potable water, or for direct use as boiler feed water. Multistage
evaporation was one of the first techniques used for desalting sea water
for drinking water and is still being used. An efficient multistage
evaporator design using flash evaporation was evaluated in this study
(see Section 5.2.4).
5.1.2 Tower Distillation
Tower distillation could be used for water purification,
and the distillate would be of potable or boiler feed quality (Ref. 13).
However, unless there are special circumstances (e.g., a supply of
excess heat or steam available and a need for potable water) this is
not a cost-effective approach for obtaining discharge or service water.
The cost for distilled water is higher than condensate water from a
multistage flash evaporator.
5.1.3 Brine Concentration
Brine concentration uses a type of falling film evapora-
tor wherein the feed brine falls in a film inside a tube bundle, and a
portion of this film is then vaporized (Refs. 14 and 15). There is also
a vapor-compression thermodynamic cycle in which the vapor is com-
pressed and introduced to the shell side of the tube bundle. The
81
-------
temperature differential between the vapor and the brine film causes
the vapor to condense as pure water. The concentrated brine solution
is withdrawn to a disposal pond. The condensate meets the discharge
or power plant reuse water quality criteria. The results of a techni-
cal and economic evaluation on the potential use of a brine concentrator
for treating the scrubber liquor are given in Section 5.2. 5.
5.1.4 Spray Drying
Spray dryers are similar to flash evaporators. A
material is sprayed into a vacuum chamber, the liquid is evaporated,
and the particulate matter is collected in the base of the dryer. This
equipment is generally used for removing moisture from solid par-
ticles that are considered the product rather than vice versa (Ref. 13).
Consequently, this specific process was not examined as a potential
water treatment operation although the condensate would be a high-
quality water. However, a combination of a modified multistage
evaporation and spray drying called multistage flash evaporation was
evaluated (see Section 5.2.4).
5.1.5 Vacuum Freezing
In the vacuum freezing vapor-compression process,
latent heat of fusion is given up when precooled feed water is intro-
duced into a chamber under low pressure. The feed water is pumped
into the system through a deaerator and then through heat exchangers
where it is cooled to the freezing temperature by cold brine and cold
product water leaving the system. The deaerated and cooled feed
water is subsequently pumped into the freezer where a simultaneous
boiling-freezing process occurs. The low pressure causes the water
to boil and part of it vaporizes. The vapor extracts heat of vaporiza-
tion from the remaining feed water, part of which then freezes. The
ice-brine slurry is transferred from the freezer to a washer to re*
move adhering brine and in turn is sent to a melter. This process is
82
-------
reported to be in a development stage for use in desalting sea water
(Refs. 16 and 17). However, an investigation into the adaptability
of this process for producing discharge or service water from desul-
furization slurry liquor indicated that additional developmental tests
must be made before using this technique.
5.1.6 Filtration or Centrifugation
Filtration or centrifugation, which is a mechanical
separation process that can remove particles from a feed water source,
is ineffective on dissolved ingredients (Refs. 18 through 20). Conse-
quently, this process must be considered for use only in conjunction
with other processes capable of removing the dissolved constituents
from the water.
5.1.7 Ultrafiltration
Ultrafiltration or microfiltration uses membranes of
pore size ranging from 0.001 to 10 microns to achieve separation
of impurities in waste water on the basis of molecular size only.
This filtration process removes smaller particles from the liquor
to be treated than normal filtration or centrifugation, but it does
not reduce the concentration levels of dissolved constituents such
as the chloride salts. Although Ultrafiltration processes are oper-
ational, the application is for removal of suspended solids, coliform
bacteria, and similar particle type contaminants in sewage waste
water.
5.1.8 Reverse Osmosis
Osmosis oc.curs when pure water and a salt solution are
on opposite sides of a semipermeable membrane and pure water dif-
fuses through the membrane diluting the salt solution (Refs. 16, 17,
and 21 through 26). The osmotic process can be reversed by .
applying pressure [>30 kg/sq cm ฃ>420 psi)] to the salt solution.
83
-------
This process has been successfully used for desalting sea water and
can be used to treat scrubber liquors for discharge or reuse. The
results of an economic evaluation of this technique following a pre-
treatment of lime-soda softening are discussed in Section 5.2.2.
5.1.9 Ion Exchange
Demineralization or ion exchange is based on the re-
moval of the impurities from water by means of synthetic resins that
have an affinity for dissolved or ionized salts. The exact types of
resins to be used will depend upon the specific constituents to be re-
moved (Refs. 16, 17, and 26 through 29). An economic analysis was
made of the ion exchange process in conjunction with the lime-soda
softening process (see Section 5.2.3). Work is under way by many
researchers to develop resins that will selectively remove specific
ions and trace metals from water solutions. Five examples for heavy
metal removal are:
a. Chelex-100 resin is reported to concentrate cadmium
(Cd), cobalt (Co), chromium (Cr), copper (Cu), lead
(Pb), manganese (Mn), nickel (Ni), and zinc (Zn) in
preference to calcium (Ca), magnesium (Mg), potas-
sium (K), and sodium (Na). A pH controlled solution
of 5 to 6 is required.
b. Amberlite EX 318 resin at a pH of >4 selectively con-
centrates Cd, Co, Cu, iron (Fe), mercury (Hg), Ni, Pb,
vanadium (V), and Zn.
c. Titanium arsenate selectively removes Cd, Co, Cu, Mn,
Ni, Pb, tin (Sn), and Zn.
d. Permutit S 1005 resin at a pH of 5 to 9 is reported to
remove Cd, Co, Cu, Mn, molybdenum (Mo), Ni, Pb,
V and Zn from a water solution.
e. Rohm and Haas EX 243 is being used to remove boron
(B) from a magnesium chloride brine solution.
84
-------
5.1.10 Selective Absorbent
Laboratory work is progressing on identifying absorbents
that have an affinity for specific ions or heavy metals (Refs. 29 through
31). An example is tannery hair chemically treated to form the sulfhy-
dryl and used as packing in a column. Silver (Ag), Cd, Cu, Hg, and Zn
will be absorbed from a water solution passed through the column.
Activated carbon is widely used for removing organic
impurities from waste water, and work is under way to determine if
activated carbon can be used effectively for removing inorganic sub-
stances. Manganese is reported to be removed from water by the
spray aeration of the water to oxidize the manganese and form a pre-
cipitate; this is followed by filtration and absorption with activated
carbon. However, the use of selective absorbents has not progressed
sufficiently for it to be considered an economically available process
for treating scrubber liquor.
5.1.11 Chemical Precipitation (including lime-soda
softening)
Chemical precipitation is an all encompassing term for
reacting ingredients together to form a chemical compound that will
precipitate from the water solution (Refs. 32 through 43). A commonly
used example of chemical precipitation is the reaction of lime and soda
ash with calcium and magnesium salts to reduce the hardness of water.
Section 5.2. 1 presents the results of an evaluation of using a lime-soda
softening operation for treating scrubber liquors. All chemical pre-
cipitation processes require a separation operation to remove the pre-
cipitates from the liquid.
Considerable research is also under way to develop
chemical reactions that will selectively remove heavy metals from a
water solution. For example, in tests at the Central Contra Costa,
County, California, Sanitary District Treatment Plant it was shown
85
-------
that a process using the addition of calcium hydroxide to water to
attain a pH of 11.5 followed by settling, recarbonation with carbon
dioxide to lower the pH to a range of 9ป 5vto 10.5 and subsequently
filtration, reduces the concentration of Ag, barium (Ba), Cd, Co,
Cr, Cu, Mn, Ni, Pb, and Zn in the water. In a similar process, the
Orange County Water District, Santa Ana, California, is using a treat-
ment process consisting of operations such as coagulation with ferric
sulfate and a lime-soda softening to reduce the concentration of heavy
metals in their sewage water reuse treatment plant.
At the EPA Water Supply Research Laboratory, National
Environmental Research Center, Cincinnati, Ohio, laboratory tests
have confirmed the reduction in concentration of arsenic (As), Ba, Hg,
and selenium (Se) as the result of coagulation with ferric sulfate and a
lime softening operation at a pH of 11.
Some other chemical reactions include:
a. Thioacetamide in water at a pH of 1 is reported to
precipitate Cd, Cu, and Pb, and at a pH of 8 to pre-
cipitate Cr, Mo, titanium (Ti) and Zn.
b. Metal sulfides such as CdS2 at a pH of 2. 5 will pre-
cipitate Co, Cr, Cu, Mo, Ni, and V from water.
c. Chelation with ammonium pyrrolidine dithiocarbonate
at a pH of 4. 5 to 6 is reported to remove Cd, Co, Cu,
Mn, Mo, Ni, Pb, and Zn.
d. Dibromo-oxime (dibromo-hydroxy-quinoline) with
acetone in brine water at a pH of 5 to 8 is reported
to precipitate Co, Cr, Cu, Mn, Pb, and Zn.
e. The addition of aluminum sulfate will precipitate As,
Cr, Cu, Pb, Se, and Zn from a water solution at a
pH of 6.8 to 7.
f. Calcium hydroxide at a pH of 9. 5 will cause precipi-
tation of Co, Cr, Cu, Mn, Ni, Pb, and Zn from water,
and dolomite limestone of 40 to 70 percent calcium
hydroxide and 30 to 60 percent magnesium hydroxide
added to water will precipitate Pb and Sn.
86
-------
g. Coagulation with ferric sulfate at a pH of 9 to 11 will
remove As from the water, and at a pH of 6 to 9 Se
will be removed. At a pH of about 7 to 8 and followed
by air injection, filtration, and magnetic attraction of
the ions, it is reported that antimony (Sb), Cr, Cu,
Fe, Hg, Ni, and Pb can be removed from water.
h. Starch xanthate mixed with a cationic polymer (such
as polyvinylbenzyl trimethyl-ammonium chloride) into
a water solution containing heavy metals is reported
to precipitate both the polymer and the heavy metals
as a cohesive floe. This process will remove Ag, Cd,
Cr, Cu, Fe, Hg, Mn, Ni, Pb, and Zn.
5.1.12 Electrochemical; Fluidized Bed
Laboratory tests are being made on a method of
removing trace heavy metals from a water-lime solution with a pH
of 7 to 8 by applying a low voltage direct current field across a
fluidized bed of conductive particles. The metallic impurities plate
out on the tin granules that are used as the collector particles.
Arsenic, Cd, Cr, Cu, Hg, and Zn were reported to be reduced in
concentration by this operation. This process will not remove the
chlorides from the water; a reverse osmosis or other process must
be used in conjunction with this method (Ref. 44).
5.1.13 Rotating Bipolar Electrodes
A process is being studied in the laboratory in which
bipolar electrodes are rotated in a central chamber containing water.
The cations in the water are attracted toward the negatively charged
surface and will flow into a side chamber and be attracted by a sta-
tionary electrode. The anions are similarly attracted into a side
chamber on the opposite side of the central chamber. This process
jnay have application for the purification of a small side stream, but
more laboratory and development work will be required before the
technique becomes operational for use in treating scrubber liquora
(Ref. 45).
87
-------
5.1.14 Electrodialysis
Electro dialysis uses electric current to drive charged
ions through one or more ion-permeable membranes. The passage of
current through the solution causes an increase in the concentration of
one species on one side of the membrane and a corresponding reduc-
tion on the other side. It is necessary to ensure the feed water does
not contain suspended particles or they will clog the pores of the mem-
brane; therefore, removal of suspended solids is accomplished by a
separation pretreatment. Tests, which were conducted with sea water,
have shown that a potable water can be obtained. However, further
work is necessary to develop this concept before it can be considered
fully operational for use with scrubber water (Refs. 46 through 48).
5.1.15 Foam Separation
A process has been tested in the laboratory in which it
is reported that sodium hydroxyphenyl butylbenzyl sulfonate as a sur-
face active agent is mixed with water and then flowed countercurrent
in a tower with nitrogen gas with a pH of 5. 5 to 8 to remove Co, Cr,
and Ni from the water. A subsequent separation operation is needed
with this process (Ref. 29).
5.1.16 Solar Distillation
Techniques and equipment to utilize solar energy to sup-
ply heat to evaporate water and subsequently condense the vapor as
pure water is being investigated. One solar still concept being tested
is a large covered pond which uses solar energy for heating and
evaporating the water in the pond. The vaporized water condenses
on the underside of the pond roof and runs off through troughs to a
collection basin. This technique shows promise for hot, dry areas
such as the southwestern U.S., but it has obvious limitations in the
colder climates and locations with many days of cloud cover. With
88
-------
the current U.S. energy shortage, more effort is being expended by
government agencies and private research laboratories to develop
practical concepts to harness solar energy. However, additional
research and testing is still needed before a concept employing solar
energy is considered usable for treating scrubber liquor in all areas
of the U.S. (Ref. 49)>
5. 1. 17 Analysis of Review
An analysis of the numerous water treatment processes,
techniques, and equipment reveals that a lime-soda softening process
and a separation operation to remove the precipitates will reduce the
concentration of As, B, Ca, Cd, Cr, Cu, Fe, Hg, Mg, Pb, Se, Zn,
and sulfates in the scrubber liquors; however, the process will not
reduce the concentration of certain soluble ingredients (i.e. , sodium
and chloride salts). This process is currently operational and will
not require a breakthrough or advancement in the state of the art.
In plants where brackish water is acceptable for service
water, the lime-soda softening would be adequate. However, in plants
where brackish water is unacceptable, an additional operation such as
reverse osmosis or ion exchange will be necessary. A pH adjustment
may be required in some cases. An evaporation/condensation process
will also produce acceptable water. Adaptations of this latter process
for treating scrubber liquor are the multistage flash evaporator and
the brine concentrator techniques; either would produce a condensate
capable of being reused in the power plant or discharged.
It is recommended that before a full-scale plant for
treating scrubber liquors is installed, pilot plant testing be performed
on the specific scrubber liquor to be treated to acquire specific data
to properly size the equipment and to check out its operation. This
study reviewed the literature and held technical discussions
89
-------
with power plant operators and organizations active in water
treatment processing and equipment manufacturing. No work was
done in the laboratory to treat any of the scrubber liquors.
5.2 SELECTED WATER TREATMENT PROCESSES
Processes adaptable for treating scrubber water to
reduce the concentration of ingredients were identified in Section 5.1.
As indicated, many processes are available that could reduce the con-
centration of one or more of the constituents; however, their develop-
ment status ranges from, those still in laboratory testing to those that
are in full-scale operation (Refs. 13 through 49). In reviewing the
different processes, five appear to have been developed to the degree
that they can be considered an off-the-shelf process and readily adap-
table for use in treating the scrubber system bleed streams. Table 5-2
lists these processes and their capability for meeting the water qual-
ity criteria. Each of these is discussed in the following Sections.
5.2.1 Liime^-Soda Softening Process
The concept of using currently available processes to
treat scrubber liquors at the potential discharge point is fulfilled by
this process if the plant is designed to accept brackish water (defined
for this study as water having 5000 mg/t or greater of chloride). In
this process lime (CaO) and soda ash (Na-CO,) are reacted with the
major species in the liquor to precipitate the Ca and Mg together with
the heavy metals. The pH is adjusted with carbon dioxide and the
solution is filtered or centrifuged with the. precipitates being sent to the
disposal site and the product water fed to the reverse osmosis or ion
exchange operation for chloride removal if the power plant cannot
accept brackish water for service water.
90
-------
Table 5-2. CAPABILITY OF WATER TREATMENT PROCESSES
TO MEET REQUIREMENTS
Water treatment
process (in
decreasing order of
cost effectiveness)
Lime -soda softening
Lime -soda softening
plus reverse
osmosis
Lime -soda softening
plus ion exchange
Multistage flash
evaporation
Brine concentration
Criteria
EPA- proposed
public water
supply intake --
October 1973
No
Yes
Yes
Yes
Yes
Nonpo table service water
Brackish
water
acceptable
Yes
Not
applicable
Not
applicable
Not
applicable
Not
applicable
Brackish
water not
acceptable
No
Yes
Yes
Yes
Yes
o
Brackish water for this study is defined as water having 5000 mg/ฃ
or more of chloride.
The literature indicates that the lime-soda softening
process can reduce the concentration of the Ca and Mg salts to less
than 1000 mg/4. Data showed that about an 80 to 90 percent reduction
in concentration of As, Cd, Mn, Pb, and Se has been achieved by this
process and, depending upon the source of the data analyzed, it was
reported that 30 to 90 percent of hexavalent chromium can be removed.
Boron will also be reduced in concentration during the softening pro-
cess; insoluble calcium and magnesium borate salts are formed and
*
subsequently removed with other precipitates.
91
-------
The pH of the liquor at the potential discharge points
ranges from 3 to 10.7 for the four scrubbing systems studied and ex-
ceeds the criteria of 5 to 10 (Table 2-1). If the pH of the water is low
at the time of bleed to the water treatment plant, the pH will be in-
creased by adding lime. After precipitation and prior to filtration or
centrifugation there is a pH adjustment with carbon dioxide, if neces-
sary, to lower the pH.
5.2.2 Pretreatment Process (Lime-Soda Softening)
Plus Reverse Osmosis
The use of reverse osmosis to treat the product water
from the lime-soda softening pretreatment process will remove sodium
chloride salts and other undesirable constituents and will result in
a high-quality product readily capable of meeting both the service water
and the EPA-proposed public water supply intake criteria.
The three basic designs for commercial reverse osmosis
modules are as follows:
a. The tubular type that uses small diameter (about 1/2-
in.-diameter) porous or perforated tubes. The reverse
osmosis membrane is installed on either the internal
or external side of the tube and bundles of the tubes are
joined to a feed water header system.
b. The hollow fiber type that uses many capilliary tubes
in a bundle configuration.
c. The spiral wound type that uses a flat sheet of mem-
brane to cover each side of a flat sheet of porous
water-conducting material. The membrane is sealed
on the two long sides and one end to form an envelope.
The other side of the membrane and porous water-
conducting material is sealed to a perforated tube that
receives the product water.
92
-------
All three types are designed to operate at pressures of about 30 to
45 kg/sq cm (430 to 640 psi) and are being used commercially.
Reverse osmosis was initially developed to produce
potable water from brackish supplies, but because multivalent ions
such as Ca, Mg, and sulfate are as readily rejected as are the mono-
valent ions (e.g., Cl), this concept has been adapted for boiler feed
water and treatment of waste waters for reuse. The literature implies
that the reverse osmosis treatment would be the only operation needed
to treat the scrubber water for either reuse or discharge. However,
reverse osmosis equipment manufacturers and system users recom-
mended that because of the high concentration of the major species in
the scrubber water, a lime-soda softening process should precede the
reverse osmosis operation. The pretreatment would reduce the con-
centration of the major species to a level compatible with current
operational practices and experience, eliminate any objectional col-
loidal matter present in the scrubber liquor, and prevent precipitation
of large deposits of the Ca and Mg sulfate salts on the membrane sur-
face. Furthermore, this high concentration would have a greater ten-
dency to cause an increase in equipment downtime for cleaning and
repair and a predicted shorter membrane life. A higher reverse
omosis investment as well as higher operating costs could result than
if pretreatment were used followed by reverse osmosis to remove the
final traces of the cations and anions.
The percent water loss from the pretreatment and re-
verse osmosis system was estimated to be about 25 to 30 percent of
the feed water to the water treatment facility. This water would con-
tain extremely high concentrations of the undesirable ingredients
(e.g. , As, Hg, Pb, and Se) and should be disposed of with the waste
products from the scrubber.
93
-------
5.2.3 Pretreatment Process (Lime-Soda Softening)
Plus Ion Exchange
A review of the literature and technical discussions
held with ion exchange manufacturers indicated that the use of the
ion exchange process by itself would not be an economical approach
as a scrubber liquor water treatment process. The manufacturers
recommend that a pretreatment operation to reduce the high con-
centration of constituents be undertaken before flowing the water over
the ion exchange resins. Otherwise, there would be excessive equip-
ment down time for flushing and clean up of the resin beds. However,
by using a pretreatment consisting of lime-soda softening and filtra-
tion or centrifugation, the constituent concentration levels will be re-
duced to values that can be handled by an ion exchange process to pro-
duce a product capable of meeting the EPA-proposed public water sup-
ply intake criteria.
Numerous resins of various chemical compositions are
available or are being tested for general water treatment and for spe-
cific element removal. These resins are, by definition, insoluble
solids containing fixed cations and anions capable of reversible ex-
change with mobile ions of the opposite sign in the solutions with which
they are brought into contact. Ion exchange was used for years in water
softening and limited impurity removal; however, industry is now
developing newer resins for different applications. For example,
Rohm and Haas has developed a deionization process for use in desali-
nation work. The process is based upon the use of weak electrolyte
ion exchange resins for treating brackish waters of 500 to 5000 ppm
total dissolved solids. The process consists of three units in series:
the alkalization unit, dealkalization unit, and carbonation unit. The
influent to the first unit is passed through a bed of weak base anion
exchange resin in which the anionic constituents are converted to the
94
-------
bicarbonate salts of Na, Ca, and Mg. In the second unit, which
contains a weak acid cation exchange resin, the bicarbonate salts are
converted to carbonic acid. The third unit, which also contains a weak
base anion exchange resin but in the free base form, absorbs the car-
bonic acid from the effluent of the second unit. After the exhaustion
cycle is complete, the first unit is regenerated to the free base form
while the second unit is converted back to the hydrogen form. Since
the third unit is already in the bicarbonate form, the flow pattern is
simply reversed for the next cycle. The third unit becomes the alka-
lization unit and the first unit becomes the carbonation unit. An adap-
tation of this process could be used to remove the chlorides from the
scrubber liquor.
Other ion exchange resin beds must also be incorporated
into the overall process to remove specific trace metals that may be in
excessive concentration. Specific resins are available for selective
heavy metal removal. For example, Rohm and Haas1 Amberlite EX 318
resin is used to concentrate Cd, Co, Cu, Fe, Hg, Ni, Pb, V, and Zn,
and EX 243 is used to remove B. Other manufacturers market resins
also capable of removing trace elements.
There is a water loss incurred in conjunction with the
operation of an ion exchange process. The amount is a function of the
degree of backwashing and rinsing during the regeneration cycle and
can be determined precisely only by conducting pilot plant tests using
the exact liquor and with the specific manufacturer's resin. However,
it is estimated that the degree of loss would be 25 to 30 percent of the
feed water to the treatment facility.
5.2.4 Multistage Flash Evaporation
Multistage flash evaporation involves the process of
progressively heating and flash evaporating the scrubber liquor in a
series of stages under progressively lower pressures. Each stage of
95
-------
the evaporator is configured with a heat exchanger for preheating the
incoming liquid and a vacuum chamber for flash evaporation. The
vapor is used for preheating the liquid and is condensed in this oper-
ation. The condensate or product water from each stage is collected
for reuse in the power plant or discharged. A portion of the residual
concentrated scrubber water is mixed with influent liquor for recycling
through the evaporator; the remaining portion is transported to a dis-
posal area. The product from this operation is of high quality and
suitable for boiler feed water or for drinking water as evidenced by
the fact that this process has been widely used as an acceptable
technique for producing potable water from sea water.
There is a loss of water with the concentrated slurry
mix. The exact quantity of water lost is a function of the specific
design of the evaporator and the equipment operating characteristics.
For example, equipment that is operated to achieve a total dissolved
solids (TDS) content in the discharge concentrate of 100,000 ppm,
using a 10,000 ppm TDS feed water and <50 ppm TDS product con-
densate, has a water loss of 10 to 12 percent of the water fed to the
evaporator.
5.2.5 Brine Concentration
A brine concentrator using the concept of a falling film
evaporator can produce a product that can meet the reuse or discharge
water quality criteria. This type of concentrator flows the feed water
in a film inside a tube bundle, and a portion of this film is vaporized.
A vapor compression thermodynamic cycle is part of the system in
which the vapor is compressed and introduced to the shell side of the
tube bundle. The temperature differential between the vapor and the
water film causes the vapor to condense as pure wate.r. The concen-
trated brine solution is subsequently withdrawn for disposal. It is-
96
-------
estimated that with a feed water of about 10,000 ppm total dissolved
solids, a discharge concentrate of 100,000 ppm, and a product of
< 50 ppm, there would be a loss of about 10 percent of the feed water.
5. 3 ESTIMATED WATER TREATMENT COSTS
In order to make a comparison of the economics of the
five selected water treatment processes discussed in Section 5.2,
engineering cost estimates were prepared for each process. Both
capital and operating costs were considered, and all costs were based
on 1974 dollars (Ref. 50 through 62). The results of the cost estimates
are shown in Figures 5-1 through 5-5. The capital costs for each
process were based on average annual charges of 18 percent of the
original investment over a 30-year life including straight line deprecia-
tion, parts replacement, insurance, taxes, and cost of capital. Opera-
ting costs were based on 4560 operating hours per year (30 year average)
and included labor, maintenance, repair, and operating power charges.
Two types of coal were assumed to be burned:
a. Type 1 coal has a thermal content of 29,700 joules/gm
(12, 500 Btu/lb) and is burned at the rate of 0. 38 kg
(0. 84 Ib) of coal to produce 1 kWh of electricity.
b. Type 2 coal has a thermal content of 18,600 joules/gm
(8,000 Btu/lb) and is burned at the rate of 0.42 kg
(0. 92 Ib) of coal per kWh produced.
The capital and operating cost comparisons for water
treatment contained in Sections 5. 3. 1 through 5. 3. 5 are based on a
flow rate variation of 200 to 700 gpm.
97
-------
PRETREATMENT PROCESS USED: LIME-SODA SOFTENING
1974 DOLLARS
1000 MW POWER PLANT SCRUBBER
LAND COSTS NOT INCLUDED
TO CONVERT gpm TO cu m/hr MULTIPLY BY 0.227
>
> ~ 2
Z
BRINE
CONCENTRATOR
MULTISTAGE
FLASH
EVAPORATOR
PRETREATMENT
PLUS ION
EXCHANGE
PRETREATMENT
PLUS REVERSE
OSMOSIS
LIME/ SODA
SOFTENING
'0 500 1000
WATER TREATMENT PLANT CAPACITY, gpm
Figure 5-1. Estimated capital investment
98
-------
PRETREATMENT PROCESS USED: LIME-SODA SOFTENING
1974 DOLLARS
1000 MW POWER PLANT SCRUBBER
COSTS BASED ON 30-YEAR ANNUALIZED COSTS, 4560 OPERATING
HOURS PER YEAR AVERAGE OVER 30 YEARS
1.25
1.0
0.75
O
at
m
o.
LJ
O
go. so
0.25
TO CONVERT
FROM
TO
MULTIPLY
BY
gpm
gal
i ii i
cu
liter
0.227
3.785
MULTISTAGE FLASH
EVAPORATOR OR BRINE
CONCENTRATOR
PRETREATMENT PLUS
ION EXCHANGE
PRETREATMENT PLUS
REVERSE OSMOSIS
LIME-SODA SOFTENING
0 500 1000
WATER TREATMENT PLANT CAPACITY, gpm
Figure 5-2. Estimated water treatment costs per gallon
99
-------
PRETREATMENT PROCESS USED: LIME-SODA SOFTENING
1974 DOLLARS
1000 MW POWER PLANT SCRUBBER
COSTS BASED ON 30-YEAR ANNUALIZED COSTS, 4560 OPERATING
HOURS PER YEAR, AVERAGE OVER 30 YEARS
0.5
2 0.
Q
O
Of
Q.
Of.
Ul
Q.
0.3
0.2
0.1
TO CONVERT
FROM
TO
MULTIPLY
BY
gpm
cu m/hr
i i i i
I
0,227
MULTISTAGE
FLASH EVAPORATOR
OR BRINE
CONCENTRATOR
PRETREATMENT
PLUS ION EXCHANGE
PRETREATMENT
PLUS REVERSE
OSMOSIS
LIME-SODA
SOFTENING
i i I
0 500 1000
WATER TREATMENT PLANT CAPACITY, gpm
Figure 5-3. Estimated water treatment costs per kilowatt hour
for power output of 1000 MW
100
-------
10
Q.
Z
UJ
2
CD
1
at
in
a.
i
iii
o
PRETREATMENT PROCESS USED: LIME-SODA SOFTENING
1974 DOLLARS
1000 MW POWER PLANT SCRUBBER
COAL: TYPE No. 1 12,500 Bty/lb, 0.84 Ib/kWh
TYPE No. 2 8,000 BtM/lb, 0.92 Ib/kWh
COSTS BASED ON 30-YEAR ANNUALIZED COSTS, 4560 OPERATING
HOURS PER YEAR AVERAGE OVER 30 YEARS
TO CONVERT
FROM
TO
MULTIPLY
BY
gpm cu m/hr 0.227
BtiVlb joules/gm 2.235
Ib/kWh kg/kWh 0.4536
'^/Million Btu ml 11 s/M joule 0.00948
0 500 1000
WATER TREATMENT PLANT CAPACITY, gpm
COAL TYPE No. 1
BRINE CONCENTRATOR OR MULTISTAGE FLASH EVAPORATOR
PRETREATMENT PLUS ION EXCHANGE
PRETREATMENT PLUS REVERSE OSMOSIS
COAL TYPE No. 2
G33E3 BRINE CONCENTRATOR OR MULTISTAGE FLASH EVAPORATOR
PRETREATMENT PLUS ION EXCHANGE
PRETREATMENT PLUS REVERSE OSMOSIS
CZZZZ3 LIME-SODA SOFTENING
Figure 5-4.
Estimated water treatment costs per million
Btu heat input for power output of 1000 MW
101
-------
PRETREATMENT PROCESS USED: LIME-SODA SOFTENING
1974 DOLLARS
1000 MW POWER PLANT SCRUBBER
COAL: TYPE No. 1 12,500 Btu/lb, 0.84 Ib/kWh
TYPE No. 2 8,000 Btu/lb, 0.92b/kWh
COSTS BASED ON 30-YEAR ANNUALIZED COSTS, 4560 OPERATING
HOURS PER YEAR AVERAGE OVER 30 YEARS
Q
hi
OL
CO
_i
O
u
O
Of.
O
Of.
Ill
a.
8
u
1.0
0.75
0.50
0.25
TO CONVERT
FROM
TO
MULTIPLY
BY
gpm cu m/hr 0.227
Btu/lb joules/gm 2.235
Ib/kWh kg/kWh 0.4536
$/short ton $/metric ton 1.105
I
I I I
0 500 1000
WATER TREATMENT PLANT CAPACITY, gpm
COAL TYPE No. 1
BRINE CONCENTRATOR OR MULTISTAGE FLASH EVAPORATOR
PRETREATMENT PLUS ION EXCHANGE
PRETREATMENT PLUS REVERSE OSMOSIS
LIME-SODA SOFTENING
TYPE NO. 2
BRINE CONCENTRATOR OR MULTISTAGE FLASH EVAPORATOR
PRETREATMENT PLUS ION EXCHANGE
PRETREATMENT PLUS REVERSE OSMOSIS
I/////I LIME-SODA SOFTENING
Figure 5-5.
Estimated water treatment costs per ton of
coal burned for power output of 1000 MW
102
-------
5.3.1 Lime-Soda Softening
Capital investment charges vary from $0.47M to
$0.9M, and total average treatment costs over 30 years, including
annualized capital charges and operating costs, vary from 0.51 cents
per gallon to 0.25 cents per gallon over the same range. These
costs convert to the following:
a. From 0.053 to 0.092 mills per kWh produced. The
power was assumed to be produced at an average
annual rate of 4560 million kWh.
b. From 0.0047 to 0.0081 mills per megajoule (0.51 to
0. 88 centers per million Btu) heat input from Type 1
coal and from 0.0067 to 0.0116 mills per megajoule
(0. 73 to 1.26 cents per million Btu) heat input for
Type 2 coal.
c. From 0. 14 to 0.24 dollars per metric ton (0. 13 to
0.22 dollars per short ton) of coal burned for Type 1
coal and 0. 13 to 0. 22 dollars per metric ton (0. 12 to
0.20 dollars per short ton) of coal burned to Type 2
coal.
5.3.2 Lime-Soda Pretreatment Plus Reverse Osmosis
The capital investment varies from 3/4 to 2 million
dollars. Athough currently used membranes have a service life of
about 3 years (depending upon the specific operating conditions and
quality of the feed water), industry sources are predicting a membrane
life of 5 to 6 years for membranes now being evaluated; therefore,
membrane life was assumed to be 5 years. Labor expenses were
assumed to be $200,000 per year.
The average water treatment costs for the lime-soda
softening pretreatment plus reverse osmosis over a 30-year period.
range from 0.90 to 0. 51 cents per galloon. These costs convert to:
a. From 0. 11 to 0.22 mills per kWh produced. Power
was assumed to be produced at an average annual
rate of 4560 million kWh.
103
-------
b. From 0.011 to 0.019 mills per megajoule (1.2 to 2.0
cents per million Btu) heat input for Type 1 coal and
from 0.014 to 0.027 mills per meagjoule (1.5 to 2.9
cents per million Btu) heat input for Type 2 coal.
(See Section 5. 3 for a definition of Type 1 and 2 coals.)
c. From 0. 31 to 0. 57 dollars per metric ton (0. 28 to
0. 52 dollars per short ton) of coal burned for Type 1
coal and 0.28 to 0.53 dollars per metric ton (0.25 to
0.48 dollars per short ton) of coal burned for Type 2
coal.
5.3.3 Lime-Soda Softening Plus Ion Exchange
The initial investment ranges from $1. 1 to 2. 5 million.
The resin life was estimated to be 5 years. The estimated average
water treatment costs over a period of 30 years for the flow rate
range assumed above are as follows:
a. From 1. 1 to 0.62 cents per gallon of water treated.
b. From 0. 14 to 0.27 mills per kWh produced. Power
was assumed to be produced at a constant average
annual rate of 4560 million kWh.
c. From 0.011 to 0.023 mills per megajoule (1.2 to 2. 5
cents per million Btu) heat input for Type 1 coal and
from 0.019 to 0.034 mills per megajoule (2.0 to 3. 6
cents per million Btu) for Type 2 coal. (See Sec-
tion 5.2.2.2 for a definition of Type 1 and 2 coals.)
d. From 0. 39 to 0. 67 dollars per metric ton (0. 35 to
0.61 dollars per short ton) of coal burned for Type 1
coal and 0. 35 to 0. 63 dollars per metric ton (0. 32 to
0. 57 dollars per short ton) of coal burned for Type 2
coal.
These costs are about 20 to 30 percent higher than the
pretreatment process plus reverse osmosis operation that will provide
water of comparable quality.
In reviewing the quality of the water produced and the
investment and treatment costs, it appears as if the lower cost pre-
treatment and reverse osmosis operation is a more cost-effective
104
-------
approach than pretreatment and an ion exchange operation to provide
water of sufficient quality for reuse or discharge.
5.3.4 Multistage Flash Evaporator
The estimated investment for a multistage flash evap-
oration facility ranges from about $1.4 to 3.7 million (Figure 5-1).
These values are 60 to 90 percent more than for comparable quality
water obtained from the pretreatment process plus a reverse osmosis
operation, and the average water treatment costs are estimated to
be 1-1/4 to 1-2/3 times higher (Figures 5-2 through 5-5).
Consequently, it does not appear to be cost effective
to use a multistage flash evaporator to treat scrubber water when the
lime-soda softening and reverse osmosis process will provide the
quality of water needed at a lower cost.
5.3.5 Brine Concentrator
Brine concentrators have been operational only a rela-
tively short time compared to other available water treatment pro-
cesses. The estimated initial investment and water treatment costs
are compared to the other processes in Figures 5-1 to 5-5. The
economic analysis was made on the same basis as previously described,
except that the labor was reduced to only a part-time operator on the
recommendation of the manufacturer, thus, the labor was estimated
at $50,000 per year. It is expected that in the future the brine con-
centrator costs can be reduced. However, at the present time this
process is technically compatible with the requirement to produce a
product with reduced concentrations of ingredients to meet the water
quality criteria, but it does not appear to be as cost effective to em-
ploy this process for low cost, off-the-shelf water treatment for
scrubber liquor when other processes such as lime-soda softening
and reverse osmosis could accomplish the same function at a lower
cost.
105
-------
REFERENCES
1. Jones, J. W., et al. Disposal of By Products from Non-
Regenerable Flue Gas Desulfurization Systems. (Presented
at American Society of Civil Engineers Annual and National
Environmental Engineering Convention. Kansas City, Missouri.
October 21-25, 1973.)
2. Rossoff, J., et al. Disposal of By-Products from Non-
Regenerable Flue Gas Desulfurization Systems - a Status
Report. (Presented at the EPA Control Systems Laboratory
Symposium on Flue Gas Desulfurization. Atlanta, Georgia.
November 4-7, 1974.)
3. Rossoff, J., and R. C. Rossi. Disposal of By-Products from
Non-Regenerable Flue Gas Desulfurization Systems: Initial
Report. Environmental Protection Agency. Publication
Number EPA 650/2-74-037-a. May 1974.
4. Federal Water Pollution Control Act Amendments of 1972 -
92nd Congress-Second Session. United States Code
Congressional and Administrative News, Number 10. West
Publishing Company. 1972.
5. Public Health Service Drinking Water Standards 1962. U.S.
Public Health Service. Publication Number 956. 1962.
6. Water Quality Criteria. Report of the National Technical
Advisory Committee to the Secretary of the Interior.
April 1968.
7. McKee, J. D., and H. W. Wolf. Water Quality Criteria.
Publication Number 3-A. State Water Resources Control
Board, California. April 1971.
8. Proposed Criteria for Water Quality. Environmental
Protection Agency. October 1973.
107
-------
9. Steam Electric Power Plant Generating Point-Source Category,
Effluent Guidelines and Standards. Federal Register. 39,
Number 196: October 8, 1974.
10. Personal Communication with TVA and Bechtel Corporation
personnel associated with the EPA/TVA Shawnee Steam Plant
Flue Gas Desulfurization Scrubbing Test Project. TVA
personnel: P. Stone, K. Metcalfe, and J. Barkely; Bechtel
personnel: L. Sybert and S. Almaula. Subject: Operations
of EPA/TVA Shawnee Steam Power Plant.
11. Personal communication. P. L. Baldwin and N. J. Ozenberger,
Arizona Public Service Company, Cholla Station, Joseph City,
Arizona. Subject: Operations of Cholla Station.
12. Personal communication. S. L. Pernick and R. D. O'Hara,
Duquesne Light Company, Phillips Station, South Heights,
Pennsylvania. Subject: Operations of Phillips Station.
13. Perry, J. H. , Editor. Chemical Engineering Handbook.
New York, McGraw-Hill Book Company.
14. Brine Concentrator Specification RCC 225T. Resources
Conservation Company. Bulletin. 1974.
15. Personal communication. O. Kirchner, Resources Conservation
Company, Renton, Washington. Subject: Brine concentrator
designs and applications.
16. Eckenfelder, W. W., and L. K. Cecil, Editors. Progress in
Water Technology Applications of New Concepts of Physical -
Chemical Waste Water Treatment. Nashville, Tennessee,
International Association on Water Pollution Research and
American Institute of Chemical Engineers. September 1972.
17. Cecil, L. K., Editor. Complete Water Reuse. American
Institute of Chemical Engineers, National Conference on
Water Reuse. April 1973.
18. Personal communication. T. E. Bird, Western Sales Manager,
The Bird Company, South Walpole, Massachusetts. Subject:
Centrifugation and filtration equipment.
19. Personal communication. E. H. Dewey, Field Engineer,
Hoffman Air and Filtration Division, Clarkson Industries,
Syracuse, New York. Subject: Filtration clarification and
vacuum distillation equipment.
108
-------
20. Personal communication. C. Novotny, Technical Director,
Industrial Filter and Pump Manufacturing Company, Eicero,
Illinois. Subject: Removal of suspended solids.
21. Cruver, J. E., and I. Nusbaum. Application of Reverse
Osmosis to Wastewater Treatment. Journal of Water Pollution
Control Federation. 46: 2, 1974.
22. Fan, L. T., et al. Analysis of a Reverse Osmosis Water
Purification System and its Optimization Based on the Plug
Flow Model for the Bulk Flow in a Membrane Unit. Institute
for Systems Design and Optimization, Kansas State University,
Manhattan, Kansas. 1969.
23. Personal communication. W. McCorvalle, Technical
Representative, Monsanto Company, St. Louis, Missouri.
Subject: Hollow fibers for use in membrane separation processes.
24. Personal communication. C. Miller, Applications Engineer,
Polymetrics Division of Technical Equities Corporation, Santa
Clara, California. Subject: Costs and operations of reverse
osmosis water treatment process.
25. Lepper, F. R. Reverse Osmosis: Larger Plants and Wider
Use. Power Engineering. May 1973.
26. Personal communication, L. Winer, Regional Manager,
Barnstead Company Division of Sybron Corporation, Los
Angeles, California. Subject: Reverse osmosis and ion
exchange processes.
27. Bresler, S. A., and E. F. Miller. Economics of Ion Exchange
Techniques for Municipal Water Quality Improvement. Journal
of American Water Works Association. 64; 11, 1972.
28. Trace Metals in Water Supplies: Occurrence, Significance and
Control. Proceedings of the 16th Water Quality Conference.
American Water Works Association, Illinois Environmental
Protection Agency, and University of Illinois. February 1974.
29. Cadman, T. W., and R. W. Dellinger. Techniques for Removing
Metals from Process Wastewater. Chemical Engineering.
April 15, 1974.
30. Giusti, O. M. , et al. Activated Carbon Absorption of
Petrochemicals. Journal Water Pollution Control Federation.
46: 5, 1974.
109
-------
31. Wyman, G. D. Remove Organics from Feedwater Make Up.
Power. July 1972.
32. Personal communication. B. B. Berger, Director of Water
Resources Research Center, University of Massachusetts.
Subject: Wastewater treatment processes.
33. Personal communication. D. G. Argo, Assistant District
Engineer, Orange County Water District, County of Orange,
Santa Ana, California. Subject: Reclamation and reuse of
county wastewater.
34. Wesner, G. M. Water Factory 21 and Waste Water Reclamation
and Sea Water Barrier Facilities. Orange County Water District,
Santa Ana, California. February 1973.
35. Wesner, G. M. , and D. G. Argo. Report on Pilot Waste Water
Reclamation Study. Orange County Water District, Santa Ana,
California. July 1973.
36. Personal communication. W. Zabban, Chief Engineer, The
Chester Engineers, Coraopolis, Pennsylvania. Subject:
Wastewater treatment processes.
37. Betz Handbook of Industrial Water Conditioning. Betz
Laboratories, Trevose, Pennsylvania. 1973.
38. James, G. V. Water Treatment. International Scientific
Series. Edinburgh, Scotland, T and A. Constable, Ltd.
39. Personal communication. A. Poole, Director of Water and
Waste Utilities, Springbrook Water Reclamation Center,
Naperville, Illinois. Subject: Reclamation and reuse of
city wastewaters.
40. Personnal communication. J. S. Kneale, The Permutit
Company, Paramus, New Jersey. Subject: Water treatment
processes.
41. Personal communication. R. Carrillo, Sales Representative,
Zimmite Corporation, Thousand Oaks, California. Subject:
Water treatment chemicals.
42. Wing, R. E. Heavy Metal Removal from Waatewater with Starch
Xanthate. Northern Regional Research Laboratory, U.S.
Department of Agriculture, Peoria, Illinois. March 1974.
110
-------
43. Personal communication. D. W. Miller, Assistant Director,
U.S. Department of Agriculture, Agricultural Research Service,
Northern Regional Research Laboratory, Peoria, Illinois.
Subject: Heavy metal removal from wastewaters.
44. Laboratory Demonstration of the Atomics International Process
for Removal of Heavy Metals from Water. Publication
Number 47. California State Water Resources Control Board
Report AI-73-13.
45. Benner, P. E. U.S. Patents, 448, 026. Removal of Ions from
an Ionized Liquid. June 3, 1969.
46. Christodoulou, A. P., et al. Parametric Economic and
Engineering Evaluation Study of the Electrodialysis Process for
Water Desalination. U.S. Department of Interior. 1969.
47. Bennett, G. F., Editor. Water - 1972. American Institute of
Chemical Engineers Symposium Series Number 129. 69; 1973.
48. Bennett, G. F., Editor. Water - 1973. American Institute of
Chemical Engineers Symposium Series Number 136. 70; 1974.
49. Androsky, A. Uses of the Sun in the Service of Man. The
Aerospace Corporation, El Segundo, California. Publication
Number ATR-74(9970)-1. October 1973.
50. Blecker, H. G., et al. How to Estimate and Escalate Costs of
Wastewater Equipment. Chemical Engineering, Deskbook Issue.
October 21, 1974.
51. Personal communication. J. L. Willa, Vice-President of Sales,
Lilie-Hoffman Cooling Towers, Inc., St. Louis, Missouri.
Subject: Cooling tower operations, costs, blowdown, and losses.
52. Boies, D. B., et al. Technical and Economic Evaluations of
Cooling Systems Blowdown Control Techniques. Environmental
Protection Agency. Publication Number EPA 660/2-73 -026.
November 1973.
53. Porter, K. S. , and A. G. Boon. Cost of Treatment of Waste
Water. Symposium on Trent Research Programs. Water
Pollution Research Laboratory of the Department of the
Environment, University of Nottingham, England. 1971.
Ill
-------
54. Farrow, J. C. , et al. Estimating Construction Costs of Waste
Water Treatment Systems. Journal of American Association of
Textile Chemists and Colorists. 2: 3, February 1970.
55. Hinomoto, H. Unit and Total Costs - Functions for Water
Treatment Based on Koenig's Data. Water Resources Research.
7: 5, 1971.
56. Ackerman, W. C. Cost of Water Treatment in Illinois. Technical
Letter Number 11. Illinois State Water Survey. 1968.
57. DiGregorio, D. Cost of Wastewater Treatment Processes.
Dorr-Oliver Inc.
58. Prehn, W. L. , et al. Desalting Cost Calculating Procedures.
U.S. Department of Interior Research and Development.
Publication Number 555. May 1970.
59. Eckenfelder, W. W., and D. L. Ford. Economics of Wastewater
Treatment. Chemical Engineering. August 25, 1969.
60. Smith, R. Cost of Conventional and Advanced Treatment of
Wastewater. Journal of Water Pollution Control Federation.
40: 9, 1968.
61. Personal communication. R. Cookingham, Sales Representative,
Gregory Pump Company, Alhambra, California. Subject: Costs
for slurry pumps.
62. Eckenfelder, W. W., and J. L. Barnard. Treatment - Cost
Relationship for Industrial Wastes. Chemical Engineering
Progress. 67_: 9, 1971.
63. Handbook for Analytical Quality Control in Water and Wastewater
Laboratories. Analytical Quality Control Laboratory, National
Environmental Research Center, Cincinnati, Ohio. 1972.
64. Methods for Chemical Analysis of Water and Wastes. Analytical
Quality Control Laboratory, National Environmental Research
Center, Cincinnati, Ohio. 1971.
112
-------
APPENDIX A
DESCRIPTION OF CHEMICAL ANALYSIS TECHNIQUES
A.I INTRODUCTION
This appendix describes the analytical techniques (Refs.
63 and 64) used to determine the concentration of constituents in the
scrubber liquors. The constituents are divided into (a) major chem-
ical species (calcium, sodium, sulfate, and chloride), (b) trace metal
species, and (c) additional chemical species. Other water quality
tests are also described.
Consideration was given to the constituent's range of
concentration and to the corresponding costs of the analyses to obtain
*r
data having high precision or high accuracy. Although the basis for
selecting the proper analytical technique was minimizing any inter-
ference from other species, the presence of chemical species inter-
fering with a particular analysis was fully acknowledged. Only when
the interference was considered significant were corrections applied.
A. 2 MAJOR CHEMICAL SPECIES
A. 2. 1 Calcium Determination
The method selected from among several has an accu-
racy of 4 percent and was one in which calcium oxalate was precipitated
Precision is defined as the relationship between a measured value and
the statistical mean of measured values, and accuracy is the relation-
ship between the true value and the measured value.
113
-------
and filtered from the solution, the filter cake was redissolved in HC1,
and the solution was titrated against KMnO, to a characteristic purple
end point. Correction was then made for excess permanganate at the
characteristic end point.
Alternative techniques using a specific ion electrode and
atomic absorption spectrophotometry were eliminated because they had
lower accuracies resulting from interferences, primarily from the
sulfate ions.
A. 2. 2 Sodium Determination
Atomic absorption spectrophotometry was selected in
preference to a method that determines sodium gravimetrically by
separating the calcium salts from the sodium salts in the scrubber
liquors by fractional precipitation. The selected technique uses a
Perkin-Elmer 303 instrument with a graphite furnace at 3302A wave-
length. For every test the liquor concentration required multiple dilu-
tions to achieve a level suitable for optimum analytical determination.
Any errors in results tend to be high because of the effects of interfer-
ence and the high concentration of sodium in the scrubber liquors; how-
ever, the errors are less than 10 percent.
A. 2. 3 Sulfate Determination
Standard nephelometry techniques are used for this task.
A barium sulfate precipitate was formed by the reaction of the sulfate
ion with a barium chloranilate reagent. The resulting turbidity was
determined by a spectrophotometer and compared to a curve from stan-
dard sulfate solutions. Although multiple dilutions are necessary to
bring the concentration to a range of optimum reliability, the resulting
error is less than 10 percent.
114
-------
A.2.4 Chloride Determination
A solid-state electrode was used to determine the
electrode potential of chloride ions in scrubber liquors. The results
were compared against a standard curve. This method has a precision
of about 1 percent and an accuracy greater than 5 percent.
A. 3 TRACE METAL SPECIES
Since most trace metal species are highly sensitive to
atomic absorption spectrophotometry, this technique was used. A
Perkin-Elmer 303 instrument, with deuterium background corrector,
and a graphite tube heater, for increased sensitivity, were used to
determine soluble concentrations for the following elements: alumi-
num, antimony, boron, beryllium, cadmium, chromium, copper, co-
balt, iron, magnesium, manganese, molybdenum, nickel, lead, silicon,
silver, tin, vanadium, and zinc. Precision and accuracy are dependent
upon the specific element, its relative concentration, and the extent of
interference. The precision and accuracy of the measurements of con-
centrations of all elements that exceed water quality reuse criteria
ranged between 5 and 20 percent.
Mercury was also determined using this technique; how-
ever, the mercury is reduced to the elemental state with stannous chlo-
ride and the absorption produced from the resulting mercury vapor is
measured. This method has a precision of about 20 percent and an
accuracy probably greater than 50 percent.
Arsenic was determined by the Gutzeit method that re-
acts arsine with mercurous bromide to produce Hg-As; the unknown is
compared colorimetrically against standards. For this application this
technique has a precision of about 25 percent and an accuracy probably
greater than 50 percent.
A fluorimetric technique that has a sensitivity down to
micrograms per liter was used to determine selenium. It has a preci-
sion of about 10 percent and is accurate to 100 percent.
115
-------
A. 4 ADDITIONAL CHEMICAL SPECIES
If these species existed in sufficiently high concentration,
they would affect the water quality of the liquors; however, most of
them were present in trace quantity and their importance in defining
water quality was considered insignificant. Therefore, precise or
complete analyses were not considered of particular value to this pro-
gram. The concentration of many of them is controlled by interaction
with one or more of the major species. Their expected concentration
was determined using an Aerospace Corporation developed computer
program (part of the work for a related EPA contract) that calculates
the maximum concentration of each minor species in chemical equi-
librium with the major species in the system. The experimental deter-
minations were always in agreement with the expected concentration
value a.
A.4.1 Carbonate Determination
Total carbonate was determined gravimetrically by the
absorption of CO., evolved from an acidified sample. This technique
does not have high sensitivity and could not detect concentrations in
scrubber liquors at values less than 10 mg/t carbonate. (The expected
value determined by the Aerospace computer program was typically
less than 1 mg/i for all samples analyzed.)
A.4. 2 Sulfite Determination
Total sulfite was determined using a specific ion elec-
trode and no significant interferences were observed. The oxidation of
the sulfite ion to sulfate in the scrubber liquor was found to be a very
rapid reaction. Liquor protected from the atmosphere will typically
reveal concentrations of several hundred milligrams per liter of the
sulfite ion; however, a brief atmospheric exposure will cause oxidation
*
and reduce these concentrations by one or more orders of magnitude.
116
-------
The reported sulfite measurements were for samples analyzed
immediately upon arrival in the laboratory. No specific action was
taken to inhibit oxidation other than to ensure that the samples were
transported from the power plant scrubber to the analytical laboratory
in sealed containers. The exposure to air during sampling, filtering,
and measuring, however, resulted in the sulfite values reported. It is
presumed that these concentrations would probably more closely repre-
sent the oxidation state of liquor in the event of their potential discharge.
A.4. 3 Phosphate Determination
The phosphate analysis was determined by spectrophoto-
metry methods, using ammonium molybdate to form the molybdenum
blue complex. Total range of phosphate content varied from 0. 5 mg/S.
in an acid liquor (pH = 4.3) to 0.01 mg/i in a base liquor (pH = 10.4).
A.4.4 Nitrogen Determination
Total nitrogen was determined by the Kjeldahl method
which converts organic nitrogen components to ammonia. The ammo-
nia is then distilled and the amount determined by titration. This
method has a precision of about 10 percent, and accuracy at the levels
of the concentrations determined is about 25 percent.
A.4. 5 Fluoride Determination
The fluoride ion was determined by the specific ion
electrode using a Beckman Model 76 digital pH meter. There are no
significant interferences in the scrubber liquors. This method has a
precision of about 5 percent; an accuracy of 20 percent is attainable at
the low levels measured.
117
-------
A. 5 OTHER WATER QUALITY TESTS
A. 5. 1 Chemical Oxygen Demand
Chemical oxygen demand was determined by reacting
the organics and sulfites present with potassium dichromate and mea-
suring the reduced chromium by spectrophotometry. While a precision
of 25 percent is attainable, accuracy depends on the sample history
(i.e. , degree of exposure to atmospheric oxygen) and is about 100 per-
cent for routine analysis.
A. 5. 2 Total Alkalinity
Total alkalinity is determined by titrating a 25 mt sam-
ple with standard acid to a pH of 4. 0. The Beckman Model 76 digital
pH meter is used as the indicating instrument. Total alkalinity is ex-
pressed as milligrams per liter calcium carbonate, but is actually a
determination of the buffering capacity of the liquor due to a number of
weak acid species (i.e., carbonate, sulfite, borate, arsenite, selenite,
and silicate). Precision is about 5 percent and accuracy is estimated
to be 25 percent.
A. 5. 3 Total Dissolved Solids Determination
The total dissolved solids are determined gravimetri-
cally by evaporating a 10 to 25 ml sample overnight in a tared weighing
bottle under vacuum at 120ฐF. Since two of the major constituents
(calcium and sodium sulfates) form stable hydrated salts and are very
hygroscopic in the anhydrous state, prolonged drying and minimal ex-
posure of the dried residue are mandatory. The precision is about 2
percent, and the accuracy is about 5 percent.
A. 5.4 Total Conductance Determination
This measurement, which was made with a General
Radio Impedance Bridge Type 1650A, gives an estimate of the total
118
-------
ionic strength of the liquor. Precision is about 1 percent, and accuracy
is estimated to be about 2 percent.
A.5.5 pH Determination
This parameter is measured with a Beckman Model 76
digital pH meter to a precision of 0.2 percent and an accuracy of 1
percent.
A.5. 6 Turbidity Determination
Turbidity measurements were made by nephelometry in
which light absorption is compared to standards that were prepared
using a formazin mixture that is a mixture of hydrazine sulfate and
hexamethylene tetramine in a water solution. Most scrubber liquor
samples were obtained as slurries and were virtually opaque. Chem-
ical analyses were performed on filtered samples, and turbidity values
were also obtained on these filtered samples. Only a few samples were
obtained as filtrate; their results are reported as received rather than
after filtering.
119
-------
APPENDIX B
COSTING BACKGROUND DATA
B. 1 INTRODUCTION
This appendix describes the methodologies used in
preparing the cost estimates for the various water treatment processes
analyzed in this study. The rationale used for both capital and oper-
ating costs is presented.
B.2 CAPITAL AND OPERATING COSTS
The cost data obtained from the various sources listed
in the reference section were converted, in all cases, to equivalent
1974 dollars. This conversion was made by use of an appropriate eco-
nomic indicator such as the Plant Cost Index published monthly in
Chemical Engineering. The estimated capital and operating costs for
each water treatment process are included with the description of each
process and are based on treating 200 to 1000 gpm. Total annualized
costs were compiled as the total of annualized capital costs and oper-
ating costs.
B. 2.1 Capital Cost Estimates
The capital costs for the various water treatment pro-
cesses were based on an equipment operational life of 30 years. The
annualized capital charges were estimated on this basis and include
elements such as insurance, cost of capital, depreciation, replace-
ments, and taxes. Certain specific equipment items, such as reverse
121
-------
osmosis membranes, identified as having a service life less than 30
years were depreciated on an accelerated schedule, e.g. , five years.
The itemized approach for computing annualized capital charges is
shown in Table B-l. It should be noted that the cost of land was not
included in the capital cost estimates.
The estimated capital costs for various water treatment
processes can be illustrated by the approach used on lime-soda ash
softening. From data obtained from the references, installed costs
were generated for the clarifier-reactant tank, the chemical storage
hopper and feed system, the filter, two surge tanks, five pumps (plus
five spares), piping for the complete facility, and building and site
improvements. The pumps and piping were assumed to be fabricated
with stainless steel because of the corrosive nature of the water being
handled. These costs were derived as a function of flow rate. An
additional 10 percent was added for interest expense during construc-
tion. The average annualized capital charges were computed on the
total of these amounts at the rate of 18 percent, as described in Table
B-l. This same procedure was used in arriving at the capital cost
estimates for the other water treatment processes.
B. 2. 2 Operating Cost Estimates
The operating cost estimates were computed on the
basis of 4560 hours per year (190 days) as an average over 30 years.
Again, using the lime-soda treatment as an example, the chemical
costs, utility power, materials, operating labor, and maintenance
labor were computed as a function of flow rate. An overhead rate of
20 percent was applied to the total of these costs. As stated above,
annualized capital and operating costs were combined to obtain total
annual costs.
122
-------
Table B-l. AVERAGE ANNUALIZED CAPITAL CHARGES
(30-YEAR LIFE - 1974 DOLLARS)
Straight line depreciation is assumed
Cost elements
Percentage of
capital investment
Depreciation based on 30-year life
Replacements items: 20% of depreciation
Insurance
Subtotal
3.33
0.67
0.50
4.50
Cost of capital:
Taxes
50% debtฎ 11%
11% (0.5)
50% equity @15%
15% (0.5)
Assume Federal tax to be the
same as equity
Other taxes at 80% of Federal
tax
5.50
7.50
7.50
6.00
Subtotal 26.50
On a 30 year basis, the average annual
capital charge on initial investment is
= 17.75%
Total
17..7 5
123
-------
B. 3 WATER TREATMENT COST CURVES
The cost curves presented in Section 5.2 of this report
were based on the operation of a scrubber at a 1000 MW station. The
cost parameters are plotted as a function of flow rate for each of the
five processes considered to be readily adaptable to treating scrubber
bleed water.
B. 3. 1 Capital Investment
The data plotted in the curves shown in Figure 5-1 re-
present the costs for materials and labor to install the water treatment
processing equipment, including contractors' fees, engineering design,
construction supervision, and interest expenses during construction at
a rate of 8 percent annually. These curves do not include land costs,
start-up costs, or brine disposal. The cost estimates were prepared
by contacting manufacturers, the treatment industry, and government
experts, as well as conducting a thorough literature search.
B.3.2 Treatment Costs per Gallon
The estimated water treatment costs per gallon (Figure
5-2) were prepared for each process by combining the annual capital
and operating costs as a function of the total annual flow rate. The
annualized capital charges were calculated as described in Section
B. 2. 1. The annual operating costs were estimated as described in
Section B. 2. 2.
B.3.3 Treatment Costs per Kilowatt Hour
The process treatment costs per kilowatt hour (Figure
5-3) were computed from the total annual costs consisting of annualized
capital charges as described in Section B. 2. 1, plus annual operating
costs, for 4560 operating hours per year. The average annual energy
9
output was assumed to be 4. 56 x 10 kWh.
124
-------
B.3.4 Treatment Costs per Million BTU
Heat Input
Water treatment costs as a function of Btu input (Figure
5-4) were computed from the total annual costs for each treatment
process and the weight and Btu content of the two types of coal con-
sidered. The coal quantities computed were based on a 1000 MW out-
put and 4560 average operating hours per year.
B.3.5 Water Treatment Costs per Ton of Coal
The total annual costs for each water treatment process
were computed as a function of weight of coal burned. Consideration
was given to the characteristics of type 1 and 2 coal for a 1000 MW
station. The same average annual operating period of 4560 hours was
used for this calculation as for the other operating curves.
125
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/2-76-024
2.
3. RECIPIENT'S ACCESSION NO.
4. TITLE ANDSUBTITLE
Reuse of Power Plant Desu Ifurization Waste Water
5. REPORT DATE
February 1976
6. PERFORMING ORGANIZATION CODE
7 AUTHOR(S)L.J.Bornstein, R.B. Fling, F.D.Hess,
R. C. Rossi, and J. Rossoff
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORflANIZATION NAME AND ADDRESS
The Aerospace Corporation
2350 East El Segundo Boulevard
El Segundo, California 90245
10. PROGRAM ELEMENT NO.
1BB392: ROAP R21AZU-025
11. CONTRACT/GRANT NO.
Grant R-802853-01-0
12. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Industrial Environmental Research Laboratory
Research Triangle Park, NC 27711
13. TYPE OF REPORT ANC
Final; 1/74-4/75
NO PERIOD COVERED
14. SPONSORING AGENCY CODE
15 SUPPLEMENTARYNOTESFOI. technical details contact Fred Roberts, EPA, Environmental
Research Laboratory, 200 SW 35th St. , Corvallis, Oregon 97330
16. ABSTRACT The repOrf giV6s results of an assessment of the potential reuse of liquor
from nonregenerable flue gas desulfurization systems by applying available water
treatment processes. Although scrubbers normally operate in a closed-loop mode,
this study investigated liquor reuse if a scrubber purge became necessary as a
result of off-design or other operating conditions. Chemical characterizations were
performed on liquors from four different scrubbers; these were assessed for use as
power plant service water or for direct discharge. Treatment is required for either
use; but, in most cases (for economic reasons), treatment for discharge is not
recommended. Chemical precipitation (e.g. , lime-soda softening, filtration, and
pH control) is adequate for most service water usage cases. Cases involving high
chloride content in the purge liquor would require an additional treatment such as
reverse osmosis, as would all cases for direct discharge. The applicabilities of
available treatment systems are given, in addition to scrubber flow diagrams, water
balance, and treatment costs for a range of liquor flow rates.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
Air Pollution
Waste Water
Liquids
Regeneration
(Engineering)
Reclamation
Water Treatment
Electric Power Plants
Flue Gases
Desulfurization
Scrubbers
Chemical Analysis
b.IDENTIFIERS/OPEN ENDED TERMS
Air Pollution Control
Stationary Sources
Reuse
c. COSATI Field/Group
13B
10B
21B
07A
3. DISTRIBUTION STATEMENT
19. SECURITY CLASS (This Report)
Unclassified
21. NO. OF PAGES
136
Unlimited
20. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
126
------- |