TVA
EPA
Tennessee Valley
Authority
Division of Energy Demonstrations
and Technology
Chattanooga, Tennessee 37401
EOT-103
United States
Environmental Protection
Agency
Industrial Environmental Research
Laboratory
Research Triangle Park NC 27711
EPA-600/7-80-063
March 1980
Application of Membrane
Technology to Power
Generation Waters
Interagency
Energy/Environment
R&D Program Report
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the INTERAGENCY ENERGY-ENVIRONMENT
RESEARCH AND DEVELOPMENT series. Reports in this series result from the
effort funded under the 17-agency Federal Energy/Environment Research and
Development Program. These studies relate to EPA's mission to protect the public
health and welfare from adverse effects of pollutants associated with energy sys-
tems. The goal of the Program is to assure the rapid development of domestic
energy supplies in an environmentally-compatible manner by providing the nec-
essary environmental data and control technology. Investigations include analy-
ses of the transport of energy-related pollutants and their health and ecological
effects; assessments of, and development of, control technologies for energy
systems; and integrated assessments of a wide range of energy-related environ-
mental issues.
EPA REVIEW NOTICE
This report has been reviewed by the participating Federal Agencies, and approved
for publication. Approval does not signify that the contents necessarily reflect
the views and policies of the Government, nor does mention of trade names or
commercial products constitute endorsement or recommendation for use.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/7-80-063
March 1980
Application of Membrane
Technology to Power
Generation Waters
by
T.L. Don Tang, Tien-Yung J. Chu,
and Ralph D. Boroughs
TVA Project Director: Hollis B. Flora II
Tennessee Valley Authority
Division of Energy Demonstrations and Technology
Chattanooga, Tennessee 37401
EPA Interagency Agreement No. D8-E721-BE
Program Element No. INE624A
EPA Project Officer: Theodore G. Brna
Industrial Environmental Research Laboratory
Office of Environmental Engineering and Technology
Research Triangle Park, NC 27711
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Washington, DC 20460
-------
DISCLAIMER
This report was prepared by the Tennessee Valley Authority and has
been reviewed by the Office of Energy, Minerals, and Industry, U.S.
Environmental Protection Agency, and approved for publication. Approval
does not signify that the contents necessarily reflect the views and
policies of the Tennessee Valley Authority or the U.S. Environmental
Protection Agency, nor does mention of trade names or commercial products
constitute endorsement or recommendation for use.
ii
-------
PREFACE
The scope of this study has changed since its inception. Because
electrodialysis removes only a small fraction of the impurities per
pass, an early decision was made to focus on reverse osmosis and ultra-
filtration. Pilot reverse osmosis and ultrafiltration modules were
purchased and tested on various waste streams. For the most part, these
tests were successful, and results were close to those that would have
been predicted by the equipment vendors. Problems that occurred were
generally related to pretreatment.
As conceptual design of reverse osmosis systems for application to
power plant waters began, it became evident that pretreatment and reject
stream disposal were often the largest part of the problem. Pretreatment
such as pH and redox control, softening, and suspended solids removal
can often produce water which is suitable for reuse, even though it may
not meet water quality standards for discharge. Furthermore, concentrate
disposal under zero discharge conditions can become expensive since an
evaporator and storage ponds are generally required. Given the problem
of reject disposal and using more modest product water quality require-
ments based on reuse, electrodialysis began to look more attractive.
Electrodialysis is capable of producing a highly concentrated reject
stream (up to 25-percent solids), thus eliminating the need for
evaporators.
Alternatives to reverse osmosis treatments, including electrodialy-
sis and the various technologies normally associated with reverse osmosis
pretreatment, were included in a broader project scope. Because of
limited resources, it was necessary to limit its depth, particularly in
relation to alternative treatments. These options may be more deeply
explored in subsequent studies.
iii
-------
ABSTRACT
Three membrane technologies (reverse osmosis, ultrafiltration, and
electrodialysis) for wastewater treatment and reuse at electric generating
power plants were examined. Recirculating condenser water, ash sluice
water, coal pile drainage, boiler blowdown and makeup treatment wastes,
chemical cleaning wastes, wet S02 scrubber wastes, and miscellaneous
wastes were studied. In addition, membrane separation of toxic sub-
stances in wastewater was also addressed. Waste characteristics, appli-
cable regulations, feasible membrane processes, and cost information
were analyzed for each waste stream. A users' guide to reverse osmosis
was developed and is provided in an appendix.
Treatment of power plant waters with membrane technologies to
attain total water reuse and zero effluent discharge is technically
feasible. Membrane technologies are not suited to remove materials
which are unstable and apt to precipitate as they are concentrated;
however, they excell in separating materials which are not susceptible
to conventional wastewater treatment (including many toxic pollutants
and dissolved solids). Thus, membrane technologies complement rather
than compete with conventional technologies. For dissolved solids
control, membrane technologies are viable alternatives to distillation.
Distillation is more expensive but requires less pretreatment than
reverse osmosis or electrodialysis. Electrodialysis is estimated to be
the cheapest but does not produce product water of comparable quality to
either reverse osmosis or distillation. To encourage wide acceptance of
membrane technologies, demonstrations are needed to develop more defini-
tive information on costs and reliability.
iv
-------
CONTENTS
Page
Preface iii
Abstract iv
Figures vii
Tables viii
Acknowledgments xi
List of Conversion Factors xii
1. Introduction 1
2. Conclustions 2
3. Recommendations 4
4. Membrane Separation Technologies 5
Reverse osmosis 5
Ultrafiltration 11
Electrodialysis 13
Similarities and differences 16
Applications to power generation 17
5. Regulations Applicable to Power Plant Waters 21
Dissolved and suspended materials 21
Radioactive waste 25
6. Recirculating Condenser Cooling Water 27
Characteristics of wastewater 27
Alternative treatments 30
Recommendations 35
7. Ash Sluice Waters 38
Characteristics of wastewater 38
Alternative treatments 45
Recommendations 46
8. Coal Pile Drainage 49
Characteristics of wastewater 49
Alternative treatments 55
Recommendations 58
9. Boiler Makeup Water Treatment Wastes 62
Characteristics of wastewater 62
Alternative treatments 62
Recommendations 68
10. Boiler Slowdown 72
Characteristics of wastewater 72
Alternative treatments 72
Recommendations 73
11. Chemical Cleaning Wastes 74
Characteristics of wastewater 74
Alternative treatments 74
Recommendations 91
12. Wet Sulfur Dioxide Scrubber Wastes 92
Characteristics of wastewater 92
Alternative treatments .- 93
Recommendations 93
13. Radioactive Wastewaters from Power Reactors 94
Characteristics of wastewater 94
Alternative treatments 96
Recommendations 98
-------
CONTENTS
(continued)
Page
14. Miscellaneous Wastes 99
Characteristics of wastewater 99
Alternative treatments 99
Recommendations 100
15. Integrated or Combined Wastes 101
References 102
Appendixes
A. User's Guide to Reverse Osmosis 110
B. Membrane Separation of Toxic Substances in
Wastewater 120
VI
-------
LIST OF FIGURES
Number
1
2
3
4
Bl
Schematic diagram of membrane separation processes .
Flow diagram for cooling tower sidestream softening
and desalting with electrodialysis
Flow diagram for cooling tower blowdown treatment
by reverse osmosis for reuse
Flow diagram for reverse osmosis treatment of
coal pile drainage . .
Cellulose acetate membrane rejection of phenol as
a function of oH
Page
6
36
37
61
126
B2 Cellulose acetate membrane rejection of three
clases of linear alkyl compounds as a function
of the total number of carbons 127
VII
-------
LIST OF TABLES
Number Page
1 Membrane Separation Processes and Principal
Driving Forces 7
2 Steam-Electric Power Generating Point Source
Category Nonthermal Limitations 22
3 List of 65 Toxic (Priority) Pollutants 24
4 Typical Control Limits for Recirculating Waters. . . 28
5 Results of Reverse Osmosis Studies on Cooling
Tower Slowdown Treatment 33
6 Characteristics of Once-Through Ash Pond Discharges. 39
7 Relationships Between Plant Operation Conditions and
pH Values of Ash Pond Effluents at TVA Coal-
Fired Power Plants 42
8 Results of Reverse Osmosis Studies on Fly Ash Pond
Effluent from TVA Plant A 47
9 Results of Reverse Osmosis Studies on Concentrated
Fly Ash Pond Effluent from TVA Plant A 48
10 Coal Analysis, Dry Basis 50
11 General Chemical Characteristics of Coal Pile
Drainages Collected from Two TVA Steam Plants ... 51
12 Range of Trace Elements in U.S. Coal 53
13 Trace Metal Concentrations in Coal Pile Drainages . . 54
14 Boiler Wash Wastes 56
15 Results of Chemical Treatment on Boiler Wash Water. . 57
16 Operating Parameters for Spiral-Wound Reverse Osmosis
Study at 75% Recovery at Mocanaqua, Pennsylvania. . 59
17 Chemistry Analyses for Reverse Osmosis Treatment
of Acid Mine Drainage 60
18 Well Water, Brine, and Product Analyses—Burbank
Public Service Department 63
19 Characteristics of Raw Waters 64
Vlll
-------
LIST OF TABLES
(continued)
Number Page
20 Water Quality Requirements for Combustion Turbine
Injection . 65
21 Cost Comparison of Three Processes 66
22 Basis sf Calculations for Table 21 67
23 Results of the Burbank Public Service Department
Comparison and Analysis of Operating Costs of
Demineralizer and Reverse Osmosis Systems 69
24 Cost Comparison of a Boiler Feedwater Treatment
System with Ion Exchange Only and With a
Combination of Reverse Osmosis and Ion Exchange. . 70
25 Comparison of Condensate Makeup Costs With and
Without Reverse Osmosis Unit 71
26 Typical Plant Equipment Which May Require
Periodic Cleaning 75
27 Characteristics of Chemical Cleaning Wastewater--
Acid Phase Composite 76
28 Characteristics of Chemical Cleaning Wastewater--
Alkaline Phase Composite 77
29 Characteristics of Chemical Cleaning Wastewater—
Passivation Drain 78
30 Results of Reverse Osmosis Studies on Chemical
Cleaning Waste 81
31 Characteristics of Composite Alkaline-Phase
Cleaning Waste 82
32 Performance of Union Carbide 3NJR Ultrafiltration
Module in Polishing Suspended Solids from the
Pretreated Boiler Cleaning Wastes 83
33 Performance of DuPont Spiral-Wound Reverse Osmosis
Module in Treating Pretreated Boiler Cleaning
Wastes 84
34 Performance of UOP-ROGA 4100 Spiral-Wound Reverse
Osmosis Module in Treating Pretreated Boiler
Cleaning Wastes 85
IX
-------
LIST OF TABLES
(continued)
Number Page
35 Performance of Osmonic P.V. 192-97 Spiral-Wound
Reverse Osmosis Module in Treating Pretreated
Boiler Cleaning Wastes 86
36 Summary of Projected Cost Estimate for Nickel
Recovery by Reverse Osmosis (Closed-Loop)
System (1972) 87
37 )perating and Maintenance Costs Per 1000 Gallons of
Permeate for Nickel Recovery by Reverse Osmosis
(Closed-Loop) System (1972) 88
38 Breakdown of Operating Costs for New England
Plating (1977) 89
39 Credits Realized for Reverse Osmosis Operation
at New England Plating 90
Al Rejection of Metal Perchlorates on Various
Commercially Available Membranes 117
Bl Classification of Priority Pollutants 128
B2 Rejection of Organics by Cellulose Acetate
Membranes 132
B3 Rejection of Organics by Cellulose Acetate
Butyrate Membranes 145
B4 Rejection of Organics by NS-100 Membranes 146
B5 Rejection of Organics by Polyamid Membranes 150
B6 Rejection of Heavy Metals by Reverse Osmosis
Membranes 152
-------
ACKNOWLEDGMENTS
This study was initiated by TVA as part of the project entitled
"Advanced Waste Heat Control" and is supported under Federal Inter-
agency Agreement No. EPA-IAG-D8-E721-BE and TV-41967A between TVA and
EPA for energy-related environmental research. Thanks are extended to
the EPA Project Officers, Julian W. Jones and Theodore G. Brna, and to
the TVA Project Director, H. B. Flora II. Appreciation is also extended
to R. J. Ruane and J. M. Wyatt for their assistance in the project.
XI
-------
LIST OF CONVERSION FACTORS
Customary Units
Length 1 inch, (in)
1 foot, (ft)
1 yard, (yd)
Area 1 in2
1 ft2
1 yd2
1 acre
1 hectare, (ha)
Volume 1 gallon, (gal)
1 yd3
1 acre-ft
1 liter, (£)
Weight 1 ounce, (02)
1 pound, (lb)
1 gram
Pressure 1 pound/in2, (psi)
1 atmosphere, (atm)
1 bar
Flow 1 gal/day, (gpd)
1 gal/min, (gpm)
Flux 1 gal/(ft2-day)
1 m£/(cm2'day)
1 cm/day
1 £/m2-day)
International System of Units, (SI)
2.54 x 10~2 meters (m)
0.3048 m
0.9144 m
6.45 x 10"4
9.29 x 10"2
0.8361 m2
m
= 4.047 x 103 m2
= 1.0 x 104 m2
3.785 x
0.7646 m3
1.233 x 103
1.0 x
10"3 m3
nr
10 3 m3
2.8 x 10"2 kilograms (kg)
0.4536 kg
1.0 x 10"3 kg
6.895 x 103 Newtons/m2 (N/m2)
1.013 x 10s N/m2
1.000 x 10s N/m2
4.381 x 10"8 m3/second (ms/s)
6.3 x 10"5 m3/s
4.716 x 10"7 m/s
1.157 x 10"7 m/s
1.157 x 10"7 m/s
1.157 x 10"8 m/s
xn
-------
SECTION 1
INTRODUCTION
The electric power industry, the largest industrial water user in
the United States, is facing a great challenge—to achieve zero pollutant
discharge in the near future. The most practical approach to meeting
this goal is to reuse water within the power plant.
Of the many technologies applicable in treating water for reuse,
the membrane technologies are exceptional in their ability to efficiently
separate and concentrate solutes from water. However, these technologies
are relatively new and often misunderstood. This report reviews the
three leading membrane separation technologies, the regulations which may
motivate their increased use, and the various power plant water uses to
which membrane technologies may be applied. Where appropriate, membrane
technologies have been compared to more conventional technologies in
order to define those areas where membrane processes are most applicable.
-1-
-------
SECTION 2
CONCLUSIONS
Reverse osmosis, electrodialysis and ultrafiltration are proven
technologies. Nevertheless they are relatively new and rapidly changing
technologies that must be used with caution if successful application is
to be assured. In many cases the successful use of membrane technologies
requires an unusual degree of finesse in the conventional processes
required for pretreatment. Thus, membrane technologies do not typically
supplant conventional technologies, but serve to complement them.
Membrane technologies are not suitable for removing materials which
are unstable and likely to precipitate as they are concentrated. These
materials will foul membranes and should be removed by conventional tech-
niques if possible. Thus, it is inappropriate to consider membrane tech-
nology as a method for preventing scale deposition
With proper pretreatment, reverse osmosis and electrodialysis can
be used to remove and concentrate dissolved salts from various streams.
In this role, membrane technology must compete with distillation. Dis-
tillation is typically more expensive but requires less extensive pre-
treatment than reverse osmosis or electrodialysis. Electrodialysis is
estimated to be least expensive but does not produce product water of
comparable quality to either reverse osmosis or distillation.
In considering electrodialysis, reverse osmosis, or distillation
for removing dissolved solids, the need for dissolved solids control should
be critically examined. When the need is established, membrane technology
should be considered as a viable alternative to distillation. Closed-loop
systems to which this concept could be applied include cooling systems,
ash sluice systems, and S02 scrubber systems.
When heavy metals cannot be satisfactorily removed by pH adjustment,
precipitation, adsorption, and suspended solids removal, reverse osmosis
should be considered as a polishing step. Brine may be recycled to the
precipitation step. This concept may be applicable to coal pile drainage
and chemical cleaning wastes.
Ultrafiltration has been proven competitive in treating oily wastes
and as a pretreatment for reverse osmosis.
Ultrafiltration and reverse osmosis have been shown to be effective
in treating various radioactive wastes.
-2-
-------
While reverse osmosis systems design is not an exact science, certain
classes of problems can be handled in a straightforward manner. This
includes processing solutions of inorganic ions to concentrate or purify,
providing that adequate pretreatment is available. Pilot testing of the
pretreatment step is advisable, but pilot testing of the reverse osmosis
system is not generally necessary. Exceptions include situations where
adequate pretreatment is not practical and where organic solutes or
weakly ionized species are important.
Reverse osmosis is an acceptable process for removing inorganic and
organic priority pollutants. Electrodialysis is feasible for removing only
inorganic priority pollutants. Ultrafiltration holds no promise for
removing priority pollutants from wastewaters.
-3-
-------
SECTION 3
RECOMMENDATIONS
The following recommendations for future work are proposed:
1. An expanded "user's guide" is needed to acquaint potential users
of membrane processes with the technology, to eliminate myths
associated with membrane systems, and to encourage the applica-
tion of membrane technology to water reuse in power plant
operation.
2. Large-scale demonstrations are needed of both electrodialysis
and reverse osmosis for controlling total dissolved solids in
cooling water. This would provide the operating experience and
cost data necessary for acceptance of these concepts by the
power industry. Preliminary pilot studies and proper attention
to conservative design should assure successful demonstrations.
3. Much work needs to be done in characterizing radwastes to deter-
mine their physical form. Decontamination factors for ultrafil-
tration and reverse osmosis modules need to be better established.
The use of ultrafiltration to prevent fouling of ion exchange
resins in nuclear plant water systems needs to be investigated.
4. Pilot plant work is needed to demonstrate the feasibility of
boric acid and lithium-7 recovery from pressurized water reactor
coolant using electrodialysis.
5. Further studies are recommended to perfect techniques for remov-
ing metals from coal pile drainage and chemical cleaning wastes.
These studies might consider reverse osmosis as a polishing step.
6. If the technologies of brine concentration by reverse osmosis and
electrodialysis are demonstrated as recommended for cooling water,
experience from that study may be transferable to other closed
loops such as ash sluice and S02 scrubber systems. Meanwhile,
both of these systems need to be examined thoroughly through the
use of parametric testing.
-4-
-------
SECTION 4
MEMBRANE SEPARATION TECHNOLOGIES
The three major membrane separation processes for water and wastewater
treatment are reverse osmosis, ultrafiltration, and electrodialysis. Each
of these processes separates the feedwater into a concentrated stream and
a depleted (purified) stream. Schematic diagrams of these processes are
shown in Figure 1. Thermodynamically, each process has to overcome the
free energy of mixing and may be compared to competitive processes on the
basis of cost and energy efficiency.
A membrane separation system is based on the selective transport of
species through a semipermeable membrane. The ability of some membranes to
selectively pass certain species can be understood by assuming that species
passing through the membrane must first dissolve in the membrane and then
diffuse through it. The structure of the membrane then determines the
relative ease with which various species dissolve in and pass through the
membrane.
The selectivity of some membranes (typically the more open or porous
materials) seems to be based primarily on solute size. Where this is the
case, it may be assumed the mechanism is primarily that of a molecular sieve,
controlled by the relative size and shape of the pores as compared to the
size and shape of the solute molecules. This can be thought of as a
special case of the solution-diffusion model where the chemical process
of dissolution becomes relatively unimportant compared to the physical
process of diffusion.
The driving force for transport through the membrane can be pressure,
concentration, or electrical gradients. Table 1 shows a list of membrane
separation processes, their functions, and the principal driving forces.1
REVERSE OSMOSIS
Fundamentals
Osmosis is a phenomenon that occurs between two solutions of differ-
ent concentrations that are separated by a semipermeable membrane, through
which water flows spontaneously from the less concentrated to the more con-
centrated solution. When a certain pressure is applied to the side of the
more concentrated solution, the direction of water flow is reversed; that
is, the water flows from the more concentrated to the less concentrated
solution. This process is called "reverse osmosis." In the literature,
it is also called "hyperfiltration." Reverse osmosis is a comparatively
-5-
-------
DESALTED PRODUCT WATER
Cl
FEED WATER
ELECTRODIALYSIS
CONCENTRATED
A= ANION-PERMEABLE MEMBRANE
C=CATION-PERMEABLE MEMBRANE
PRESSURIZED SOLUTION
OF A AND B
CONCENTRATED A
MEMBRANE
REVERSE OSMOSIS OR M SOLUTION OF B
ULTRAFILTRATION V
Figure 1. Schematic diagram of membrane separation processes.
Source: Lacey, R. E. Membrane Separation Processes. Chemical
Engineering. September 4, 1972, pp. 56-57. Reprinted
by permission of the publisher.
-6-
-------
TABLE 1. MEMBRANE SEPARATION PROCESSES AND PRINCIPAL DRIVING FORCES
Principal
Process Function of membrane driving force
Reverse osmosis Selective transport of water Pressure
Ultrafiltration Discriminates on the basis of Pressure
molecular size, shape, and
flexibility
Electrodialysis Selective ion transport Electrical potential
gradient
Dialysis Selective solute transport Concentration
Gel permeation chroma- Retards high-molecular-weight Concentration
tography solute penetration
Liquid permeation Selective transport of liquids Concentration
Separation in a Selective ion transport while Electrical poten-
battery separator retaining colloids tial gradient
Oxygen determination Control rate of depolarization Partial pressure or
by electrode concentration
Specific ion determi- Selective transport of an ion Concentration
nation by electrode complex (activity)
Source: Weber, W. J., Jr. Physicochemical Processes for Water Quality Control.
Wiley-Interscience, 1972. p. 308. Printed by permission of the
publisher.
new development for the separation, fractionation, or concentration of
substances in an aqueous solution or gaseous mixture. Because no heat is
added and no phase change is involved in product recovery, reverse osmosis
is an effective process requiring relatively low energy input.
Reverse osmosis feedwater is pressurized by pumping and then chan-
neled along the membrane surface. As product water (permeate) passes
through the membrane, the feedwater becomes progressively more concen-
trated. A continuous flow must be maintained along the feedwater/
concentrate side of the membrane to prevent an excessive buildup of dis-
solved solids. The concentrate stream leaving the system is often referred
to as the "reject," especially in water purification applications.
The physical configuration must provide for distributing feedwater
and collecting the concentrate and permeate. For membrane development
work on a laboratory scale, the "plate and frame" arrangement is common.
The frame supports a number of parallel sheets of membrane material and
provides manifolds for distributing and collecting the water. Permeate
compartments alternate with feed-concentrate compartments.
The most common configurations for commercially available membrane
modules are tubular, spiral wound, and hollow fiber.
The tubular system typically uses porous or perforated tubes to support
the membrane. Bundles of tubes are joined to a feedwater header system.
Tubular devices enjoyed some degree of commercial success during the late
-7-
-------
1960's. A number of vendors were successful in obtaining installations,
especially in nonwater applications, such as chemical separations and
food and drug processing. The tubular systems will handle larger particu-
late matter with less plugging than other designs, permitting feedwaters
with low water quality to pass with a less stringent pretreatment require-
ment. Tubular units have a packing density of about 100 square feet* of
membrane surface area per unit cubic foot of volume. This is an improve-
ment over the plate and frame design, but the packing density of the
tubular configuration is still considerably less than that of other con-
figurations. Tubular units use a fluid velocity of 3 to 4 feet/second
to maintain turbulence, thus preventing the buildup of dissolved salts
at the membrane surface.
In the spiral-wound system, a large flat sheet of membrane is used
to cover each side of a flat sheet of porous, water-conducting backing
material. The membrane is sealed with adhesive on the two long edges
and one end to form an "envelope." The other ends of the membrane are
sealed to a perforated tube which receives the product water. A series
of these envelopes are attached and wound around the central tube,
separated from each other by a spacer. These spacers, typically an open-
mesh material, provide a path for the feed-to-concentrate flow and are
designed to promote turbulence. Packing density of this design is about
300 square feet of membrane per cubic foot of pressure vessel. Spiral-
wound modules are more vulnerable to plugging from particulates in the
feedwater than are tubular modules.
The hollow-fiber systems use hairlike capillary tubes of aromatic
polyamide or cellulose triacetate membrane. Huge numbers of these micro-
fibers, in a bundle configuration, have their ends potted in resin, which
acts as a header to collect product water. A porous feed-distributor
tube passes along the bundle axis, dead-ending just short of the tube
sheet for the hollow fibers. Feed solution is introduced through this
tube and passes radially outward through the fiber bundle, which is
interspersed with cloth layers to maintain the bundle configuration and
promote orderly flow. Product water passes inward through the fiber walls
and is collected at the header. Concentrate is collected from around
the periphery of the bundle.
Hollow-fiber systems have about 5,000 square feet of membrane surface
per cubic foot of pressure vessel, by far the highest packing density of
any membrane configuration. However, the flow per square foot of membrane
area is lower than in other membrane systems. Hollow-fiber systems are
more vulnerable to plugging from particulates in the feedwater than are
tubular or spiral-wound systems.
Cellulose acetate membranes were the first to gain commercial accep-
tance; and there is now a variety of membranes based on cellulose acetate
but with differing substructures, degrees of acetylation, and methods of
*For convenience, customary units are used in this report. To convert
customary units to SI units, use the list of conversion factors given
on page xii.
-8-
-------
fabrication. Noncellulosic membranes of commercial importance include
aromatic polyamide (DuPont's B-9), poly-ether/amide (Universal Oil Prod-
uct's PA-300), and laminated polyethylenimine cross-linked with toluene
2,4-diisocyanate (North Star's NS-100). Dynamically formed membranes of
zirconium oxide and polyacrylic acid are also important but have not been
packaged and marketed as modules since they may be formed during the
operation itself.
The appendix to this report entitled "A User's Guide to Reverse
Osmosis" introduces the major factors which must be considered in speci-
fying or buying reverse osmosis equipment. It focuses on the most readily
available membranes, cellulose acetate and polyamide. For many industrial
applications, however, high temperature membranes, such as the PA-300 or
the dynamic membrane, should be considered. New membranes are being devel-
oped by several companies. Before applying new membrane materials, pilot
testing is necessary to determine the effects on the membrane of various
chemical and physical environments. Membrane lifetimes can only be
determined after years of tests.
In choosing the operating pressure for a reverse osmosis system,
tradeoffs must be considered between increased production at higher pres-
sures and the resultant penalties of higher pumping cost and shorter mem-
brane life. For brackish water desalting, pressures of 400 to 600 psi
are common while pressures in the neighborhood of 600 to 1,200 psi are
generally used for sea water.
Development
The osmotic phenomenon was known more than two centuries ago when
Abbe' Nollet's experiments on diffusion through animal membranes were
published in 1748. Successful tests with artificially prepared membranes
were performed by Traube in 1867.2 About another century later, Reid,
Breton, and others at the University of Florida under the sponsorship of
the Department of the Interior studied desalinization with cellulose ace-
tate membranes by using a trial-and-error approach to selecting the proper
membranes for rejecting strong electrolytes. In the early 1950's, their
work showed that the cellulose acetate membrane possessed characteristics
of salt and water permeability that made it potentially attractive as a
membrane for desalinization.3'4 Loeb, Sourirajan, and others at the Uni-
versity of California followed up the research and developed a technique
of casting a modified cellulose acetate membrane which gave water fluxes
on the order of 10 to 20 gal/(ft2-day), with salt rejections of 95 percent
or better under a driving pressure of 600 to 800 psi. Development of
this technique represented a significant advance in the technology and
paved the way for the ultimate development of practical desalinization.2'5
In the early 1960's, the Office of Saline Water supported several projects
in research and development of reverse osmosis membranes, modules, and
systems, which further advanced the technology for practical application.
In late 1962, Aerojet-General Corporation at Azusa, California,
first studied the application of reverse osmosis to treating municipal
wastewater. The bench-scale study was conducted on filtered, municipal,
secondary effluent using 3-in.-diameter flat plate test cells and operating
-9-
-------
at pressures of 750 to 1500 psi. Results of the study showed that high-
quality water could be produced from secondary effluent by reverse osmosis.
In 1967, the Orange County Water District, California, operated a plate-
and-frame reverse osmosis unit on a pilot scale for treating municipal
wastewater.6 Although they obtained results that corroborated the earlier
study, plate-and-frame modules are no longer commercially available.
Spiral-wound and tubular membrane modules were pilot-tested by Gulf
Environmental Systems Company (now the Fluid Systems Division of Universal
Oil Products) in 1969 at Pomona, California. The tests on the Gulf spiral-
wound modules indicated that the modules could process secondary effluent
with high organic and inorganic removals, but some form of pretreatment
or additional treatment of the secondary effluent was needed to prevent
excessive clogging of the concentration-side spacers with suspended solids.
Tests on the Universal Water Corporation tubular unit showed good removal
efficiencies in more than 90 percent of the tests for all parameters except
nitrate nitrogen, which was removed with 80.6 percent efficiency. Flushing
the module for one hour with an 0.75 percent solution of "Biz" detergent
increased the flux from 18.0 to 20.6 gal/(ft2-day"), compared with an
original flux of 23.9 gal/(ft2'day).6
In a laboratory study in 1969, Aerojet-General Corporation used con-
ventional cellulose acetate membranes in flat plate cells and tubular
configurations. Raw sewage, digester supernatant, and activated-carbon-
treated secondary effluent from Orange County and Pomona sewage treatment
plants were tested and found to exhibit very similar wastewater constituent
rejections.6
In early 1970, the Gulf Environmental Systems Company modified the
pilot spiral-wound reverse osmosis units to incorporate provisions for
chemical and physical cleaning. The modified units were tested on five
different feed streams: (1) primary effluent with and without sand fil-
tration, (2) sand-filtered activated sludge effluent, (3) chemically
clarified primary effluent with sand filtration, (4) chemically clari-
fied primary effluent with sand filtration and activated carbon treatment,
and (5) activated-carbon-treated activated sludge effluent. Results of
these studies indicated that spiral-wound reverse osmosis units can be
successfully operated on primary and activated sludge wastewater effluents
with only moderate pretreatment and chemical cleaning. The activated
carbon pretreatment is unnecessary on sand-filtered, activated sludge
effluent or chemically clarified, sand-filtered, primary effluent.6
A pilot demonstration program was conducted by the Eastern Municipal
Water District in Hemet, California, from March 6, 1970, to June 25, 1976,
to determine the efficiencies and costs of pretreatment operations and
to compare the performance of a selected group of reverse osmosis units
manufactured by major firms. The pretreatment included (1) chemical clari-
fication with alum in a sludge-blanket-type clarifier, (2) filtration in
pressure units with single media, 0.45- to 0.55-mm sand, (3) granular
carbon adsorption with 8 by 12 mesh carbon, (4) diatomaceous earth fil-
tration in a 30-in.-diameter unit, and (5) chlorination with sodium hypo-
chlorite. The reverse osmosis units tested included Aerojet-General
Corporation's tubular unit, DuPont's hollow-fiber modules, Gulf's spiral-
wound unit, Raypak's tubular unit, and Universal Water's tubular modules.
-10-
-------
In these tests calcium deposits, organic slimes, and scaling were believed
to be the mechanisms involved in membrane fouling.6
Pilot plant work at Lebanon, Ohio, from the late 1960's to the early
1970's tested different types of reverse osmosis units, including plate-
and-frame, hollow-fiber, and tubular modules. Lime-clarified raw waste-
water, secondary effluent, and primary effluent were treated. A tubular
reverse osmosis unit can successfully treat municipal wastewater when
membrane cleaning techniques are routinely practiced. When new membrane
tubes were first used, flux decline took place, probably as a result of
membrane compaction. The commercial bioenzyme Biz was found to be an
effective chemical cleaner, and when primary effluent was treated, the
insertion of turbulence promoters into each tube resulted in a 42 percent
increase in flux.6
Application
Early development of reverse osmosis was directed toward seawater
desalinization. In recent years, application of reverse osmosis has
accelerated dramatically in both the depth of treatment technology and
the breadth of application fields. Membranes of higher water flux and
salt rejection under a lower driving pressure were developed for specific
applications. Reverse osmosis is not only used for seawater and brackish
water desalinization but also for domestic and industrial water and waste-
water treatment applications. In municipal applications, reverse osmosis
is used to remove total dissolved solids for potable water supply and to
polish treated sewage effluents for reuse or ground water recharge. In
industrial applications reverse osmosis is used for both process water
purification and wastewater treatment, which in many cases can be con-
solidated into a single purpose, water reuse.
ULTRAFILTRATION
Fundamentals
Ultrafiltration is a membrane process that separates collodial mate-
rials and high-molecular-weight solutes such as proteins, soluble oil,
microorganisms, polymers, clays, and natural gums from their solvent,
usually water. Ultrafiltration is similar to reverse osmosis in being
driven by pressure, using membranes made from similar materials, and
operating in the same mode as far as the feed, permeate, and rejection
streams are arranged. It differs from reverse osmosis in being unable
to separate dissolved inorganic salts, using more porous membranes, and
operating under lower pressure.
Because ultrafiltration is used to separate colloidal and high mole-
cular weight materials, fouling of the membranes is often expected. To
control fouling, both the membrane material and its physical configuration
can be selected to facilitate cleaning.
Tubular configurations are generally used where fouling is expected
to be severe. High fluid velocities and large diameter (up to one inch)
tubes combine to create turbulence and scour the membrane walls. Tubular
-11-
-------
configurations also facilitate periodic mechanical cleaning and lend them-
selves to sponge ball cleaning systems similar to the Amertap system used
for cleaning power plant condensers.
Spiral-wound ultrafiltration units are somewhat more vulnerable to
fouling than tubular units. Where fouling of spiral-wound units can be
controlled by chemical cleaning or flushing, they provide a significant
savings over tubular units.
Hollow-fiber configurations are generally more vulnerable to fouling
than the tubular- or spiral-wound configurations. To counteract this,
one company (Romicon) has developed an ultrafiltration system using hollow
fibers large enough to permit turbulent flow and rigid enough to permit
backwashing.
Where fouling is expected, some method of chemical cleaning is usually
planned in conjunction with mechanical cleaning, flushing, or backwashing.
Pilot testing of the cleaning program is usually needed to establish effi-
cacy and cost. When strong chemical agents are used, it is important to
select a membrane material that will withstand repeated cleanings. Two
membrane materials which can tolerate high temperatures and strong chemi-
cals over a wide pH range are polysulfone and zirconium (IV) oxide.
Other chemically resistant ultrafiltration membranes have been developed
using proprietary materials.
Although susceptible to hydrolysis at high and low pH, cellulose
acetate membranes are widely used for ultrafiltration. These membranes
are produced in the same way as reverse osmosis membranes of cellulose
acetate. Control of the annealing temperature is used to produce mem-
branes of different porosities. Higher temperatures produce tight
membranes with low water fluxes and high solute rejection.
Manufacturers sometimes specify the porosity of a membrane in terms
of the "molecular weight cutoff," but this term tends to be misleading
since solute passage is determined not only by a molecule's size but by
its shape and charge distribution as well. This is illustrated by the
poor rejection of organics exhibited by reverse osmosis membranes; small
charged species are rejected while larger organics pass through. Never-
theless, in principal it is possible to establish a molecular weight cutoff
for a given membrane by trying a wide variety of organic solutes. Typical
molecular weight cutoffs range from 1,000 to 80,000. Membranes at the
lower end of this range can exhibit some rejection of salts, blurring
the distinction between reverse osmosis and ultrafiltration.
Pressures of 20 to 100 psi are typical for ultrafiltration systems;
in contrast, reverse osmosis systems typically employ pressures from 400
to 1,200 psi.
Development
Application of ultrafiltration, or some processes similar to ultra-
filtration, goes back to a distant past when pressurized filtration was
practiced. Early workers used cellophane or porous cellulose nitrate
-12-
-------
membranes almost exclusively. Reproducibility was poor, and adsorption
on pore walls and plugging were common.1 The systematic study of the
process and development of the membranes used did not start until inves-
tigation of reverse osmosis was under way. In the last few years, there
has been a rapid development of the ultrafiltration process, especially
in the area of industrial water and wastewater treatment.
Application
Ultrafiltration is suitable for applications in removing colloidal
and high molecular weight organic solutes from water, especially for sys-
tems in which the high rejection of ionic and low molecular weight organic
solutes is not required. Industrial applications in electroplating waste
treatment, oil-water separation, chemical recovery from waste streams,
food processing, and pharmaceutical waste treatment and water reuse have
been successfully demonstrated. A recent study by Bhattacharyya et al.7
demonstrated that ultrafiltration can be used to treat laundry and shower
wastes with good solute removal at high water recovery, producing water
suitable for nonpotable human contact uses.
ELECTRODIALYSIS
Fundamentals
Unlike reverse osmosis and ultrafiltration, which are pressure-driven,
electrodialysis separation is induced by electric currents. Cation- and
anion-selective membranes are alternately placed across the current path.
The electrically attracted cations are allowed to pass through the cation
membranes and the anions through the anion membranes, thus increasing and
decreasing ionic concentrations in alternating spaces between the mem-
branes. Reject streams, containing high concentrations of ions, and pro-
duct streams, containing low concentrations of ions, can then be separated.
The active sites in ion-selective membranes are similar to those in
ion-exchange resins. By making the polymer structure of the membrane
sufficiently dense, passage of undesired ions and water is impeded while
ions of the preferred charge diffuse through from one active site to
another. A variety of membrane materials are now available. Desirable
properties include physical strength, high selectivity, low electrical
resistance, and resistance to chemical attack.
There are three different physical configurations commonly used for
electrodialysis: the unit cell configuration and two plate-and-frame
configurations, one using sheet flow and the other using a tortuous flow
path. Each of these systems is described briefly as follows:
Unit cells are formed by sealing together anion and cation exchange
membranes, making an envelope. A number of these envelopes are placed
side by side in an array, separated by spacers which allow water to flow
between envelopes. Electrodes are placed at each end of the array, and
the entire assembly is contained in a tank provided with manifolds for
-13-
-------
feedwater distribution. Depleted water is collected in an overflow trough.
Concentrated brine is produced within the unit cell envelopes and collected
by way of small tubes which are sealed into the envelopes. This arrangement
generates a very concentrated brine since water enters the brine cells
only by diffusion through the membranes. The unit cell is less complex
than the plate-and-frame configuration but suffers from poorer flow dis-
tribution. The cells cannot be disassembled to permit cleaning. If
precipitation of dissolved salts occurs within a cell, it will swell and
must be discarded.
Plate-and-frame configurations are engineered to provide more uniform
flow conditions across the membrane but are inherently more complex. The
frame holds a large number of parallel membranes pressed between gaskets
to create a seal. Each compartment of the stack is provided with ports
for water flow in and out, with alternate compartments manifolded together.
The entire assembly can be dissassembled so that each membrane can be
individually cleaned or replaced.
Sheet flow configurations use thin membranes which are closely spaced
in order to minimize electrical resistance. Mesh or fabric spacers keep
the membranes apart and provide for fairly even flow distribution in spite
of the tendency for thin membranes to lose their shape.
To provide for greater turbulence and better control of flow patterns,
the tortuous path configuration has been developed. In this system, flow
along the membrane face is constrained to flow in a zig-zag pattern by a
spacer which provides long thin flow channels. The high turbulence attained
requires the use of thicker membranes and greater distances between them.
The increased resistance attributable to greater distance is offset by the
lessened resistance of the boundary layer. This is the layer of flow
closest to the membrane, where turbulent eddies are inhibited by the
membrane wall. These eddies enhance the diffusion of solute and reduce
the electrical resistance of the solution.
Solutions can become depleted in the boundary layer when the diffu-
sion of ions from the bulk solution does not meet the demand created by
the flow of ions through the membrane. This usually occurs first at the
anion membrane and results in the electrolysis of water. Current is then
wasted in transporting OH and H ions. A key parameter in designing an
electrodialysis system is the "limiting current density" at which water
dissociation begins to occur. This is a function of the particular solute
as well as the physical arrangement and flow rate.
Development
In 1903, Morse and Pierce introduced electrodes into the inner and
outer chambers of a laboratory dialyzer and found that electrolytic impur-
ities were more readily removed from gelatin when a voltage was applied.
Schwein used electrodialysis in purifying sugar extracts around 1900.
Later on, Pauli applied engineering design principles to reduce concen-
tration polarization to a minimum in 1924. In 1940, Meyer and Strauss
developed ion-selective membranes which were capable of operating against
-14-
-------
a concentration gradient with reasonable current efficiencies. -1 Substan-
tial commercial development of electrodialysis systems started in the
early 1960's, when the Office of Saline Water was strongly supporting a
program for desalinization of brackish water, followed by the studies on
wastewater application. From December 1962 through July 1963, secondary
sewage effluent from the city of Clinton, Massachusetts, pretreated by
cartridge filters and granular-activated carbon adsorption, was treated
by two laboratory-scale electrodialysis stacks. Fouling of the mem-
branes was detected, especially the anion membranes that absorbed
organic materials, resulting in increased resistance and decreased
capacity.6'8
In 1967, an electrodialysis pilot plant was tested in Orange County,
California. The influent was municipal wastewater, with 15 to 20 percent
industrial wastes treated by trickling filters, chemical filtration, dual-
media filtration, and granular carbon sorption. The removal efficiencies
were 22 to 30 percent for total dissolved solids. Membrane fouling was
found to result primarily from the fraction of chemical oxygen demand
represented by methylene-blue-active substances. It was concluded that
if the feedwater were adequately pretreated the performance of an elec-
trodialysis plant treating reclaimed water could be expected to parallel
a similar plant treating brackish water of the same mineral content in
terms of operating characteristics and costs.6'9
The Federal Water Pollution Control Administration operated a 60,000-
gallons per day (gpd) electrodialysis pilot plant at Lebanon, Ohio in 1967
using municipal secondary effluent treated by diatomaceous filtration and
granular carbon adsorption as the feedwater. A turbidity of less than 0.1
Jackson Turbidity Units (JTU), combined with periodic disinfection, was
found to be the best way to control fouling. The membranes were not
highly selective for any particular ion.10
In 1969, the Sanitation Districts of Los Angeles County operated a
15,000-gpd electrodialysis unit to treat municipal wastewater pretreated
by activated sludge and granular carbon sorption. The pH in the concen-
trated stream was maintained at 3.5 by injection of sulfuric acid. Average
reduction was about 34 percent for total dissolved solids, 15 percent
for chemical oxygen demand, 43 percent for ammonia, 50 percent for nitrate,
and 23 percent for phosphate. The membranes were cleaned every second day
by flushing with an enzyme detergent, tap water, and acid.6'11
From March 1970 through June 1971, a 50,000-gpd electrodialysis pilot
plant in Santee, California, treated municipal wastewater that had received
activated sludge treatment followed by chemical clarification, dual-media
filtration, and granular carbon adsorption. Removal efficiencies were
50 percent for total dissolved solids, 32 percent for chemical oxygen
demand, 32 percent for nitrate-nitrogen, 22 percent for phosphate-
phosphorus, 66 percent for calcium, and 68 percent for chloride. Flush-
ing with acid, washing with enzymatic detergent, and injecting air were
found to be helpful in mitigating fouling and increasing resistance. The
deposit on the cation membranes was found to contain a higher proportion
of inorganic matter than the deposit found on the anion membranes.6
-15-
-------
Application
Electrodialysis has been actively used to desalt whey; to "sweeten"
citrus juices; to desalt fishmeal wastewater; to recover chemicals such
as carboxylic acid, pulping waste chemicals, and chromates; to denitrify
agricultural runoff waters; and to desalt brackish water, which is by
far the principal application. Its application in various salt removal
or brine concentration processes such as cooling tower effluent reduction
has also been demonstrated.12 In Japan, the method of manufacturing salt
has completely changed from the salt field method to the electrodialysis
process since the end of 1971.6 However, in the area of wastewater treat-
ment, the profitability of the electrodialysis equipment has not been
sufficient to support a large number of manufacturers and suppliers, as
compared with reverse osmosis and ultrafiltration.
SIMILARITIES AND DIFFERENCES
Similarities
Reverse osmosis, ultrafiltration, and electrodialysis have several
similarities:
1. Permeation of materials through semipermeable membranes
achieves separation.
2. Feed, reject, and product streams are arranged in such a
fashion that continuous operation is attainable.
3. There is no phase change.
4. No heat is needed for the separation process.
5. Boundary-layer effects or concentration polarization tend to
occur.
6. Membrane fouling is possible.
Differences
There are several differences between reverse osmosis and ultrafil-
tration versus electrodialysis:
1. Reverse osmosis and ultrafiltration are driven by pressure,
whereas electrodialysis is driven by electrical potential.
2. Reverse osmosis and ultrafiltration separate colloidal materials
and high-molecular-weight solutes, whereas electrodialysis can
only transfer ions, leaving colloidal and high-molecular-weight
solutes in the diluted stream.
-16-
-------
3. The power required for reverse osmosis and ultrafiltration is
approximately proportional to the product water flow, but is
not very sensitive to variations in the dissolved solids con-
centration of the feedwater. In contrast, the power required
for electrodialysis is nearly proportional to the amount of
salt removed but is only slightly sensitive to the flow rate.
4. Reverse osmosis and ultrafiltration can typically remove from
90 to 99 percent of the solute in one stage. Electrodialysis
can typically remove from 25 to 60 percent per stage.
5. Electrodialysis can concentrate brine to between 10 and 25 per-
cent solids (100,000 to 250,000 ppm) in one stage with feed
solutions as low as 1,000 ppm. In contrast, ultrafiltration
and reverse osmosis can concentrate feedwaters by a factor of
less than 10 per stage and reverse osmosis is limited to a
maximum concentration of approximately 70,000 ppm.
Contrasting reverse osmosis and ultrafiltration, reverse osmosis is
designed to remove dissolved ions whereas ultrafiltration is designed to
remove organics and particulates. Of course, reverse osmosis will remove
many organics and particulates at the expense of fouling the membrane.
To control fouling, ultrafiltration systems are generally designed for
ease of cleaning by either mechanical or chemical means. Because of the
more porous membranes used in ultrafiltration, typical pressures are
between 20 and 100 psi. In contrast, typical reverse osmosis pressures
range from 400 to 1,200 psi.
APPLICATIONS TO POWER GENERATION
Reverse Osmosis
Reverse osmosis is being used to produce high-quality boiler feed-
water. When reverse osmosis is used as a pretreatment process ahead of
mixed-bed ion exchange units, the demineralizers produce five to ten
times more deionized water between regenerations; manpower requirements
for operation and maintenance are reduced; and the reliability of water
quality is improved. The chemical requirements are also lowered by 90
to 95 percent, and the life of ion exchange resins is extended,13 which
reduces costs and minimizes waste disposal problems. Raw water with a
minimum concentration of total dissolved solids of 350 to 650 ppm is one
criterion, in terms of costs, to justify using reverse osmosis as a pre-
treatment for deionization instead of a complete ion exchange system.
However, if costs of wastewater treatment to meet the future stringent
standards are considered, reverse osmosis application could be justified
more easily.
Reverse osmosis has recently been applied in water and wastewater
treatment at power plants. Applications included production of boiler
feedwater, treatment of wastewater at steam plants, and treatment of
liquid radwaste at nuclear plants.
-17-
-------
In late 1971, Carolina Power and Light Company made provisions for
the addition of a 53-gpm reverse osmosis system for radwaste application
at the Brunswick Steam Electric Plant. This reverse osmosis system was
provided to treat medium and low chemical priority wastewater associated
with floor drains and demineralizer regeneration wastes.14 The design
of this system was based on pilot studies conducted at their H. B.
Robinson Nuclear Plant.
A 250-gpm product capacity reverse osmosis system was operated at
the Harrison Power Plant in 1972, mainly for supply of condensate makeup
water. The station is jointly owned by the Potomac Edison Company, West
Perm Power Company, and Monongahela Power Company. It was reported that
the reverse osmosis unit offered advantages by reducing chemical opera-
ting cost and complexity of control and required less floor space than a
multistep ion exchange system.15
A 2000-gpd prototype Westinghouse reverse osmosis unit was installed
and operated at the Ginna Power Station of the Rochester Gas and Electric
Company in 1972 to treat the plant's laundry waste. The reverse osmosis
unit was capable of removing a minimum of 90 percent of the biological
oxygen demand (BOD), phosphate, and radionuclides while reducing the
waste volume ratio by a minimum of 100 to I.16
A 200-gpm reverse osmosis system was installed in series with the
existing 200-gpm makeup demineralizers at the Willow Glen Power Station
of Gulf States Utilities Company in 1973. The advantages of reverse
osmosis includes (1) lower overall operating costs, (2) greater overall
capacity, (3) reduced manpower requirements, (4) removal of organic matter,
(5) longer resin life, and (6) better quality effluent 17
In 1974, a 2500-gpm reverse osmosis (ROGA spiral-wound module) pilot
plant was tested at the H Power Station of Karaso Electric Power Company
in Japan to treat the power plant wastewater for reuse. Wastewater, elec-
trostatic precipitator washing water, oily waste, and other continuous
wastewater discharges were tested. It was concluded that the reverse
osmosis process following the coagulation and filtration pretreatment
was extremely reliable for power plant wastewater treatment.18
A 900-gpm, spiral-wound reverse osmosis system and other pretreat-
ment processes were added upstream of a vapor-compression evaporator in
1976 at the San Juan Power Station owned jointly by Public Service Company
of New Mexico and Tucson Gas and Electric Company. The reverse osmosis
unit was selected to reduce the volume of waste going to the vapor-
compression evaporator by a factor of five. The product waste from the
reverse osmosis process was recycled for use as demineralizer input, as
desulfurization-system makeup water, and in other station applications.
This installation has reduced the normal cost of producing boiler makeup
and has increased demineralizer-system capacity at little additional cost.19
A 75-gpm reverse osmosis unit was installed at North Lake Power Sta-
tion of Dallas Power and Light Company in 1976 to produce boiler makeup
water. It was reported that the installation was still free of significant
problems after one year of operation.17
-18-
-------
In 1977, a 400-gpm reverse osmosis system with cellulose-triacetate
hollow-fiber membranes was placed in service at Coffeen Power Station of
Central Illinois Public Service Company. The reverse osmosis unit was
able to treat the lake water and to supply sufficient quality and quan-
tity of boiler makeup water at the plant.20
When reverse osmosis is used to treat wastewater, the permeate is
generally reused rather than discharged, because its quality is superior
or comparable to available makeup water. Product water may be used for
makeup to virtually any plant water system. This report considers the
possible use of reverse osmosis to treat recirculating cooling water,
ash sluice water, coal pile drainage, boiler blowdown, metal cleaning
wastes, wet sulfur dioxide scrubber wastes, radioactive wastes, and oily
and miscellaneous wastes.
Ultrafiltration
Floor washings from power plants that contain dilute concentrations
of suspended solids, colloidal materials, oil, grease, and other constitu-
ents can be treated by ultrafiltration, after which the product water
can be reused for ash transport water, additional floor washing require-
ments, and other applications. The resultant concentrate can be reduced
in volume for ease of storage or disposal. Cooling tower blowdown can
be treated with ultrafiltration, either alone or with reverse osmosis
depending on the blowdown water quality. The high-quality water result-
ing from ultrafiltration plus reverse osmosis could be suitable for boiler
feedwater, whereas the lower-quality water resulting from ultrafiltra-
tion alone can be used for washing operations. Boiler blowdown can also
be treated with ultrafiltration, followed by reverse osmosis and ion
exchange to provide product water for boiler feed.
A Romicon ultrafiltration pilot system was installed at the Ravens-
wood Power Station of Consolidated Edison Company to evaluate membrane
flux and nonreactive silica rejection on untreated municipal water supply
and deionized municipal water. The results showed that the ultrafiltra-
tion system was a very attractive pretreatment for boiler makeup
demineralizers.2 *
At nuclear plants, ultrafiltration may be used to remove colloidal
materials from radioactive waste streams. This would not only remove
some of the radioactive materials but would also prevent fouling of ion-
exchange resins or evaporators located downstream. Fouled resin is expen-
sive to replace, and being radioactive it is also expensive to dispose.
Fouled evaporators are difficult to maintain and have lower decontamina-
tion efficiency. This results in higher radiation doses to employees
and to the downstream population.
Ultrafiltration of radioactive equipment drain wastes has been suc-
cessfully demonstrated at the Tsuruga22 generating plant (Japan). An
EPRI sponsored program at the R. E. Ginna Nulcear Plant of Rochester Gas
and Electric is now in progress to demonstrate the use of ultrafiltration
as a pretreatment for radwaste evaporators.14
-19-
-------
In summary, ultrafiltration can be used as a pretreatment to reverse
osmosis for either water or wastewater treatment to produce high quality
water for reuse, or it can be used alone for wastewater treatment to pro-
duce lower-quality water for suitable reuse purposes.
Electrodialysis
Electrodialysis has been shown to be applicable to cooling tower
effluent reduction when it is used to remove and concentrate dissolved
solids from a sidestream.12 Similarly, electrodialysis may be applicable
to the control of total dissolved solids in other recirculating waters
such as ash sluice water.
Electrodialysis has been promoted for use in preparing boiler feed-
water, but is at a disadvantage compared with reverse osmosis because it
does not remove particulates and organics which foul ion exchange polish-
ing resins. Also, several electrodialysis stages are necessary to produce
water comparable with reverse osmosis permeate.
Separation of solute anions and cations can be accomplished in a
special three-compartment electrodialysis stream. Feedwater enters the
central compartment where the cations and anions begin moving toward their
respective electrodes. At the cathode, hydroxides are formed by the elec-
trolysis of water with hydrogen given off as a byproduct. The anode reac-
tion forms hydrogen ions (H ) giving off oxygen as a byproduct. A system
of this type has been proposed for separating valuable lithium-7 from
nuclear reactor waste streams.23 The same concept24 is also used in a
proposed regenerable SC>2 scrubber system where the spent scrubber liquor,
a solution of alkali and sulfate salts, is split into an alkali stream
and a sulfuric acid stream. The alkali stream can then be reused in the
scrubber. The sulfuric acid is recovered for other uses.
-20-
-------
SECTION 5
REGULATIONS APPLICABLE TO POWER PLANT WATERS
The principal wastewaters from power generation are cooling waters,
ash sluice waters, coal pile drainage, demineralizer regeneration wastes,
boiler blowdown, metal cleaning wastes, sulfur dioxide scrubber wastes,
radioactive wastewaters, and miscellaneous wastes such as oily wastes
and floor drainage. In general, the characteristics of wastewaters vary
with the types of plants and are governed by the sources and types of
fuel used, plant configuration, process operation conditions, and feed-
water quality.25 In planning for pollution control, management must be
cognizant of a wide variety of regulatory requirements applicable to
these wastewaters and power plant discharges. These include regulations
on discharges of dissolved and suspended materials as well as regulations
on thermal discharges and releases of radioactivity. Thermal discharges
are not considered in this study.
DISSOLVED AND SUSPENDED MATERIAL
Pursuant to the Federal Water Pollution Control Act of 1972 (PL 92-
500), as amended by the Clean Water Act of 1977 (PL 95-217), and other
Federal laws, and commonly referred to as the Clean Water Act (CWA),
the U.S. Environmental Protection Agency (EPA) has established effluent
limitations, guidelines, and new source performance standards for the
point-source category of steam-electric power generating plants.26 A sum-
mary of the applicable standards, set forth in 40 CFR part 423, is pre-
sented in Table 2. In general, limitations are set for free and total
residual chlorine in cooling water discharges; for chromium, zinc, phos-
phorus, and other corrosion-inhibiting materials in cooling tower blow-
down; and for suspended solids, oil, and grease in ash sluice water,
boiler blowdown, metal cleaning wastes, and other low-volume wastewaters.
Suspended solids are also limited in coal pile drainage and in effluents
from construction activities. Concentrations of copper and iron are
limited in boiler blowdown and metal cleaning wastes. In all sources,
the pH is limited to the range of 6.0 to 9.0 (except once-through cooling
water), and no polychlorinated biphenyls (PCB's) should be detectable.
Through the National Pollutant Discharge Elimination System (NPDES) permit
issuing program, these standards are applied.
The various state and local governments may have their own water
quality and effluent standards for discharges, which are equal to or more
stringent than those established by EPA.
-21-
-------
TABLE 2. STEAM-ELECTRIC POWEK GENERATING POINT SOURCE CATEGORY NONTHERMAL LIMITATIONS
,a,b
Waste Streams and Pollutants
All waste streams
pH (except once-through cooling)
PCB's
Best Practicable Control
Technology Currently
Available
(BCT)
maxc avg
6.0-9.0
No Discharge
Best Available
Technology Economically
Achievable
(BAT)
c d
max avg
6.0-9.0
No Discharge
New Source
Performance
Standards
(NSPS)
c d
max avg
6.0-9.0
No Discharge
Low Volume Waste Streams
T.S.S.
Oil Grease
Ash Transport Water
T.S.S.
Oil and Grease
Bottom Ash Transport Water
T.S.S.
Oil and Grease
Fly Ash Transport Water
T.S.S.
Oil and Grease
Metal Cleaning Wastes
T.S.S.
Oil and Grease
Copper (Total)
Iron (Total)
Boiler Slowdown
T.S.S.
Oil and Grease
Copper (Total)
Iron (Total)
Once-Through Cooling Water
Free Available Chlorine8
Cooling Tower Slowdown
Free Available Chlorine8
Zinc
Chromium
Phosphorus
Other Corrosion Inhibitors
Area Run-off Subcategory
T.S.S.
pH
Pretreatment Standards1
Free Available Chlorine
Total Residual Chlorine
100 mg/£
20 mg/£
100 mg/£
20 mg/£
30 mg/£
15 mg/£
30 mg/£
15 mg/£
100 mg/£
20 mg/£
30 mg/£
15 mg/£
100 mg/£
20 mg/£
30 mg/SL
15 mg/SL
100 mg/SL
20 mg/H
1 mg/£
1 mg/£
100 mg/i
20 mg/t
0.5 mg/£
0.5 mg/£
30 mg/£
15 mg/£
1 mg/4
1 mg/£
30 mg/SL
15 mg/i
1 mg/£
1 mg/£
0.2 rog/£
0.2 mg/£
<50 mg/£
6.0-9.0
No Limitation
No Limitation
e e
100 mg/£ 30 mg/£
20 mg/£ 15 mg/SL
100 mg/£ 30 mg/£
20 mg/£ 15 mg/l
100 mg/£ 30 mg/£
20 mg/SL 15 mg/SL
1 mg/£ 1 mg/£
1 mg/£ 1 mg/JZ
100 mg/SL 30 mg/£
20 mg/SL 15 mg/SL
1 mg/l 1 mg/£
1 mg/£ 1 mg/£
0.5 mg/£ 0.2 mg/£
0.5 mg/£ 0.2 mg/£
1 mg/£ 1 mg/£
0.2 mg/£ 0.2 mg/£
5 mg/£ 5 mg/£
Case by Case Limit
<50 mg/£
6.0-9.0
No Limitation
No Limitation
f f
100 mg/£ 30 mg/£
20 mg/£ 15 mg/£
No Discharge
No Discharge
100 mg/£ 30 mg/£
20 mg/l 15 mg/l
1 mg/£ 1 mg/l
1 mg/£ 1 mg/SL
100 mg/l 30 mg/£
20 mg/SL 15 mg/SL
I mg/SL 1 mg/SL
1 mg/SL 1 mg/SL
0.5 mg/SL 0.2 mg/£
0.5 mg/£ 0.2 mg/£
No
Discharge of
Corrosion
Inhibitors
<50 mg/£
6.0-9.0
No Limitation
No Limitation
"low multiplied by the concentration limitation.
,
Where waste streams from various sources are combined for treatment or discharge, the quantity of each pollutant
attributable to each waste source shall not exceed the specified limitation for that waste source.
b. All sources must meet State Water Quality Standards by 1977 [Section 301(b)(l)(c)].
c. Maximum for any one day.
d. Average of daily values for 30 consecutive days.
e. Allowable discharge equals flow multiplied by concentration divided by 12.5.
f. Allowable discharge equals flow multiplied by concentration divided by 20.0.
g. Limits given are maximum and average concentration.
Neither free available chlorine nor total residual chlorine may be discharged from any unit for more than 2 hours in
one day and not more than one unit in any plant may discharge free available or total residual chlorine at any one
unless the utility can demonstrate that the units in a particular location cannot operate at or below this level of chlorine.
h. Limits were remanded by the Fourth Circuit Court of Appeals in July 1976.
i. Applies only to sources which discharge to publicly owned treatment works. These discharges are also subject to the
provisions of 40 CFR part 128.
-22-
-------
Under Section 307 of the CWA, EPA was also required to publish efflu-
ent standards for toxic pollutants, with special attention to the 65 priority
pollutants listed in Table 3. These, as well as other potentially toxic
substances, are likely to be the focus of regulatory action in the future,
and the power generation utility industry could shortly be faced with
meeting potentially strict discharge limitations for the toxic pollutants
found in its wastewater discharges. For instance, EPA has recently stated
(45 Fed. Reg. 16,832) that:
The Clean Water Act and a modified consent decree in NRDC
v. Costle, 12 ERG 1833 (D.D.C. 1979), require that EPA
develop guidelines to control toxic substances in industrial
effluents. Section 307(a) of the Act identifies 65 toxic
pollutants; they are listed in Table 1 of Committee Print
95-30 of the Committee on Public Works and Transportation,
House of Representatives.
Section 304 requires that EPA determine the best available
technology (BAT) to control toxic pollutants from existing
point sources. BAT will consist of the most effective
technology which can still be economically achieved by
the affected industries. EPA will also determine best
conventional technology (BCT) which industries can use on
conventional pollutants which do not require BAT.
Under Section 306 of the Act, EPA is establishing new
source performance standards (NSPS) for new plants. Under
Sections 307(b) and 307(c), EPA will set pretreatment stan-
dards for both existing and new sources which discharge
into municipal waste treatment systems. These sets of
standards will in most cases require technologies equivalent
to BAT.
Major issues raised in setting effluent guidelines are:
(1) Identification of the major pollutants discharged to
and from treatment systems;
(2) Determination of major technology options to control
these pollutants;
(3) Determination of the capital and annual costs of the
technology options; and
(4) Determination of the resulting economic impacts.
As a consequence, EPA is developing new guidelines for many indus-
tries, including the steam-electric power generating point-source category.
Final modifications to 40 CFR part 423 are expected in December 1980.
Also, EPA is considering a major revision of water quality standards.
This regulation may require states to adopt water quality standards for
some toxic pollutants covered by ambient water quality criteria. One
effect of this will be that dischargers (both municipal and industrial)
-23-
-------
TABLE 3. LIST OF 65 TOXIC (PRIORITY) POLLUTANTS
1. Acenaphthene
2. Acrolein
3. Acrylonitrile
4. Aldrin/Dieldrin
5. Antimony and compounds*
6. Arsenic and compounds
7. Asbestos
8. Benzene
9. Benzidine
10. Beryllium and compounds
11. Cadmium and compounds
12. Carbon tetrachloride
13. Chlordane
14. Chlorinated benzenes
15. Chlorinated ethanes
16. Chloralkyl ethers
17. Chlorinated naphthalene
18. Chlorinated phenols
19. Chloroform
20. 2-chlorophenol
21. Chromium and compounds
22. Copper and compounds
23. Cyanides
24. DDT and metabolites
25. Dichlorobenzenes
26. Dichlorobenzidine
27. Dichlorcethylenes
28. 2,4-dichlorophenol
29. Dichloropropane and
dichloropropene
30. 2,4-dimethylphenol
31. Dinitrotoluene
32. Diphenylhydrazine
33. Endosulfan and metabolites
34. Endrin and metabolites
35. Ethylbenzene
36. Fluoranthene
37. Haloethers
38. Halomethanes
39. Heptachlor and metabolites
40. Hexachlorobutadiene
41. Hexachlorocyclohexane
42. Hexachlorocyclopentadiene
43. Isophorene
44. Lead and compounds
45. Mercury compounds
46. Naphthalene
47. Nickel and compounds
48. Nitrobenzene
49. Nitrophenols
50. Nitrosamines
51. Pentachlorophenol
52. Phenol
53. Phthalate esters
54. Polychlorinated biphenyls
55. Polynuclear aromatic
hydrocarbons
56. Selenium and compounds
57. Silver and compounds
58. 2,3,7,8-Tetrachlorodibenzo-
p-dioxin
59. Tetrachloroethylene
60. Thallium and compounds
61. Toluene
62. Toxaphene
63. Trichloroethylene
64. Vinyl chloride
65. Zinc and compounds
*The term "compounds" shall include organic and inorganic compounds.
-24-
-------
may have to install treatment technology beyond that required by best
practical wastewater treatment technology or best available technology
(BAT) guidelines. A final rule is expected in May 1981.
In the same vein EPA is preparing additional quality criteria for
toxics substances released into water. The states will refer to this
guidance when they establish water quality standards. This guidance
will be an important agency decision in the area of toxics. September
1980 is the anticipated date of its release.
Where toxics do pose a problem, membrane technologies can be used
to recondition water for internal reuse. The reconditioned water may
but need not necessarily meet discharge standards. Nevertheless, an
assessment of membrane technology for treating wastewater streams for
removal of toxic substances has been prepared and appears in Appendix B.
The results of this assessment indicate the technical feasibility of
removing priority pollutants from wastewater by membrane processes. Since
water reuse is a national goal of high priority, the balance of this report
shows how membrane technologies can assist in meeting this goal.
RADIOACTIVE WASTES
The release of radioactive wastes from nuclear power plants is regu-
lated by the Nuclear Regulatory Commission (NRC). The Code of Federal
Regulations, Title 10 Part 20, (10 CFR part 20) sets maximum concentra-
tions of various radionuclides that will be permitted in the air and water
outside of restricted areas. It also sets maximum dose rates to people
both inside and outside restricted areas. At the same time, each nuclear
plant must reduce radioactive releases below these levels to the extent
practicable. This requirement is set forth in 10 CFR part 50 which
requires that releases be "as low as is reasonably achievable" (ALARA).
Appendix I of 10 CFR part 50 gives guidelines for determining ALARA com-
pliance. The key requirement is that applicants shall "include in the
radwaste system all items of reasonably demonstrated technology that,
when added to the system, sequentially and in order of diminishing return,
can, for a favorable cost-benefit ratio, effect reductions in dose to
the population reasonably expected to be within 50 miles of the reactor."
In calculating the benefits from dose reductions, a value of $1000/man-rem
is to be used.
Several regulatory guides have been published by the NRC to aid in
implementing Appendix I of 10 CFR part 50. United States Nuclear Regu-
latory Commission Reports27'28 can be used to estimate releases. Dis-
persion in aquatic environments is estimated using Regulatory Guide
1.113.29 Dose rates can then be calculated using Regulatory Guide
1.109.30 Finally, the cost-benefit analysis is performed in accordance
with Regulatory Guide 1.110.31
Since Regulatory Guide 1.110 deals only with "reasonably demonstra-
ted technology" and includes reverse osmosis data from only one system
(the laundry waste system at R. E. Ginna Nuclear Plant), designers who
rely on this guide would be inhibited from innovative uses of membrane
technology. The guide does not prohibit innovation but does recommend
-25-
-------
costs and decontamination factors that are to be used when determining
the cost-benefit ratio for purposes of Appendix I of 10 CFR part 50. If
other numbers are used, the burden of proof falls on the applicant.
-26-
-------
SECTION 6
RECIRCULATING CONDENSER COOLING WATER
CHARACTERISTICS OF WASTEWATER
This discussion will not address once-through cooling systems, since
the discharge from these systems is essentially natural surface water,
altered only by the addition of chlorine or other biocides and occasionally
by pH control within the range of 6 to 9.32 Corrosion inhibitors are
not generally used in once-through cooling water. With few exceptions,
thermal effluent regulations now require recirculating cooling systems
on all new power plants.
Cooling water quality is a function of the material flows in and
out of the recirculating cooling water system. Sources of dissolved and
suspended material include makeup water, chemical additives, air contami-
nates, and corrosion products. Outflows of dissolved and suspended mate-
rials include blowdown, droplets (drift) entrained in the cooling tower
exhaust air and sludges and brines from various water treatment processes.
Sulfur dioxide and other chemicals as well as particulates are inad-
vertently scrubbed from the air in the cooling tower. By assuming an
air analysis and 100-percent scrubbing efficiency, conservative estimates
of this effect can be made. Similarly, by assuming a typical corrosion
rate, the concentration of corrosion products, such as copper, zinc, and
nickel, can be estimated.33 Although these inadvertent sources are small
in magnitude, they can result in significant concentrations in the recir-
culating cooling water system if blowdown flows are small.
Makeup water is generally the largest source of incoming dissolved
and suspended solids. If the cooling systems water volume is to remain
constant, makeup flows must equal the losses from evaporation, drift,
and blowdown. Usually untreated surface water is used for makeup, although
ground waters and treated wastewaters are sometimes used. Where available
waters are extremely hard, makeup water is sometimes softened.
Management of cooling water chemistry is necessary to prevent the
precipitation of scale on heat transfer surfaces. Potential sealants
include calcium carbonate, calcium sulfate, calcium phosphate, and vari-
ous magnesium and silica compounds. Typical operating limits for control
of scaling are given in Table 4. Control methods include the use of sof-
tening, pH control, sequestering agents, and control of blowdown.
To prevent scaling from calcium, magnesium, or silica salts, softening
may be applied to either the cooling tower makeup water or a small portion
-27-
-------
TABLE 4. CONTROL LIMITS FOR RECIRCULATING WATERS
Constraint
Remarks
pH and hardness
Langelier Saturation Index
a,b
I = pH - pH =0.0 to 1.0
Sow S
Ryznar Stability Index
I = 2pH - pH = 6.0 to 7.0
S3T- S
pH = measured pH
pH = pH at saturation
with CaCO
Sulfate and calcium
(Cgo )(CCa) = 200,000, where IS
(Cgo )(CCa) = 460,000, where IS
4
(Cgo )(CCa) = 730,000, where IS
4
Magnesium and silica
= 0.01
= 0.05
= 0.1
(CMg)(CSi) = 8'5°5
CSi0
"SO,
concentration of
SO, in mg/£
C., = concentration of
Ca in mg/JH
IS = ionic strength
= concentration of
Mg in mg/£
= concentration of
Si02 in mg/A
a. Chen, Y.S., J. L. Petrillo, and F. B. Kaylor "Optimal Water Reuse in
Recirculating Cooling Water Systems for Steam-Electric Generating
Systems." In Proceedings, Second National Conference on Complete
Water Reuse, Chicago, Illinois, May 1975. Sponsored by American
Institute of Chemical Engineers, New York. pp. 528-541.
b. Ryznar, J. W. "A New Index for Determining Amount of Calcium
Carbonate Scale Formed by a Water." Journal American Water Works
Association, Vol. 36, No. 4, April 1944. pp. 472-486.
c. Selmeczi, J. G. and J. P. Miller. "Supersaturation of Calcium
Sulfate in Water." Paper presented at American Chemistry Society,
Great Lakes Regional Meeting. Purdue University. West Lafayette,
Indiana. June 4, 1974. 27 pp.
d. Nelson, E. R. "Water Recycle/Reuse Possibilities: Power Plant
Boiler and Cooling Systems." EPA-660/2-74-089. December 1974.
51 PP.
-28-
-------
(sidestream) of the recirculating cooling water flow. Comparing makeup
and sidestream softening, sidestream softening provides for the cooling
system to preconcentrate the flow, thus permitting a smaller softening
system. Nevertheless, where makeup waters are very high in hardness,
makeup softening may be advantageous. At the site of the proposed Sun-
desert Nuclear Plant, using agricultural runoff as makeup, recent pilot
plant tests showed that superior results can be obtained by combining
the makeup with a sidestream before softening.34
Control of pH is helpful in preventing calcium carbonate scale and
controlling corrosion. For control of carbonate scale acid is usually
added to lower the alkalinity and to convert carbonate ions to bicarbonate.
Sulfuric acid is usually used because of its low cost, but the increased
threat of calcium sulfate scaling must be considered. Although maintaining
a low pH helps control carbonate scale, too low a pH will result in increased
corrosion and should be avoided. It is seldom necessary or advisable to
reduce the pH below 7.
Some antiscaling agents act to distort the crystal structure of scale
deposits, slowing their growth. Commonly used agents include polyphospho-
nates, polyacrylates, and natural organics. Sequestering agents such
as sodium hexametaphoshate can combine with potential sealants to prevent
their deposition. These treatments are generally designed to delay the
deposition of scale, causing it to be soft, amorphous, and more easily
removed should deposition occur. Additives of this type would presumably
interfere with the operation of softening systems.
Typically, the major outflow of dissolved solids is through the blow-
down. By increasing the makeup and blowdown flows, the concentration of
all major ions can be reduced. This in turn reduces the potential for
scale deposition. Other outflows of both water and dissolved solids may
occur as sludges or brines from various treatment schemes. With proper
design, these flows may be adequate to maintain a mass balance without
blowdown. It is clear that the total mass of dissolved solids leaving
the system must (on an average) equal that entering. Thus, some pro-
vision must be made to remove this dissolved solids loading.
High concentrations of dissolved solids can lead to increased cor-
rosion rates because high conductivity increases the rate of electrolytic
corrosion and because high concentrations of particular ions can cause
corrosion. Chlorides are particularly corrosive to stainless steels,
and sulfates can create cracking in some concretes.
Corrosion control is sought through the use of chemical additives,
pH control, sacrificial anodes, and corrosion resistant materials. Common
additives are based on zinc, chromium and phosphate compounds.
Biocides are commonly used to control the growth of slime on heat
transfer surfaces. Chlorine is the most widely used, but a number of
proprietary biocides are also available.
-29-
-------
ALTERNATIVE TREATMENTS
Reverse osmosis can be used to produce high quality water from cool-
ing tower blowdown in a single stage and to produce a concentrated brine
by using several stages. In this role, reverse osmosis is similar to
evaporative brine concentrators. In contrast, electrodialysis requires
several stages to produce high quality water, but only one stage to pro-
duce a highly concentrated brine. Because electrodialysis is capable of
producing a more concentrated brine than reverse osmosis, brine disposal
costs can be reduced. The primary purpose of each of these salt removal
methods is to control the total dissolved solids concentration in the
cooling system and to avoid corrosion.
Conventional blowdown will serve the same purpose if it can be dis-
charged. Conventional methods of treating blowdown for discharge should
be considered. These include pH control and aeration to precipitate metals
and the use of higher blowdown rates to reduce their concentration. Coagu-
lation, settling, and filtration can be used for suspended solids removal.
(Ultrafiltration is not cost effective in this role.) Where conventional
treatments fail to produce an acceptable discharge, water reuse must be
considered.
One water reuse option is to allow sidestream softening sludge and
cooling tower drift to serve as the only blowdown, resulting in very high
concentrations of dissolved solids. Corrosion resistant materials can
be used to reduce the need for controlling dissolved solids.
Electrodialysis
By treating a sidestream of the recirculating cooling water, elec-
trodialysis can be used to maintain dissolved solids at a low level in
the cooling system while producing a low volume of concentrated brine.
Brine must generally be stored on site, although other disposal methods
are sometimes available. Pretreatment by softening and filtration is
necessary to prevent scaling of the membranes. The softening step also
serves to prevent scaling within the cooling system.
According to Jordan et al.,12 numerous case studies have shown that
electrodialysis is less expensive than alternative systems for salt removal
and concentration. A study by Blackburn35 shows that at salt removal
rates and concentration ranges applicable to power plant cooling towers,
electrodialysis is one of the most cost-effective processes for salt removal.
In fact, for waters with dissolved solids concentrations less than 10,000
ppm (one percent), his data shows electrodialysis to be much less expensive
than evaporative brine concentrators.
In spite of this, vapor-compression distillation has been selected
over electrodialysis for cooling water brine concentration at several
recent plants designed for zero liquid discharge.36 Apparently, the lack
of experience with electrodialysis installations of this type has been
an important factor. Utilities require reliable technologies with proven
operability. Although electrodialysis has been used for desalting water
supplies in many communities throughout the United States, large-scale
-30-
-------
use of electrodialysis to concentrate brine has been primarily limited
to Japan, where it is used to concentrate salt from sea water.37
A large-scale demonstration of electrodialysis to remove salt from
cooling water would be useful to answer questions about the reliability
and cost of such systems in the United States. Design of a large-scale
system can be accomplished after key parameters, such as cell resistance
and limiting current density, are established in small-scale tests.
Reverse Osmosis
Reverse osmosis can be used to treat cooling water for discharge or
for reuse. Because reverse osmosis permeate is generally of high quality,
it would usually be reused. The brine, however, is more concentrated
than the feed and often must be stored indefinitely or disposed at high
cost. If this is the case, overall cost is reduced by concentrating the
brine as much as possible, thus reducing its volume.38
Brine concentration by reverse osmosis is usually limited first by
calcium sulfate deposition and then by osmotic pressure increases. To
minimize the blowdown flow, cooling water will usually be concentrated
to the maximum extent possible without the precipitation of scale. What-
ever the potential scale forming species may be, further concentration
in the reverse osmosis process will lead to scaling of the membranes
unless precautions are taken. Typically, softening would be used to pre-
vent calcium and magnesium scaling. Scale inhibiting chemicals may also
be used. Otherwise, water recovery must be limited to ensure that solu-
bility limits are not exceeded in the brine.
Osmotic pressure also limits water recovery. System pressure must
substantially exceed the osmotic pressure in order to maintain practical
premeate productivity rates. Reverse osmosis modules designed for brack-
ish water desalting are limited to operating pressures below 400 to 600 psi
depending on the module; and practical brine concentrations are in the
neighborhood of 15,000 to 20,000 ppm. Higher concentrations, up to about
70,000 ppm, are feasible in modules designed for sea water desalting.
Cooling water is, for the most part, concentrated natural waters,
and there is a large data base on the treatment of natural waters. Never-
theless, treatment of cooling water is different in that chemicals may
have been added to reduce scaling or corrosion and corrosion products
may be present. Also, the problem of brine concentration and disposal
is more significant.
Reverse osmosis has been proven to be applicable in treating cooling
water for reuse.39'33
Chian and Fang's study40 indicated that cellulose acetate and aro-
matic polyamide membranes satisfactorily removed chemicals from cooling
tower blowdown. At 67 percent product recovery, permeates from both mem-
branes contained a smaller amount of chemicals than those found in munici-
pal water used for the makeup feedwater. A higher product water recovery
should be used in practice to an extent limited only by the solubility
of calcium sulfate.
-31-
-------
TVA conducted a preliminary study on treating cooling tower blowdown
by reverse osmosis. One blowdown sample obtained from a steam plant in
the southwest was filtered through 1- to 3-|Jm fiber filters for suspended
solids removal, and then the blowdown was tested by four reverse osmosis
modules: DuPont Spiral-Wound Polyamide membrane, Dow Hollow-Fiber
Cellulose Triacetate membrane, Roga Spiral-Wound Cellulose Triacetate
membrane, and DuPont Hollow-Fiber Polyamide membrane. The testing results
based on 10 percent volume recovery are shown in Table 5 and indicate a
high percentage removal of chromium, zinc, and phosphate, commonly used
corrosion inhibitors.
Kasarek39 proposed a calcium reduction system in addition to the
reverse osmosis system to facilitate further water recovery. The reverse
osmosis system consists of a holding pond to equalize the flow and to
cool the tower blowdown; a blend tank to control temperature and pH and
to feed antiscalent; a cartridge filter; and the reverse osmosis units.
The calcium reduction system consists of a chemical feed to inhibit and
counteract the antiscalent in the reject stream and cause the calcium
sulfate to precipitate. After clarification, the sludge is blown down,
and the supernatant is pumped back to the blend tank for further water
recovery. The sludge can be dried in an evaporation pond to achieve a
complete zero discharge. For a water production rate of 144,000 gpd at
a feed recovery of 80 percent, the capital cost was estimated to be
$200,000 in 1977. The operating and maintenance cost was $2.05 per
thousand gallons of product water, assuming $3.00 per thousand pounds of
steam, $0.04 per thousand gallons of cooling water, and $0.02 per kWh of
electricity. Including the capital investment, with an interest rate of
10 percent and 1 percent of investment as taxes, the total yearly cost of
the system was $2.47 per thousand gallons of product water.
Overall costs of a brine concentration/water purification system
should include the costs of ultimate brine disposal as well as credits
for high quality water produced for reuse. Where waste brine must be
stored on site or evaporated, costs are reduced when volume reduction by
reverse osmosis is increased to its practical limits. A study by Dow38
illustrates the advantages of high recovery when ponding or deep well
injection is used for brine disposal. Compared to the vapor-compression
evaporative brine concentration process, reverse osmosis is less costly
in both capital and operating cost. Reverse osmosis power requirements
range from 10 to 30 kWh per thousand gallons while vapor compression
evaporation requires from 70 to 90 kWh per thousand gallons. In terms
of 1976 dollars, operating and maintenance costs of reverse osmosis and
vapor-compression evaporators are estimated at 45 to 50 cents per thou-
sand gallons and $2.35 per thousand gallons respectively. Equipment and
installation costs are estimated at $.90 per gallon per day for reverse
osmosis and $7 per gallon per day for evaporators. (These figures are
based on estimates made for TVA by Sheppard T. Powell Associates.)
At the San Juan generating station,19 the Public Service Company of
New Mexico found the use of reverse osmosis more economical than evapo-
rators alone. Reverse osmosis is being used to preconcentrate cooling
tower blowdown prior to treatment by a vapor-compression evaporative
brine concentrator. The reverse osmosis system is designed to concen-
trate the blowdown by a factor of five, reducing the required evaporator
-32-
-------
TABLE 5. RESULTS OF REVERSE OSMOSIS STUDIES ON COOLING TOWER SLOWDOWN TREATMENT
Membrane type
Feed composition
Operation
Permeate Rejection Pressure
(gal/min) (%) (psig)
u>
u>
I
DuPont Spiral Wound Alkalinity (total), 58.1 mg/£ as CaC03
Conductivity, 4700 |Jmho
TDS, 4000 mg/£
COD, 24 mg/£
Cr, 0.01 mg/£
Zn, 0.2 mg/£
PO.-P (total), 0.49 mg/£
Kjeldahl-N, 0.95 mg/£
(N02 + N03)-N, 17 mg/£
pH, 6.3 units
Dow Hollow Fiberb Alkalinity (total), 43 mg/£ as CaC03
Conductivity, 5000 pmho
TDS, 4300 mg/£
COD, 19 mg/£
Cr, 0.01 mg/£
Zn, 0.25 mg/£
PO,-P (total), 0.67 mg/£
Kjeldahl-N, 0.78 mg/£
(N02 + N03)-N, 18 mg/£
pH, 6.3 units
0.45
0.45
0.45
0.45
0.45
0.45
0.45
0.45
0.45
0.45
0.27
0.27
0.27
0.27
0.27
0.27
0.27
0.27
0.27
0.27
84.3
82.1
85.8
95.8
>80
85
93.9
78.9
80
—
90.7
95.6
95.3
89.5
>80
92
98.5
48.7
87.2
-
400
400
400
400
400
400
400
400
400
400
200
200
200
200
200
200
200
200
200
200
-------
TABLE 5 (continued)
ft
Membrane type Feed composition
Roga Spiral Wound Alkalinity (total), 70 mg/£ as CaC03
Conductivity, 4100 pmho
IDS, 8300 mg/£
COD, 21 mg/l
Cr, 0.01 mg/£
Zn, 0.37 mg/2
PO.-P (total), 1.34 mg/£
Kjeldahl-N, 0.45 mg/£
(N02 + N03)-N, 27 mg/£
pH, 6.5 units
DuPont Hollow Fiber Alkalinity (total) , 72 mg/£ as CaC03
Conductivity, 8900 (Jmho
IDS,
COD, 35 mg/£
Cr, 0.01 mg/£
Zn, 0.89 mg/£
PO.-P (total), 1.74 mg/£
Kjeldahl-N, 0.55 mg/H
(N02 + N03)-N, 55 mg/£
pH, 6.4 units
Permeate
(gal/min)
1.1
1.1
1.1
1.1
1.1
1.1
1.1
1.1
1.1
1.1
1.6
1.6
1.6
1.6
1.6
1.6
1.6
1.6
1.6
1.6
Rejection
(%)
91.4
85.2
80.7
33.3
>80
83.8
96.3
71.1
64.8
-
86.5
89.2
-
94.3
>80
96.6
97.1
87.3
84.5
"
Operation
Pressure
(psig)
400
400
400
400
400
400
400
400
400
400
400
400
400
400
400
400
400
400
400
400
,Feedflow rate for all modules was 4 gal/min.
The Dow module was a low pressure module designed for testing purposes.
-------
size by a similar factor. Operating costs for the reverse osmosis system
are about $1.00 to $1.20 per thousand gallons of treated water. Consider-
ing the combination of reverse osmosis and evaporation together, the opera-
ting cost is estimated at $1.70 to $1.80 per thousand gallons.
To enhance user confidence in reverse osmosis, a User's Guide to
Reverse Osmosis has been included as an appendix to this report. Most
of the problems reported with reverse osmosis systems in the past have
been caused by poor pretreatment and design, but are not inherent to the
technology. The appendix discusses the major factors which must be con-
sidered in specifying or buying a reverse osmosis system with special
attention to potential problem areas.
In selecting a reverse osmosis system for treating cooling water,
the temperature tolerance of the membranes should be carefully checked
against expected water temperatures. Some type of auxiliary cooling may
be necessary as pretreatment to the reverse osmosis unit. Other pretreat-
ment needs are discussed in the appendix.
RECOMMENDATIONS
When a cooling system is required to operate with zero liquid dis-
charge and dissolved solids must be controlled, electrodialysis is
apparently the most cost-effective approach to removing and concentrating
dissolved solids for storage. The recommended system would treat a side-
stream of the recirculating cooling water. Softening and filtration would
precede electrodialysis. The electrodialysis unit would produce a salt-
depleted stream which would be returned to the cooling system and a con-
centrated stream which would be stored in lined ponds.
A combination of reverse osmosis and evaporative brine concentration
can be used instead of electrodialysis. Using reverse osmosis serves to
preconcentrate the brine and substantially reduce the cost of the evapo-
ration step. This system may be superior to electrodialysis where large
volumes of permeate or distillate quality water are needed in other plant
systems.
Flow diagrams of the recommended systems are shown in Figures 2 and
3. Note that hot water feed is preferred for electrodialysis because it
improves performance, while cool water is often desirable for reverse
osmosis in order to prevent membrane hydrolysis.
Large-scale demonstrations of these concepts are recommended to
provide cost data and demonstrate operability. Of the two systems,
electrodialysis appears to offer the most promise.
-35-
-------
RECIRCULATION
EVAPORATION
*
DRIFT
\ 7
COOLING
TOWER
MAKEUP
WATER RECOVERY
PRETREATMENT
ELECTRODIALYSIS
PLANT
SLUDGE TO
LANDFILL
BRINE TO
EVAPORATION POND
Figure 2. Flow diagram for cooling tower sidestream softening and desalt-
ing with electrodialysis. Adapted from Jordan, Mcllhenny &
Westbrook "Cooling Tower Effluent Reduction by Electrodialysis."
Presented at the American Power Conference, Chicago, Illinois,
April 22, 1976. Used by permission.
-36-
-------
RECIRCULATION
CONDENSATE ,
EVAPORATION
7
REVERSE OSMOSIS
PLANT
TO BOILER
MAKEUP
POLISHING
PERMEATE
DISTILLATE
BRINE
BRINE
CONCENTRATOR
COOLING
TOWER
MAKEUP
SLOWDOWN
PRETREATMENT
SLUDGE TO
LANDFILL
CONCENTRATE TO
EVAPORATION PONDS
Figure 3. Flow diagram for cooling tower blowdown treatment by reverse
osmosis for reuse.
-37-
-------
SECTION 7
ASH SLUICE WATERS
CHARACTERISTICS OF WASTEWATER
Ash produced from a pulverized coal-fired power generation plant
includes fly ash and bottom ash, with a weight ratio of about four to
one. Fly ash generally is collected in hoppers below mechanical collec-
tors and electrostatic precipitators, conveyed pneumatically to a water
ejector, and sluiced to a settling pond. Bottom ash is collected in the
bottom of the furnace and periodically removed from the furnace by water,
through a clinker grinder, and piped to a settling pond.41
In some cases where space is at a premium, ash is dewatered in
dewatering bins and shipped off site to be landfilled.
The quantity and characteristics of ash pond effluent are affected
by the type of fuel used, the ash content of the fuel, the mode of firing,
the design of the combustion chamber, the fraction of fly ash vs. bottom
ash, the quantity and quality of sluice water used, and the performance
of the settling pond." The primary contents of fly ash and bottom ash
are metal oxides of silica, calcium, magnesium, iron, and aluminum; oxi-
dation products of sulfur and phosphorus; and carbon residuals.42 4S
The pyritic material, consisting primarily of iron sulfides that are
removed from the coal before combustion, may also be disposed of into
the ash pond. Metallic and sulfuric compounds present in the ash may be
released and become constituents of the sluice water.41
TVA operates ash ponds that receive fly ash, bottom ash, or both.
Specific data on the chemical characteristics of ash ponds are summarized
in Table 6. The pH of these effluents varies from 3.3 to 12; half of
the effluents are alkaline. The acidic or alkaline characteristics depend
on the content of sulfur trioxide and alkaline metal oxides in the ash
materials and on the buffering capacity of the water used for sluicing.
Factors that affect the ash characteristics include source of the coal,
methods of firing, ash fusion temperature, and efficiency of equipment
for collecting fly ash.41
The operating conditions for TVA's 12 coal-fired power plants are
summarized in Table 7. For the plants that use pulverized coal, the pH
of ash pond effluents is mainly affected by the source of coal. Ash pond
effluents from plants that receive coal from western Kentucky and southern
Illinois are alkaline, whereas those from plants that receive coal from
eastern Tennessee, eastern Kentucky, and Virginia are neutral or acidic.41
-38-
-------
TABLE 6. CHARACTERISTICS OF ONCE-THROUGH ASH POND DISCHARGES
«a
Flow, gpm
Total alkalinity,3
mg/S. as CaC03
Phen. alkalinity,3
mg/S, as CaC03
Conductivity,3
[j mhos /cm
Total hardness,
mg/S, as CaC03
pH, units
Dissolved solids,
Suspended solids,
mg/S,
Aluminum, mg/S.
Ammonia, mg/S, as N
Arsenic, mg/S,
Barium, mg/S,
Beryllium , mg/S,
Cadmium, mg/£
Calcium, mg/S,
Chloride, mg/S.
Chromium, mg/S,
Copper, mg/S,
Cyanide, mg/S,
Iron, mg/S,
Lead, mg/S,
Magnesium, mg/S,
Manganese, mg/S.
Mercury, mg/S,
Nickel, mg/S,
Total phosphate,
mg/2 as P
Selenium, mg/S,
Silica, mg/2
Silver, mg/S,
Sulfate, mg/£
Zinc, mg/S,
Fly ash
pond
6667.3
18.7
<1
811
315
4.2
517
48.3
7.9
0.75
0.011
0.2
0.01
0.038
126
7
0.072
0.33
<0.01
2.3
0.066
14
0.49
0.0003
0.08
0.03
0.002
13
<0.01
346
1.4
Plant A
Bottom ash
pond
17798.7
71.7
<1
322
141.5
7.1
168.7
57
3.2
0.11
0.007
0.1
<0.01
0.001
38
7
0.007
0.07
<0.01
5.2
0.017
6.0
0.17
0.0005
0.06
0.07
0.002
7.4
<0.01
45
0.08
Plant B
Fly ash
pond
NA
84.3
38
788
329
9.3
524.3
70.7
1.6
0.07
0.029
0.1
<0.01
0.001
152
6
0.013
0.03
<0.01
1.4
0.015
3.6
0.12
0.0008
0.05
0.06
0.015
7.1
<0.01
214
0.05
Bottom ash
pond
NA
55.7
5
223.5
92.5
8.1
136.3
57
2.2
0.07
0.014
0.1
<0.01
0.002
50
7
0.009
0.06
<0.01
4.7
0.018
6.2
0.40
0.0009
0.06
0.06
0.007
6.4
<0.01
102
0.13
Plant C
Eastern
Outlet
7730
70.7
<1
495
212.5
7.1
342.3
39.3
1.5
0.11
0.013
0.2
<0.01
0.006
78
11
0.006
0.05
0.01
1.7
0.021
10
0.20
0.0034
0.05
0.04
0.010
7.4
0.01
158
0.13
Average values of weekly grab samples; all other numbers are average values of
quarterly grab samples.
-39-
-------
TABLE 6 (continued)
Flow, gpm
Total alkalinity,3
mg/Jd as CaC03
Phen. alkalinity,
mg/Jd as CaC03
Conductivity,
(Jmhos/cm
Total hardness,
mg/Jd as CaC03
pH, units3
Dissolved solids,
Suspended solids,
Aluminum, mg/Jd
Ammonia, mg/Jd as N
Arsenic, mg/Jd
Barium, mg/Jd
Beryllium , mg/Jd
Cadmium, mg/Jd
Calcium, mg/£
Chloride, mg/Jd
Chromium, mg/Jd
Copper, mg/Jd
Cyanide, mg/Jd
Iron, mg/Jd
Lead, mg/Jd
Magnesium, mg/Jd
Manganese, mg/Jd
Mercury, mg/Jd
Nickel, mg/Jd
Total phosphate,
mg/Jd as P
Selenium, mg/Jd
Silica, mg/Jd
Silver, mg/Jd
Sulfate, mg/Jd
Zinc , mg/Jd
Plant C
Western
Outlet
1651.7
69
<1
329.5
129
7.4
217
40
3.4
0.09
0.022
0.14
<0.01
0.002
37
11
0.009
0.06
0.01
6.0
0.017
10
0.18
0.0070
0.06
0.12
0.003
6.7
0.01
99
0.14
Plant D
8224
61.7
5
265
126.5
8.5
157.3
19
1.4
0.06
0.034
0.2
<0.01
0.001
31
3
<0.005
0.03
<0.01
0.32
0.016
8.3
0.02
0.0002
0.06
0.03
0.070
4.0
0.01
57
0.03
Plant E
5857.5
141
105.5
819
288
11.2
389.5
8
2.5
0.06
0.028
0.2
<0.01
0.001
126
6
0.017
0.08
<0.01
0.16
0.017
0.3
0.01
0.0002
<0.05
0.01
0.007
7.0
0.01
147
0.05
Plant F
30616.5
96.5
82.5
915
304
10.9
408.5
23
1.7
0.17
0.008
0.2
<0.01
0.001
107
5
0.033
0.03
<0.01
0.22
0.013
1.57
0.01
0.0003
0.05
0.02
0.014
6.0
<0.01
160
0.05
Plant G
7391
46.3
20
366
198
9.6
263
19.3
1.7
0.12
0.030
0.2
<0.01
<0.001
73
4
0.011
0.05
0.01
0.53
0.014
2.4
0.02
0.0024
<0.05
0.07
0.010
4.4
<0.01
182
0.05
Average values of weekly grab samples; all other numbers are average values of
quarterly grab samples.
-40-
-------
TABLE 6 (continued)
Flow, gpm
Total alkalinity,3
mg/S. as CaCOa
Phen. alkalinity,3
mg/S, as CaC03
Conductivity,3
pmhos/cm
Total hardness,3
mg/S. as CaC03
pH, units
Dissolved solids,
mg/S,
Suspended solids,
rag/A
Aluminum, mg/S,
Ammonia, mg/S, as N
Arsenic, mg/£
Barium, mg/£
Beryllium , mg/S.
Cadmium, mg/£
Calcium, mg/S.
Chloride, mg/S.
Chromium, mg/S.
Copper, mg/£
Cyanide, mg/S,
Iron, mg/S.
Lead, rng/JU
Magnesium, mg/S,
Manganese, mg/JK
Mercury, mg/£
Nickel, mg/S,
Total phosphate,
mg/S, as P
Selenium, mg/S,
Silica, mg/S,
Silver, mg/S,
Sulfate, mg/S,
Zinc, mg/S,
Plant H
2692.3
67.7
11.3
392
115.5
8.6
270
14.3
1.6
0.34
0.123
0.2
<0.01
0.001
50
14
0.006
0.04
<0.01
0.56
0.015
7.4
0.06
0.0004
0.05
0.12
0.017
4.9
<0.01
98
0.05
Plant I
16872
121.7
91
652.5
211.5
10.7
259
19
1.5
0.07
0.36
0.2
<0.01
<0.001
84
6
0.017
0.06
<0.01
0.26
0.012
1.2
0.05
0.0003
0.05
0.06
0.012
7.1
<0.01
81
0.08
Plant J
14062.3
40
4.5
311.5
103
6.2
187.3
39
2.6
0.05
0.041
0.2
<0.01
0.001
34
5
0.005
0.11
<0.01
2.4
0.015
6.7
0.38
0.0003
0.05
0.06
0.004
6.4
<0.01
119
0.07
Plant K
22447.3
99.7
63.3
472.5
174
10.8
247.7
19.3
1.8
0.06
0.033
0.2
<0.01
0.001
76
10
0.019
0.05
<0.01
0.39
0.017
1.6
0.02
0.0003
0.06
0.05
0.010
6.7
<0.01
83
0.05
Plant L
14494
86
44
359.5
178
10.2
212.7
15
2.0
0.52
0.032
0.1
<0.01
0.001
54
6
0.009
0.06
<0.01
0.56
0.017
2.6
0.03
0.0003
<0.05
0.06
0.010
5.7
<0.01
80
0.04
Average values of weekly grab samples; all other numbers are average values
of quarterly grab samples.
-41-
-------
TABLE 7. RELATIONSHIPS BETWEEN PLANT OPERATION CONDITIONS AND
pH VLUES OF ASH POND EFFLUENTS AT TVA COAL-FIRED
POWER PLANTS3
Constituent
Coal source
Method of
firing
Ash content
of coal, %
Plant D
E. Kentucky
Tangential
15.5
Plant H
Virginia
E . Kentucky
E. Tennessee
Tangential
15
Plant J
E. Kentucky
E. Tennessee
Tangential
19.1
Plant E
W. Kentucky
Circular
Tangential
15.3
Fly ash of
total ash, %
Bottom ash of
total ash, %
Sluice water-
to-ash ratio,
75
25
67
33
75
25
67
33
gal/ton
pH value of
raw water
pH value of
ash pond
effluent
10,770
7.5
8.6b
11,425
7.0
8.9b
9,520
7.6
6.3b
9,585
7.0
11. lb
.All values based on average values during 1974.
Combined bottom and fly ash pond.
,Fly ash pond only.
Bottom ash pond only.
Source: Chu, T.-Y. J., P. A. Krenkel, and R. J. Ruane. Characterization
and Reuse of Ash Pond Effluents on Coal-Fired Power Plants. Paper
presented at 49th Annual Water Pollution Control Federation Con-
ference, Minneapolis, Minnesota, Oct. 2-8, 1976.
-42-
-------
TABLE 7 (continued)
Constituent
Coal source
Method of
firing
Ash content
of coal, %
Plant F
W. Kentucky
S. Illinois
Opposed
16.3
Plant G
W. Kentucky
Tangential
15.7
Plant I
W. Kentucky
Tangential
Horizontal
14
Plant K
S. Illinois
W. Kentucky
Circular
15.6
Fly ash of
total ash, %
Bottom ash of
total ash, %
Sluice water-
to-ash ratio,
80
20
80
20
70
30
75
25
gal/ ton
pH value of
raw water
pH value of
ash pond
affluent
19,490
7.4
11. 2b
12,345
7.3
9.6b
42,430
7.4
11. 2b
17,265
7.6
10. 8b
values based on average values during 1974.
Combined bottom and fly ash pond.
Fly ash pond only.
Bottom ash pond only.
Source: Chu, T.-Y. J. , P. A. Krenkel, and R. J. Ruane. Characterization
and Reuse of Ash Pond Effluents on Coal-Fired Power Plants. Paper
presented at 49th Annual Water Pollution Control Federation Con-
ference, Minneapolis, Minnesota, Oct. 2-8, 1976.
-43-
-------
TABLE 7 (continued)
Constituent
Coal source
Method of
firing
Ash content
of coal, %
Plant L
W. Kentucky
N. Alabama
Horizontal
Tangential
16
Plant B
W. Kentucky
Vertical
14.8
Plant C
W. Kentucky
S. Illinois
Cyclone
11
Plant A
W. Kentucky
Cyclone
18.8
Fly ash of
total ash, % 75
Bottom ash of
total ash, % 25
Sluice water-
to-ash ratio,
gal/ton 15,730
pH value of
raw water 7.5
pH value of
ash pond ,
effluent 10.1
50
50
7.5
8.'od
30
70
23,065
7.4
7.1b
30
70
9,810°
12,380(
7.7
7.2
, All values based on average values during 1974.
Combined bottom and fly ash pond.
,Fly ash pond only.
Bottom ash pond only.
Source: Chu, T.-Y. J., P. A. Krenkel, and R. J. Ruane. Characterization
and Reuse of Ash Pond Effluents on Coal-Fired Power Plants. Paper
presented at 49th Annual Water Pollution Control Federation Con-
ference, Minneapolis, Minnesota, Oct. 2-8, 1976.
-44-
-------
The high concentrations of suspended solids in ash ponds that
receive fly ash probably are caused by low-density, hollow-sphere ashes
(cenospheres) that are not removed by natural settling. The amount of
cenospheres generated at coal-fired plants can be as much as 4 to 5 per-
cent by weight, or 15 to 20 percent by volume, of the fly ash. The prin-
cipal constituents in cenospheres are similar to those in fly ash.42
Because trace elements also are present in the cenospheres, they can
contribute to the total concentration of trace elements in the ash pond
discharges.42
In a recent study for EPA,47 Radian Corporation reviewed different
options for recycling ash sluice waters. Because bottom ash is not very
reactive, recycle of bottom ash sluice water typically poses no serious
problems. Fly ash is much more reactive, and because ashes from different
coals have different characteristics, it is difficult to predict the char-
acteristics of recirculated fly ash sluice water. Depending on the par-
ticular situation, recirculating fly ash sluice waters may require either
no treatment or one of several treatment methods. The simplest treatment
allows time for ash to react with water and for supersaturated species
to precipitate (under control) in a large reaction tank before entering
the sluice lines. Alternatively, softening of the recirculating water
will reduce the potential for calcium and magnesium scaling. Whichever
option is chosen, pilot testing would be required to determine design
criteria for a full-scale recirculating system and to determine the
recirculating water characteristics.
Blowdown can be used to reduce the total dissolved solids concen-
tration, or brine concentration by reverse osmosis, evaporation, or elec-
trodialysis may be used. Again, the need for dissolved solids control
can be established by pilot testing.
ALTERNATIVE TREATMENTS
Reverse osmosis, electrodialysis, and vapor-compression distillation
can be used to control dissolved solids by concentrating a blowdown or
sidestream from recirculating ash sluice systems. Dissolved solids can
also be controlled by increasing the makeup flow and discharging blowdown.
Conventional methods of rendering blowdown acceptable should be con-
sidered. These include control of pH and redox potential to precipitate
metals and the use of higher blowdown rates to reduce their concentration.
Suspended solids can generally be controlled by settling or filtration.
Electrodialysis
Dissolved salt can be removed from a sidestream of an ash sluice
system by electrodialysis. This system would be essentially similar to
that proposed in Section 2 for cooling water. Cost comparisons made in
Section 2 would appear valid for ash sluice waters also, but much needs
to be learned about the characteristics of ash sluice waters when recir-
culation is practiced.
-45-
-------
Reverse Osmosis
Reverse osmosis can be used to concentrate blowdown from ash sluice
systems. The system would be much like that described in Section 2 for
concentration of cooling tower blowdown, and the cost advantages would
be similar. Careful pH control may be necessary to prevent membrane
fouling by metal oxides.
Laboratory tests on treating once-through ash pond effluent were
conducted by TVA. Fly ash pond effluent from TVA Plant A was tested
because of its acid characteristics and its relatively high concentra-
tions of several trace metals. The major constituents in the ash pond
effluent are suspended solids, sulfate, calcium, magnesium, and silica;
therefore, spiral-wound and tubular modules were selected to treat this
wastewater. The fly ash pond effluent was run through a 1- to 3-(Jm fiber
filter to remove suspended solids; it was then run through each module
of the Roga spiral-wound cellulose triacetate membrane, the DuPont spiral-
wound polyamide membrane, and the UOP Fluid Science tubular cellulose
acetate membrane individually. The data of rejection on total dissolved
solids from each module are shown in Table 8. The Roga spiral-wound module
appears better for treating this type of wastewater, and additional experi-
ments were conducted with this module to treat fly ash pond effluent which
had been concentrated by a factor of 1.7. The experimental results are
shown in Table 9.
Although treatment of once-through ash pond water is not expected
to be economical, these results show satisfactory flux and rejection per-
formance and indicate no special problems should be expected from this
type of waste treatment.
RECOMMENDATIONS
Treatment of recirculating ash sluice water by reverse osmosis or
electrodialysis is feasible. At each plant, the need for dissolved solids
control should be critically examined. If control is needed, alternative
methods should be explored, including reverse osmosis and electrodialysis.
Large-scale demonstrations of reverse osmosis or electrodialysis
for concentrating ash sluice blowdown should be deferred until a better
understanding of recirculating ash sluice systems is attained and the
need for dissolved solids control in these streams is established.
Experience gathered in demonstrating electrodialysis and reverse
osmosis to control dissolved solids in cooling waters should be trans-
ferable to ash sluice pond water treatment.
Dry ash handling systems should be given consideration for reducing
water requirements and discharges.
-46-
-------
TABLE 8. RESULTS OF REVERSE OSMOSIS STUDIES ON FLY ASH
POND EFFLUENT FROM TVA PLANT A
Membrane
Feed composition
Operation
Permeate Rejection pressure Feed flow
(gal/min) (%) (psig) (gal/min)
Roga Spiral Wound Conductivity, 852.5 pmho/cm 1.2
pH, 4.3
DuPont Spiral Wound Conductivity, 840 pmho/cm
pH, 4.5
0.45
UOP Fluid Science Conductivity, 827.5 |Jmho/cm 0.16
Tubular pH, 4.3
93
92.6
79.3
400
400
400
4.0
4.0
4.0
-------
TABLE 9. RESULTS OF REVERSE OSMOSIS STUDIES ON CONCENTRATED
FLY ASH POND EFFLUENT FROM TVA PLANT A
oo
i
Permeate
Membrane Feed composition (gal/min)
Roga Spiral Wound Conductivity, 1450 (Jmho/cm
TDS, 1500 mg/2
Al, 9.6 mg/2
Ca, 280 mg/2
Cd, 0.29 mg/2
Cr, 0.25 mg/2
Cu, 6.9 mg/2
Fe, 0.56 mg/2
Pb, 0.042 mg/2
Mg, 51 mg/2
Mn, 1.5 mg/2
Se, 0.003 mg/2
Si, 27.8 mg/2
S04, 815 mg/2
Zn, 6.65 mg/2
pH, 4.7 mg/2
1.15
1.15
1.15
1.15
1.15
1.15
1.15
1.15
1.15
1.15
1.15
1.15
1.15
1.15
1.15
1.15
Rejection
(%)
•
96.7
95.8
96.8
97.9
93.2
96.2
>91
>90
96.5
96
>90
88.5
96.9
99.1
-
Operation
pressure
(psig)
400
400
400
400
400
400
400
400
400
400
400
400
400
400
400
400
Feed flow
(gal/min)
4.0
4.0
4.0
4.0
4.0
4.0
4.0
4.0
4.0
4.0
4.0
4.0
4.0
4.0
4.0
4.0
-------
SECTION 8
COAL PILE DRAINAGE
CHARACTERISTICS OF WASTEWATER
To ensure uninterrupted generation of electricity, an outdoor coal
reserve is maintained at each power plant. A 90-day coal supply is
customarily maintained. Coal piles are typically 8 to 12 m (25 to 40
ft) high and cover an area of 10 to 40 ha (25 to 100 acres). Normally,
600 to 1800 m3 (780 to 2340 yd3) of coal storage is required for every
megawatt of rated capacity.4'
Coal pile drainage results from percolation of rainfall through
stored coal. The water quality of the drainage is affected by the
leaching of oxidation products of metallic sulfides associated with the
coal. The sulfide-bearing minerals that predominate in coal are pyrite
and marcasite, both iron sulfide ores. Marcasite is unstable and degrades
into pyrite. The oxidation of pyrite results in production of ferrous
iron and acidity. This ferrous iron then undergoes oxidation to the
ferric state. Ferric iron then hydrolyzes to form insoluble ferric
hydroxide, thus producing more acidity.48
The following characteristics of the coal pile runoff were reported
for a power plant at Springfield, Illinois: suspended solids—2200 to
21,000 mg/£, with an average of 10,300 mg/£; total iron--50 to 80 mg/A,
with an average of 61 mg/A; and total aluminum—240 mg/JH.49
TVA conducted an intensive study on coal pile drainage.50 The
study programs were established at two coal-fired steam plants. Plant J
has a rated capacity of 1700 MW, with a 90-day coal supply amounting to
about 9.6 x 10* m3 (1.26 x 106 yd3), or 1.1 x 109 kg (1.2 x 106 tons).
Plant E has a rated capacity of 1400 MW, with a 90-day coal supply
amounting to about 8.6 x 10s m3 (1.13 x 106 yd3), or 9.88 x 108 kg (1.08
x 106 tons). Coal for plant J is mined in eastern Tennessee and Kentucky,
and coal for plant E is mined in western Kentucky. A typical analysis
of coal from both plants is shown in Table 10.
Both of the TVA coal pile drainage systems investigated exhibited
highly acidic drainages. Acidity was quite variable in both cases
(Table 11), but pH was limited to a rather tight band (2.3 to 3.1).
Means (arithmetic) are similar; 21 of the 33 values fall between 2.6 and
3.0. Values of pH of coal pile drainage reported by EPA26 exhibit a
slightly broader range of 2.1 to 3.0. Anderson and Youngstrom51 report
a pH of 2.2 to 5.8 for hourly pH measurements over a 3-week period.
Matsugu52 reports a pH of 2.4 to 3.0 for 67 grab samples of coal pile
-49-
-------
TABLE 10. COAL ANALYSIS, DRY BASIS
Constituent Plant J Plant E
Total moisture, % 3.8 4.2
Volatile matter, % 34.1 37.7
Ash % 17.2 15.0
Fixed carbon, % 48.7 47.3
Total sulfur, % 2.1 3.9
Energy, Btu/lb 12,270 12,450
Ash analysis
CaO, % of ash 1.4 4.2
MgO, % of ash 1.1 1.1
-50-
-------
TABLE 11. GENERA! CHEMICAL CHARACTERISTICS OF COAL PILE DRAINAGES
COLLECTED FROM TWO TVA STEAM PLANTS
pH
Range 2.3-3.1
Mean 2.8
Number 2)
of
samples
2 Range 2.3-3.1
1 Mean 26
Number 12
of
samples
Range 2.5-2.7
Mean 2.63
Number 14
of
samples
Acidity
(mg/£ as
CaC03)
300-7100
3447
20
700-4800
1648
12
300-1400
710
14
Sulfate
1800-9600
5160
18
1800-6200
3050
12
PLANT E
870-5500
2300
14
Dissolved
solids
PLANT J
2500-16,000
7790
20
PLANT E
270-8200
3905
12
(DISCRETE STORM)
1200-7500
2700
14
Suspended
solids Iron
8.0-2300 240-1800
430 943
20 21
38-680 23-590
282 350
12 12
69-2500 62-380
650 150
14 14
Manganese
8.9-45
28
21
1.8-12
4.5
12
0.88-5.4
2.3
14
Source: Cox, D. B. , T.-Y. J. Chu, and R. J. Ruane. Quality and Treatment of Coal Pile Runoff. In:
Proc., Seventh Symposium on Coal Mine Drainage Research, NCA/BCR Coal
Conf. and Expo IV, Louisville, Kentucky, October 13-20, 1977.
pp. 232-255.
-------
leachate. For these same samples, acidity varied from 10 to 120
milliequivalents/liter (meq/£). Thus, the pH of coal pile drainage, at
least for eastern coal, is generally in the relatively narrow range of
2.1 to 3.1.
Concentrations of suspended solids at plant J ranged from 8 to 2300
mg/£, with a mean of 430 mg/£. At plant E, where direct runoff was
collected as a single composite sample for each storm event, the mean
and range of suspended solids concentrations were somewhat lower. Much
higher values for suspended solids were reported by Matsugu.52
Iron concentrations at plant E ranged from 23 to 590 mg/£, with a
mean of 354 mg/JL Concentrations of iron at plant J were higher, with a
range of 240 to 1800 mg/£ and a mean of 943 mg/S,. Values for iron
reported by EPA26 ranged from 0.17 to 93,000 mg/£, with a mean of 19,500
rag/A. A somewhat narrower range of 10 to 5300 mg/Jd and a lower mean of
1150 were reported by Anderson and Youngstrom.51
Manganese concentrations reported by Anderson and Youngstrom51
ranged from 4.5 to 72.0 mg/£, with a mean of 17.1 mg/£. Somewhat lower
concentrations, ranging from 3.4 to 12 mg/JH, with a mean of 6.9 mg/2,
were reported by Matsugu.52 Levels at plant J (Table 12) were com-
parable to those presented by Anderson and Youngstrom.51 Values for plant
E were somewhat lower.
Trace elements of environmental concern in coal that have been
identified by EPA53 are presented in Table 12. Results of analyses for
selected trace elements are presented in Table 13. Of these elements,
lead, barium, and titanium were low or consistently below the limits of
detection. Most means of trace element concentrations at plant J are
three to eight times as high as those at plant E.47
Concentrations of copper at plant J are higher than the criteria
set by EPA54, but are lower than those reported by EPA26 or Anderson and
Youngstrom.51 Concentrations for plant E are lower still.
The mean zinc concentrations of 6.46 mg/£ at plant J and 2.42 mg/S,
at plant E are similar to the means of 5.9 mg/A reported by EPA26 and
3.67 rng/A reported by Anderson and Youngstrom.51 The criterion estab-
lished by EPA54 for a public water supply is 5 mg/SL,
Cadmium concentrations at both plants were below water quality
criteria.54
Mean concentrations of 260 mg/2 at plant J and 43 rag/A at plant E
were found for aluminum.
Concentrations of nickel are also above levels found in surface
water.55
Chromium concentrations are well below established water quality
criteria54 at both plants.
-52-
-------
TABLE 12. RANGE OF TRACE ELEMENTS IN U.S. COALS
Major Elements Minor Elements
Element Concentration Range (Weight %) ElementConcentration Range (ppm)
Al
Ca
Cl
Fe
K
Mg
Na
Si
Ti
0.43 -
0.05 -
0
0.32 -
0.02 -
0.1 -
0
0.58 -
0.002-
3.04
2.67
0.56
4.32
0.43
0.25
0.20
6.09
0.32
As
B
Be
Br
Cd
Co
Cr
Cu
Ga
Ge
Hg
La
Mn
Mo
Ni
Pb
Sb
Sc
Se
Sn
U
V
Y
Zn
Zr
0.
1.
0
4
0.
0
0
1.
0
0
0.
0
6
0
0.
4
0.
10
0.
0
<10
0
<0.
0
8
5 - 106
2 - 356
- 31
- 52
1 - 65
- 43
- 610
8 - 185
- 61
- 819
01- 1.6
- 98
- 181
- 73
4 - 104
- 218
2 - 9
- 100
4 - 8
- 51
-1000
-1281
1 - 59
-5600
- 133
Source: Wewerka, E. M. et al. Environmental Contamination From Trace
Elements in Coal Preparation Wastes - A Literature Review and
Assessment. EPA-600/7-76-007, August 1976. 61 pp.
-53-
-------
TABLE 13. TRACE METAL CONCENTRATIONS IN COAL PILE DRAINAGE
(values in mg/£)
Cu
Minimum 0 . 42
Maximum 1 . 4
Mean 0.86
V °
Number of 21
samples
Minimum 0.07
Maximum 0.46
Mean 0.18
V
Number of 12
samples
Zn
2.3
16
6.46
0
21
1.1
5.1
2.42
0
12
Cd
<0.001
<0.001
21
21
<0.001
0.003
0.002
6
12
Al
66
440
260
0
21
20
92
43
0
11
Ni
0.
4.
2.
0
21
0.
0.
0.
0
12
Cr
PLANT J
7 <0.005
5 0.011
5 0.006
12
18
PLANT E
15 <0.005
49 0.011
34 0.01
8
12
Hg
<0.0002
0.0025
0.0004
13
21
0.0005
0.0072
0.003
0
11
As
0.005
0.36
0.15
0
20
0.006
0.046
0.02
0
9
Se
0.001
0.03
0.005
0
20
<0.001
0.006
0.002
3
10
Be
0.03
0.07
0.04
0
19
<0.01
0.03
0.01
8
11
N- indicates number of samples below detection limits.
Source: Cox, D. B., T.-Y. Chu, and R. J. Ruane. Quality and Treatment of Coal Pile Runoff
In: Proc., Seventh Symposium on Coal Mine Drainage Research, NCA/BCR Coal Conf. and
Expo IV, Louisville, Kentucky, Oct. 13-20, 1977- pp. 232-255.
-------
Levels of beryllium are well below the established criterion for
waters of a hardness of 245 to 300 mg/£.55
Mercury concentrations were an order of magnitude higher at plant E
than at plant J. Levels at both plants exceeded the established water
quality criterion.54
Arsenic levels in drainage from plant J ranged from 0.005 to 0.36
mg/£> with a mean of 0.15 mg/£. These values generally exceeded estab-
lished criteria, whereas those concentrations found at plant E generally
did not. Concentrations of selenium behaved similarly. This is signifi-
cant since selenium and arsenic exhibit antagonistic toxicities.56
The Clean Water Act of 1977 requires EPA to establish effluent limi-
taions for 65 toxic (priority) pollutants. Elements on this list which
have been found in significant concentrations in coal pile drainage include
arsenic, selenium, nickel, copper, zinc, and mercury.
ALTERNATIVE TREATMENTS
Most metals can be precipitated by the conventional process of pH
adjustment and removed as suspended solids. Theoretically, the solubili-
ties of arsenic and selenium are not affected by pH adjustment,57_but a
number of other processes have been proposed for their removal.58 62
TVA has tested the ability of several reverse osmosis membranes to
remove trace metals, and the results are presented in the appendix. Rejec-
tions were generally better than 80 percent. Adjustment of pH followed
by suspended solids removal is required as pretreatment for reverse osmosis.
In pretreating one waste by pH adjustment, TVA found that this process
alone would produce acceptable effluent. Even the arsenic concentration
was significantly reduced. The waste was boiler fireside wash water from
an oil-fired generating unit. This waste is similar to coal pile drainage
in its pH (2.4) and in its high concentration of metals, notably iron.
Characteristics of this waste are presented in Table 14 and results of
pretreatment studies in Table 15.
There is no information in the literature on treatment of coal pile
drainage with membrane processes. However, reverse osmosis treatment of
mine drainage, which has characteristics similar to those of coal pile
drainage, has been studied for many years. The EPA has been sponsoring
and conducting research since 1966 on the use of reverse osmosis for
treatment of mine drainage. Potentials for recovery of valuable heavy
metals are also being investigated.63 Blackshaw and Pappano64 reported
on a study concerning the technical and economic feasibility of using
reverse osmosis for treatment and purification of acid mine drainage.
The reverse osmosis unit was rated at 60,000 gpd product water output,
featuring feedflow and waste brine flow controllers that can be preset
to desired flow rates, depending on recovery requirements and reverse
osmosis modular-tube arrays. Wilmoth and Scott65 investigated an appli-
cation of reverse osmosis to a ferrous-iron acid mine drainage in Mocanaqua,
Pennsylvania. The operating parameters and chemical data are shown in
-55-
-------
TABLE 14. BOILER WASH WASTE
Concentration
(rag/A) Unless
Parameter Otherwise Indicated
Total Dissolved Solids
Total Suspended Solids
S04
Cl
Si02
Dissolved Fe
Suspended Fe
Ca
Mg
Na
K
Al
As
B
Ag
Be
Ba
Cd
Cu
Hg
Mn
Ni
Pb
Se
Ti
V
Zn
Kjeldahl Nitrogen - N
Nitrite Plus Nitrate Nitrogen - N
Local Phosphate - P
pH
Conductivity
Turbidity
6100
1300
6000
71
100
1300
230
30
34
5.5
0.1
7.4
0.09
0.32
0.01
<0.01
7.9
<0.001
8.6
<0.0002
15
55
0.01
0.008
<1.0
20
4.3
0.45
0.16
0.61
2.4 (Units)
4550 (fjmho/cm)
47 (JTU)
-56-
-------
TABLE 15. RESULTS OF CHEMICAL TREATMENT ON BOILER WASH WATER
Parameters
pH, unit
As, mg/£
Ba, mg/Jd
Ca, mg/£
Cu, mg/£
Fe, mg/A
Mg, mg/SL
Mn, mg/JH
Ni, mg/£
V, mg/£
Concentration
Initial
2.4
0.09
7.9
30
8.6
1530
34
15
55
20
Final
7
<0.005
0.7
730
0.35
2.9
-
2
0.53
<0.1
8
<0.005
1.3
800
0.05
0.2
20
0.56
0.22
0.2
9
<0.005
1.0
950
0.03
0.05
24
0.15
0.10
0.3
10
<0.005
1.1
980
0.04
0.14
21
0.02
0.06
0.4
All concentrations of metals are total metals.
-57-
-------
Tables 16 and 17, respectively. A contract study for EPA66 tested spiral-
wound reverse osmosis systems on acid mine drainage discharges at four
locations: Norton, West Virginia; Morgantown, West Virginia; Ebensbury,
Pennsylvania; and Mocanaqua, Pennsylvania. The water quality characteris-
tics of those sites were quite different. At all sites, the limiting
factor in high recovery operation was calcium sulfate insolubility. Neu-
tralization of the product water was needed in all cases to elevate pH,
and in some cases, to remove residual iron and manganese. Neutrolosis,
which is the blending of neutralized brine supernatant from the reverse
osmosis unit back into the feed to the unit, has been shown to be a promis-
ing process for obtaining maximum recoveries and reducing the brine disposal
problem. The pretreatment at all sites consisted of lO-pm filtration.
Ultraviolet disinfection, acid injection, or both were needed at some
sites to prevent iron oxidation and precipitation.
Kaup67 proposed a combined treatment process with reverse osmosis
and ion exchange for acid mine drainage. After pretreatment for fouling
control, 75 percent of the wastewater will be recovered by reverse osmosis
and the remaining 25 percent concentrate will be treated by ion exchange.
Cost information for treating coal pile drainage with reverse osmosis
is unavailable from the literature.
Wilmoth66 reported that estimates of reverse osmosis treatment costs
for acid mine drainage ranged from $0.75 to $2.00 per thousand gallons
in June 1976, excluding treatment and disposal of the highly polluting
waste-brine stream.
RECOMMENDATIONS
More study needs to be made of conventional means of removing heavy
metals from coal pile drainage. 'Particularly, methods using lime/limestone
on soda ash for pH adjustment should be investigated.
Reverse osmosis can be used to remove residual heavy metals after
the bulk has been removed by pH adjustment and suspended solids removal.
A suggested flow diagram is shown in Figure A. To help prevent calcium
sulfate scale, either caustic soda or soda ash is chosen as the base for
pH adjustment. This will allow greater water recovery and lower brine
disposal costs. Brine is recycled so that those species on the threshold
of precipitation (primarily metal oxides) can be removed under control
in the sedimentation basin. Further water recovery is thus possible.
Design and optimization of such a system would require a considerable
development effort, including pilot plant work and analysis. Optimization
of the precipitation and sedimentation processes would be desirable. In
fact, reverse osmosis pretreatment may remove heavy metals to the extent
that the reverse osmosis itself is not required.
-58-
-------
TABLE 16. OPERATING PARAMETERS FOR SPIRAL-WOUND
REVERSE OSMOSIS STUDY AT 75% RECOVERY
AT MOCANAQUA, PENNSYLVANIA
Constituent Value
Raw water feed flow, gal/rain 6.02
Product waterflow, gal/min 4.50
Brine water discharged, gal/min 1.52
Brine water recycled, gal/min 4.26
Minimum brine/product flow ratio
(Tubes 1 and 2), ratio/module 5:1
Maximum brine/product flow ratio
(Tube 3), ratio/module 12:1
Water recovery, % 74.8
Recovery of blended feed, % 43.8
Feed pressure, psig 602
Feed water temperature, °F 62.6
Tube 1 flux, gal/ft2-day at
600 psi net and 77°F 19.56
Tube 2 flux, gal/ft2-day at
600 psi net and 77°F 19.52
Tube 3 flux, gal/ft2-day at
600 psi net and 77°F 18.77
Length of run, h 1672
aAll values are means from 73 data sets.
Source: Wilmoth, R. C., and R. B. Scott, Water Recovery from
Coal Pile Drainage. Proc. 3rd National Conference on
Complete Water Reuse, Cincinnati, Ohio, June 1976.
Reprinted by permission of the author.
-59-
-------
TABLE 17. CHEMISTRY ANALYSES FOR REVERSE OSMOSIS TREATMENT
OF ACID MINE DRAINAGE
Raw feed Waste brine Effluent
Conductivity
(|jmho/cm)
Acidity
1000 3600 17
240 810 32
, as CaC03)
pH (unit)
Ca (mg/£)
Mg (mg/A)
Total iron (mg/£)
Total Fe2+ (mg/A)
Na (mg/£)
Al (mg/je)
Mn (rog/£)
S04 (mg/£)
Alkalinity
(mg/£, as CaC03)
TDS (mg/A)
Source: Wilmoth, R. C.,
3.4
130
90
77
64
-
12
12
750
0
1300
and R. B. Scott.
2.9
490
310
330
250
-
44
24
2800
0
4500
Water Recove
4.5
0.5
0.6
0.5
0.5
-
0.2
0.08
2.2
0
5
:ry from Coal Pile
Drainage. Proc., 3rd National Conf. on Complete Water Reuse,
Cincinnati, Ohio, June 1976. Reprinted by permission of the
author.
-60-
-------
COAL PILE DRAINAGE
CAUSTIC SODA
OR SODA ASH
BRINE RECYCLE
BACKWASH
PRODUCT
FOR REUSE
FILTRATION
REVERSE
OSMOSIS
CARTRIDGE
FILTERS
SEDIMENTATION
SLUDGE
METAL OXIDES
MIX
TANK
•ACID
-ANTISCALANT
Figure 4. Flow diagram for reverse osmosis treatment of coal pile drainage.
-61-
-------
SECTION 9
BOILER MAKEUP WATER TREATMENT WASTES
CHARACTERISTICS OF WASTEWATER
The regenerants from boiler makeup water demineralizers typically
contain about 10-20 percent solids, mostly sodium sulfate, and include
all of the ions present in the water supply, but concentrated manyfold.
The volume of regenerant waste produced depends on the size of the units
in the system and the manner in which the rinse water is handled. Some
multiple-bed ion-exchange demineralizers reduce the amount of rinse waste
by using rinse-recycling techniques.68 The frequency of regeneration,
which will determine the amount of regeneration chemicals used and the
wastes produced, depends on the quality of feedwater and other factors
and varies from daily to weekly. If a reverse osmosis process is used
ahead of the ion exchange system, savings in overall annual costs can be
achieved. The quality of the raw feedwater varies from one plant to
another, and thus the extent of pretreatment required also varies.
Because a large amount of acid (typically sulfuric) and caustic (typi-
cally sodium hydroxide) are used in ion exchange regeneration, compared
with the small quantity of chemicals used for pH control and scale pre-
vention in reverse osmosis systems, the use of reverse osmosis will cut
down the total dissolved solids in the wastewater.
ALTERNATIVE TREATMENTS
Both reverse osmosis and electrodialysis69 have been suggested as
roughing demineralizers for boiler makeup water. Both generally require
pretreatment to prevent membrane scaling and fouling and are followed by
polishing demineralizers. Reverse osmosis is generally preferred because
it produces high quality water, free of colloids, in one stage. Electro-
dialysis reduces salt content by about 30 to 50 percent per stage but
does not remove colloids. Colloids often cause fouling of the polishing
demineralizers.
Rowland et al.70 first reported the use of reverse osmosis as a precur-
sor to a demineralizer system for the Burbank (California) Public Service
Department. The process resulted in a cost savings of more than one-third
compared with the conventional system of ion exchange only. There were
five reasons for these savings: (1) when preceded by reverse osmosis,
the demineralizers produce five to ten times more deionized water between
regenerations; (2) manpower requirements for operation and maintenance
are reduced to less than 50 percent of that for an ion exchange system
only; (3) reliability of product water quality is improved; (4) chemical
-62-
-------
requirements are lowered by 90 to 95 percent, reducing costs and minimizing
waste disposal problems; and (5) the life of the ion exchange resin is
extended. The typical brine and product analyses for well-water feedstock
are shown in Table 18.
TABLE 18. WELL WATER, BRINE, AND PRODUCT ANALYSES—
BURBANK PUBLIC SERVICE DEPARTMENT
Well-water Concentration of consitituent
constituent
Silica
Calcium
Magnesium
Sodium
Carbonate
Bicarbonate
Chloride
Sulfate
Phosphate
Nitrate
Total salines (approximate)
Total hardness (as CaC03)
fpH = 5.8
pH = 6.1
CpH = 5.1
Source: Rowland, H. , I. Nasbaum,
Q
Well water
24
56
13
33
0
90
23
163
Trace
4
365
193
and F. J. Jester
Brine
70
186
43
110
0
231
69
582
0
9
1185
641
Consider
(mg/£) in
Product
5
2
Trace
5
0
19
4
8
Trace
2
35
5
Reverse
Osmosis for Producing Feedwater. Power, December 1971.
Reprinted with permission from Power Magazine, Copyright
McGraw-Hill, Inc., 1971.
Wadlington13 reported that a reverse osmosis unit installed ahead
of an existing ion exchange system reduced the chemical regenerant
requirements, increased the resin life, reduced total chemical waste, and
provided a capability for treating increases in total dissolved solids
with little or no increase in operating cost, which is unlike ion exchange
treatment alone. The savings in chemical regenerants alone were enough
to pay back the reverse osmosis capital cost in 1.5 years.
Skrinde et al.71 compared the costs of using these three processes
to treat a specific source of water, which is characterized in Table 19,
to meet the required quality water shown in Table 20. Process II (softening
followed by ion exchange) and process III (reverse osmosis followed by ion
exchange) have the common characteristics of providing a pretreatment step
ahead of a final polish by demineralization. Process II consists of first-
stage softening by the addition of lime and soda ash, thus precipitating
hardness. This results in hardness reduction by replacing calcium and
magnesium ions with sodium ions, while the total dissolved solids remain
relatively unchanged. Softening is followed by ion exchange to remove the
-63-
-------
TABLE 19. CHARACTERISTICS OF RAW WATER?
Analysis
Well Water City water
Total dissolved solids, mg/S.
Total hardness, as CaC03, mg/S.
Alkalinity, as CaC03, mg/S,
Noncarbonate hardness, as CaC03, mg/S.
Silica, as Si02, mg/S,
Sulfate, as S04, mg/S,
Chloride, as Cl, mg/S,
Fluoride, as F, mg/S,
Calcium, as Ca, mg/S.
Magnesium, as Mg, mg/S,
Sodium, as Na, mg/S.
Iron, as Fe, mg/2
Odor
Potassium, as K, mg/S,
Color, units
Carbon dioxide, as C02, mg/£
Bicarbonate, as CaC03, mg/S.
Carbonate, as CaC03, mg/£
Hydroxide, as CaC03, mg/S.
pH, units
pHs
Ryznar index (2 pHs-pH)
Langlier index (pH-pHs)
Hydrogen sulfide, as HaS, mg/S,
Turbidity, JTU
Organic carbon, mg/S.
360.0
240.0
161.0
79.0
11.0
49.5
24.0
0.7
59.0
22.0
17.8
0.1
None
1.75
0.5
0.0
149.0
12.0
0.0
7.9
7.5
7.1
+0.4
_
1.8
0.64
401.0
262.0
136.0
126.0
19.4
64.2
21.4
0.6
55.2
25.5
12.3
0.1
None
2.2
0.7
0.0
127.0
10.1
0.0
7.5
7.7
7.9
-0.2
0.4
1.7
-
Source: Skrinde, R. T., W. M. Steeves, L. S. Shields, and T. L. Tang.
Economic and Technical Evaluation of Reverse Osmosis for
Industrial Water Demineralization. Paper presented at
Industrial Water and Pollution Conf. and Exposition, Chicago,
Illinois, March 14-16, 1973.
-64-
-------
TABLE 20. WATER QUALITY REQUIREMENTS FOR COMBUSTION
TURBINE INJECTION
Analysis Concentration
(mg/A)
Dissolved solids 90
Suspended solids 9
Alkalinity 3
Silica 9
Sulfide 1
Phosphate 1.5
Chloride 6
Iron 0.01
Copper 0.005
Calcium 10
Sodium and potassium (combined) 2
Source: Skrinde, R. T., W. M. Steeves, L. S. Shields, and T. L. Tang.
Economic and Technical Evaluation of Reverse Osmosis for
Industrial Water Demineralization. Paper presented at Indus-
trial Water and Pollution Conf. and Exposition, Chicago,
Illinois. March 14-16, 1973.
remaining hardness and sodium, producing a final product virtually free of
contaminants. The basic difference in pretreatment between processes II
and III is in the relative amount of total dissolved solids removed by the
pretreatment process. The reverse osmosis in process III would remove
90 to 95 percent of the total dissolved solids, and the softening in proc-
ess II would remove a maximum of only 60 percent of the hardness and an
average of about 30 percent of the total dissolved solids. Reverse
osmosis produces a much higher quality water than lime-softening units,
making the demineralization that follows much more efficient and inexpen-
sive. The lime-softening process produces substantial amounts of sludge,
increasing the cost of its disposal. On the basis of both cost and opera-
tional criteria, process III was preferred to process II.
A capital cost analysis71 indicated that processes I (ion exchange
only) and III were quite similar (see Tables 21 and 22). A straight ion
exchange system is preferred for smaller installations. It was determined
that operating costs for process I would be somewhat higher than for process
III. On an overall annual cost basis, which included capital depreciation
and operation and maintenance costs, process III, consisting of reverse
osmosis followed by an ion exchange polishing step, was judged to be the
most economically sound method for boiler makeup water treatment.
Depending on the source of boiler makeup water, which could be the
city water supply, well water at the plant site, or surface waters, the
quality of feedwater to a reverse osmosis-ion exchange system might differ
significantly from one source to another. Suspended solids and organic
matter may be more significant in the surface water than in ground water,
-65-
-------
TABLE 21. COST COMPARISONS OF THREE PROCESSES
Capital cost ($) process for
Item
Chemical feed systems
Filters, 5 to 10 pm
Reverse osmosis units (incl. pumps)
Lime softener
Prefilters
Carbon filters
Degasifier
Pumps
Demineralizer system
Water storage
Subtotals
Contingency and technical direction
of installation (10%)
Total capital costs
Chemical feed
Cartridge replacements, 5 to 10 Mm
Power for high-pressure pumping
Reverse osmosis membrane service
contract
Lime softening
Other power requirements
Demineralizer regeneration chemicals
Demineralizer resin replacement
Operation labor
I
5,000
-
-
-
10,000
12,000
10,000
31,000
340,000
60,000
468,000
46,000
514,000
637
-
-
-
-
1,460
78,489
17,659
14,600
II
_
-
-
80,000
10,000
-
10,000
22,000
280,000
60,000
462,000
46,000
508,000
-
-
-
-
5,886
1,300
58,867
14,716
23,725
III
7,000
6,000
220,000
-
10,000
-
10,000
17,000
200,000
60,000
530,000
53,000
583,000
5,169
1,682
6,216
22,075
-
960
5,886
5,886
10,950
Total operating costs
Present worth of operating costs
Comparative cost
Overall unit cost, $/1000 gal.
112,845 104,494 58,824
1,867,600 1,729,391 973,546
2,381,600 2,237,391 1,556,546
1.33 1.26 0.96
Source: Skrinde, R. T., W. M. Steeves, L. S. Shields, and T. L. Tang.
Economic and Technical Evaluation of Reverse Osmosis for
Industrial Water Demineralization. Paper presented at
Industrial Water and Pollution Conference and Exposition,
Chicago, Illinois, March 14-16, 1973.
-66-
-------
TABLE 22. BASIS OF CALCULATIONS FOR TABLE 21
Lime softening
Demineralizer series regeneration
Alternate 1
Alternate 2
Alternate 3
Demineralizer resin regeneration
Alternate 1
Alternate 2
Alternate 3
Energy
Chemicals
H2S04
NaOH
Na2S03
Polyphosphate
Filter cartridge replacement, 5 to 10 pro
Reverse osmosis membrane replacement
Operation labor
Alternate 1
Alternate 2
Alternate 3
General maintenance
Operating cost escalation
(compounded annually)
Annual interest rate used for calculating
present work of operating costs
Bond redemption period
$4.00/100,000 gal
$4.00/100,000 gal
$40.00/100,000 gal
$53.33/100,000 gal
$4.00/100,000 gal
$10.00/100,000 gal
$12.00/100,000 gal
$0.006l/kWh
$0.02/lb
$0.04/lb
$0.08/lb
$0.l6/lb
$1.00/100,000 gal
$15.00/100,000 gal
$10.00/h
3.0 man-hours/day
6.5 man-hours/day
4.0 man-hours/day
Not included
5%
10%
30 years
Source: Skrinde, R. T., W. M. Steeves, L. S. Shields, and T. L. Tang.
Economic and Technical Evaluation of Reverse Osmosis for
Industrial Water Demineralization. Paper presented at Indus-
trial Water and Pollution Conf. and Exposition, Chicago,
Illinois, March 14-16, 1973.
-67-
-------
and on the other hand, the concentration of total dissolved solids may
be rat ti higher in some well waters than in the waters of streams and lakes.
In general, a filtration system (such as a dual-media filter) is needed
to remove the suspended solids, followed by a carbon bed for removing
the residual suspended solids, organic materials, and chlorine. Chlorine
is harmful to some membranes. Antiscalant addition, pH control, and car-
tridge filters are also common pretreatment needs.
After reverse osmosis, degasifiers generally are used to remove the
carbon dioxide to relieve the burden to the anion-ion exchange units.
Cost comparisons between the system of ion exchange only and the
combination of reverse osmosis and ion exchange are shown in Table 23
for the Burbank Public Service Department in California.70
Truby72 reported a cost comparison of a boiler feedwater treatment
system with ion exchange only and with a combination of reverse osmosis
and ion exchange, as shown in Table 24.
Table 25 compares condensate makeup costs with and without a reverse
osmosis unit.
Skrinde et al.71 compared costs for three processes. Table 21 shows
the calculated costs based on criteria shown in Table 22 for a 240-gal/min
boiler makeup water treatment system. The overall unit cost is $0.96
per thousand gallons of product water for the recommended reverse osmosis-
ion exchange combination systems, which is lower than the cost for two
other processes.
RECOMMENDATIONS
Reverse osmosis is a proven technology for removing the bulk of thfc
dissolved solids loading before polishing demineralizers. Its use
reduces the total discharge of dissolved solids since little chemical
treatment need be used. Presently it is economically attractive to use
reverse osmosis for feedwaters with dissolved solids concentrations above
250 to 500 ppm. If penalties were charged for discharging dissolved solids,
reverse osmosis would be attractive for feedwaters with lower concentrations.
-68-
-------
TABLE 23. RESULTS OF THE BURBANK PUBLIC SERVICE DEPARTMENT
COMPARISON AND ANALYSIS OF OPERATING COSTS OF
DEMINERALIZER AND REVERSE OSMOSIS SYSTEMS
Operating history
Cost element May- June 1970 May- June 1971
Demineralizer, 36-in. diam.
Product water, gal 746,300 862,857
Regenerations 67 28
Operating labor
(1 hr/regeneration) , $ 361 151
Chemicals, $ 781 326
Demineralizer, 30-in. diarn.3
Product water, gal 35,790 266,000
Regenerations 6 6
Operating labor
(1 hr/regeneration), $ 32 32
Chemicals, $ 38 38
Feedwater cost ($53/acre-ft) , $ 127 40b
Regeneration water
Quantity, gal 236,000 118,000
Cost, $ 39 19
Reverse osmosis system
Operating labor (1 hr/day) , $ 328
Electric power ($1.20/day), $ 73
Chemicals ($0.29/day), $ 18
Feedwater cost ($53/acre-ft) , $ 233
Total cost, $ 1378 1258
Cost/1000 gal deionized water, $ 1.77 1.16
yi
Only in operation when 36-in. -diameter unit is being regenerated.
Not all product water flows through the reverse osmosis system before
deionization; some raw water is mixed with the reverse osmosis product
before the demineralizer because maximum reverse osmosis system through-
put is less than the makeup requirements.
Source: Rowland, H. , I. Nasbaum, and F. J. Jester. Consider Reverse
Osmosis for Producing Feedwater. Power, December 1971.
Reprinted with permission from Power Magazine. Copyright
McGraw-Hill, Inc., 1971.
-69-
-------
TABLE 24. COST COMPARISON OF A BOILER FEEDWATER TREATMENT SYSTEM
WITH ION EXCHANGE ONLY AND WITH A COMBINATION OF
REVERSE OSMOSIS AND ION EXCHANGE
Cost item
Ion exchange only
($/yr) ($/1000 gal)
Combination of
reverse osmosis
and ion exchange
($7yr)($/1000 gal)
Capital
Labor
Maintenance
Pretreatment chemicals
Demineralization chemicals
Resin
Modules
Waste chemicals
Energy
160,230
96,000
50,000
46,460
172,680
32,310
-
81,300
12,430
1.015
0.608
0.317
0.294
1.094
0.205
-
0.515
0.079
168,430
96,000
50,000
46,440
15,880
8,950
30,880
2,570
40,270
1.067
0.608
0.317
0.294
0.101
0.057
0.196
0.016
0.255
Total
651,410
4.13
459,420
2.91
Cost based on water containing 960 mg/£ total dissolved solids.
Source: Truby, R. Reverse Osmosis for Boiler Feedwater Treatment. Power
Engineering, December 1976. Reprinted by permission.
-70-
-------
TABLE 25. COMPARISON OF CONDENSATE MAKEUP COSTS
WITH AND WITHOUT REVERSE OSMOSIS UNIT
Without reverse osmosis
Pretreatment
Cation-anion makeup demineralizer
(estimated 40,000-gal service run)
Mixed-bed makeup demineralizer
One year's operation, 1.5 x 107 gal $83,700
With reverse osmosis (2-yr old membranes)
Pretreatment
Operating membrane (replacement every 2 years)
Cation-anion makeup demineralizer
(200,000-gal service run)
Mixed-bed makeup demineralizer
One year's operation, 1.5 x 107 gal $25,200
With reverse osmosis (new membranes)
Pretreatment
Operating membrane (replacement every 2 years)
Cation-anion makeup demineralizer
(1,000,000-gal service run)
Mixed-bed makeup demineralizer
One year's operation, 1.5 x 107 gal $12,000
Source: Wadlington, M. Chemical Regenerant Savings Can Pay for a
Reverse Osmosis Unit. Industrial Water Engineering,
June-July 1976. Reprinted by permisssion.
-71-
-------
SECTION 10
BOILER SLOWDOWN
CHARACTERISTICS OF WASTEWATER
The quantity of boiler blowdown varies with allowable boiler water
solids content, condenser leakage, makeup water quality, and boiler opera-
ting conditions. Old plants operating at low pressures can tolerate con-
centrations of total dissolved solids in boiler water as high as 2500
rag/£> with an average blowdown rate of 10 percent. Modern boilers can
operate with concentrations of total dissolved solids of less than 50
rog/£> which will have a blowdown range of 0.1 to 1 percent of steamflow,
or about 20 to 200 gpd/MW. The chemical characteristics of boiler blow-
down depend on the construction materials, the feedwater chemical control
program, and the type of plant operation. Generally, the boiler blowdown
has a basic pH, a low concentration of total dissolved solids, and very
small amounts of copper, iron, nickel, and chromium; it can be reused in
other operations of the power plant with or without treatment.32
According to a TVA study,25 boiler design pressure has the most signi-
ficant effect on the quality of boiler feedwater, condensate, and blow-
down. Scale of calcium and magnesium salts and deposits of iron oxide,
elemental copper, and copper oxide are frequently found in boilers, even
when operating with very pure feedwater. Chemicals such as sodium phos-
phates, EDTA, and NTA are used to prevent these salts from precipitating
on boiler surfaces. To prevent corrosion, boiler water must be neutra-
lized with alkalies such as sodium carbonate, sodium hydroxide, or ammonia.
Chemical deoxygenation is commonly used to prevent corrosion of metallic
surfaces.
The quantities of ammonia, phosphate, iron, and copper in boiler
blowdown may be significant,26 depending on the kind of chemical addi-
tives and the type of boiler used.
ALTERNATIVE TREATMENTS
Normally (for modern high pressure boilers) the quality of boiler
blowdown is better than that of the raw makeup water supply. Thus, the
blowdown is recycled back to the inlet to the makeup water treatment sys-
tem, which could use the reverse osmosis-ion exchange combination process
as mentioned in Section 9. The blowdown from a low-pressure, drum-type
boiler generally contains higher concentrations of total dissolved solids
and can be discharged into neutral and alkaline ash ponds to precipitate
iron and copper. This technique is being practiced by TVA. For those
-72-
-------
power plants with acidic ash ponds or where the ash is dry-handled, the
boiler blowdown has to be treated with other methods. To achieve the
goal of water reuse, the lower quality blowdown can be treated by a
reverse osmosis-ion exchange boiler makeup treatment system with equal
or higher cost-effectiveness than that for the high-pressure boiler
blowdown.
The pretreatment needed generally includes pH control (lowering the
pH by adding H2S04), antiscalent addition (sodium hexametaphosphate),
and cartridge filtration. Degasifier and ion exchange units may be needed
to treat the product water from the reverse osmosis system before it is
reused as the boiler makeup water.
The process recommended for boiler blowdown treatment is essentially
similar to the boiler makeup water treatment system proposed in Section
9, except that the prefilters (such as dual-media filters) are excluded.
Cost should be similar to those discussed in the section on boiler makeup
water.
RECOMMENDATIONS
Boiler blowdown can be returned to the boiler makeup water treatment
system, which may use reverse osmosis as a roughing demineralizer.
-73-
-------
SECTION 11
CHEMICAL CLEANING WASTES
CHARACTERISTICS OF WASTEWATER
Periodic cleaning of metal surfaces in steam generators is required
to remove the accumulated deposits, which raise water wall temperatures,
reduce heat transfer, and lower cycle efficiency. A wide range of
cleaning compounds and neutralizing agents is used. The frequency of
cleaning and the quantity and quality of pollutants discharged depend on
plant reliability requirements, construction materials, use of corrosion
inhibitors, antifoulants, and biocides, compliance with equipment and
water quality control requirements, and mode of operation of unit and
equipment. Required cleaning frequencies for various equipment and the
corresponding waste volume are typified in Table 26.32
Metal cleaning processes used by TVA are either circulation or
soaking methods. Waste streams from the circulation method consist of
(1) an acid waste in which the iron is chelated with cleaning solvent
such as citric acid or hydroxyacetic-formic acid and (2) a passivation
drain. Waste streams from the soaking method consist of (1) an acid
waste in which the iron is not chelated with cleaning solvent such as
hydrochloric acid, (2) an alkaline copper waste in which the copper is
strongly complexed with ammonia, and (3) a passivation drain. Waste
solutions from these processes contain large amounts of metallic species,
nitrogen and phosphate compounds, and organic materials. The characteris-
tics of the chemical cleaning wastes depend on the method of cleaning,
type of boiler and accessory equipment cleaned, and the types of cleaning
chemicals. Characteristics of the wastes resulting from cleaning processes
at several TVA plants are presented in Tables 27, 28, and 29. Character-
istics of the wastes for cleanings at the same plant may vary. The volume
of wastewater resulting from cleaning processes at TVA plants ranges from
about 0.3 to 0.9 x 106 gal per unit per cleaning.73
ALTERNATIVE TREATMENTS
Results of a recent study conducted by TVA showed that the acid-
and neutralization-phase cleaning wastes can be treated to meet the
effluent limitation guidelines (1 mg/JH copper and 1 mg/Jfc iron) by a
conventional chemical precipitation process. Treatment of the alkaline-
phase cleaning waste is complicated by the strong copper-ammonia complexes.
-74-
-------
TABLE 26. TYPICAL PLANT EQUIPMENT WHICH MAY REQUIRE PERIODIC CLEANING
Equipment
Materials
Waste volume per
cleaning or wash
(gal/MW)
Typical
frequency
per year
Boiler, steam generators
Feedwater heaters
Condenser tubes
Air preheaters
Boiler fireside
Miscellaneous equipment,
piping
Stainless steel
Carbon steel
Inconel
Carbon steel
Stainless steel
Admiralty brass
90/10, 80/10, 70/10 Cu-Ni
Monel
Arsenical copper
Admiralty brass
90/10 Cu-Ni
Stainless steel
Carbon steel
50-400
0.2-1
Only
preoperational
100-1500
100-1000
4-24
2-8
Carbon steel
Stainless steel
Copper alloys
Source: Wright, J. H., and S. J. Dea. Water Pollution Control in the Power Generation
Industry. Industrial Wastewater Management Handbook, ed. H. S. Azad, Copyright
McGraw-Hill, 1976. Reprinted by permission from McGraw-Hill Book Company.
-------
TABLE 27. CHARACTERISTICS OF CHEMICAL CLEANING WASTEWATER--ACID PHASE COMPOSITE
Constituent
Waste Volume
pH
Suspended solids
Total organic
carbon
Chemical oxygen
demand
Oil & grease
Phenols
Silica
Ammonia nitrogen
Organic nitrogen
Nitrate & nitrite
nitrogen
Phosphorus
Sulfate
Aluminum
Arsenic
Barium
Beryllium
Cadmium
Calcium
Chromium
Copper
Iron, total
Lead
Magnesium
Manganese
Mercury
Nickel
Potassium
Selenium
Silver
Sodium
Tin
Zinc
Iron, ferrous
Unit
millions
of
gallons
units
mg/JZ
mg/£
mg/£
mg/fi
mg/S.
mg/X.
mg/£
mg/S.
mg/.e
mg/S,
mg/fi
mg/S.
mg/ll
mg/S.
mg/S.
mg/S,
mg/S.
mg/S.
tog/11
mg/S.
mg/i
mg/S.
mg/S.
mg/S.
mg/S.
mg/S.
mg/S.
mg/S.
mg/i
mg/i
mg/S.
mg/i
Aa
0.175-
0.250
3.3
57
4600
9900
23
0.050
19
325
225
-
1.2
-
-
0.008
-
-
<0.001
16
<0.005
0.69
4200
-
-
19
-
110
-
-
-
-
-
0.94
"
AC
0.175-
0.250
10.1
51
10000
16000
16
0.013
-
6200
6400
49
1.1
<4
1.8
0.41
0.2
<0.01
0.12
20
15
0.7
4800
1.1
1.6
37
<0.0004
8.3
3.1
<0.008
0.03
840
1.5
2.5
4800
cb
0.217
0.8
8
240
1200
<5
0.065
66
140
0.06
0.07
30
<1
6.5
0.060
<0.1
<0.01
0.010
42
1.5
2.2
1300
0.4
8.7
6.9
<0.0002
77
1.4
<0.004
0.02
31
<1
5.9
"
Plant
Gb
0.099
0.7
120
90
1500
11
0.070
120
80
140
<0.01
50
10
6.6
0.010
0.4
<0.01
0.051
70
6
7.6
3820
3.8
6.5
29
<0.0002
260
2.3
<0.002
0.02
74
7.3
170
"
Ib
0.087
0.7
18
1800
1200
7.6
0.035
240
220
75
<0.01
35
<1
7.0
0.030
0.1
<0.01
0.032
53
1.1
18
1420
0.86
5.7
10
0.0002
170
1.5
<0.002
0.07
40
<1
34
Kb
0.070
0.5
35
220
1900
20
0.020
31
290
10
<0.01
50
<1
8.2
0.055
0.3
<0.01
<0.10
64
<8.3
13
3720
5.2
8.8
28
0.0002
300
1.8
<0.002
0.03
49
2.8
53
Lb
0.090
0.7
33
120
1500
23
0.025
-
150
870
-
45
-
-
0.035
-
-
0.001
74
0.005
47
2780
<0.010
-
22
-
150
-
-
-
-
-
24
~
Circulation method with hydroxyacetic-formic acid.
Soaking method with hydrochloric acid.
Circulation method with citric acid followed by addition of NH4OH.
Source: Steiner, G. R., C. L. McEntyre, and T.-Y. J. Chu. Treatment of Metal Cleaning Wastes at
TVA Power Plants. Paper presented at 84th National Meeting of American Institute of
Chemical Engineers, Atlanta, Georgia, February 26-March 1, 1978. 44 pp.
-76-
-------
TABLE 28. CHARACTERISTICS OF CHEMICAL CLEANING WASTEWATER--
ALKALINE PHASE COMPOSITE
Constituent
Waste Volume
Dissolved solids
Suspended solids
Chemical oxygen
demand
Oil & grease
Silica
Ammonia nitrogen
Organic nitrogen
Nitrate & nitrite
nitrogen
Phosphorus
Bromide
Chloride
Fluoride
Sulfate
Aluminum
Arsenic
Barium
Beryllium
Cadmium
Calcium
Chromium
Copper
Iron
Lead
Magnesium
Manganese
Mercury
Nickel
Potassium
Selenium
Silver
Sodium
Tin
Zinc
Unit
millions
of
gallons
mg/£
mg/£
mg/A
mg/£
mg/£
mg/2
mg/H
mg/£
mg/£
mg/£
mg/£
mg/£
mg/£
mg/£
rag/A
mg/£
mg/£
mg/£
mg/£
mg/£
mg/A
mg/A
mg/£
mg/£
mg/£
mg/£
mg/A
mg/£
mg/A
mg/£
mg/£
mg/£
mg/A
C
0.217
340
8
24
<5
7.2
700
40
0.04
10
52
60
1.5
<1
<0.2
0.048
<0.1
<0.01
<0.001
3.0
<0.005
100
1.7
<0.010
2.9
0.03
<0.0002
0.52
70
<0.002
<0.01
3.7
<1
0.06
G
0.165
1400
71
120
<5
14
2000
<10
0.51
30
<5
-
6.1
<1
<0.2
<0.005
<0.1
<0.01
<0.001
0.4
<0.005
790
4.9
<0.010
0.67
0.04
<0.0002
2.5
220
<0.002
0.02
15
<1
0.54
Plant
I
0.116
1400
47
61
<5
50
2100
270
0.21
2.0
-
350
6
<1
0.4
0.012
<0.1
<0.01
<0.001
<1.0
<0.005
510
2.9
<0.010
0.1
0.03
0.0006
0.94
<0.1
<0 . 002
<0.01
0.1
<1
0.17
K
0.140
1900
200
84
<5
13
4200
2600
0.46
400
-
110
5.6
18
0.3
<0.010
<0.1
<0.01
<0.001
4.9
<0.005
680
2.2
<0.010
0.51
0.05
0.0006
3.3
230
<0.002
0.04
12
<1
0.35
L
0.120
1600
7
160
-
-
4700
3600
-
275
<5
180
-
-
-
0.030
-
-
-
-
-
590
<0.05
-
-
-
-
0.62
260
-
-
-
-
0.19
Source: Steiner, G. R., C. L. McEntyre, and T.-Y. J. Chu. Treatment of
Metal Cleaning Wastes at TVA Power Plants. Paper presented at
84th National Meeting of American Institute of Chemical Engineers,
Atlanta, Georgia, February 26-March 1, 1978. 44 pp.
-77-
-------
TABLE 29. CHARACTERISTICS OF CHEMICAL CLEANING
WASTEWATER—PASSIVATION DRAIN
oo
I
Plant3
Constituent
Waste volume
PH
Dissolved solids
Suspended solids
Chemical oxygen
demand
Oil & grease
Ammonia nitrogen
Organic nitrogen
Nitrate & nitrite
nitrogen
Phosphorus
Copper
Iron
Sodium
Hydrazine
Unit
millions
of
gallons
units
mg/2
mg/£
mg/£
mg/4
mg/A
mg/£
mg/£
mg/A
mg/A
mg/£
mg/A
mg/£
A
0.110
9.5
70
10
92
<5
80
110
-
-
0.13
4.6
-
0.025
C
0.072
11.9
4000
88
68
-
3.2
4.3
<0.01
-
0.65
3.2
150
G
0.033
12.2
4500
140
68
<5
12
5
0.03
770
25
28
1800
I
0.029
12.0
5400
20
46
-
22
38
0.50
850
1.7
4.4
1.3
K
0.035
10.7
2500
10
68
21
5
0.01
420
0.4
2.3
960
L
0.030
11.9
5300
12
76
<5
32
<0.00
0.03
980
2.8
1.3
2400
0.001
Plant A uses circulation method; all other plants use soaking method.
Source: Steiner, G. R., C. L. McEntyre, and T.-Y. J. Chu. Treatment of
Metal Cleaning Wastes at TVA Power Plants. Paper presented at
84th National Meeting of American Institute of Chemical Engineers,
Atlanta, Georgia, February 26-March 1, 1978. 44 pp.
-------
Treatment of hydrochloric acid cleaning wastes from TVA plants73 by
adding lime to reach a pH of about 10 decreased the concentration of
dissolved iron to less than 1.0 mg/£. It also removed the lower concen-
trations of copper (2.2 to 47 mg/l) present in the waste to 1.0 rag/A.
In a similar study, hydroxyacetic-formic acid wastes were treated with
lime. The concentration of dissolved iron reached 1.0 mg/JH at pH 10.1
and 0.33 mg/£ at pH 12.0. Citric acid wastes were treated with hydrogen
peroxide followed by pH adjustment to 13.0 with caustic soda, which
reduced the dissolved iron concentrations to below 1.0 mg/£. Waste from
the acid stages of the chemical cleaning process can be treated in an
alkaline ash pond that has a pH of about 9, where the high concentrations
of iron in the wastes will precipitate and the dissolved iron can be
reduced to an "equivalent" concentration of 1.0 mg/Jd in which the dilution
factor is excluded.
The theoretical minimum solubility for copper hydroxides occurs at
pH values of about 8.9 with a resulting dissolved copper concentration
of 0.004 mg/2. However, adjustment of alkaline cleaning waste to this
pH does not reduce the dissolved copper in the waste to less than 1.0
mg/£ because the copper is complexed with ammonia.73 Copper will be
removed from solution to levels less than 1.0 mg/£ if the copper-ammonia
complex is broken and the waste is still alkaline. Treatment of alkaline
wastes containing 730 mg/£ of copper and 27,000 mg/£ of ammonia nitrogen
with sodium sulfide reduces the dissolved copper level to below 1.0 mg/£.
However, contact of any acid with sodium sulfide produces hydrogen
sulfide gas. If the pH of the waste is adjusted to below 9.0, the
residual sulfide in the waste would be converted to hydrogen sulfide.
Due to the potential safety hazard from hydrogen sulfide, treatment of
alkaline wastes with sodium sulfide is not recommended.73 Studies with
lime and caustic soda treatment indicated that the concentration of
dissolved copper in the alkaline chemical cleaning wastes can be reduced
to 1.0 mg/£ if the pH of the wastes is raised to a value of 11.5 to 12.0
and sufficient reaction time is allowed. Agitation of the wastes by
circulation or aeration at high pH values speeds removal of ammonia,
which, in turn, accelerates removal of dissolved copper.73 The ion
exchange process has also been tried to treat the alkaline chemical
cleaning wastes. It was found to be impractical because of the low
breakthrough capacity and low regeneration efficiency of the resins,
which would not provide cost-effective treatment of the wastes.73
Studies on treating the alkaline chemical cleaning wastes in TVA ash
ponds showed that the chelated copper-ammonia bonds can be broken by
dilution with ash pond water, which allows precipitation of the copper
at alkaline pH values. Additional copper is removed by adsorption on
fly ash. A detention time of up to 10 hours is required for breaking
the copper-ammonia bond and for subsequent chemical precipitation of the
copper. 4
Reverse Osmosis
Studies by TVA on treating chemical cleaning wastes by reverse
osmosis, especially the alkaline wastes, were begun because the metal
ions, especially copper, in their complex forms are difficult to reduce
-79-
-------
to the level required by the EPA guidelines with conventional chemical
treatment processes. TVA also hoped to meet the goal of water conser-
vation by treatment and reuse of the overall chemical cleaning wastes.
The alkaline-phase drain cleaning waste generally has a pH of 9 to
12. Experiments were conducted using DuPont hollow-fiber and DuPont
spiral-wound modules to treat straight alkaline wastes and Roga spiral-
wound, Dow hollow-fiber, and Fluid Science tubular modules to treat
neutralized wastes. All alkaline and neutralized wastes were filtered
through 1- to 3-pm fiber filters to remove suspended solids before the
membrane tests. During the neutralization process, the waste was first
adjusted with 36 N hydrochloric acid to a pH of 6.0, where some copper
precipitates; it was then filtered and readjusted to a pH of 4.0.
Results of chemical cleaning waste treatment studies shown in Table 30
indicate that cellulose acetate membrane modules have poorer rejection
on copper complex and total dissolved solids than do polyamide membrane
modules, but all membranes have poor rejection on total ammonia. The
DuPont hollow-fiber module appears most promising. The studies showed
that the polyamide membranes are able to treat the waste under conditions
of low volume recovery. However, it would be economically desirable to
maximize recovery. Therefore, the DuPont modules were selected to treat
the alkaline-phase cleaning waste for high volume recovery. The membrane
fouling problems resulting from the copper precipitates in the unstable
cleaning waste were considered a possible key factor preventing the
operation at the high volume recovery. However, the membrane processes
were still considered practicable if a simple and inexpensive pretreatment
process could be found for this waste. To investigate this possibility,
a study was conducted to neutralize the alkaline-phase cleaning waste
with acid-phase cleaning waste to a pH of 6. After a settling process,
the clarified solution was treated by ultrafiltration to remove the
residual fine particles. Then the ultrafiltration permeate was adjusted
to a pH of 5 by hydrocloric acid addition before treatment by reverse
osmosis.
The characteristics of composite alkaline-phase cleaning waste
(drain plus flush) used for this study is shown in Table 31. A pH of 6
or more is necessary for precipitation of iron from the cleaning wastes
to meet the effluent limits, but the neutralization process cannot
remove copper to the required l-mg/£ concentration. Therefore, the
combined treatment of acid- and alkaline-phase cleaning wastes was
considered as an effective pretreatment process to reduce the strength
of the wastes and to remove iron (from the acid-phase cleaning wastes)
before the membrane processes. After the sludge settling, the clarified
solutions were treated by ultrafiltration (Union Carbide 3NJR module)
for colloidal particle removal.
The test results are shown in Table 32. The ultrafiltration permeates
were then adjusted to a pH of 5 by adding hydrochloric acid. These pre-
treated boiler cleaning wastes were treated by three spiral-wound reverse
osmosis membranes--DuPont, UOP-ROGA, and Osmonics. The test results are
presented in Tables 33, 34, and 35.
-80-
-------
TABLE 30. RESULTS OF REVERSE OSMOSIS STUDIES ON CHEMICAL CLEANING WASTES
o
Flux Rejection Pressure Feedflow
Membrane type Feed composition (gal/min) (%) (psig) (gal/min)
DuPont Hollow Fiber Alkaline copper waste
Cu, 2200 mg/2 Cu, 99.98
NH3-N, 10,500 mg/2 0.85 NH3-N, 20.39 350 4.2
TDS, 4500 mg/2 TDS, 95.84
pH, 11.1
DuPont Spiral Wound Alkaline copper waste
Cu, 1700 mg/2 Cu, 98.82
NH3-N, 9450 mg/2 0.30 NH3-N, 40.74 400 3.8
TDS, 4300 mg/2 TDS, 87.44
pH, 10.0
Roga Spiral Wound Neutralized copper waste
Cu, 820 mg/2 Cu, 83.62
NH3-N, 4100 mg/2 0.35 NH3-N, 58.8 400 4.15
TDS, 2600 mg/2 TDS, 79.04
pH, 4.35
UOP Fluid Science Tubular Neutralized copper waste
Cu, 2000 mg/2 0.06 Cu, 78.45 550 3.75
pH, 4.0
Dow Hollow Fiber Neutralized copper waste
Cu, 410 mg/2 0.01 Cu, 74.95 200 4.1
pH, 4.0
aRejection = 1 - (concentration in permeate)/(concentration in feed).
-------
TABLE 31. CHARACTERISTICS OF COMPOSITE ALKALINE-
PHASE CLEANING WASTE3
Constituent Concentration
pH 10.6
Dissolved solids 1200
Suspended solids 10
Total organic carbon 400
Chemical oxygen demand 50
Oil and grease 3
Silica 10
Ammonia nitrogen 2900
Organic nitrogen 1435
Nitrite and nitrate nitrogen 0.6
Phosphorus 1
Antimony 0.1
Bromide 53
Chloride 63
Fluoride 1
Sulfate
Aluminum <0.2
Arsenic 0.016
Barium <0.1
Beryllium <0.01
Cadmium <0.001
Calcium <1
Chromium 0.01
Copper 313
Iron 0.3
Lead <0.01
Magnesium 0.3
Manganese 0.01
Mercury <0.0002
Nickel 0.6
Potassium 173
Selenium <0.001
Silver 0.07
Sodium 1
Tin <1
Zinc 7
a
.One drain plus two flushes.
All values in milligrams per liter except pH which
is in units.
-82-
-------
TABLE 32. PERFORMANCE OF UNION CARBIDE 3NJR ULTRAFILTRATION MODULE IN
POLISHING SUSPENDED SOLIDS FROM THE PRETREATED BOILER CLEANING WASTES
Suspended solids concentrations (mg/Jg) in
Volume recovery Feed Permeate
0.01
0.5
0.95
6
9
57
2
2
3
Q
V /V-,, where V = volume of permeate and V^. = initial volume of
p x p a
feed solution.
Tables 27 through 31 show that the suspended solids in the chemical
cleaning wastes vary from 7 to 200 mg/2. Before feeding into the membrane
system, these wastes were treated by settling, with or without chemical
assistance, followed by filtration'such as dual-media filtration, pH
control, antiscalent addition, and cartridge filters. Depending on the
type of reuse, chemical cleaning wastes which have been treated by
reverse osmosis may or may not need further treatment.
Cost Estimates
Literature information is unavailable on cost for the membrane
process of treating power plant chemical cleaning wastes. To determine
the magnitude of treatment costs, the cost of treating other industrial
wastes, which have been studied extensively in both technical and economic
terms and which have similar wastewater characteristics as those of
chemical cleaning wastes, were sought.
The wastewaters from the plating industry possess various metal
ions, acid, caustic, and other chemical ingredients such as cyanides and
phenols, which are also the constituents of the chemical cleaning wastes.
The concentrations of each individual constituent may differ, but the
occurrences of the constituents, as a group, are quite compatible,
except that generally the chemical cleaning wastes have much higher
concentrations of suspended solids than do the plating wastes.
Golomb75 studied closed-loop reverse osmosis treatment of nickel-
plating solutions and projected cost estimates (Table 36). Three levels
of operation with dragout rates of 15, 50, and 500 gpd were considered.
A flux rate of 10.7 gal/(ft2-day) at 450 psi pressure was used. Capital
costs were based on straight-line amortization over a 5-year period at
an interest rate of 8.5 percent. Operation was assumed for 240 days per
year, with three 8-hour shifts per day. Operating labor costs of $8.00
per hour and a membrane life of 18 months were assumed. Total operating
and maintenance costs per thousand gallons of permeate are shown in
Table 37.
-83-
-------
TABLE 33. PERFORMANCE OF DUPONT SPIRAL-WOUND REVERSE OSMOSIS MODULE IN TREATING PRETREATED BOILER CLEANING WASTES
a b
i
oo
Constituent
Permeate flow,
gal/min
Feed temperature,
°C
pH, units
Conductivity ,
pmhos/cm
NH3-N, mg/£
Org.-N, mg/£
TOC, mg/£
COD, mg/£
Si02, mg/£
P04-P, mg/£
Cl, mg/£
S04, mg/£
F, mg/£
Ca, mg/£
Mg, mg/£
Na, mg/£
Br, mg/£
Br03, mg/£
Ag, mg/£
As, mg/£
Ba, mg/£
Cr, mg/£
Cu, mg/£
Fe, mg/£
K, mg/£
Ni, mg/£
Zn, mg/£
V /V_ = 0.01 V /V_ = 0.125
p F p F
Feed Permeate Feed Permeate
0.24 - 0.22
26.2 - 26.2
5.01 5.49
20,100 2510
2400 160 2200 180
1700 540
25 5.2
220 47 -
6.5 0.3
0.23 0.02
4000 - 920
1 <1 -
110 22.5
45 3.2
4.1 0.8
9.5 0.5
2.2 0.2
14 0.4
0.06 <0.01
<0.002 <0.002
0.7 <0.01
<0.005 <0.005
36 1.8 36 2.0
0.62 0.07
120 11 -
30 1.8
7.2 0.45
VVF = °"25
Feed Permeate
0.21
-
-
V /VF = 0.375 V j\
Feed Permeate Feed
0.18
_
4.91
Jf = 0.5
Permeate
0.16
-
5.41
25,200 3300
2400 280
-
-
-
9.5 0.1
-
-
-
-
-
-
-
-
-
-
-
-
-
41 2.5
-
-
-
"" ~
3700
-
35
430
18
- -
-
- -
- -
_
-
_
- -
20
_
_
_
_
50
2.6
190
- -
15
440
-
6.2
52
0.4
-
-
-
-
-
-
-
-
1.6
-
-
-
-
4.0
0.19
23
-
1.1
VVF
= 0.95
Feed Permeate
_
-
4.90
29,200
4800
2400
35
530
23
0.1
8400
1
150
76
12
14.2
13
26
0.1
0.003
-
0.13
72
7.7
270
110
18
0.14
-
5.35
4400
620
580
7.4
92
0.6
0.02
1500
<1
70
5.9
1.0
2.1
0.7
2.4
<0.01
<0.002
-
0.011
5.4
0.37
28
4.9
1.4
Operation pressure = 400 psig; feedflow rate = 3.9 gal/min;
.feed solution.
Dashes indicate no data is available.
V = volume of permeate;
V_ = initial volume of
-------
TABLE 34. PERFORMANCE OF UOP-ROGA 4100 SPIRAL-WOUND REVERSE OSMOSIS MODULE
IN TREATING PRETREATED BOILER CLEANING WASTES*' )C
Constituent
Permeate flow,
gal/min
Feed tempera-
ture, °C
pH, units
Conduct ivi ty ,
|jmhos/cm
NH3-N, mg/A
Org.-N, mg/A
TOC, mg/£
COD, mg/£
I Si02, mg/A
°° P04-P, mg/A
I Cl, mg/A
S04, mg/A
F, mg/A
Ca, mg/A
Mg, »g/A
Na, mg/A
Ag, mg/A
As, mg/A
Ba, mg/A
Cr, mg/A
Cu, mg/A
Fe, mg/A
K, mg/A
Ni, mg/A
Zn, mg/A
VVF
Feed
^
25
5.0
11,800
1900
1200
40
320
11
0.07
9500
<1
88
15
2.3
-
0.07
<0.002
<1
0.45
45
0.37
120
36
11
=0.01 V /V = 0.125
Permeate Feed Permeate
1.07 - 1.01
25.5
5.1 5.2
3810
540 1800 700
90
8 - -
120
0.4
0.01
1600
<1 - -
23
3 - -
0.48
-
<0.01
<0.002
<0.1
0.01
4.6 65 7.2
<0.05
25
3.5
1.2
V /V =0.25 V /V
Feed Permeate Feed
0.96
25.7 - 25.8
5.2 5.04 5.2
13,200 4350
2200 600 3400
- - -
_
-
13 0.8
-
- - -
_
_
-
-
-
_
- _ _
-
- - -
63 6.4 69
-
-
_
- - -
F = 0.375 V /V
Permeate Feed
0.93
25.9
5.2
15,600
800 3000
2700
50
400
15
-
-
-
-
-
-
-
-
-
-
-
7.6 94
0.54
170
-
- -
F = 0.05 V /VF = 0.625
Permeate Feed Permeate
0.89 - 0.84
26
5.03 5.18
5450
650 3200 850
720
8
140
0.7
-
- - -
-
_
- - -
_
.
...
- - -
- - -
-
12 98 9
<0.05
33
-
_ — _
V /V-, = 0.75 V /V
Feed Permeate Feed
0.72
26.3 - 27.3
5.14 5.13 5.10
20,000 7700
4200 1400 5900
-
_
-
16 1.7
.
- -
-
- -
- - -
-
- - -
- -
- -
- -
- -
_
- -
-
-
_ — —
F = 0.875 V /VF
Permeate Feed
0.65
-
5.03
29,000
1300 7200
-
100
-
-
-
19,000
6
280
50
9.9
34
0.09
-
2.2
-
210
1.7
500
150
38
, = 0.95
Permeate
0.48
-
5.4
15,400
2200
-
17
-
-
-
8600
<1
130
7
1.2
6.7
0.01
-
0.3
-
30
0.14
80
13
3.7
Two modules in series.
Operation pressure = 400 psig; feedflow rate = 3.9 gal/min; V
''Dashes indicate no data is available.
= volume of permeate; Vp = initial volume of feed solution.
-------
TABLE 35. PERFORMANCE OF OSMONIC P.V. 192-197 SPIRAL-WOUND REVERSE OSMOSIS MODULE
IN TREATING PRETREATED BOILER CLEANING WASTES3'
I
00
Constituent
Permeate flow,
gal/min
Feed tempera-
ture, °C
pH, units
Conductivity,
M mhos /cm
NH3-N, mg/Z
Org.-N, mg/Z
TOC, mg/Z
COD, mg/Z
Si02, mg/Z
P04-P, mg/Z
Cl, mg/Z
S04, mg/Z
F, mg/Z
Ca, mg/Z
Mg, mg/Z
Na, mg/Z
Br, mg/Z
Br03, mg/Z
Ag, mg/Z
As, mg/Z
Ba, mg/Z
Cr, mg/Z
Cu, mg/Z
Fe, mg/Z
K, mg/Z
Ni, mg/Z
Zn, mg/Z
Vp/Vj
Feed
_
23.2
5.02
9700
1800
1600
12
230
20
0.04
4300
2
98
27
3.1
4.3
7.7
13
0.02
0.004
0.54
0.053
38
0.59
120
33
7.1
r = 0.01
Permeate
0.12
-
5.4
2100
240
200
10
560
0.9
0.02
580
2
12
1.8
0.1
0.7
0.2
9.6
<0.01
<0.002
<0. 1
<0.005
0.58
0.08
9
0.31
0.09
V /V,, = 0.125
P F
Feed Permeate
0.12
25.2
5.2 4.5
13,600 2090
2000 280
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
41 0.58
-
-
-
— -
V /V,. = 0.25
P F
Feed Permeate
0.11
26.5
5.2 4.52
14,500 2560
2600 330
-
-
-
26 1.0
-
-
-
-
-
-
-
-
-
-
-
-
-
46 0.68
-
-
-
— ~
V /VF = 0.375
Feed Permeate
0.10
27.0
5.2 4.51
14,150 3050
2800 400
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
52 0.74
-
-
-
~ ~
VVF
= 0.5
Feed Permeate
_
27.2
5.3
15,400
3700
-
55
320
19
-
7100
-
-
-
-
-
29
13
-
-
-
-
67
1.4
230
-
14
0.10
-
4.66
3600
420
-
10
90
1.7
-
1400
-
-
-
-
-
1.0
12
-
-
-
-
0.99
<0.05
16
-
0.17
V /Vp = 0.625 V /Vp = 0.75 V /Vp = 0.875
Feed Permeate Feed Permeate Feed Permeate
0.08 - 0.07 - 0.06
27.5 - 28 29
5.37 4.62 5.25 4.64 5.19 4.9
21,500 4300 25,000 6050 28,000 9000
4200 600 5300 800 6400
------
------
- - -
46 3.0
- . _
------
------
_ _ .
------
------
------
------
- - - -
------
------
------
- - - -
78 1.4 92 1.8 130 2.7
------
- - - -
------
.
VVF
= 0.95
Feed Permeate
_
29
5.16
33,300
7500
2000
60
1200
92
0.15
-
8
310
37
9
20
87
32
0.11
0.004
1.2
0.02
160
2.4
530
180
102
0.05
-
4.75
11,500
1800
420
15
350
13
0.02
-
<1
50
3.5
0.1
2.1
1.0
22
0.02
<0.002
<0. 1
<0.005
3.9
<0.05
61
1.6
2.1
-Operation pressure = 400 psig; feedflow rate = 3-9 gal/min;
Dashes indicate no data is available.
V — volume of permeate;
V., = initial volume of feed solution.
-------
TABLE 36. SUMMARY OF PROJECTED COST ESTIMATE FOR NICKEL RECOVERY
BY REVERSE OSMOSIS (CLOSED-LOOP) SYSTEM (1972)
o»
a b
Parameter Toronto installation Projected scale
Flux rate at 450 psi, gal/ (ft2 -day)
Permeation rate, gpd
Nominal reverse osmosis unit capacity, gpd
Installed capital cost , $
Amortized capital cost per 1,000 gal permeate, $
Operating and maintenance cost per 1,000 gal
permeate , $
Total cost per 1,000 gal permeate, $
Total cost per day, $
Recovered value per day as Ni, $
Recovered value per day as total salts , $
Savings in deionized water usage per day, $
Total savings as total salts plus water per day, $
Net savings per day, $
Payback on capital investment
10.7
1,065
1,000
7,000
6.96
0.70
7.66
8.15
9.46
16.62
0.49
17.11
8.96
39 months
10.7
3,550
3,500
9,000
2.68
3.40
6.08
21.60
31.50
55.30
1.65
56.95
35.35
13 months
Projected scale
10.7
35,500
35,000
37,000
1.03
2.05
3.08
109.30
315.00
553.00
16.50
569.50
460.20
4 months
Dragout rate = 15 gpd; evaporation rate = 150 gpd.
Dragout rate = 50 gpd; evaporation rate = 500 gpd.
p
Dragout rate = 500 gpd; evaporation rate = 5,000 gpd.
Capital and maintenance costs can vary widely depending on the design of the unit.
Source: Golomb, A. An Example of Economic Plating Waste Treatment by Reverse Osmosis. In: Proc., 6th
International Water Pollution Research Conf., June 18-23, 1973. Reprinted by permission.
-------
McNulty et al.76 conducted field tests on reverse osmosis treatment
of copper cyanide rinse waters and estimated the operating cost for a
closed-loop system. These estimates have been revised by TVA to correct
an arithmetic error and are presented in Table 38. The total operating
cost is calculated to be $2.89/day. The credit realized from the chemicals
recovery is shown in Table 39, which indicates total credits of $2.65/shift.
These figures show that the operating cost is almost entirely offset by
the credits resulting from closed-loop operation.
In applying metal plating industry experience to boiler cleaning
wastes, several key differences should be kept in mind. The metal plat-
ing industry can often reuse the concentrate as is, and take credit for
chemicals saved. Concentrated boiler cleaning waste appears to have no
ready market. Also, boiler cleanings occur only once every one to five
years, thus large storage areas are necessary if treatment facilities
are not to be left idle much of the time.
TABLE 37
TABLE 37. OPERATING AND MAINTENANCE COSTS PER 1000 GALLONS OF PERMEATE
FOR NICKEL RECOVERY BY REVERSE OSMOSIS (CLOSED-LOOP) SYSTEM (1972)
Dragout rate
Item 15 gpd 50 gpd 500 gpd
Membrane replacement
Power
Labor
Materials
$0.50
0.10
Nil
0.10
$0.50
0.10
2.70
0.10
$0.50
0.10
1.35
0.10
Totals $0.70 $3.40 $2.05
Source: Golomb, A. An example of Economic Plating Waste Treatment
by Reverse Osmosis. In Proc., 6th International Water
Pollution Research Conf., June 18-23, 1973.
Reprinted by permission.
-88-
-------
TABLE 38. BREAKDOWN OF OPERATING COSTS FOR NEW ENGLAND PLATING (1977)
Power
Major power requirement is for the high pressure pump (at 50% efficiency)
(2.67 gal) (400 Ib) (231 in.3) ft (hp-s) min 1 _ . 9, .
min InT3gal '(12 in.)'(550 ft-lb) (60 s) (0.5) l'a np
(1.25) . (0.746 kW) . (8 hr) . ($.036) _ $0.27
hp * shift * kWh shift
Module Replacement
$720 1 module $1.44
' • - ^~ -J- _- i_i__
module 500 days day
Maintenance costs (5% of capital investment per year)
.05 t $8500 capital cost _ $1.16
year 365 days/year day
Deionized water (at $2.00 per 1000 gal)
(0.023 gal PI water) . (8 hr) . (60 min) . $2.00 _ $0.022
min * shift * hr * 1000 gal ~ shift
TOTAL COST PER DAY
One shift per day $0.27 + 1.44 + 1. 16 + 0.02 = $2.89/day
Three shifts (0.27 + 0.02) • 3 + 1.44 + 1. 16 = $3.47/day
COST PER 1000 GAL
A u-r*. j $2.89 1 min 1 shift 1000 gal $3.01
One shift per day - - ' 480~mIK * 1000 ial = 1000 gal
-ru u-r. A $3.47 1 min 1 day 1000 gal $1.20
Three shifts per day - • * 1440 min ' 1000 gal = 1000 gal
This table is based on the work of K. J. McNulty, R. L. Goldsmith, A. Gollan,
S. Hossain, and D. Grant. Reverse Osmosis Field Test: Treatment of Copper
Cyanide Rinse Waters. EPA-600/2-77-170, August 1977. Several errors in the
original have been corrected, and the labels have been clarified.
-89-
-------
TABLE 39. CREDITS REALIZED FOR REVERSE OSMOSIS OPERATION
AT NEW ENGLAND PLATING
Value of Recovered Plating Solution (at $1.83/gal)
Loss rate without reverse osmosis system = 0.0013 gal/min.
Recovery rate with reverse osmosis system = 99.96%.
$1.83 . (0.0013 gal) . ( „ . (480 min) = $1.14
gal min w. *»•«>; ghift ghift
Water and Sewer Credits
62 gal t $0.50 _ $0.03
shift ' 1000 gal shift
Chemical Treatment Credits
(0.0013 gal) (480 min) (8.9 02 CuCN) (1 Ib) 0.35 Ib CuCn
min ' shift ' gal '16 oz shift to treat«ent
(0.35 Ib CuCN) $0.22 1 Ib NaOH $0.50 8 Ib C12 _ $1 43
shift ' Ib NaOH * 1 Ib CuCN Ib C12 ' 1 Ib CuCN ~ shift
Total Credits = $2.65/shift
Adapted from: McNulty, K. J., R. L. Goldsmith, A. Gollan, S. Hossain,
and D. Grant. Reverse Osmosis Field Test: Treatment of
Copper Cyanide Rinse Waters. EPA-600/2-77-170, August
1977.
-90-
-------
RECOMMENDATIONS
Reverse osmosis can be used to remove residual heavy metals from
chemical cleaning wastes after the bulk has been removed by pH adjust-
ment and precipitation. The design and optimization of such a system
will require additional analyses and extensive pilot plant work.
If boiler cleaning wastes were treated at a central facility serving
a large number of plants, this would allow for better utilization of treat-
ment facilities. After pH adjustment, precipitation and reverse osmosis,
treated water could perhaps be reused to makeup fresh cleaning solutions.
A large facility of this type may also be able to market precipitated
metals for reuse.
More research is needed to perfect conventional methods of treatment.
These would include the use of one or more of the following processes:
pH adjustment, aeration, adsorption, chemical precipitation coagulation,
settling, or filtration.
-91-
-------
SECTION 12
WET SULFUR DIOXIDE SCRUBBER WASTES
CHARACTERISTICS OF WASTEWATER
A study by TVA25 indicated that the water requirements in a typical
closed-loop nonregenerable scrubber system are about 2500 gpm (9463
A/rain) per thousand-MW capacity to replace water lost through evaporation
and sludge disposal and in washing the mist eliminator to remove accumu-
lated solids. Types of wastewater that might be discharged from nonre-
generable scrubber systems include washwater from the mist eliminator,
moisture included in the sludge, and occasional bleedoff from the recir-
culating system. Water is lost to sludge disposal at a rate of about
0.3 x 106 gal (1.1 x 106 L) per day for each percent sulfur in the coal
for each thousand-MW capacity of the power plant.77 Periodic bleeding
from the closed-loop scrubber system may be necessary to avoid scaling
from the buildup of dissolved solids and to discharge excess water that
may occur as a result of changes in operating conditions or system
repairs.
The Aerospace Corporation characterized and assessed scrubber
liquor discharges from four different closed-loop nonregenerable scrubber
systems: pH — 3.04 to 10.7; lead — 0.01 to 0.4 mg/£; mercury — 0.0004 to
0.07 mg/£; selenium--<0.001 to 2.2 mg/£; chloride— 420 to 4800 mg/Jfc;
sulfate--720 to 10,000 mg/£; arsenic— <0. 004 to 0.3 mg/£; boron— 8 to 46
mg/£; cadmium — 0.004 to 0.11 mg/Jfc; chromium — 0.01 to 0.5 rag/£; iron— 0.02
to 8.1 mg/£; manganese—0.09 to 2.5 mg/£; and silver— 0.005 to 0.6
mg/4.77
Mist eliminator wash water, as observed during pilot plant studies
at TVA, is generally acidic, with a pH of less than 3, and has relatively
high concentrations of iron (0.07 to 13 rag/A) and sulfate (700 to 1200
Regenerable desulfurization scrubber systems often require a par-
ticulate scrubber to remove chlorides and fly ash from the flue gas. In
addition, the particulate scrubber removes up to half of the S03 and
between 5 to 20 percent of the S02 in the flue gas. Particulate scrubber
water is recirculated, but blowdown is necessary to maintain a mass
balance in the system. Blowdown from this system is typically high in
chlorides and sulfates and very low in pH.
There are many proposed flue gas desulfurization systems, each with
its own characteristic wastes. It is beyond the scope of this report to
characterize them each in detail.
-92-
-------
ALTERNATE TREATMENTS
In general, scrubber liquors become unsuitable for reuse because of
high scaling potentials, high dissolved solids (particularly chlorides),
or low pH. Of these, membrane technologies are suited to control of
dissolved solids but not to alleviating scaling or pH problems. Control
of pH and scaling is a necessary prerequisite to either reverse osmosis
or electrodialysis.
If membrane technologies are used for dissolved solids control,
this will impact the choice of bases for pH adjustment to a neutral
yalue, since the use of lime or limestone would lead to membrane scaling.
For equivalent neutralization, lime is cheaper than caustic or soda ash
by a factor of 5 to 10. This is important since large quantities of
acid must be neutralized.
After neutralization, softening and other pretreatments as appro-
priate, either reverse osmosis or electrodialysis may be used to remove
and concentrate dissolved solids for subsequent disposal. Pretreatment
needs are discussed in appendix A.
Evaporative brine concentration has advantages over membrane processes
in not requiring extensive pretreatment. Vapor compression evaporators
marketed by Resources Conservation Corporation reportedly operate routinely
on supersaturated solutions without scaling problems.79
Electrodialysis has been suggested as a method of regenerating S02
scrubber liquor."'24'80 Several variations on this method exist, and
no attempt is made to evaluate them fully in this report since scrubber
technology is a very complex and growing area. In general, an alkaline-
sulfate-sulfite solution is split into separate alkaline and sulfuric
acid streams. The alkaline stream is then reused to absorb additional
sulfur oxides. Sulfuric acid or SQ% is recovered for other uses. A
three-chamber electrodialysis unit is used. Feedwater enters in the
center. Cations such as sodium migrate to the cathode compartment,
passing through a cation exchange membrane. At the cathode, cations
join with hydroxide ions (OH ) formed by electrolysis of water. Likewise,
anions, such as sulfate and sulfite, p^ass through the anion membrane
where they join with hydrogen ions (H ) produced by the dissociation of
water molecules at the anode.
Regenerable concepts such as this produce less waste material than
the so-called throwaway systems. However, lime-limestone scrubbers can
be designed for zero liquid discharge, although vast quantities of
sludge are produced and must be stored.
RECOMMENDATIONS
Reverse osmosis and electrodialysis should be considered as alter-
natives to evaporators for the removal and concentration of dissolved
solids from scrubber loops, giving consideration to the differing pre-
treatment requirements. Further studies of scrubber systems should be
conducted including those using electrodialysis to regenerate the alkaline
scrubber solution.
-93-
-------
SECTION 13
RADIOACTIVE WASTEWATERS FROM POWER REACTORS
Because the treatment of radioactive liquid wastes is a complex
subject, this section will only attempt to introduce the problem, and
place special emphasis on applications of membrane technology.
CHARACTERISTICS OF WASTEWATER
There are two primary mechanisms that produce radionuclides in power
reactors—fission and activation. Fission of uranium produces a broad
spectrum of radionuclides with mass numbers primarily from 72 to 160.81
These nuclides then decay to other nuclides. All reactors using uranium-235
yield a similar distribution of fission products, but the quantity released
to the reactor coolant varies depending on the number of defects in the
fuel cladding.
The reactor coolant and any associated additives or impurities can
be activated by neutron bombardment in the reactor core. The primary
activation product of water is N16, with smaller contributions from C14,
N13, N17, O19, and F18. In pressurized water reactors (PWRs), boron is
added to the coolant to control the rate of neutron capture, and lithium
is added to control pH. Corrosion products also appear in the coolant.
Each nuclide produces its own characteristic family of activation products
in quantities determined by the nuclide's "cross section," its concentra-
tion, and the neutron flux. For a further discussion of neutron activation,
see Glasstone and Sesonske.82
From the reactor coolant, radionuclides may take several pathways
depending on the plant design and operation. Two types of power reactors
predominate in the United States—boiling water reactors (BWRs) and pres-
surized water reactors (PWRs). In BWRs steam for the turbogenerator is
produced in the reactor, whereas in PWRs steam is generated in a heat
exchanger, which separates the primary reactor coolant from the secondary
coolant.
In BWRs the liquid radioactive wastes are customarily grouped in
four categories—high-purity, low-purity, chemical, and detergent waste.27
High-purity wastes include those collected by closed drains from equip-
ment, including pump and valve leakoffs, demineralizer backwash, and ultra-
sonic resin cleaning. These liquids are normally reused as primary reactor
coolant after appropriate treatment. Low-purity wastes consist of valve
and pump leaks and other miscellaneous wastes which are collected by open
drains, floor drains, or sumps. Chemical wastes are highly concentrated
-94-
-------
wastes, primarily spent ion exchange regenerants and laboratory wastes.
(Ion exchange is typically used for reactor coolant cleanup, condensate
cleanup, and treatment of more dilute radioactive wastes. Resin from
these systems may be regenerated or thrown away.) Detergent wastes con-
sist of laundry, personal shower, and equipment decontamination wastes.
Approved methods of calculating the activities and quantities of these
wastes are given in the BWR-GALE code.27 To these estimates, conservative
designers would add factors to enhance reliability and ease of operation.
PWR waste streams are usually categorized somewhat differently.83'28
Clean wastes, which originate from leaks in the primary coolant systems,
are collected by closed drains. After treatment these wastes are suitable
for reuse as primary coolant. Dirty wastes are collected from sumps,
floor drains, and other open drains. Floor drain wastes from the turbine
building are usually segregated from floor drain wastes from the reactor
containment building. Chemical wastes originate (as in BWRs) from spent
ion exchange regenerant and laboratory wastes. Again, as in BWRs, ion
exchange is used for condensate cleanup, reactor coolant cleanup, and
treatment of more dilute radioactive wastes. PWR chemical wastes may be
treated separately or combined with floor drain and other dirty wastes.
Steam generator blowdown is sometimes used (in addition to or in lieu of
condensate polishing) to control steam purity while allowing the use of
corrosion inhibitors (usually phosphates) in the steam generator. The
properties of this waste are functions of condenser in-leakage, primary-
to-secondary coolant leakage, and the amount of corrosion inhibitor used.
The volumes and activities of PWR radioactive wastes may be estimated by
using the PWR-GALE code.28 These estimates should be tempered by experience.
Comparing waste streams from PWRs and BWRs, there are many similarities
but some significant differences. Chemical wastes originate from laboratory
wastes and spent ion exchange regenerant in either type of reactor. The
amount and activity of spent ion-exchange regenerant depends more on indi-
vidual plant design than on the type of reactor. Detergent wastes are
also common to both reactor types. Superficially, the high-purity waste
from BWRs appears similar to clean waste from PWRs; however, the PWR waste
is typically higher in activity and generally contains between 10 and
2000 mg/£ of boric acid as boron. Also, both the lithium and boron present
in the PWR coolant generate tritium by neutron absorption and subsequent
alpha decay. Tritium is a long-lived radionuclide which presents special
problems. Low-purity wastes from BWRs are similar to the dirty wastes
from PWRs, provided that the PWR dirty wastes are not mixed with chemical
wastes. However, PWR floor drains from the containment building are likely
to be higher in activity than BWR floor drains, whereas the PWR turbine
building will be lower in activity.
In treating PWR wastes, boric acid recovery becomes important. Early
in the fuel cycle, the PWR primary coolant may contain about 2000 mg/£
boric acid (as boron). As reactivity decreases, boric acid is purposely
removed from the coolant. Radionuclides are typically removed from the
reactor coolant by a mixed-bed demineralizer saturated in borate ions.
The effluent boric acid is reconcentrated by evaporation. When the boric
acid concentration in the coolant is very low, reconcentration becomes
uneconomical; borate is then removed by an ion exchanger in hydroxide
form.
-95-
-------
Lithium is used in PWR's to control reactor coolant pH. Because
lithium-7 is less prone to generate tritium as an activation product, it
is used instead of the more common lithium-6 isotope. Unfortunately,
lithium-7 is difficult to separate from naturally occurring lithium, which
is a mixture of isotopes. Thus, recovery and reuse of lithium-7 is desired.
Although radioactive wastes are closely monitored and well charac-
terized on the basis of their radioactive elements, less is known about
the chemical and physical form of these species. This is a handicap in
designing efficient waste treatment schemes. Recent studies indicate
significant decontamination factors can be attained by ultrafiltration,22>84
indicating that some of the radioactive species are not carried as
dissolved ions but possibly as colloids or adsorbed onto colloids.
Sometimes the nonradioactive constituents of a waste pose significant
problems, as when ion exchange resins are fouled by colloids. Also, evap-
orator performance is often limited by foaming attributed to organics,83
and condensate demineralizers are often loaded with ions and colloids
from in-leakage of cooling water at the condenser.
ALTERNATIVE TREATMENTS
Reverse osmosis can be used as a roughing demineralizer to increase
the time between demineralizer regenerations. This lowers the costs of
regenerant chemicals or replacement resins. It can also reduce the cost
of solid radioactive waste disposal in the form of spent resins and evapo-
rator bottoms, whether this results in a net savings depends on several
factors, including the available methods for disposing of reverse osmosis
reject and ion exchange regenerant, the cost of resin and resin disposal,
the cost of regenerant chemicals, and the decontamination factors and
personal doses associated with each alternative.
In studying the cost of treating steam generator blowdown, Westing-
house (in an unpublished report) found that reverse osmosis would be more
economical than nonregenerable ion exchange if the concentration of total
dissolved solids in blowdown exceeded about 80 mg/£. After adjusting
the Westinghouse numbers to account for site-specific conditions, TVA
found the break-even point to be 15 mg/Jfc, largely because evaporators
were already available to process the reject water from reverse osmosis.
Using 80 mg/JU as a rough guideline and realizing that a plant-by-plant
assessment may be necessary, some generalizations about the applicability
of reverse osmosis to radioactive waste demineralization can be made.
Reverse osmosis is not applicable to the high-purity or clean wastes,
which are very low in dissolved materials (other than boric acid). Further-
more, reverse osmosis is not useful in treating the highly concentrated
chemical wastes, with osmotic pressures in the neighborhood of 500 psi.
Three potential applications for reverse osmosis are steam-generator blow-
down, floor drain wastes, and detergent wastes. These wastes can have
dissolved solids contents as high as 80 mg/£, although each waste should
be evaluated separately.
-96-
-------
The ion loading in steam generator blowdown comes primarily from
condenser in-leakage and water treatment chemicals. (The radioactivity
comes from primary-to-secondary coolant leakage.) Typical blowdown rates
are 0.06 percent of the condensate flow for phosphate treatment and 0.45
percent for all-volatile treatment.28 Although the condensate may have
a relatively low level of dissolved material, this material is concentrated
into a much smaller blowdown stream. Total dissolved solids in the blow-
down can reach fairly high levels during condenser in-leakage, especially
for a seawater-cooled plant. Under these circumstances, reverse osmosis
becomes very attractive.
Floor drain wastes gather water leaked from various radioactive sys-
tems, combined with other leaks and drains originating from uses of cooling
water or potable water. Thus, floor drain waste will have characteristics
intermediate between reactor coolant and river water, plus miscellaneous
materials (both dissolved and suspended) from other sources. Depending
on the individual site and predicted water quality, reverse osmosis may
be economical in treating these wastes.
For radioactive wastes, reverse osmosis is best known for treating
laundry waste. Westinghouse installed a tubular reverse osmosis station
at R. E. Ginna Nuclear Plant and documented its performance,85 resulting
in good acceptance of the concept. To prevent calcium phosphate fouling,
deionized or softened water is recommended for laundry makeup water. In
fact, the reverse osmosis permeate can be recycled for this purpose.
If deionized water is used in the laundry systems, this makes another
system somewhat more attractive than reverse osmosis for treating the
waste; carbon filtration can be used to remove organics, followed by ion
exchange to remove inorganics.
Each system must be thoroughly evaluated before one could definitely
conclude that reverse osmosis is the superior solution to the laundry
waste problem. In making such comparisons, care should be taken to con-
sider the pretreatraent needs of any proposed reverse osmosis system as
stated in the Appendix—User's Guide to Reverse Osmosis. Post-treatment
required for radioactive waste streams usually includes ion exchange
polishing of the permeate and evaporation of the reject. Evaporation is
then followed by solidification and storage.
In many radioactive systems, suspended solids are removed by precoat
filters or powdered ion exchange resins. When the pressure across these
systems drops excessively, the filter-aid or ion exchange resin is wasted,
and creates large volumes of solid radioactive wastes. As an alternative,
ultrafiltration could be used to remove colloids and suspended solids.
Reject streams could be concentrated to low volumes by reject staging or
recycle. The Tsuruga Power Station of the Japan Atomic Power Company
Ltd.22 is using this approach to pretreat wastes from equipment drains.
Other possible uses may include condensate demineralizers and floor drain
wastes.
Ultrafiltration may also be used to improve the performance of radio-
active waste evaporators by pretreating wastes to remove organic materials.
This approach is currently being investigated by TERA (Teknekron Energy
Resource Analysts) Corporation under contract to EPRI.
97-
-------
Electrodialysis has been proposed as a method of treating reactor
coolant from PWR's to separate lithium-7 from boric acid.23 In this proc-
ess, lithium ions would be attracted to the cathode through a cation selec-
tive membrane, then combined with hydroxide ions generated at the cathode
by electrolysis of water. Similarly, borate ions would pass through an
anion selective membrane to the anode and combine with hydrogen ions.
Subsequent chemical treatments would probably be necessary to separate
the desired lithium-7 from other alkali and metals. Boric acid concentrate
could presumably also be reused, after purification if necessary. Reverse
osmosis might be used to purify boric acid since boric acid will pass
through membranes, leaving highly ionized species and colloids behind.
(This application of reverse osmosis was suggested by Joseph Markind,
now with TERA Corporation.)
RECOMMENDATIONS
In planning future nuclear plants, reverse osmosis should be evalu-
ated for roughing demineralizer service in radioactive systems. This
evaluation should take into account the problems of waste disposal, the
radiation dose commitments, and the operating and maintenance costs
associated with each alternative.
Similarly, reverse osmosis should be evaluated for applications to
detergent radioactive wastes, and ultrafiltration should be considered
for removal of colloids and suspended solids.
The use of reverse osmosis in laundry and roughing demineralizer
applications is proven technology, but the existing experience should be
better documented. The efficiency of past processes should also be improved.
Ultrafiltration as a pretreatment process should be demonstrated to
provide a basis for estimating performance parameters such as removal
efficiencies, decontamination factors, volume recovery factors, and required
flow-through rates. With the proper information in hand, detailed evalua-
tions can then be made of ultrafiltration as compared with precoat filters.
The proposed method of recovering lithium and boric acid from PWR
coolant deserves further study.
A more detailed review of membrane applications to radioactive wastes
has been prepared by the Walden Division of Abcor, Inc.14
-98-
-------
SECTION 14
MISCELLANEOUS WASTES
CHARACTERISTICS OF WASTEWATER
Miscellaneous wastes can be divided into two groups—oily wastes
and nonoily wastes. The oily wastes include wastes from floor drains,
oil unloading or transfer stations, oil pumping stations, oil storage
tank drainage, oil purification system drains, and equipment base drains.
The nonoily wastes include water treatment wastes (excluding ion exchange
regeneration wastes), laboratory and sampling streams, and cooling tower
basin cleanings.86 Generally, the nonoily wastes consist of moderate
amounts of suspended solids and high amounts of dissolved solids, whereas
the oily wastes contain suspended solids, dissolved solids, and oil and
grease.
ALTERNATIVE TREATMENTS
Markind et al.87 investigated the use of the reverse osmosis for
concentrating waste-soluble oil coolants. They concluded that reverse
osmosis can be used to concentrate soluble coolants to give a permeate
that can be reused or dumped to sewers. The economics appeared favorable
for the reverse osmosis system as compared with a hauling operation or
conventional, physicochemical treatment system.
Bansal's laboratory and pilot studies88 demonstrated that ultrafil-
tration with temperature- and abrasion-resistant inorganic (by Union
Carbide) membranes can treat oily and latex processing solutions and pro-
duce a permeate acceptable for reuse or discharge. The waste oil emul-
sions can be concentrated to 25 percent or higher. The oil can be
concentrated further by sulfuric acid treatment, which yields an oil of
70 percent concentration.
Wang et. al.89 reviewed the state-of-the-art of alternative commercial
techniques for separating emulsified oil from water, which included air
flotation, precipitate flotation, adsorption flotation, magnetization,
coalescence, chromatography, layer filtration, absorption-adsorption,
centrifuge-hydrocyclone, sedimentation, chemical coagulation, heating-
evaporation-distillation, crystallization or freezing, biological oxida-
tion, clarification-flotation, and filtration-extraction-filtration.
The oil removal efficiencies of these techniques vary from less than 50
percent to greater than 99 percent. Their costs also vary a great deal.
The degree of pretreatment needed before ultrafiltration is determined
by the feed quality tolerable to ultrafiltration, the product quality
-99-
-------
required, and the overall costs incurred. In the ultrafiltration process
studied by Bansal88 only free oil and suspended solids had to be removed
before ultrafiltration. With a feed oil concentration as high as 55,600
mg/JU, permeates of 5 to 14 mg/Jl oil were obtained, resulting in 99.97 to
99.99 percent oil rejection. This demonstration seems to cover the range
of the oil wastes under consideration in the power plant operation and
shows the applicability of the ultrafiltration process.
Generally, devices for removing free oils, such as stilling tanks
or belt skimmers, or units for removing suspended solids, such as Sweco
or basket filters, are needed before ultrafiltration to treat the oily
and nonoily miscellaneous wastes. If Union Carbide's inorganic membranes
are used, no pH control is required. Without further treatment, the
product water from the ultrafiltration modules can be used for ash sluicing,
chemical cleaning, and sulfur dioxide scrubbing.
A cost analysis conducted by Bansal90 indicates that for a 50,000-gpd
unit, the operating cost varies from $2.05 to $3.63 per thousand gallons
of waste oil in 1975. The first figure is for a recovery of 90 percent
and a membrane life of three years, whereas the second figure is for a
95 percent recovery and a membrane life of two years. If the concentrated
oil is recovered and used as fuel, a net profit of $1.95 to $4.37 per
1000 gal can be realized.
Markind e_t al.87 reported on an economic study of using reverse osmo-
sis for treating waste coolant oils in 1975. All systems were amortized
at 15 percent interest for five years. If the concentrate were burned
as fuel (credit of $1.00 per 106 Btu or $0.003 per gal of feedwater),
total costs would be $0.01/gal for a 500-gpd system. The highest cost
would result from the disposal of concentrate at $0.06/gal, which would
increase the total costs from $0.0l6/gal for a 100,000-gpd system to
$0.09/gal for a 500-gpd system. A competitive physicochemical treatment
system would give a cost of $0.0165/gal for a 100,000-gpd system.
Ultrafiltration and chemical methods of removing oil from waste streams
were compared in a panel discussion at the 38th International Water
Conference.91 It was agreed that ultrafiltration can be cost effective
in treating oils, but that case by case pilot studies are needed since
oily wastes are variable and difficult to characterize. Ultrafiltration
was said to produce a more consistent effluent quality than chemical treat-
ment and to be less sensitive to operator errors. It was emphasized that
pilot studies are necessary to develop cleaning procedures and schedules.
RECOMMENDATIONS
Ultrafiltration is a competitive technology for treating oily wastes.
Pilot studies are necessary on a case-by-case basis to determine operating
characteristics as a function of cleaning methods, water recovery rates,
and waste characteristics.
-100-
-------
SECTION 15
INTEGRATED OR COMBINED WASTES
Power plants are complex and generate a variety of differing waste
streams. Many different flow patterns have been devised41'92'34'46 cas-
cading waters and combining streams for treatment. Plants are now operating
with cascading water reuse and zero liquid discharge.93'79
To permit maximum water reuse, conventional treatments such as soften-
ing, pH control, precipitation, and suspended solids removal are used.
If dissolved solids control is not used, the only outflows will be through
such incidental losses as entrained water in sludge and cooling tower
drift. Deliberate removal and concentration of dissolved solids is often
desired to control corrosion. In this event, reverse osmosis, electro-
dialysis, and distillation should each be considered. Distillation is
apparently the most expensive but requires less pretreatment. Electro-
dialysis is apparently the least expensive,35 but it does not produce
water which is comparable in quality to reverse osmosis permeate or evapo-
rator distillate. Both reverse osmosis and electrodialysis suffer from
a lack of large-scale experience in this application within the U.S.
-101-
-------
REFERENCES
1. Weber, W. J., Jr. Physicochemical Processes for Water Quality
Control. Wiley-Interscience, New York, New York, 1972. 640 pp.
2. El Paso Environmental Systems, Inc. Synopsis of Reverse Osmosis
Technology. El Paso, Texas, 1977.
3. Breton, E. J., Jr. Water and Ion Flow Through Imperfect Osmotic
Membranes. U.S. Dept. of Interior, Office of Saline Water, Research
and Development Progress Report N16, 1957.
4. Reid, C. E., and E. J. Breton, Jr. Water and Ion Flow Across
Cellulosic Membranes. Journal Applied Polymer Science, 1, 1959.
pp. 133 and f.
5. Sourirajan, S., and S. Loeb. Sea Water Research. Dept. of Engineering,
University of California at Los Angeles. Report No. 58-65, 1958.
6. Gulp, R. L., G. M. Wesner, and G. L. Gulp. Handbook of Advanced
Wastewater Treatment. 2d ed., Van Nostrand Reinhold Co., New York,
New York, 1978. 640 pp.
7. Bhattacharyya, D., A. B. Jumawan, Jr., S. 0. Witherap, and R. B.
Grieves. Ultrafiltration of Complex Wastewaters, Recycling for
Nonpotable Use. Journal of the Water Pollution Control Federation, 50,
May 1978. pp. 846-861.
8. Smith, J. D., and J. L. Eisenmann. Electrodialysis in Wastewater
Recycle. In: Proceedings, 19th Industrial Waste Conference, Purdue
University, 1964, pp. 738-760.
9. Orange County Water District. Pilot Wastewater Reclamation and
Injection Study. Report prepared by Montgomery-Toups, December 1967.
10. Brunner, C. A. Pilot-Plant Experiences in Demineralization of
Secondary Effluent Using Electrodialysis. Journal of the Water
Pollution Control Federation, October 1967. pp. R1-R15.
11. Dryden, F. D. Mineral Removal by Ion Exchange, Reverse Osmosis and
Electrodialysis. In: Proceedings, Wastewater Reclamation and Reuse
Workshop, Lake Tahoe, California, June 25-27, 1970. p. 179.
12. Jordan, D. R., W. F. Mcllhenny, and G. T. Westbrook. Cooling Tower
Effluent Reduction by Electrodialysis. In: Proceedings, American
Power Conference, Chicago, Illinois, April 22, 1976. pp. 980-987.
-102-
-------
13. Wadlington, M. Chemical Regenerant Savings Can Pay for a Reverse
Osmosis Unit. Industrial Water Engineering, June/July 1976.
pp. 17-21.
14. Markind, J., and Tarn Van Tran. A study of Reverse Osmosis Appli-
cability to Light Water Reactor Radwaste Processing. Walden Division,
Abcor, Inc., Wilimington, Massachusetts. 1978. 132 pp.
15. Hughes, T. R., and R. D. Windon. Operating Experiences with the
Reverse Osmosis System at Harrison Power Station. In: Proceedings,
37th International Water Conference, Engineers' Society of Western
Pennsylvania, Pittsburgh, Pennsylvania, October 26-28, 1976.
pp. 1-11.
16. Minard, P. G., R. R. Stana, and E. DeMerritt. Two Years Experience
with a Reverse Osmosis Radioactive Laundry Water Concentration.
In: Proceedings, The Second National Conference on Complete Water
Reuse, Chicago, Illinois, May 4-8, 1975. pp. 936-941.
17. Hollier, M. Operating Experiences with Reverse Osmosis Demineralizing
for Boiler Feedwater and Makeup Treatment Systems at Willow Glen Power
Station. In: Proceedings, 38th International Water Conference, Engi-
neers' Society of Western Pennsylvania, Pittsburgh, Pennsylvania,
November 1-3, 1977. pp. 277-282.
18. Kakimoto, A. K., Y. Kakimoto, H. Karaso, 0. Abe, K. Yabe, M. Imaizumi.
Wastewater Reclamation by the Reverse Osmosis Process at a Power
Station. In: Proceedings The Second National Conference on Complete
Water Reuse, Chicago, Illinois, May 4-8, 1975. pp. 942-955.
19. Dascher, R. E. and Roy Lepper. Meeting Water-Recycle Requirements
at a Western Zero-Discharge Plant. Power Vol. 121, (8), 1977.
pp. 23-28.
20. Toomey, A. E. Utility Boiler Feedwater Treatment at CIPSCO. In: Pro-
ceedings, 38th International Water Conference, Engineers' Society of
Western Pennsylvania, Pittsburgh, Pennsylvania, November 1-3, 1977.
pp. 97-110.
21. Mitchell, D., and D. G. Parker. Differing State Water Quality Stand-
ards and Industrial Wastewater Treatment Costs. In: Proceedings,
31st Industrial Waste Conference, Purdue University, May 4-6, 1976.
pp. 150-156.
22. Kurihara, Y., M. Oda, K. Okugawa, T. Sunaoka, M. H. Kleper, J. Markind,
R. L. Goldsmith, and J. M. Ryan. Treatment of Equipment Drain Wastes
in a Nuclear Power Plant with an Electro Magnetic Filter and Ultrafil-
tration System. In: Proceedings, 38th International Water Conference.
Engineers' Society of Western Pennsylvania, Pittsburgh, Pennsylvania,
1978. pp. 135-148.
23. Grant, P. J. and J. R. Schepers. Power Plant Water Purification,
Chemical Recovery, and Waste Treatment: A Novel Technique with
Multiple In-plant Applications. In: Proceedings, American Power
Conference, Chicago, Illinois, April 24-26, 1978. pp. 976-985.
-103-
-------
24. McRae, Wayne A. Commercial Applications of Ion Exchange Membranes
in Resource Conservation. Presented at the Seminar on Membrane
Separation Technology, Clemson University, Clemson, South Carolina,
August 18, 1977. 16 pp.
25. Chu, T.-Y.J., and R. J. Ruane. Wastewater Treatment for Coal-Fired
Electric Generating Stations. In: Proceedings, 6th Water and
Wastewater Equipment Manufacturers Association Conference, St. Louis,
Missouri, April 1978. pp. 311-330.
26. Environmental Protection Agency. Development Document for Effluent
Limitations Guidelines and New Source Performance Standards for the
Steam Electric Power Generating Point Source Category.
EPA-440/l-74-029a, October 1974. 840 pp.
27. U.S. Nuclear Regulatory Commission. Calculation of Releases of
Radioactive Materials in Gaseous and Liquid Effluents from Boiling
Water Reactors (BWR-GALE code). NUREG-0016, USNTIS PB 251 717,
April 1976. 138 pp.
28. U.S. Nuclear Regulatory Commission. Calculation of Releases of
Radioactive Materials in Gaseous and Liquid Effluents from Pres-
surized Water Reactors (PWR-GALE code). NUREG-0017, USNTIS PB 251
718, April 1976. 146 pp.
29 U.S. Nuclear Regulatory Commission. Estimating Aquatic Dispersion
of Effluents from Accidental and Routine Reactor Releases for the
Purposes of Implementing Appendix I. Regulatory Guide 1.113,
revised April 1977. 90 pp.
30. U.S. Nuclear Regulatory Commission. Calculation of Annual Doses to
Man from Routine Releases-of Reactor Effluents for the Purpose of
Evaluating Compliance with 10 CFR Part 50, Appendix I. Regulatory
Guide 1.109, October 1977. 80 pp.
31. U.S. Nuclear Regulatory Commission. Cost-Benefit Analysis for
Radwaste Systems for Light-Water-Cooled Nuclear Power Reactors.
Regulatory Guide 1.110, March 1976. 89 pp.
32. Wright J. H., and S. J. Dea. Water Pollution Control in the Power
Generation Industry. Industrial Wastewater Management Handbook,
ed. H. S. Azad, McGraw-Hill, 1976. pp. 10-1 through 10-41.
33. Chen, Y. S., J. L. Petrillo, and F. B. Kaylor. Optimal Water Reuse
in Recirculating Cooling Water Systems for Steam Electric-Generating
Stations. In: Proceedings, 2nd National Conference on Complete Water
Reuse, Chicago, Illinois, May 4-8, 1975. pp. 528-541.
34. Carrol, J. W., L. R. LePage, and B. A. Milnes. Use of Agricultural
Runoff Water for Power Plant Cooling. In: Proceedings, 39th Inter-
national Water Conference. Engineers' Society of Western Pennsylvania,
Pittsburg, Pennsylvania, October 31-November 2, 1978. pp. 309-322.
-104-
-------
35. Blackburn, James W. Removal of Salts from Process Waste Waters.
In: Chemical Engineering, Deskbook Issue. October 17, 1977. pp.
33-41.
36. Electrical World. Tower Slowdown Recovers Feed Water. February 1,
1976, pp. 56-57.
37. Nishiwaki, Tadashi. "Concentration of Electrolytes Prior to Evapora-
tion With an Electromembrane Process." Chapter 6 In: Industrial
Processing with Membranes. Lacey and Loeb, editors. Wiley-
Interscience, New York, New York, 1972. pp. 83-106.
38. LeGros, P. G., C. E. Gustafson, B. P. Shepard, and W. F. Mcllhenney.
Systems Analysis of Brine Disposal From Reverse Osmosis Plants.
National Technical Information Service. 1970. PB 198 937.
Springfield, Virginia 22161. 201 pp.
39. Kasarek, L. J. Significantly Increased Water Recovery from Cooling
Tower Slowdown Using Reverse Osmosis. Paper presented at the 84th
National Meeting of the American Institute of Chemical Engineers,
Atlanta, Georgia, Feb. 26-March 1, 1978. 23 pp.
40. Chian, E.S.K., and H.H.P. Fang. RO Treatment of Power Plant Cooling
Tower Slowdown for Reuse. In: Proceedings, American Institute of
Chemical Engineers Symposium, Water 1975, 71 (151), 1975. pp. 82-86.
41. Chu, T.-Y.J., P. A. Krenkel, and R. J. Ruane. Characterization and
Reuse of Ash Pond Effluents in Coal-Fired Power Plants. Journal of
the Water Pollution Control Federation Conference, 50. Nov. 1978.
pp. 2494-2508.
42. Bolton, N. E., R. I. Van Hook, W. Fulkerson, W. S. Lyon, A. W. Andren,
J. A. Carter, and J. F. Emery. Trace Element Measurements at the
Coal-Fired Allen Steam Plant. NSI-EP-43, Oak Ridge National Laboratory,
Oak Ridge, Tennessee, 1973. 136 pp.
43. Stephens, N. T., A. K. Furr, C. C. Thompson, and R. L. Charter. Trace
Metals in Effluents from Coal-Fired Furnaces. Paper presented at the
4th Annual Environmental Eng. and Sci. Conf. University of Louisville,
Louisville, Kentucky, 1974. pp. 701-717.
44. O'Connor, J. T., M. H. Virshbo, C. J. Cain, C. J. O'Brien. The
Composition of Leachates from Combustion Byproducts. Paper presented
at the ASCE National Environmental Eng. Div. Conf., Symposium on
Wastewater Effluent Limits, University of Michigan, Ann Arbor,
Michigan, 1973. 22 pp.
45. Weeter, D. W., J. E. Niece, A. M. DiGioia, Jr. Environmental Management
of Residues from Fossil-Fuel Power Stations. Paper presented at the
47th Water Pollution Control Federation Annual Conference, Denver,
Colorado, 1974. 22 pp.
46. Noblett, James G. and Peter G. Christman. Water Recycle/Reuse Alter-
natives in Coal-Fired Steam-Electric Power Plants. EPA-600/7-78-055a & b.
March 1978. Volumes I & II. 171 pp. and 576 pp.
-105-
-------
47. Cox, D. B., T.-Y.J. Chu, and R. J. Ruane. Quality and Treatment of
Coal Pile Runoff. In Proceedings: Seventh Symposium on Coal Mine
Drainage Research. NCA/BCR Coal Conference and Expo IV, Louisville,
Kentucky, October 13-20, 1977. pp. 232-255.
48. Federal Water Quality Administration. Oxygenation of Ferrous Iron.
U.S. Govt. Printing Office, Washington D.C., 1970.
49. Browne, F. X., and D. Wynes. Power Plant Wastewater Treatment
Design and Operation. In: Proc., 32d Annual Purdue Ind. Waste
Conference, Purdue University, 1977. pp. 968-980.
50. Stone and Webster Engineering Corporation and Ionics Incorporated.
Stone and Webster/Ionics S02 Removal and Recovery Process, Phase I
Final Report. EPA-650/2-75-045. May 1975.
51. Anderson, W. C., and M. P. Youngstrom. Coal Pile Leachate Quantity
and Quality Characteristics. In: Proceedings, 6th Symposium on Coal
Drainage Research, NCA/BCR Coal Conference and Expo III, 1976.
52. Matsugu, R. S. Ontario Hydro Research Division Report. Toronto,
Ontario, 1976.
53. Wewerka, E. M., J. M. Williams, P. L. Wanek, and J. D. Olsen.
Environmental Contamination from Trace Elements in Coal Preparation
Wastes - A Literature Review and Assessment. EPA-600/7-76-007,
August 1976. 61 pp.
54. Environmental Protection Agency. Quality Criteria for Water.
EPA-440/9-76-023, 1976.
55. National Academy of Science. Water Quality Criteria 1972. U.S.
Government. Printing Office', Washington, D.C., 1973. 594 pp.
56. Levander, 0. A. Metabolic Interrelationships Between Arsenic and
Selenium. Presented at the NIEHS International Conference on Environ
mental Arsenic, Fort Lauderdale, Florida, October 5-8, 1976.
57. Mesmer, R. E. and C. F. Baes, Jr. The Hydrolysis of Cations: A
Critical Review of Hydrolic Species and Their Stability Constants in
Aqueous Solution. Report No. ORNL-NSF-EATC-3, Oak Ridge National
Laboratory, 1975. 489 pp.
58. Cherkinski, S. N. and F. I. Ginzburg. Purification of Arsenious
Waste Waters. Water Pollution Abstract, 14, 1941. pp. 315-316.
59. Berezman, R. I. Removal of Inorganic Asenic from Drinking Water
Under Field Conditions. Water Pollution Abstract, 29, 1965. p. 185.
60. Lee, J. Y. and R. G. Rosehart. Arsenic Removal by Sorption Process
from Waste Waters. Canadian Mining and Metallurgical Bulletin
(Canadian Institute of Mining). 65 (11), 1972. pp. 33-37.
-106-
-------
61. Bellack, E. Arsenic Removal from Potable Water. Journal American
Water Works Association, 63, 1971. pp. 454-458.
62. Hannah, S. A., M. Jelus, and J. M. Cohen. Removal of Uncommon Trace
Metals by Physical and Chemical Treatment Processes. Journal of the
Water Pollution Control Federation, 49 (11), 1977. pp. 2297-2309.
63. Wilmoth, R. C., and R. D. Hill. Mine Drainage Pollution Control
via Reverse Osmosis. Mining Engineering, 25 (3), March 1973.
pp. 45-47.
64. Blackshaw, G. L., and A. W. Pappano. Potable Water from Acid
Mine Drainage by Reverse Osmosis. In: Proceedings, 5th Annual
Environmental Engineering and Science Conference, Louisville,
Kentucky, March 3-4, 1975. pp. 561-571.
65. Wilmoth, R. C., and R. B. Scott. Water Recovery from Acid Mine
Drainage. In: Proceedings, 3d National Conference on Complete Water
Reuse, Cincinnati, Ohio, June 27-30, 1976. pp. 441-446.
66. Wilmoth, R. C. Applications of Reverse Osmosis to Acid Mine
Drainage Treatment. EPA-670/2-73-100, December 1973. 159 pp.
67. Kaup, E. G. Reclamation of Acidic Mine Drainage Waters by Ion
Exchange and Reverse Osmosis. In: Proceedings, American Institute
of Chemical Engineers Symposium, Water 1973, 70 (136), 1973.
pp. 557-562.
68. Rice, J. K., and Strauss, S. D. Water Pollution Control in Steam
Plants. Power, A Special Report, 121 (4), April 1977. pp. S1-S20.
69. Katz, W. E. Electrodialysis Preparation of Boiler Feed and Other
Demineralized Waters. In: Proceedings American Power Conference,
vol. 33. 1971. pp. 830-841.
70. Rowland, H., I. Nasbaum, and F. J. Jester. Consider RO for
Producing Feedwater. Power, December 1971. pp. 47-48.
71. Skrinde, R. T., W. M. Steeves, L. S. Shields, and T. L. Tang.
Economic and Technical Evaluation of Reverse Osmosis for Ind. Water
Demineralization. Paper presented at Industrial Water and Pollut.
Conf. and Expo. Chicago, Illinois, March 14-16, 1973. 14 pp.
72. Truby, R. Reverse Osmosis for Boiler Feedwater Treatment. Power
Engineering, December 1976. pp. 58-60.
73. Steiner, G. R., C. L. McEntyre, and T.-Y.J. Chu. Treatment of
Metal Cleaning Wastes at TVA Power Plants. Paper presented at
84th National Meeting of American Institute of Chemical Engineers,
Atlanta, Georgia, February 26-March 1, 1978. 44 pp.
74. Chu, T.-Y.J., G. R. Steiner, and C. L. McEntyre. Removal of Complex
Copper-Ammonia Ions from Aqueous Wastes with Fly Ash. In: Proceedings,
32nd Industrial Waste Conference, Purdue University, May 10-12, 1977.
pp. 359-376. _10?_
-------
75. Golomb, A. An Example of Economic Plating Waste Treatment by Reverse
Osmosis. In: Proceedings, 6th International Water Pollution Research
Conference, June 18-23, 1973. pp. B/2/3/1-B2/3/11.
76. McNulty, K. J., R. L. Goldsmith, A. Gollan, S. Hossain, and D. Grant.
Reverse Osmosis Field Tests: Treatment of Copper Cyanide Rinse Waters
EPA-600/2-77-170, August 1977. 88 pp.
77. Bornstein, L. J., R. B. Fling, F. D. Hess, R. C. Rossi, and J. Rossoff.
Reuse of Power Plant Desulfurization Wastewater. EPA-600/2-76-024,
EPA, Washington, B.C., 1976. 125 pp.
78. Schultz, J. J., T. M. Kelso, J. L. Graham, J. F. Metcalfe, and N. D.
Moore. The TVA Widows Creek Limestone Scrubbing Facility, Part III,
Pilot-Plant and Prototype Operating Experience. In: Proceedings,
Flue Gas Desulfurization Symposium, New Orleans, Louisiana, May 14-17,
1973. pp. 495-538.
79. Grimm, C. and J. Rios. Operating Experience with a Water Management
System Designed for Zero Waste Water Discharge at the Colstrip Steam
Electric Station. In: Proceedings, 39th International Water
Conference. Engineers' Society of Western Pennsylvania, Pittsburgh.
October 31-November 2, 1978. pp. 323-341.
80. Meliere, K. A., R. J. Gartside, W. A. McRae, and T. F. Seamans.
Description and Operation of the Stone and Webster/Ionics S02 Removal
and Recovery Pilot Plant at The Wisconsin Electric Power Company,
Valley Station, In Milwaulkee. Prepared for: EPA Flue Gas
Desulfurization Symposium, Atlanta, Georgia, November 7, 1974.
pp. 1109-1126.
81. Glasstone, S. and A. Sesoaske. Nuclear Reactor Engineering. Van
Nostrand Reinhold Co., New York, 1967. p. 95.
82. Glasstone, S. and A. Sesonske. ibid. Chap. 2. pp. 30-108.
83. Godbee, H. W. Use of Evaporation for the Treatment of Liquids in
the Nuclear Industry. Oak Ridge National Laboratory, September
1973. 97 pp.
84. Herald, W. R. and R. C. Roberts. Development of Ultrafiltration and
Inorganic Adsorbents for Reducing Volumes of Low Level and Intermediate
Level Liquid Wastes: April-June 1978. Mound. Laboratory, Miamisburg,
Ohio, July 1978.
85. Stana, B. R., and J. W. Koenst. Evaluation of a Westinghouse
Reverse Osmosis Radioactive Laundry Water Concentrator. Westing-
house Electric Corporation, Pittsburgh, Pennsylvania, 1973. 49 pp.
86. Rosain, R. M. The Design of Wastewater Treatment Facilities to
Meet Power Plant Effluent Guidelines. In: Proceedings, 31st
Industrial Waste Conference, Purdue University, May 5-6, 1976.
-108-
-------
87. Markind, J., J. S. Neri, and R. R. Stana. Use of Reverse Osmosis
for Concentrating Waste Soluble Coolants. In: Proceedings, American
Institute of Chemical Engineers Symposium, Water 1975, 71 (151),
1975. pp. 70-75.
88. Bansal, I. K. Concentration of Oily and Latex Wastewaters Using
Ultrafiltration Inorganic Membranes. In: Proceedings, 3d National
Conference on Complete Water Reuse, Cincinnati, Ohio, June 27-30,
1976. pp. 274-285.
89 Wang, L. K., J. Y. Yang, and D. B. Dahm. Separation of Emulsified
Oil from Water. Chemistry and Industry, July 5, 1975. pp. 562-564.
90. Bansal, I. K. Ultrafiltration of Oily Wastes from Process Industries.
Paper presented at the 68th Annual Meeting of the American Institute
of Chemical Engineers Conference in Los Angeles, California,
November 16-20, 1975. 7 pp.
91. Zabban, W., E. D. Escher, R. Helwick, E. F. Maziarz. Treatment of
Oily Wastes: Ultrafiltration and Chemical Methods. In: Proceedings,
38th International Water Conference. Engineers Society of Western
Pennsylvania, Pittsburgh, Pennsylvania. November 1-3, 1977.
pp. 234-239.
92. Kawabe, A., Y. Kakimoto, H. Karaso, 0. Abe, K. Yabe, and M. Iraaizumi.
Wastewater Reclamation Through Reverse Osmosis Process at a Power
Station. In: Proceedings, The Second National Conference on Complete
Water Reuse, Chicago, Illinois, May 4-8, 1975. pp. 942-955.
93. Printz, J., L. W. Johnson, J. T. Aronson, and H. J. Martin. Design
and Operation of a Zero Discharge System at Hayden Station. In:
Proceedings, 39th International Water Conference. Engineers' Society
of Western Pennsylvania. Pittsburgh. October 31-November 2, 1978.
pp. 343-351.
-109-
-------
APPENDIX A
USER'S GUIDE TO REVERSE OSMOSIS
For the past several years TVA has been studying possible applica-
tions of reverse osmosis and ultrafiltration to power plant waste treat-
ment. In the process, cognizance of equipment and technical literature
has been maintained. Potential power plant applications have been dis-
cussed with the plant designers and operators. In addition, some hands-on
experience has been gained by treating waste samples in TVA pilot facility
which is equipped to test modules manufactured by Abcor, Dow, DuPont,
Osmonics, Romicon, UOP-ROGA, UOP-Fluid Sciences, and Union Carbide.
As a result, TVA has developed a proposed user's guide to reverse
osmosis. Using this guide, prospective users may avoid the major pit-
falls in selecting reverse osmosis units. Normally, the user must first
adequately describe the problem (or opportunity) to vendors and then be
able to review and evaluate their proposals. Hopefully, this guide pro-
vides a comprehensive list of factors which must be considered in this
process. Although new membranes are being developed this guide emphasizes
commercial cellulose acetate and polyamide membranes. No attempt to
address the dynamic membranes formed by zirconium oxide-polyacrilic acid.
If there are other important areas that have been glossed over, comments
and discussion with interested readers are welcomed by the authors.
Many problems reported for reverse osmosis systems are not intrinsic
to the technology but are due to inadequate design of auxiliary equipment.
The consumer must share the responsibility for such problems with designers
and vendors.
Feedwater and Pretreatment Criteria
Many of the problems reported with reverse osmosis units stem from
inadequate pretreatment. Often, design of the pretreatment system is
based on an inadequate description of influent water quality. Where sur-
face water is used, care should be taken to project future changes in
water quality. In brackish water areas, future development may cause
increased water use upstream, thus degrading water quality and quantity.
Overuse of groundwater in coastal areas can cause salt water intrusion.
Inland, overpumping may cause ground water levels to drop, decreasing
water availability. Variations in weather can also affect water supply
and quality. Extended droughts result in lower supplies with increased
dissolved solids. High suspended solids are frequently associated with
flash floods. The destratification (seasonal overturn) of lakes which
occurs as temperatures drop each fall can cause major changes in water
110
-------
quality. Information on natural water quality variations is generally
available from State and Federal agencies charged with water pollution
control and resource conservation, e.g., U.S. Geological Survey, TVA,
and EPA. Industrial water and waste streams should be carefully studied
to predict the changes in water quality and quantity which occur due to
process changes. (This can be the most difficult task in implementing
any industrial waste treatment.) Water analyses should include each of
the parameters mentioned below under pretreatment requirements. In addi-
tion, the concentrations of major ions must be known.
Parameters that may require control through pretreatment include
suspended solids, silica, iron, manganese, biofouling, calcium, magnesium,
pH, and temperature. Criteria for control of iron, manganese, calcium,
and magnesium are fairly clear, although different control processes can
be used. Criteria for controlling suspended solids, biofouling, tempera-
ture, and pH vary from one membrane manufacturer to another depending on
the configuration and the membrane material.
There are several approaches to setting criteria for influent sus-
pended solids, and a good deal of controversy surrounds the issue. As a
potential user, it is best to use all of these approaches if possible so
that the maximum number of vendors can be considered. One approach is
to use the silt density index test. In this test, feed water under a
controlled pressure is forced through a .45 pro (Millipore brand) filter
for a fixed time period, usually 15 minutes. The percent plugging is
then 100% x (l-Tl/t2^t where Tx and T2 are the time intervals required
to collect 500 mi at the beginning and end of the test, respectively.
The silt density index is the percent plugging divided by the total test
period in minutes. Where the 15-minute test shows less than 45 percent
plugging, the silt density index is less than 3, and membrane fouling
should not be a problem, although higher silt densities are tolerable in
some systems. Another approach is simply to require a cartridge prefilter
rated at 5 to 10 pm. While this protects the membranes from plugging,
it does not protect the owner from high operational cost incurred from
continually replacing these filters. Also, colloidal fouling can be caused
by particles smaller than 5 pm. TVA recommends viewing these replaceable
filters as an in-situ test of feed water quality similar to the silt den-
sity index. Neither test is strictly comparable to the membrane module
since no flow is maintained along the filter to carry away concentrate
and no ionic strength variations occur. Ionic strength plays a key role
in the stability of colloids and fine particulates and is proportional
to the concentration of each ion times the square of its charge. In fact,
traditional coagulation practice is based, in part, on destabilization
of colloids by increasing the ionic strength. The zeta potential test
can be used in conjunction with the silt density index. In this test,
colloid mobility in an electric field is observed with the aid of a micro-
scope. The net charge on the colloids is inferred from their size and
speed. The higher the electric charge, or zeta potential, the more stable
the colloid suspension will be, since like charges repel. Where colloids
are stable, higher silt density indexes are tolerable. Another measure
of colloid stability is the 24-hour settling test recommended by Osmonics.
Each of these approaches is valid in the context of the manufacturer's
recommendations.
-Ill-
-------
In general, particulate removal criteria for hollow fiber modules
are more stringent than for spiral wound. Where pretreatment to meet
these stringent requirements is feasible, both membrane configurations
should be considered, since the two types are competitively priced. On
the other hand, tubular configurations are much higher in cost but are
useful where influent particulate removal is impractical or uncertain.
In these cases, periodic cleaning is required, either by chemical or
mechanical means.
Pilot plant testing of a reverse osmosis unit is always needed when
particulate levels cannot be controlled to meet the manufacturer's criteria.
Where regular cleaning is necessary, pilot plant work should include devel-
oping cleaning methods and schedules and documenting the expected produc-
tivity. Where particulate levels are to be controlled by pretreatment,
pilot testing of the membrane is not necesary (at least not to predict
particulate fouling). Nevertheless, the efficacy of the pretreatment
itself should be tested. This may be done on a bench scale by jar tests
or preferably on a larger scale. Pretreated water can then be subjected
to the usual criteria for particulate fouling. These may include silt
density, zeta potential, the 24-hour settling test, and possibly cart-
ridge filtration to determine the required frequency of replacement.
Methods of removing particulates include coagulation and filtration
and ultrafiltration. Inline coagulation is often cost effective. Suc-
cessful coagulation requires careful operation and control which, though
not difficult, does require training. Ultrafiltration is an expensive
pretreatment which can sometimes be justified on the basis of its high
reliability and relative lack of dependence on operator expertise. This
may result in fouling of the ultrafiltration membrane to protect the
reverse osmosis membrane. The two are about equally expensive on the
basis of surface area, but the ultrafiltration membrane will be cheaper
in terms of cost per unit of flow. Also, ultrafiltration membranes are
available which can tolerate strong chemical cleaning solutions.
Softening can affect particulate fouling in two ways. Lime-soda
ash softening will often trap particulates in the floe which forms. On
the other hand, ion exchange softening stabilizes colloids by reducing
the ionic strength, which can be helpful in preventing fouling.
Criteria for the control of calcium and magnesium scale are well known.
Slide rules, charts, and nomographs abound for determining the Langlier
Index, (Al, A2)* recognized criterion for CaC03 scale. In fresh waters,
published solubility constants can be used for guidance with adjustments
for ionic strength. See, for example, Chapter 28 of Fair, Geyer and Okun,
(A3) for a discussion of this procedure and a list of useful solubility
constants. Marshall and Slusher (A4) provide an algorithm for determining
CaS04 solubility in salt waters. Many firms have access to comprehensive
computer programs for determing the stability of dissolved species under
a variety of conditions. Examples in the public domain include WATEQ (A5)
and MINEQL (A6). To predict scaling, the reject water characteristics
*Numbers in parenthesis refer to the list of references at the end of this
appendix.
-112-
-------
should be estimated based on the feedwater and the volume recovery desired.
The worst case concentration at the membrane surface is then estimated
based on the vendor's experience. (Usually this is something like 10 per-
cent more concentrated than the bulk reject flow.) The following species
should be checked: CaS04, CaC03, and MgOH2.
Options to consider for control of calcium and magnesium scales include
the use of chemical dispersants, lime soda softening, or ion exchange
softening. These processes may in turn affect other pretreatroent needs.
Lime soda softening affects pH, iron, manganese, silica, and particulates.
Ion exchange softening will lower the ionic strength and stabilize particu-
lates. This effect is more dramatic than the lowering of ionic strength
alone would indicate, since at a given ionic strength divalent ions are
30 times more effective than monovalent in destabili2ing colloids. Ion
exchange is also capable of removing iron and manganese, although this
can sometimes foul the resin. Dispersants such as sodium hexametaphosphate
can also control iron and manganese fouling when used in stoichiometric
quantities. Where calcium carbonate scaling is the only problem, pH control
alone may be adequate.
Silica scaling is still poorly understood, and good predictive cri-
teria have not yet been established. Silica reacts with nearly every
cation, and deposits are often amorphous and difficult to characterize.
Some vendor guidelines suggest keeping silica below 100 ppm; others recom-
mend higher limits. Silica can be removed by coprecipitating it with
magnesium in the lime-soda softening process.
Iron and manganese fouling presents a special problem in that it
takes only a few parts per million to foul the membranes, and yet these
species are easily controlled by manipulating pH and redox potentials.
Metal cleaning wastes in industry present a similar fouling problem. It
is necessary to first destabilize the metal hydroxides by aeration and
high pH, remove the precipitate, and finally restablize the supernatant
by lowering the pH. The need for treatment is again determined by the
predicted concentrations at the membrane surface. Remember that small
errors in predicting or controlling pH or redox potential can create a
significant problem; thus, allowances should be made accordingly. Sta-
bility diagrams showing the log of concentration versus pH for a given
metal are helpful in understanding and controlling this problem. The
use of these diagrams is discussed in more detail in Fair, Geyer, and
Okun (A3) and Stumm and Morgan (A7).
In addition to its role in stabilizing or destabilizing cations, pH
is important to membrane life. The most common membrane materials, cel-
lulose acetate and polyamide, are subject to accelerated hydrolysis away
from the optimal pH of about 5.5. The polyamide material is less sensi-
tive to pH having a useful pH range of about 4 to 12. In comparison,
cellulose acetate membranes are useful between a pH of 3 and 8. At the
extremes of these ranges, membrane life decreases dramatically. New
membrane materials are being developed which operate in other pH ranges.
Certain weakly ionized species, such as carbon dioxide, ammonia, or
silica may exhibit pH dependent salt passage. If these species are
important, pH adjustment may be tailored to provide optimal rejections.
-113-
-------
Both the performance and the life of reverse osmosis membranes are
affected by temperature. Generally, the flux or the permeate rate for
both the concentrate and permeate need to be defined. The cost of con-
centrate handling should be evaluated and included in these criteria--
especially if concentrate must be evaporated and stored. The value of
recovered water and products should be investigated.
If the goal is to produce high quality water, optimal water recovery
is determined by the cost of raw water, pretreatment, and possibly brine
disposal. Where the goal is to concentrate the brine for reuse or dis-
posal, recovery is usually limited by scale formation or by osmotic
pressure increases.
Estimating Membrane Performance
TVA has tested the performance of seven commercial reverse osmosis
modules and found each to meet the manufacturer's specifications with
the exception of two tubular units. One of these tubular units was a
discontinued product, and specifications were not available. As mentioned
before, the cost of commercial tubular units prohibits their use unless
fouling is to be expected. This being the case, their performance cannot
be accurately predicted in any event without field tests.
For the spiral and hollow fiber modules, performance predictions
are often possible by making adjustments for variations from the standard
test conditions used in the manufacturer's specifications. The solution-
diffusion model is used to make these adjustments. A good discussion of
this model is provided in J. W. McCutchan's paper, "Membranes Simplified,"
presented at the Seminar on Membrane Separation Processes at Clemson
University in August 1977. It is assumed that water and the various
solutes dissolve in the membrane and diffuse through it in response to
pressure and concentration gradients. The following equations result:
Ji = A (AP - ATT)
J2 = B AC
Here Jt is the water flux, J2 is the solute flux, AP is the pres-
sure difference, ATI is the osmotic pressure, and AC is the concentration
difference—all measured across the membrane. The membrane itself can
be characterized by the coefficients A and B. Note that B varies with
each solute. Osmotic pressure can be calculated from solute concentra-
tions. A rough estimate is 10 psi per 1,000 ppm dissolved solids.
Osmotic pressure can also be inferred from measurements of boiling point
elevation or freezing point depression (A8). Sophisticated computer pro-
grams are also available which can provide good estimates of osmotic
pressure by taking into account ion pairing and ionic strength correc-
tions .
In TVA tests involving two cellulose acetate and two polyamide modules,
no significant variation was found in the permeability coefficient "A"
as pressure was varied from 50 to 400 psi. Also, "A" varied with tempera-
ture in the same manner as the reciprocal of the absolute viscosity of water.
-114-
-------
A plausible explanation for the variation in liquid viscosity with tem-
perature relates it to the energy necessary for a molecule to break free
of the liquid matrix (A9). Similarly, a water molecule must free itself
from the liquid matrix in order to dissolve in the membrane. Thus, the
same mechanism may control both of these processes.
Productivity varies in time as a function of fouling, compaction,
and hydrolysis. Where pretreatment is inadequate to prevent fouling,
experience developed from pilot testing is necessary to determine mem-
brane productivity. Compaction and hydrolysis are better understood,
and manufacturers can often provide guidelines for conservatively
estimating these losses. Compaction is a function of the pressure and
temperature history of the membrane, while hydrolysis is dependent on
the temperature and pH history. To predict productivity from standard
test conditions, refer to the fundamental equation for water flux:
Ji _ (AP-An) ,
V (AP'-AfT)
where the prime denotes the test conditions. The bars indicate average
values taken between the feed and reject conditions.
Besides productivity, the other major performance parameter is solute
passage, the ratio of solute concentration in the permeate (C ) to that
in the feed (Cf), expressed as a percentage. Rejection, a related parameter,
is simply 100 percent minus the solute passage. From the fundamental
flux equations, the following equation for solute passage, SP, can be
developed.
cp - JE - J2/Ji _ B (AC) ,
sp - P - p— - ii
Lf Lf A (AP-ATl) Cf
To predict solute passage based on tests at other conditions, the following
ratio can be used:
SP _ AC (AP'-Aff ) . Cf' ,
op ' ~ _
AC' (AP-ATI) Cf
The prime again denotes test conditions.
For the major ions found in natural waters, solute passage information
is available both in the literature and from the membrane vendors. Weakly
ionized solutes, such as ammonia and H2C03, exhibit pH dependent rejection,
and the rejections of cations and anions are interdependent. These factors
make exact predictions difficult. Nevertheless, where a membrane exhibits
high rejections of NaCl, rejections of other inorganic ions are generally
as good or better. The imprecision is usually tolerable. Where extremely
-115-
-------
high product quality is desired, ion exchange polishing is used. Since
the loading to the ion exchanges is low, this is not a costly addition
to the process.
Solute passage prediction is generally difficult for weakly ionized
inorganics and small organic molecules. If these solutes are important,
pilot testing is generally necessary.
TVA was interested in developing rejection data on trace metals.
Tests were run on a mixture of metal perchlorates to study their relative
salt passage. The trends in these data are not consistent among the mem-
branes tested, nor are they entirely consistent with similar data reported
in the literature. Nevertheless, the TVA results may be useful to others,
they are shown in Table Al.
Very few reverse osmosis installations are expected to fail because
of poor performance predictions, especially where pretreatment is adequate
to prevent fouling and the solute rejections have been studied.
Auxiliary Equipment
Many problems have occurred because of poor auxiliary hardware.
DSS Engineers, Incorporated, in a study for the Office of Water Research
and Technology (AID), has reviewed 11 commercial desalting plants, including
7 reverse osmosis plants. Anyone purchasing a reverse osmosis system
should review their report to identify the potential problems. One of
the most common problems was the failure of acid addition systems.
The corrosion resistance of metals, especially those in contact with
the liquid, should be stressed in the specifications and should be checked
during the bid evaluations and construction. This is not a simple matter.
DBS reports that "Corrosion of cbpper alloys, stainless steels, aluminum,
cast iron, and carbon steels has occurred in raw water service at one
plant or another. No universally applicable set of materials selections
seems possible because of the great variety of raw water constituents
and concentrations encountered." Incidentally, corrosion of steel will
generally cause membrane fouling downstream. Plastic pipe is corrosion
resistant, but is subject to cracking from mechanical stresses due to
vibration. Differences in thermal expansion can also cause mechanical
stresses where plastic and metal are joined.
Maintenance requirements should be assessed to ensure availability
of service and spare parts.
Other mechanical features should be checked. TVA had considerable
problems which were solved before obtaining suitable pilot plant operation.
These included a piston pump which was not matched to its load, a check
valve installed backwards, a faulty solenoid, and a few common fittings
mixed in with stainless steel. Problems of this type should be less common
on production systems.
-116-
-------
Table Al
REJECTION OF METAL PERCHLORATE ON VARIOUS
COMMERCIALLY AVAILABLE MEMBRANES
Ion Membrane
Dow Hollow-Fiber
(CTAa)
DuPont Spiral-Wound
(Polyamide)
ROGA Spiral-Wound
(CTAa)
UOP Tubular
Osmonics Spiral-Wound
(CAB)
Mg+2
Ca+2
Sr 2
Ba 2
As*3
Cd 2
Cr*6
Cu 2
Pb!2
Mn 2
Ni+2
Se 4
Zn+2
• i
K }
Na+1
BO?
94.2
97.2
83.7
92.2
90.7
95.4
97.3
94.7
>93.8
95.4
94.9
98.4
95.0
92.8
85.0
87.0
28.6
84.0
84.6
87.3
98.1
88.8
91.3
86.4
82.2
>92.3
90.5
89.0
81.1
90.0
69.2
81.8
72.7
65.5
88.9
84.1
79.2
82.8
90.7
85.2
92.6
92.6
>92.9
87.5
83.7
92.2
87.7
77.6
62.5
68.2
36.4
91.7
89.9
88.8
94.6
93.1
95.9
96.2
96.9
>95.0
95.3
95.7
97.2
95.2
87.0
76.7
76.2
68.8
92.0
89.4
87.5
93.1
91.9
94.5
96.0
96.8
>91.4
94.2
94.9
98.7
94.3
77.3
66.7
52.6
41.7
° CTA = Cellulose Triacetate
CA = Cellulose Acetate
This module is not commercially available; however, the membrane material is similar to that in
, the B-9 hollow-fiber module
Lead (Pb) concentrations in the permeate were below the minimum detectable amount.
Operating Condition
Temperature (°C): 20
Pressure (psig): Dow Hollow-Fiber, 158; DuPont Spiral-Wound, 375; Roga Spiral-Wound, 378; UOP Tubular, 373;
Osmonic Spiral-Wound, 378.
Feed Concentration (mg/£): Mg, 25; Ca, 100; Sr, 2.5; Pb, 15; As, 0.15; Cd, 2.5; Cr, 1.0; Cu, 10; Pb, 0.15;
Mn, 2.5; Ni, 12.5; Se, 0.5; Zn, 2.5; Li, 2.5; K, 25; Na, 25; B, 10.
pH of solution: 4.8
-------
Conclusion
While reverse osmosis systems design is not an exact science, certain
classes of problems can be handled in a straightforward manner. This
includes processing solutions of inorganic ions to concentrate or purify,
providing that adequate pretreatment is available. Pilot testing of the
pretreatment step is advisable, but pilot testing of the reverse osmosis
system is not generally necessary. Exceptions include situations where
adequate pretreatment is not practical and where organic solutes or weakly
ionized species are important.
-118-
-------
REFERENCES
Al. Larson, T. E. and A. M. Buswell. Calcium Carbonate Saturation
Index and Alkalinity Interpretations. Journal American Water Works
Association. Vol. 34, pp. 1667-1684, November 1942. pp. 1667-1684.
A2. BETZ Laboratories, Inc. Handbook of Industrial Water Conditioning.
Seventh edition, Trevose, Pennsylvania, 1976. p. 180. 425 pp.
A3. Fair, G. M., J. C. Geyer, and D. E. Okun. Water and Wastewater
Engineering. Vol. 2. John Wiley & Sons, 1968. 669 pp.
A4. Marshall, W. L. and R. Slusher. Aqueous Systems at High Temperature,
Solubility to 200°C of Calcium Sulfate and Its Hydrates in Sea Water
and Saline Water Concentrates, and Temperature-Concentration Limits.
Journal of Chemical and Engineering Data. Vol. 13, No. 1, pp. 83-93,
January 1968.
A5. Plummer, L. N., B. F. Jones, and A. H. Truesdell. WATEQF--A Fortran
IV Version of WATEQ, A Computer Program for Calculating Chemical
Equilibrium in Natural Waters. Water Resources Investigations 76-13,
U.S. Geological Survey, Reston, Virginia, 1976. 61 pp.
A6. Westall, J. C., J. L. Zachary, and F.M.M. Morel. MINEQL, A Computer
Program for the Calculation of Chemical Equilibrium Composition of
Aqueous Systems. Technical Note 18. Ralph M. Parsons Laboratory,
Massachusetts Institute of Technology, Cambridge, Massachusetts. 1976.
91 pp.
A7. Stumm, W. and J. J. Morgan. Aquatic Chemisty. Wiley-Interscience,
New York, 1970. 583 pp.
A8. Glasstone, S. Textbook of Physical Chemistry, 2d edition, McMillan
& Co., London. 1966. 672 pp.
A9. Glasstone, S. ibid., pp. 501-507.
A10. DSS Engineers, Incorporated. Commercial Membrane Desalting Plants,
Data and Analysis. National Technical Information Service PB-253 490.
Springfield, Virginia 22161. 335 pp.
-119-
-------
APPENDIX B
MEMBRANE SEPARATION OF TOXIC SUBSTANCES IN WASTEWATER
INTRODUCTION
The release of toxic and potential toxic compounds to the environ-
ment in industrial wastewaters has created the need to seek a scientific
but manageable approach to the identification and control of these chemi-
cals. A major problem in the development of suitable treatment technolo-
gies for toxic pollutants is the large number of compounds to be addressed
and the vast number of combinations of these substances which may be found
in any given waste stream. With the present emphasis on control of specific
compounds, the 129 priority pollutants were chosen for assessment of their
treatment technologies. These 129 priority pollutants can be classified
into 10 groups (Bl)*, as shown in Table Bl**. This study is concerned
with the potential of membrane technology for controlling priority
pollutants.
REVERSE OSMOSIS
Separation Mechanism
Reverse osmosis separation is the combined result of preferential
sorption of solvent or solute at the membrane-solution interface and the
flow of the interfacial fluid through the pores on the membrane surface.
Both the porous structure and the chemical nature of the membrane surface
together determine the solute and solvent flux through the membrane.
This flux is a function of the magnitude of preferential sorption; the
effective thickness of the membrane; the size, number, and distribution
of pores on the membrane surface; and the operating pressure, temperature,
and flow conditions in the apparatus.
The mechanism for inorganic rejection differs somewhat from the mecha-
nism for organic rejection. Theories for reverse osmosis separation of
inorganic ions in aqueous solution were proposed on the basis of electro-
static repulsions of ions at the membrane-solution interface (B2-B4).
Inorganic salt rejection occurs because the inorganic ions are repelled
from the surface of the membrane materials that have a low dielectric
^Numbers in parenthesis refer to the list of references at the end of
this appendix.
-'""Because of the large number of tables in this appendix,
figures and tables are placed after the text.
-120-
-------
with the higher-valence ions being repelled a greater distance. Conse-
quently, the water is absorbed by the membrane surface. The pressure
membrane surface. The pressure exerted on the feed solution forces the
flow of pure water through the pore if the pore in the membrane is the
correct size (twice the pure water layer). The inorganic salts are kept
in the concentrated solution.
Organic rejection is based on the polar and sieve mechanisms. The
steric effect is determined by the size and shape of the organic molecule.
For example, the approximate size of the pores in a cellulose acetate
membrane with 97 percent NaCl rejection is 20 angstroms (2 nanometers).
Usually, such a membrane will reject substantially all the organics with
a molecular weight above 200 and will reject a percentage of those having
a molecular weight between 100 and 200, depending on the shape of the
organic molecule. Polyamide and NS-100 membranes may have a smaller pore
size than cellulose acetate membranes. The polar effect relevant to
reverse osmosis separation, particularly for cellulose acetate membranes,
is the acidity (or the proton-donating characteristic) or the basicity
(or the proton-accepting characteristic) of the molecule concerned. A
quantitative measure of this polar effect is given by either the hydrogen
bonding ability or the dissociations constant of the solute. When the
acidity of the organic is less than that of pure water, the latter is
preferentially sorbed, and positive organic separations are obtainable.
When the acidity of the organic is more than that of water, the organic
is preferentially sorbed at the membrane-solution interface, and organic
separation can be positive, negative, or zero, depending on experimental
conditions.
Rejection of Organics
Tables B2 to B5 screen several hundred rejections reported in 18
publications (B5-B22). These tables include data on several varieties
of the cellulose acetate membrane, a cellulose acetate butyrate membrane,
four varieties of NS-100 membrane, and a duPont polyamide membrane. Much
of the information was tabulated by Cabasso, et al. (B23).
In considering the rejections reported from reverse osmosis experi-
ments, one must bear in mind that experimental procedures for the evalua-
tion of rejection efficiency were in most cases the so-called "short run
type," as noted by Matsuura and Sourirajan (B5-B10). The data presented
by Chian and Fang (Bll) indicate that some organic solutes cause inherent
membrane properties to change. For example, when a cellulose acetate
membrane, which has an initial salt rejection of 97.3 percent (5000 ppm
NaCl at 102 atm), was run in a series of short-term experiments while
several solutes were measured one after the other. After seven solutes
were measured in a row (1000 ppm each of ethanol, i-propanol, acetic acid,
formaldehyde, acetone, ethyl ether, glycerol), the salt (NaCl) rejection
of the membrane was reduced to 90.1 percent. Several reasons are possible
for this decline in the rejection properties of a membrane.
Factors causing reduced rejection may include plasticization of the
membrane by absorbed organic solutes that were not flushed out of the
membrane. This example was raised to point out that rejection data from
-121-
-------
the short-term experiments do not necessarily project the precise condi-
tions of organic concentration. However, polyamide membranes did not
show such rejection decreases. During organic concentration, the mem-
brane is exposed to several solutes for a relatively longer period of
time, and changes in the membrane's inherent rejection properties may
result.
The mechanism by which a membrane rejects the passage of certain
solutes while permitting water transport is still a matter of great con-
troversy in the cited literature. Nevertheless, some conclusions can be
drawn from the data in these tables.
Cellulose Acetate Membranes--
The rejection by membranes prepared from cellulose acetate is highly
dependent on the method by which the membrane is prepared and on the degree
of acetylation of the starting polymer (B24, B25). In general, a correla-
tion exists between the ability of the solute to form hydrogen bonds with
water and the ability of the solute to form bonds with membrane. Dis-
sociated solutes, polyhydric alcohols, and the paraffins are strongly
rejected by the membranes. Rejections of phenols in their undissociated
state is extremely low; the same characteristic is observed with other
organics capable of ionization (Figure Bl).
Rejection of organics increases with an increase in the degree of
branching and in the number of carbon atoms for compounds having the same
functional groups. However, some exceptions to this statement are known,
as illustrated in Figure B2, which shows a decrease in rejection of dicar-
boxylic acids with increasing carbon number. The mechanism of permeability
of organic solutes through the membrane probably involves several parameters
other than molecular size and the chemical nature of the functional groups.
NS-100 Membranes —
The NS-100 membrane, prepared from an interfacial condensation between
an amine and an isocyanate, yields better rejections than do cellulose
acetate membranes. Cadotte, et al. (B20) reported the rejections of two
versions of this membrane. They classify the preparations of the membrane
as a "standard" or a "modified" procedure. The modified procedure yields
a more highly cross-linked membrane.
Analysis of the data in Table B4 for NS-100 rejections shows that
increasing molecular size (for solutes of the same functional type) leads
to better rejections. The higher the cross-link density of membrane sur-
face, the better is the rejection. The nature of the functional group
of organic solutes is much less important in governing rejection by the
NS-100 membrane than it is for the cellulose acetate membrane.
The excellent rejection of organic solutes by the NS-100 membrane
and the wide pH range over which it can operate are attractive features.
However, the sensitivity of this material to chlorine (1 ppm or greater
leads to reduced rejection) is an impediment to its practical use. The
pH stability of the membrane allows operation in both acidic and basic
media; this permits alteration in rejection properties of those solutes
-122-
-------
that act as Lewis bases. The polyurea structure of the membrane is pro-
tonated in solutions that have pH values below 10.8. This resembles a
Lewis base characteristic of the NS-100 membrane toward higher rejections
for amines than for acidic compounds.
Polyamide Membranes--
Chian and Fang (fill) have tested the rejection properties of polya-
mide membranes toward a variety of solutes. The membranes were provided
by duPont's Perma-sep Division as hollow fibers, and flat sheet membranes
were supplied by Chemstrand. The rejection of the polyamide is quite
similar to that observed for the NS-100 membrane (see Table B5), and it
is highly sensitive to the presence of trace levels of chlorine.
In summary, the degree of organic solutes varies with the particular
solute-membrane combination. Reverse osmosis treatment of wastewater if
generally very effective for removing organic compounds that have molecular
weights above 200. Below molecular weight 200, the rejection characteris-
tics are variable, depending on type of membrane used, steric size of
the organic molecule, degree of hydrogen bonding that exists, and ioniza-
tion of the organic molecule.
Molecular weights of organic compounds in the list of priority pollu-
tants range from a low of 50 to a high of over 400. Most of the organic
priority pollutants have molecular weights between 100 and 200. The
information presented in Tables B2 to B5 indicate that cellulose acetate
membranes are not effective for removing low-molecular-weight organics,
whereas NS-100 and polyamide membranes will effectively remove a wide
class of organic priority pollutants.
Rejection of Heavy Metals
Reverse osmosis shows promise as a purification process for brackish
water. Application of reverse osmosis for removing compounds such as
calcium, magnesium, and sodium from solution is not pertinent to this
topic and will not be discussed here.
Information presented in the literature (B26-B31), indicates that
good removal of metallic species in the priority pollutants would probably
be achieved with reverse osmosis treatment of wastewater. (See Table
B6.) In addition, the bulk of the metal removal would take place in the
pretreatment operations, such as clarification or lime softening, required
to achieve satisfactory operation of the reverse osmosis process.
ULTRAFILTRATION
Separation Mechanism
The predominate mechanism in ultrafiltration separation is selective
sieving through pores. Rejection of an ultrafiltration membrane for a
certain substance depends on its molecular shape, size, and flexibility
as well as the operating conditions. A useful membrane must be able to
-123-
-------
separate distinctly at an economical rate. To accomplish this, the ultra-
filtration membrane must have a narrow molecular weight cutoff and a high
solvent flux at low pressure differentials.
Rejection of Organics
Ultrafiltration membranes generally have nominal molecular weight
cutoffs for substances ranging from 500 to 106 in molecular weight. The
major application for ultrafiltration membranes is for removing colloidal
material and large organic molecules in solution. Industrial applications
have been successful in treatment of electroplating, pharmaceutical, and
laundry wastes; oil-water separation; chemical recovery from waste streams;
waste fractionation; and food processing. Since all organic priority
pollutants have a molecular weight below 500, ultrafiltration would not
be a promising process for removing organic priority pollutants.
Rejection of Heavy Metals
Ultrafiltration will not reject dissolved inorganic salts in solu-
tion, and ultrafiltration systems are not concerned with heavy metal
removal.
ELECTRODIALYSIS
Separation Mechanism
The principle of electrodialysis separation is the transport of ions
through ion-exchange membranes as a result of an electrical driving force.
When an electric current passes through the solution compartments and ion-
exchange membranes, cations tend to migrate toward the negatively charged
electrode (cathode) and anions tend to migrate toward the positively
charged electrode (anode). The cations and anions in one set of solution
compartments can pass freely through the cation- and anion-exchange mem-
branes that form the walls of this first set of compartments. However,
once the cations and anions are in the second set of solution compartments
(the alternate compartments), cations are blocked from further transfer
because the anion-exchange membranes will not allow their passage. Simi-
larily, anions are blocked from further transfer because they are blocked
by cation-exchange membranes. An ion-depleted solution can be withdrawn
from the first set of compartments, whereas an ion-enriched solution can
be withdrawn from the second set of compartments. The principal limita-
tion of the production rates achievable in electrodialysis units is con-
centration polarization at the surfaces of the ion-exchange membranes.
Rejection of Organics
Only ionized material can be separated from water by electrodialysis;
organics and other un-ionized substances may be concentrated in the demin-
eralized product water. Therefore, the chief function of electrodialysis
in the treatment of wastewater is to remove inorganics. In the chemical
-124-
-------
industry, electrodialysis has been used to recover carboxylic acids (such
as acetic, citric, and latic acids) and pulping waste chemicals (such as
lignin products) from waste streams (B32). However, electrodialysis is
not concerned with removal of organic priority pollutants.
Rejection of Heavy Metals
Desalination of brackish waters is the main commercial use of elec-
trodialysis systems at present. Another area of interest is wastewater
demineralization. Electrodialysis has been claimed to be a successful
and economical means of removing heavy metals, such as antimony, arsenic,
cadmium, chromium, cobalt, copper, manganese, and zinc, from water and
wastewaters (B33, B34), but no data are available in support of the claim.
CONCLUSIONS
The results of this assessment, based on published literature, indi-
cate the feasibility of removing priority pollutants from wastewaters by
membrane processes. Reverse osmosis is an acceptable process for removing
inorganic and organic priority pollutants. Electrodialysis is feasible
for removing only inorganic priority pollutants. Ultrafiltration holds
no promise for removing priority pollutants from wastewaters.
-125-
-------
ioo-
- 100
7 8 9 10 II
pH OF FEED SOLUTION
Figure Bl. Cellulose acetate membrane rejection of phenol as a function
of pH. Source: Matsuura, T., and S. Sourirajan; "Reverse
Osmosis Separation of Phenols in Aqueous Solutions Using
Porous Cellulose Acetate Membranes." Journal of Applied
Polymer Science 16, 2531 (1972). Reprinted by permission.
-126-
-------
60
50
40
30
20
10
Dicarboxylic acid -
60
50
40
Dioles
23456
Number of Carbons
8
Figure B2. Cellulose acetate membrane rejection of 3 classes of linear
alkyl compounds, as a function of the total number of carbons.
Source: Cabasso, L., et al., "Evaluation of Seraipermeable
Membranes for Concentration of Organic Contaminants in Drinking
Water." U.S. Environmental Protection Agency, Report No. EPA-
670/1-75-001 (1975).
-127-
-------
Table Bl
LIST OF PRIORITY TOXIC POLLUTANTS
Compound Name Molecular Weight
I. Pesticides
1. Acrolein 56
2. Aldrin 365
3. «-BHC 291
4. P-BHC 291
5. Y-BHC 291
6. 6-BHC 291
7 . Chlordane 406
8. ODD 320
9. DDE 318
10. DDT 355
11. Dieldrin 381
12. oc-Endosulfan
13. 3-Endosulfan
14. Endosulfan sulfate
15. Endrin
16. Endrin aldehyde
17. Heptachlor 374
18. Heptachlor epoxide 389
19. Isophorone 138
20. TCDD (2,3,7,8-tetrachlorodibenzo-p-dioxin) 332
21. Toxaphene: CjoHioCle 343
517
II. Metals and Inorganics
1 . Antimony
2. Arsenic
3. Asbestos
4. Beryllium
5 . Cadmium
6. Chromium
7 . Copper
8. Cyanides
9. Lead
10. Mercury
11. Nickle
12. Selenium
13. Silver
14. Thallium
15. Zinc
-128-
-------
Table Bl. (Continued)
Compound Name
III. PCB's and Related Compounds
1. PCB-1016 (Arochlor 1016)
2. PCB-1221 (Arochlor 1221)
3. PCB-1232 (Arochlor 1232)
4. PCB-1242 (Arochlor 1242)
5. PCB-1248 (Arochlor 1248)
6. PCB-1254 (Arochlor 1254)
7. PCB-1260 (Arochlor 1260)
8. 2-Chloronaphthalene
Molecular Weight
258
201
232
267
300
328
376
163
IV. Halogenated Aliphatics
1. Methane, bromo- (methyl bromide)
2. Methane, chloro- (methyl chloride)
3. Methane, dichloro- (methylene chloride)
4. Methane, chlorodibromo-
5. Methane, dichlorobromo-
6. Methane, tribromo- (bromoform)
7. Methane, trichloro- (chloroform)
8. Methane, tetrachloro- (carbon tetrachloride)
9. Methane, trichlorofluoro-
10. Methane, dichlorodiflouro-
11. Ethane, chloro-
12. Ethane, 1, 1-dichloro-
13. Ethane, 1, 2-dichloro-
14. Ethane, 1, 1, 1-trichloro-
15. Ethane, 1, 1, 2-trichloro-
16. Ethane, 1, 1, 2, 2-tetrachloro-
17. Ethane, hexachloro-
18. Ethene, chloro- (vinyl chloride)
19. Ethene, 1, 1-dichloro-
20. Ethene, 1, 2-trans-dichloro-
21. Ethene, trichloro-
22. Ethene, tetrachloro-
23. Propene, 1, 2-dichloro-
24. Propene, 1, 3-dichloro-
25. Butadiene, hexachloro-
26. Cyclopentadiene, hexachloro-
95
51
85
208
164
253
119
154
137
121
65
99
99
133
133
168
237
63
97
97
131
166
113
111
261
273
V. Ethers
1. Ether, bis(chloromethyl)-
2. Ether, bis(2-chloroethyl)-
3. Ether, bis(2-chloroisopropyl)-
4. Ether, 2-chloroethyl vinyl-
5. Ether, 4-bromophenyl phenyl-
115
143
171
107
249
-129-
-------
Table Bl. (Continued)
Compound Name
6. Ether, 4-chlorophenyl phenyl-
7. Bis (2-chloroethoxy) methane
Molecular Weight
204
173
VI. Monocyclic Aromatics (excluding phenols, cresols, phthalates)
78
113
147
147
147
182
285
106
123
92
182
182
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
VII.
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
VIII
1.
2.
3.
4.
5.
6.
Benzene
Benzene , chloro-
Benzene, 1, 2-dichloro-
Benzene, 1, 3-dichloro-
Benzene, 1, 4-dichloro-
Benzene, 1, 2, 4-trichloro
Benzene, hexachloro-
Benzene, ethyl-
Benzene, nitro
Toluene
Toluene, 2, 4-dinitro-
Toluene, 2, 6-dinitro-
Phenols and Cresols
Phenol
Phenol, 2-chloro-
Phenol, 2, 4-dichloro-
Phenol, 2, 4, 6-trichloro-
Phenol, pentachloro-
Phenol, 2-nitro-
Phenol, 4-nitro-
Phenol, 2, 4-dinitro-
Phenol, 2, 4-dimethyl-
m-Cresol, p-chloro-
o-Cresol, 4, 6-dinitro-
. Phthalate Esters
Phthalate, dimethyl-
Phthalate, diethyl-
Phthalate, di-n-butyl
Phthalate, di-n-octyl-
Phthalate, bis (2-ethylhex\
Phthalate, butyl benzyl-
94
129
163
198
266
139
139
184
122
143
198
194
222
278
391
391
312
IX. Polycyclic Aromatic Hydrocarbons
1. Acenaphthene
2. Acenaphthylene
154
152
-130-
-------
Table Bl. (Continued)
Compound Name Molecular Weight
3. Anthracene 178
4. Benzo (a) anthracene 228
5. Benzo (b) fluoranthene 252
6. Benzo (k) fluoranthene 252
7. Benzo (ghi) perylene 276
8. Benzo (a) pyrene 252
9. Chrysene 228
10. Dibenzo (a, n) anthracene 278
11. Fluoranthene 202
12. Fluorene 166
13. Indeno (1, 2, 3-cd-)pyrene 276
14. Naphthalene 128
15. Phenanthrene 178
16. Pyrene 202
X. Nitrosamines and Other Nitrogen-Containing Compounds
1. Nitrosamine, dimethyl- (DMN) 74
2. Nitrosamine, diphenyl- 198
3. Nitrosamine, di-n-propyl- 130
4. Benzidine 184
5. Benzidine, 3, 3' -dichloro- 253
6. Hydrazine, 1, 2-diphenyl- 184
7. Acrylonitrile 53
-131-
-------
N>
I
Solute
Alcohols
Ethyl Alcohol
n-Propanol
i-Propanol
Methane1
n-Butyl Alcohol
i-Butyl Alcohol
s-Butyl Alcohol
t-Butyl Alcohol
3-Pentanol
n-Amyl Alcohol
n-Hexyl Alcohol
Cyclohexanol
n-Heptyl Alcohol
n-Octyl Alcohol
Benzyl Alcohol
Phenethyl Alcohol
Table B2
REJECTION OF ORGANICS BY CELLULOSE ACETATE MEMBRANES
Molecular
Weight
46.1
60.1
60.1
32.0
74.
74.
74.
74.
88,
88.
102.
100.
116.2
130.2
108.1
122.2
Concentration
_ ppm _
100
1000
100
II
1000
11
II
100
1000
100
it
it
n
n
it
it
it
Lp
xlO5
Pressure
atm
4.6
0.6
1.0
0.6
1.0
4.6
it
0.6
1.0
0.6
1.0
4.6
0.6
1.0
0.6
1.0
4.6
17.0
40.8
"
102
it
17.0
n
40.8
"
102
it
17.0
40.8
it
102
102
17.0
Rejection
%
10
24
13
37
18
24
37
60
44
77
52
5
-20
-8
-12
-8
16
41
34
79
40
23
11
60
9
23
5
8
5.42
5.91
it
5.77
Reference
B5
Bll
Bll
Bll
Bll
B5
B5
Bll
Bll
Bll
Bll
B5
Bll
Bll
Bll
Bll
B5
B5
B5
B5
B5
B5
B5
B5
B5
B5
B5
B5
Lp = Hydraulic permeability constant (flux water/differential hydraulic pressure, ml/cm2 sec atm)
-------
Table B2. (Continued)
I
t->
OJ
Solute
Phenols
Phenol
Substituted Phenols
Chlorophenol
m-Chlorophenol
p-Chlorophenol
p-Chlorophenol
Pyrocatechol
Resorcinol
Hydroquinone
p-Aminophenol
p-Methoxyphenol
p-Cresol
m-Cresol
2,4 Dichlorophenol
m-Nitrophenol
Molecular
Weight
94.1
128.6
128.6
128.6
128.6
110.1
110.1
110.1
109.1
124.1
108.1
108.1
163.0
139.1
Concentration
Ppm
1000
tt
100
128
it
»
ti
100
n
1000
100
II
100
II
35
100
Lp
xlO5
1.0
.6
4.6
It
II
It
4.6
1.0
0.6
1.0
0.6
4.6
4.6
4.6
Pressure
atm
102.0
40.8
17.0
Rejection
18
-6
1
"
11
II
17.0
40.8
68.0
102.0
17.0
"
n
it
40.8
40.8
102.0
"
17.0
"
17.0
it
it
120.0
17.0
22
20
21
42
20
18
1
2
-1
0
1
-0.63
7.03
-8.20
18.19
4.0
27.0
0
2
2
-34
2
6.25
5.20
II
It
II
Reference
Bll
Bll
B5, B7
B7
B7
B7
BIO
BIO
BIO
BIO
B5
B5
B7
B5
Bll
Bll
Bll
Bll
B7
B7
B7
B7
B7
B12
B7
Lp = Hydraulic permeability constant (flux of water/differential hydraulic pressure, ml/cm2 sec atm)
-------
Table B2. (Continued)
Solute
Molecular
Weight
Substituted Phenols (cont.)
p-Nitrophenol 139.1
Polyhydric Alcohols
D-Sorbitol
Dulcitol
Arabitol
Xylitol
Adonitol
1-Erythritol
Glycerol
Triethylene Glycol
Propylene Glycol
cis-and-trans-1,2-
Cyclohexanediol
trans-1,2-
Cyclohexanediol
1,6 Hexanediol
1,5 Pentanediol
1,3 Propanediol
Dextrose
182.2
182.2
152.2
152.2
152.2
122.1
92.1
150.0
76.1
116.2
116.2
118.2
104.2
76.1
Concentration
PP">
100
1000
II
100
180.2
100
n
940
Lp
xlO5
4.6
it
it
0.6
1.0
4.6
4.6
Pressure
atm
Rejection
102.0
it
100
17.0
102.0
102.0
102.0
99
99
97
97
97
94
98
85
37
68
83
86
45
54
48
99.7
5.95
Reference
20.4
"
27
11
41
"
55
"
68
Tt
17.0
-10
70
-13
78
-21
80
-28
81
-37
82
0
3.0
8.5
3.0
8.5
3.0
8.5
3.0
8.5
3.0
8.5
B13
B13
B13
B13
B13
B13
B13
B13
B13
B13
B7
B5
B5
B5
B5
B5
B5
Bll
Bll
B17
B5
B5
B5
B5
B5
B5
B12
Lp = Hydraulic permeability constant (flux of water/differential hydraulic pressure, ml/cm2 sec atm)
-------
Table B2. (Continued)
10
(Jl
I
Solute
Aromatics
Benzene
Toluene
Cumene
Styrene
Ethylbenzene
o-Xylene
ra-Xylene
p-Xylene
Propylbenzene
t-Butylbenzene
Cyclic Hydrocarbons
Cyclopentene
Cyclohexane
Cycloheptatriene
Cyclohexene
Cyclopentane
Methyl
Cyclopentane
Molecular
Weight
78.11
92.13
120.19
104.14
106.16
106.16
120.19
134.12
68.11
84.16
92.1
82.1
70.1
84.2
Concentration
PP""
100
it
100
100
It
It
It
II
It
tt
100
If
100
It
LP5
xlO5
4.6
4.6
4.6
4.6
4.6
tt
Pressure
atm
17.0
17.0
it
Rejection
17.0
3.4
6.8
33.8
17.0
17.0
6.8
33.8
67.6
102.0
17.0
it
"
tt
"
"
"
76
97
89
54
73
80
98
71
53
35
70
78
86
84
85
98
99
60
90
66
65
70
93
pH Reference
B6
B6
B6
B6
B6
B6
B6
B6
B6
B6
66
B6
B6
B6
B6
B6
B6
B6
B6
B6
B6
B6
B6
Lp = Hydraulic permeability constant (flux of water/differential hydraulic pressure, ml/cm2 sec atm)
-------
Table B2. (Continued)
Solute
LO
Hydrocarbons
1,6 Heptadiyne
1,6-Heptadiene
Isopentane
Isoprene
1-Hexyne
1,5 Hexadiene
1-Heptyne
4-Methyl-l-
pentene
2-Methyl-l-
pentene
1-Hexene
n-Hexane
Moaocarboxilic Acids
Pivalic acid
i-Butyric acid
Valeric acid
n-Butyric acid
Propionic acid
Acetic acid
Molecular
Weight
92.13
96.2
72.2
68.11
82.14
82.14
96.17
84.16
84.2
84.16
86.17
102.
88.
102.
88.
74.
60.1
Concentration
PP""
100
100
100
100
100
II
100
4-Phenylbutyric
acid
164.2
1000
It
100
Pressure
atm
17.0
17.0
17.0
n
17.0
17.0
xlO-
4.6
4.6
4.6
4.6
4.6
n
tt
4.6
tt
tt
1.0 40.8
0.6
1.0 102.0
0.6
4.6
17.0
Rejection
I
42
94
87
52
54
59
77
90
92
90
99
66
35
19
25
24
24
10
18
18
30
22
3.97
Reference
B6
B6
B6
B6
B6
B6
B6
B6
B6
B6
B6
B7
B7
B7
B7
B7
B7
Bll
Bll
Bll
Bll
B7
Lp = Hydraulic permeability constant (flux of water/differential hydraulic pressure, ml/cm2 sec atm)
-------
Table B2. (Continued)
Solute
Monocarboxilic Acids
3-Phenylpropionic
acid
Phenylacetic acid
Benzoic acid
Caprylic acid
L-Leucine
p-Aminobenzoic
acid
i
00
71 m-Aminobenzoic
acid
Anicic acid
m-Toluic
m-Hydroxybenzoic
acid
m-Nitrobenzoic
acid
o-Nitrobenzoic
acid
p-Nitrobenzoic
o-Chlorobenzoic
Molecular
Weight
(cont. )
150.2
136.1
122.1
144.2
131.2
137.1
137.1
152.1
136.1
138.1
167.1
167.1
167.1
156.6
Concentration Lp Pressure
ppm
100
it
it
it
it
1312
it
it
137
it
11
it
it
137
it
it
it
"
100
it
it
it
tt
100
II
xlO5 atm
4.6 102.0
it ii
34.0
17.0
tt it
40
"
_ ti
4.6 17.0
40
11
"
"
40
"
"
"
"
4.6 17.0
it it
tt tt
ti ti
ti ti
4.6 17.0
it it
Rejection
18
26
29
19
19
96
98
99
13
2
12
59
75
57
70
88
90
96
20
23
19
37
76
35
45
pH
3.6
4.5
5.8
4.2
3.2
2.3
1.7
3.9
3.1
2.4
2.0
1.8
Reference
E7
E7
B13
E7
E7
B18
B18
B18
E7
B18
B18
B18
B18
B18
B18
B18
B18
B18
E7
$7
E7
E7
E7
E7
E7
Lp = Hydraulic permeability constant (flux of water/differential hydraulic pressure, ml/cm sec atm)
-------
Table B2. (Continued)
Solute
oo
oo
Dicarboxylic Acids
Oxalic acid
Malonic
Succinic Acid
Adipic
Pimelic acid
Suberic
Azelaic
Aldehydes
Formaldehyde
Acetaldehyde
Propionaldehyde
n-Butylaldehyde
i-Butylaldehyde
Crotonaldehyde
Furfuryl
Benzaldehyde
Molecular
Weight
90.0
Concentration
104.
118.
146.
160.
174.2
188.2
1
,1
1
.2
30.0
44.
58.
72.
72.
70.
96.
100
100
1000
II
II
II
II
It
100
It
100
Lp
xlO5
4.6
n
n
4.6
4.6
Pressure
atm
17.0
17.0
0.6 40.8
1.0
0.6 102.0
1.0
4.6 17.0
n
17.0
106.1
Rejection
94
55
50
41
36
25
15
33
20
48
30
60
75
60
78
15
2
10
4.64
Reference
B8
B8
B8
B8
B8
B8
B8
Bll
Bll
Bll
Bll
B9
B9
B9
B9
B9
B9
B9
Lp = Hydraulic permeability constant (flux of water/differential hydraulic pressure, ml/cm2 sec atm)
-------
Table B2. (Continued)
OJ
vo
Solute
Ke tones
Acetone
Methyl Ethyl
Ketone
Cyclopentanone
Cyclohexanone
Diisopropyl
Ketone
Diisobutyl
Ketone
Benzyl Methyl
Ketone
Acetophenone
Molecular
Weight
58.1
72.1
84.1
98.1
114.2
142.2
134.2
120.1
Concentration
ppm
100
ii
1000
fl
ft
If
100
-
100
II
II
It
II
II
Lp
xlO5
4.6
-
-
1.0
0.6
1.0
0.6
4.6
-
4.6
ii
ii
ii
ti
it
Pressure
atm
17.0
102.0
100
40.8
40.8
102.0
102.0
17.0
163
17.0
ii
n
it
ii
ii
Rejection
%
22
17
47
5
23
6
30
24
18
26
39
67
59
17
17
pH Reference
B9
B13
B14
5.48 Bll
Bll
Bll
Bll
B9
B15
B9
B9
B9
B9
B9
B9
Lp = Hydraulic permeability constant (flux of water/differential hydraulic pressure, ml/cm sec atm)
-------
Table B2. (Continued)
Solute
Ethers - Noncyclic
Diethyl ether
Diisopropyl ether
Anisole
Ethyl vinyl
ether
Phenetole
Methyl benzyl
ether
Polyoxyethylene
Nonylphenyl
ether
Cyclic Ethers
Tetrahydropyran
Tetrahydrofuran
1,4 Dioxane
Propylene Oxide
Styrene oxide
Epichlorohydrin
Molecular
Weight
74.1
Concentration
102.2
108.1
72.1
122.2
122.2
661.41
925.5
1409.6
86.
72.
88.
58.
120.
100
1000
100
52
100
238
100
100
4.6
1.0
0.6
1.0
0.6
4.6
3.07
2.75
2.88
4.6
4.6
Pressure
atm
17.0
40.8
102.0
17.0
40
92.5
17.0
17.0
it
Rejection
57
12.5
30.2
10
24
84
51
55
46
26
83.9
59.5
92
47
53
49
33
26
18
5.59
Reference
B9
Bll
Bll
Bll
Bll
B9
B9
B9
B9
B9
B19
B19
B19
B9
B9
B9
B9
B9
B9
Lp = Hydraulic permeability constant (flux of water/differential hydraulic pressure, ml/cm2 sec atm)
-------
Table B2. (Continued)
Solute
Esters
n-Amyl Acetate
Methyl n-Butyrate
Ethyl acetate
Methyl acetate
Methyl benzoate
Ethyl chloro-
acetate
Methyl chloro-
acetate
Vinyl acetate
Methyl acrylate
Methyl meth-
acrylate
Molecular
Weight
130.2
102.1
88.1
74.1
136.1
122.6
108.5
86.1
86.1
100.1
Concentration
Ppm
100
II
II
1000
ii
ii
100
II
Lp
xlO5
4.6
it
Pressure
a tin
17.0
Rejection
1.0 40.8
0.6
1.0 102.0
0.6
4.6 17.0
13
3
35
40
_pH_
Reference
50
44
45
40
-3.53 5.34
20.7 "
11.14 5.38
12.36
25
B9
B9
B9
B9
Bll
Bll
Bll
Bll
B9
B9
B9
B9
B9
B9
Lp = Hydraulic permeability constant (flux of water/differential hydraulic pressure, ml/cm2 sec atm)
-------
Table B2. (Continued)
Solute
Primary Amines
Aniline
p-Anisidine
o-Anisidine
p-Toluidine
m-Toluidine
,1 o-Toluidine
i^ m-Phenylenediamine
1 o-Phenylenediamine
p-Chloroaniline
o-Chloroaniline
m-Chloroaniline
m-Nitroaniline
Secondary Amines
Di-n-Butylamine
Piperidine
Dimethylamine
Diisopropylamine
N-Methylaniline
Molecular
Weight
93.1
.2
.2
.2
.2
123.2
123.2
107.2
107.
107.
108.
108.
127.6
127.6
127.6
138.1
129.5
85.2
45.1
101
107.2
Concentration
PPM
100
1000
100
100
Lp
xlO5
4.6
1.0
0.6
1.0
0.6
4.6
4.6
Pressure
atm
17.0
40.8
102.0
34
rt
68
it
17.0
it
it
it
Rejection
I
17.0
it
-9
22
-5
17
88
9
78
4
4
8
7
7
7
15
14
14
19
21
8
63
88
16
72
12
6.62
6.4
7.3
6.4
7.5
Reference
B9
fill
Bll
Bll
Bll
B13
B13
B13
B13
B9
B9
B9
B9
B9
B9
B9
B9
B9
B9
B9
B9
B9
B9
B17
B9
Lp = Hydraulic permeability constant (flux of water/differential hydraulic pressure, ml/cm sec atm)
-------
Table B2. (Continued)
Solute
Tertiary Amines
Triethyl
Trimethyl
N,N-Dimethyl-
benzylamine
N,N-Dimethyl-
aniline
Miscellaneous
Nitrome thane
Boric Acid
i
H-»
u>
Sodium Oleate
Sodium dodecylbenzene-
sulfonate
Tetradecylbenzyl-
ammonium chloride
Urea
Molecular
Weight
101.2
59.1
135.2
121.2
61.0
43.8
304.5
349.5
704.6
60
Concentration
ppm
100
11
11
ti
—
-
-
-
-
-
-
210
188
-
1000
ft
ft
Lp
xlO5
4.6
"
it
"
_
-
-
-
-
-
-
1.07
4.03
-
1.0
0.6
1.0
0.6
Pressure
atm
17.0
"
"
"
34
68
34
68
34
68
100
40
"
100
40
it
102.0
ii
Rejection
% pH
95
77
56
33
-6 7.2
-9
43 7.0
67
97 11.0
97
99.9
99.8
93.7
45
17.56 7.22
38.02
26.76
24.37
Reference
B9
B9
B9
B9
B13
B13
B13
B13
B13
B13
B16
B19
B19
B17
Bll
Bll
Bll
Bll
2,4 Dichlorophenoxy-
acetic acid 221 35 - 102.0 92.8 B12
Lp = Hydraulic permeability constant (flux of water/differential hydraulic pressure, ml/cm2 sec atm)
-------
Table B2. (Continued)
Molecular
Solute Weight
Pesticides
Aldrin 365
Atrazine 216
Captan 301
DDE 318
DDT 354.5
Diazinon 304
Dieldrin 381
Heptachlor 373
Heptachlor epoxide 389
Lindane 291
Malathion 330
Methylparathion 263
Parathion 291
Randox 174
Trifluralin 335.3
Concentration
0.95
4.98
2.18
10.53
Lp
xlO5
0.9
0.9
0.9
0.9
Pressure
atm
41.8
41.8
41.8
41.8
Rejection
100
7.34
4.59
0.46
0.28
3.16
2.14
0.97
2.05
3.38
7.05
6.09
0.9
0.9
0.9
0.9
0.9
0.9
0.9
0.9
0.9
0.9
0.9
41.8
41.8
41.8
41.8
41.8
41.8
41.8
41.8
41.8
41.8
41.8
84.0
98.8
100
100
98.3
99.9
100
99.8
99.5
99.2
99.6
99.9
72.0
99.7
Reference
B12
B12
B12
B12
B12
B12
B12
B12
B12
B12
B12
B12
B12
B12
B12
Lp = Hydraulic permeability constant (flux of water/differential hydraulic pressure, ml/cm2 sec atm)
-------
Table B3
REJECTION OF ORGANICS BY CELLULOSE ACETATE BUTYRATE MEMBRANES
Solute
Concentration
ppm
1000
"
"
"
it
it
ti
it
"
n
ii
n
n
it
1000
It
It
It
II
II
It
It
II
II
tl
II
5000
it
Lp
xlO5
0.2
ii
"
n
"
"
tt
n
1!
tt
II
II
II
11
0.2
ii
it
n
ii
"
"
it
ii
"
"
"
"
it
Pressure
atm
40.8
102.4
40.8
102.4
40.8
102.4
40.8
102.4
40.8
102.4
40.8
102.4
40.8
102.4
40.8
102.4
40.8
102.4
40.8
102.4
40.8
102.4
40.8
102.4
40.8
102.4
40.8
102.4
Rejection
I
-9.09
1.27
1.9
14.43
41.04
57.58
12.33
26.18
41.05
50.64
16.60
6.16
39.51
2.01
90.21
96.08
30.97
38.15
10.56
-12.07
8.00
31.36
25.51
-5.19
-25.14
-10.78
99.6
99.8
PH
5.77
"
5.42
it
5.91
it
3.77
"
4.64
it
5.48
"
5.59
"
5.97
11
5.20
"
6.25
"
7.72
"
6.62
"
5.38
"
6.64
"
Reference
Bll
Bll
Bll
Bll
Bll
Bll
Bll
Bll
Bll
Bll
Bll
Bll
Bll
Bll
Bll
Bll
Bll
Bll
Bll
Bll
Bll
Bll
Bll
Bll
Bll
Bll
Bll
Bll
Methanol
Ethanol
i-Propanol
Acetic Acid
Formaldehyde
Acetone
Ethyl Ether
Glycerol
Hydroquinone
Phenol
Urea
Aniline
Methyl Acetate
Sodium Chloride
Lp = hydraulic permeability constant (flux of water/differential hydraulic
pressure, ml/cm2 sec atm)
-------
Table B4
REJECTION OF ORGANICS BY NS-100 MEMBRANES
Solute
Methanol
Ethanol
n-Propanol
Isopropanol
n-Butanol
n-Pentanol
3-Pentanol
Phenol
Membrane
Designation
Standard
Modified
52C
53J
52C
53J
Standard
Modified
52C
53J
52C
53J
Standard
Modified
Standard
Modified
52C
53J
52C
53J
Standard
Modified
Standard
Modified
Standard
Modified
52C
53J
Standard
Modified
52C
53 J
Concentration
ppm
1000
it
1000
II
II
II
II
It
II
II
II
II
1000
Lp
xlO5
1.0
0.5
1.1
1.3
1.1
1.3
1.0
0.5
1.1
1.3
1.1
1.3
1.0
0.5
1.0
0.5
1.1
1.3
1.1
1.3
1.0
0.5
1.0
0.5
1.0
0.5
1.1
1.3
1.0
0.5
1.1
1.3
Pressure
atm
54.4
tl
40.8
"
102
102
54.4
"
40.8
"
102
"
54.4
It
54.4
11
40.8
11
102.0
it
54.4
11
54.4
11
54.4
"
40.8
II
54.4
11
102.0
102.0
Rejection
%
33.2
41.0
-5.98
18.06
1.41
9.87
79.8
86.7
64.78
65.66
70.01
73.57
92.7
93.6
91.4
92.3
82.01
85.10
56.06
40.16
94.2
96.3
95.2
97.7
97.9
99.4
68.57
55.68
84.1
87.1
70.21
65.72
pH
.
-
5.77
it
n
n
-
-
5.42
n
n
"
-
-
-
-
5.91
n
n
n
-
-
-
-
-
-
6.25
11
-
-
6.25
6.25
Reference
B20
B20
Bll
Bll
Bll
Bll
B20
B20
Bll
Bll
Bll
Bll
B20
B20
B20
B20
Bll
Bll
Bll
Bll
B20
B20
B20
B20
B20
B20
Bll
Bll
B20
B20
Bll
Bll
Lp = hydraulic permeability constant (flux of water/differential hydraulic pressure, ml/cm2 sec atm)
-------
Table B4. (Continued)
Solute
p-Chlorophenol
Formaldehyde
Acetaldehyde
Benzaldehyde
Acetone
Acetophenone
Cyclohexanone
Methyl ethyl
ketone
Membrane
Designation
Standard
Modified
52C
53J
Standard
Modified
52C
53J
Standard
Modified
Standard
Modified
52C
53 J
Standard
Modified
52C
53 J
Standard
Modified
Standard
Modified
Standard
Modified
Concentration
1000
11
1000
II
II
II
II
tt
500
n
1000
II
Lp
xlQs
1.0
0.5
1.1
1.3
1.0
0.5
1.1
1.3
1.0
0.5
1.0
0.5
1.1
1.3
1.0
0.5
1.1
1.3
1.0
0.5
1.0
0.5
1.0
0.5
Pressure
atm
54.4
17.0
40.8
68.0
102.0
54.4
40.8
tt
54.4
n
102.0
it
54.0
II
40.8
54.0
tl
102.0
54.0
Rejection
81.0
66
79
80
80
83.2
37.05
38.62
54.7
70.0
49.96
50.55
75.0
80.6
87.6
95.0
80.44
76.03
93.4
96.1
78.26
77.32
96.
98.
97.
.7
.2
.3
98.7
4.64
4.64
5.48
5.48
90.8
95.5
Reference
B20
B20
B20
B20
B20
B20
fill
Bll
B20
B20
Bll
Bll
B20
B20
B20
B20
Bll
Bll
B20
B20
Bll
Bll
B20
B20
B20
B20
B20
B20
Lp = hydraulic permeability constant (flux of water/differential hydraulic pressure, ml/cm2 sec atm)
-------
Table B4. (Continued)
Solute
Acetic acid
Lactic acid
Benzoic acid
Ethylamine
Triethylamine
Aniline
oo
i
Methyl acetate
Ethyl acetate
Ethyl ether
Membrane
Designation
52C
53 J
Standard
Modified
52C
53J
Standard
Modified
Standard
Modified
Standard
Modified
Standard
Modified
52C
53J
Standard
Modified
52C
53 J
52C
53 J
Standard
Modified
52C
53J
Standard
Modified
52C
53J
52C
53J
Concentration
1000
II
II
1000
ft
tl
tf
500
tf
1000
tl
II
II
1000
II
II
II
It
II
Lp
xlO5
1.1
1.3
1.0
0.5
1.1
1.3
1.0
0.5
1.0
0.5
1.0
0.5
1.0
0.5
1.1
1.3
1.0
0.5
1.1
1.3
1.1
1.3
1.0
0.5
1.1
1.3
1.0
0.5
1.1
1.3
1.1
1.3
Pressure
atm
40.8
M
54.0
If
102.0
11
54.0
n
it
it
M
ff
It
tl
40.8
11
54.0
It
102.0
ft
40.8
tl
54.0
"
102.0
It
54.0
11
40.8
"
102.0
n
Rejection
%
68.70
48.02
71.0
80.3
9.00
69.86
87.6
89.3
66.3
82.0
86.2
92.8
99.7
98.5
31.36
51.80
91.8
95.9
14.99
31.31
36.37
28.32
89.2
91.7
23.02
44.62
95.8
96.9
90.93
82.18
65.62
77.09
3.77
3.77
6.62
6.62
ff .
5.38
ft
5.38
ft
5.59
Reference
Bll
Bll
B20
B20
Bll
Bll
B20
B20
B20
B20
B20
B20
B20
B20
Bll
Bll
B20
B20
Bll
Bll
Bll
Bll
B20
B20
Bll
Bll
B20
B20
Bll
Bll
Bll
Bll
Lp = hydraulic permeability constant (flux of water/differential hydraulic pressure, ml/cm2 sec atm)
-------
Table B4. (Continued)
Solute
Glycerol
Hydroquinone
Urea
NaCl
Membrane
Designation
52C
53J
52C
53 J
52C
53 J
52C
53 J
52C
53 J
52C
53 J
53 J
52C
53 J
Concentration
Ppm
1000
II
1000
II
It
tt
1000
5000
Lp
xlO5
1.1
1.3
1.1
1.3
1.1
1.3
1.1
1.3
1.1
1.3
1.1
1.3
1.3
1.1
1.3
Pressure
atm
40.8
ii
102.0
ii
40.8
40.8
102.0
n
40.8
tr
102.0
n
40.8
102.0
Rejection
96.87
89.15
96.18
91.69
82.23
71.93
86.29
77.38
64.15
55.37
74.32
63.38
99.71
99.26
99.24
5.20
5.20
7.72
Reference
Bll
Bll
Bll
Bll
Bll
Bll
Bll
Bll
Bll
Bll
Bll
Bll
Bll
Bll
Bll
Pesticides
Aldrin
Atrazine
Captan
DDE
DDT
Diazinon
Dieldrin
.Heptachlor
Heptachlor
epoxide
Lindane
Malathion
Methylparathion
Parathion
Randox
Trifluralin
0.95
7.34
4.59
0.46
0.28
,16
,14
0.97
2.05
3.38
7.05
6.09
4.98
2.18
10.53
1.36
1.36
1.36
1.36
1.36
1.36
1.36
1.36
.36
.36
1.36
.36
.36
.36
1.36
41.8
41.8
41.8
41.8
41.8
41.8
41.8
41.8
41.8
41.8
41.8
41.8
41.8
41.8
41.8
100
97.8
100
100
100
88.1
100
100
99.8
99.0
99.7
99.6
99.8
98.6
100
B22
B22
B22
B22
B22
B22
B22
B22
B22
B22
B22
B22
B22
B22
B22
Lp = hydraulic permeability constant (flux of water/differential hydraulic pressure, ml/cm2 sec atm)
-------
Table B5
REJECTION OF ORGANICS BY POLYAMIDE MEMBRANES
Solute
o
i
Methanol
Ethanol
i-Propanol
Acetic acid
Formaldehyde
Acetone
Ethyl ether
Glycerol
Membrane
Type
B-9
B-9
B-9
B-9
B-9
B-9
B-9
B-9
B-9
B-9
B-9
B-9
B-9
Concentration
PP
-------
Table B5 (continued)
Solute
Hydroquinone
Phenol
Urea
Aniline
Methyl Acetate
Sodium
chloride
Membrane Concentration Lp
Type ppm xlO5
B-9 551
1000 .43
_ II If
B-9 773
B-9 500-2000
1000 .43
_ ii ii
B-9 1188
1000 .43
_ ii M
B-9 440
1000 .43
_ ii ii
B-9 370
1000 .43
B-9 3850
B-9 1500
5000 .43
_ ii it
Pressure
atm
27.2
40.8
102.0
27.2
40.8
102.0
27.2
40.8
102.0
27.2
40.8
102.0
27.2
40.8
102.0
27.2
40.8
102.0
Rejection
60.49
84.49
98.61
44.67
55
80.06
88.72
34.45
55.41
89.32
47.28
78.08
82.19
57.45
54.31
44.50
93.06
90
99.13
99.58
5.20
6.25
7.72
6.62
5.38
6.64
Reference
Bll
Bll
Bll
Bll
B21
Bll
Bll
Bll
Bll
Bll
Bll
Bll
Bll
Bll
Bll
Bll
Bll
B21
Bll
Bll
Lp = hydraulic permeability constant (flux of water/differential hydraulic pressure, ml/cm2 sec atm)
-------
I
h-»
Ul
I
Solute
Arsenic
Boron
Barium
Cadmium
Chromium
Membrane Type
Table B6
REJECTION OF HEAVY METALS BY REVERSE OSMOSIS MEMBRANES
Concentration
EPJ5
Cellulose acetate
Cellulose acetate
Polyamide
Cellulose acetate
Cellulose acetate
Cellulose acetate
Polyamide
Cellulose acetate
Cellulose acetate
Cellulose acetate
Polyamide
Cellulose acetate
Cellulose acetate
Cellulose acetate
Cellulose acetate
Polymide
Cellulose
Cellulose
Cellulose
Cellulose
Cellulose
Cellulose
Cellulose
Cellulose
Cellulose
Polyamide
NS-100
NS-100
acetate
acetate
acetate
acetate
acetate
acetate
acetate
acetate
acetate
0.15
0.15
0.15
0.35
10
10
10
9.15
15
15
15
1
2.5
2.5
2.5
0.94-1.01
8.65-9.35
1
1
1
12.5
12.5
Pressure
atm
12
26.5
26.5
_
12
26.5
26.5
28.2
12
26.5
26.5
-
28.2
12
26.5
26.5
_
-
-
-
-
28.6
28.2
12
26.5
26.5
41.8
41.8
Rejection
I
90.7
90.7-93.1
88.8
38-60
28.6
36.4-68.8
65.5
97.8
92.2
82.8-94.6
98.1
78-99+
98.7-99
95.4
85.2-95.9
91.3
86-98
92.6
52-99
91-99+
98.6
95-96.9
85.1-93.2
97.3
92.6-96.2
86.4
97.6
91.3
PH
4.8
4.8
4.8
5
4.8
4.8
4.8
_
4.8
4.8
4.8
11.5
-
4.8
4.8
4.8
0.9-1.9
2.6
4.4-4.7
5.5-6.1
7.6
-
-
4.8
4.8
4.8
8
11
Reference
*
JL
*
B26
*
*
*
B31
*
*
*
B27
B31
*
*
JL
B27
B26
B27
B27
B26
B31
B31
*
*
*
B31
B31
*This study
-------
Table B6. (Continued)
Solute
Copper
Iron
Lead
Ui
oo
i
Manganese
Mercury
(organic)
Nickel
Membrane Type
Cellulose acetate
Cellulose acetate
Cellulose acetate
Cellulose acetate
Polyamide
NS-100
NS-100
Cellulose acetate
NS-100
NS-100
Cellulose acetate
Cellulose acetate
Cellulose acetate
Cellulose acetate
Polyamide
NS-100
NS-100
Cellulose acetate
Cellulose acetate
Cellulose acetate
Polyamide
Cellulose acetate
Cellulose acetate
Cellulose acetate
Cellulose acetate
Cellulose acetate
Cellulose acetate
Polyamide
NS-100
NS-100
Concentration
ppm
0.65-0.7
6.25-6.5
10
10
10
12.5
12.5
6.5
12.5
12.5
0.95-1.1
4.8-9.3
0.15
0.15
0.15
12.5
12.5
3.8
2.5
2.5
2.5
100
140
1000
1000
12.5
12.5
12.5
12.5
12.5
Pressure
atm
28.2
28.2
12
26.5
26.5
41.8
41.8
_
41.8
41.8
28.2
28.2
12
26.5
26.5
41.8
41.8
_
12
26.5
26.5
18
18
14.6
21.4
26.5
26.5
26.5
41.8
41.8
Rejection
%
94.8-97+
99.2-99.6+
94.7
92.6-96.9
82.2
99.9
99.9
98.2
100
100
97.8-99.5+
97.8-99.9
93.8+
91.4+
92.3+
100
100
100
95.4
87.5-95.3
90.5
63.2
66.1
99
99.6
94.9
83.7-95.7
89
96.3
98.8
PH
— ,
-
4.8
4.8
4.8
8
11
6.7
8
11
_
-
4.8
4.8
4.8
8
11
6.7
4.8
4.8
4.8
_
-
4-6
4-6
4.8
4.8
4.8
8
11
Reference
B31
B31
*
*
*
B31
B31
B30
B31
B31
B31
B31
*
*
*
B31
B31
B30
*
*
*
B28
B28
B29
B29
*
*
*
B31
B31
*This study
-------
Table B6. (Continued)
Solute
Selenium
Strontium
Zinc
Membrane Type
Cellulose acetate
Cellulose acetate
Polyamide
Cellulose acetate
Cellulose acetate
Polyamide
Cellulose acetate
Cellulose acetate
Cellulose acetate
Cellulose acetate
Polyamide
NS-100
NS-100
Concentration
Ppm
0.5
0.5
0.5
2.5
2.5
2.5
9.4-10
32.8-31.4
2.5
2.5
2.5
12.5
12.5
Pressure
atm
12
26.5
26.5
12
26.5
26.5
28.2
28.2
12
26.5
26.5
41.8
41.8
Rejection
%
98.4
92.2-98.7
81.1
83.7
79.2-88.8
87.3
96.9-98.6
98.8-99.5
95
87.7-95.2
90
97.9
100
pH
4.8
4.8
4.8
4.8
4.8
4.8
_
-
4.8
4.8
4.8
8
11
Reference
*
*
*
*
*
*
B31
B31
*
*
*
B31
B31
*This study
-------
REFERENCES
Bl. U.S. Environmental Protection Agency. Rationale for the Development
of BAT Priority Pollutants Parameters. EPA, Washington, D.C. 1977.
B2. Glueckawf, E. On the Mechanism of Osmotic Desalting With Porous
Membranes. Proceedings of First International Symposium on Water
Desalination, Office of Saline Water, Department of Interior,
Washington, D.C., pp. 145-150, 1967.
B3. Bean, C. P. Research and Development Progress Report No. 465.
Office of Saline Water, U.S. Department of Interior, Washington,
D.C., 1969.
B4. Matsuura, T., L. Pageau, and S. Sourirajan. Reverse Osmosis Separa-
tion of Inorganic Solutes in Aqueous Solutions Using Porous Cellulose
Acetate Membranes. J. Applied Polymer Science, 19:179-198, 1975.
B5. Matsuura, T., and S. Sourirajan. Physiocochemical Criteria for
Reverse Osmosis Separation of Monohydric and Polyhydric Alcohols in
Aqueous Solutions Using Porous Cellulose Acetate Membranes.
J. Applied Polymer Science, 17:1043-1071, 1973.
B6. Matsuura, T., and S. Sourirajan. Reverse Osmosis Separation of
Hydrocarbons in Aqueous Solutions Using Porous Cellulose Acetate
Membranes. J. Applied Polymer Science, 17:3683-3708, 1973.
B7. Matsuura, T., and S. Sourirajan. Physiocochemical Criteria for
Reverse Osmosis Separation of Alcohols, Phenols, and Monocarboxylic
Acids in Aqueous Solutions Using Porous Cellulose Acetate Membranes.
J. Applied Polymer Science, 15:2905-2927, 1971.
B8. Matsuura, T., and S. Sourirajan. Reverse Osmosis Separation of
Organic Acids in Aqueous Solutions Using Porous Cellulose Acetate
Membranes. J. Applied Polymer Science, 17:3361-3382, 1973.
B9. Matsurra, T., and S. Sourirajan. Physiocochemical Criteria for
Reverse Osmosis Separation of Aldehydes, Ketones, Ethers, Esters,
and Amines in Aqueous Solutions Using Porous Cellulose Acetate
Membranes. J. Applied Polymer Science, 16:1663-1688, 1972.
BIO. Matsuura, T., and S. Sourirajan. Reverse Osmosis Separation of
Phenols in Aqueous Solutions Using Porous Cellulose Acetate Membranes
J. Applied Polymer Science, 16:2531-2554, 1972.
-155-
-------
Bll. Chian, E.S.K., and H.H.P. Fang. Evaluation of New Reverse Osmosis
Membranes for the Separation of Toxic Organic Compounds from Waste
Water. American Institute of Chemical Engineers Symposium Series,
70(136):497, 1974.
B12. Lonsdale, H. K., B. P. Cross, F. M. Graber, and C. E. Milstead.
Permeability of Cellulose Acetate Membranes to Selected Solutes.
J. Macromolecular Science, Part B-Physics. B5(l):167-188, 1971.
B13. Anderson, J. E., S. J. Hoffman, and C. R. Peters. Factors Influen-
cing Reverse Osmosis Rejection of Organic Solutes from Aqueous
Solution. J. Physical Chemistry 76:4006-4011, 1972.
B14. Wendt, R. P., R. J. Toups, J. K. Smith, N. Leger, and E. Klein.
Using a Rotating Batch Dialyzer. Industrial and Engineering Chemistry
Fundamentals. 10:406-411, 1971.
B15. Sourirajan, S. Reverse Osmosis. Logos Press Limited, London, 1970.
580 pp.
B16. Hinden, E., G. H. Dunstan, and P. L. Bennett. Water Reclamation by
Reverse Osmosis. U.S. Department of the Interior. Federal Water
Pollution Control Administration Grant, WP-00624.
B17. Deinzer, M., R. Melton, and D. Mitchell. Trace Organic Contaminants
in Drinking Water, Their Concentration by Reverse Osmosis. Water
Research, 9:799-805, 1975.
B18. Kamizawa, C., H. Masuda, and S. Ishizaka. Study on Reverse Osmosis:
The Permeation Behavior of Surfactant Solutions Through Cellulose
Acetate Membranes. Bulletin, Chemical Society of Japan. 45:2967-2969,
1972.
B19. Kamizawa, C., H. Masuda, and S. Ishizaka. Study on Reverse Osmosis:
The Permeation Behavior of Amino Acid Solutions Through Cellulose
Acetate Membranes. Bulletin, Chemical Society of Japan. 45:2964-2966,
1972.
B20. Cadotte, J. E., C. V. Kopp, K. E. Cobian, and L. T. Rozelle. Progress
Report on in-situ-formed Condensation Polymers for Reverse Osmosis
Membranes. Office of Saline Water, U.S. Department of the Interior,
1974. 92 pp.
B21. Permasep Products Division. Engineering Design Manual, DuPont
Company, Wilmington, Delaware, 1976.
B22. Chian, E.S.K., W. N. Bruce, and H.H.P. Fang. Removal of Pesticides
by Reverse Osmosis. Environmental Science & Technology, 9(l):52-59,
1975.
B23. Cabasso, I., C. S. Eyer, E. Klein, and J. K. Smith. Evaluation of
Semipermeable Membranes for Concentration of Organic Contaminants
in Drinking Water. EPA-670/1-75-001. U.S. Environmental Protection
Agency, Cincinnati, Ohio, 1975. 182 pp.
-156-
-------
B24. Merten, U., Desalination by Reverse Osmosis. The MIT Press,
Cambridge, Massachusetts, 1966, 93 pp.
B25. Lonsdale, H. K., and H. E. Podall. Reverse Osmosis Membrane Research.
Plenum Press, New York, New York, 1972, 503 pp.
B26. Cruver, J. E., Reverse Osmosis—Where It Stands Today. Water Sewage
Works, 10:74-78, 1973.
B27. Silman, H., The Substitution of Toxic Materials. Plating, 61(4):
332-336, 1974.
B28 Matsuura, T., and S. Sourirajan. Studies on Reverse Osmosis for
Water Pollution Control. Water Research, 6:1073-1086, 1972.
B29. Hauck, R. H., and S. Sourirajan. Reverse Osmosis Treatment of
Diluted Nickel-Plating Solutions, J. Water Pollution Control
Federation, 44:1372-1383, 1972.
B30. Mixon, F. 0., The Removal of Toxic Metals from Water by Reverse
Osmosis, OSW-RDPR-73-889, National Technical Information Service,
1973.
B31. Chian, E.S.K., and H.H.P. Fang, Removal of Toxic Compounds by
Reverse Osmosis. Presented at ASME Conference, Seattle, Washington,
1974.
B32. Bovet, E. D., Industrial Waste Desalting for Byproduct Recovery.
J. American Water Works Association, 62(8):539-542, 1970.
B33. McRae, W. A., and E. J. Paris. Removal of Trace Heavy Metals from
Water by Electrodialysis. Proc. Traces of Heavy Metals in Water
Removal Process and Monitoring, Princeton University, New Jersey,
pp. 143-157, 1975.
B34 Lacey, R. E., and S. Loeb. Industrial Processing with Membranes.
Wiley-Interscience, New York, New York, 1972, 348 pp.
-157-
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/7-80-063
2.
3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE
Application of Membrane Technology to Power
Generation Waters
5. REPORT DATE
March 1980
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
T.L. Don Tang, Tien-Yung J. Chu, and
Ralph D. Boroughs
8. PERFORMING ORGANIZATION REPORT NO.
EDT-103
9. PERFORMING ORGANIZATION NAME AND ADDRESS
10. PROGRAM ELEMENT NO.
Tennessee Valley Authority
Division of Energy Demonstrations and Technology
Chattanooga, Tennessee 37401
1NE624A
11. CONTRACT/GRANT NO.
EPA Interagency Agreement
D8-E721-BE
12. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Industrial Environmental Research Laboratory
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
Final; 10/75-9/79
14. SPONSORING AGENCY CODE
EPA/600/13
is. SUPPLEMENTARY NOTES IERL-RTP project officer is Theodore G. Brna, Mail Drop 61,
919/541-2683. TVA project director is Hollis B. Flora II.
16. ABSTRACT
The report gives results of an examination of three membrane technologies
(reverse osmosis, ultrafiltration, and electrodialysis) for wastewater treatment and
reuse at electric power generating plants. Recirculating condenser water, ash sluice
water, coal pile drainage, boiler blowdown and makeup treatment wastes, chemical
cleaning wastes, wet SO2 scrubber wastes, and miscellaneous wastes were studied.
Membrane separation of toxic substances in wastewater was also studied. Waste
characteristics, applicable regulations, feasible membrane processes, and cost in-
formation were analyzed for each waste stream. A user's guide to reverse osmosis
was developed and is provided as an appendix. Power plant water treatment with
membrane technologies to attain total water reuse and zero effluent discharge is
technically feasible. Membrane technologies are not suited to remove materials that
are unstable and apt to precipitate as they are concentrated; however, they excel in
separating materials not susceptible to conventional wastewater treatment (e.g. ,
very soluble toxics and dissolved solids). Thus membrane technologies complement
rather than compete with conventional technologies. For dissolved solids control,
membrane technologies are viable alternatives to distillation. Distillation is more
costly, but requires less pretreatment than reverse osmosis or electrodialysis.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Pollution Electrodialysis
Waste Water Toxicity
Water Treatment Dissolved Organic
Membranes Matter
Electric Power Plants
Osmosis Distillation
Filtration
Pollution Control
Stationary Sources
Membrane Separation
Reverse Osmosis
Ultrafiltration
Dissolved Solids
13B
06T
08H
11G
10B
07D 13H,07A
13. DISTRIBUTION STATEMENT
Release to Public
19. SECURITY CLASS (This Report)
Unclassified
21. NO. OF PAGES
170
20. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
158
------- |