PE82-108440
Feasibility of Commercialized Water Treatment
Techniques for Concentrated Waste Spills
TRW Environmental Engineering. Div.
Redondo Beach, CA
Prepared for
Municipal Environmental Research Lab
Cincinnati, OH
Sep 81
U.S. DEPARTMENT OF COMMERCE
National Technical Information Service
NT1S®
-------
EPA-600/2-81-213
September 1981
FEASIBILITY OF COMMERCIALIZED WATER TREATMENT TECHNIQUES
FOR CONCENTRATED WASTE SPILLS
by
Masood Ghassemi, Kar Yu, and Sandra Quinlivan
TRW Environmental Engineering Division
Redondo Beach, California 90278
Contract No. 68 03 2560
Project Officer
Frank Freestone
Oil and Hazardous Materials Spills Branch
Municipal Environmental Research Laboratory-Cincinnati
Edison, New Jersey 08817
MUNICIPAL ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATL_OHIQ_45268_
REPRODUCED BY
NATIONAL TECHNICAL
• INFORMATION SERVICE
U.S DEPAtTIH*! Of CO»»f«C£
SPDMGUHD »» 22161
-------
TECHNICAL REPORT DATA
(Please read Intmicnons on the reverse before completing)
1 REPORT NO
EPA-600/2-81-213
3. RECIPIENT'S ACCESSION >O
ORD Report
CCESSION >O k
10844 0
4. TITLE AND SUBTITLE
Feasibility of Commercialized Water Treatment
Techniques for Concentrated Waste Spills
5 REPOR- DATE
September 1981
8 PERFORMING ORGANIZATION CODE
7 AUTHOHIS)
M. Ghassemi, K. Yu, and S. Quinlivan
8. PERFORMING ORGANIZATION RcPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
TRW, Inc.
One Space Park
Redondo Beach, CA 90278-
10. PROGRAM ELEMENT NO.
11 CONTHACT7GRANT NO
68-03-2560
12 SPONSORING AGENCY NAME AND ADDRESS
Municipal Environmental Research,Laboratory - Cin., OH.
Office of Research and^Development
U.S. Environmental Protection Agency
Cincinnati, OH-45268-, . . ' . • »«<
13. TYPE OF REPORT AND PERIOD COVERED
Task Final-
14. SPONSORING AGENCY CODc
EPA/600"/! 4
15. SUPPLEMENTARY NOTES
Project Officer: Frank Freestone, (201) 321-6632
16 ABSTRACT
The suitability and economics of using reverse osmosis, ultrafiltration, ion exchange,
wet air oxidation, high purity oxygen activated sludge-process, ultraviolet-ozone oxida-
tion, and coagulation/precipitation for on-site treatment of concentrated wastes were
evaluated. Published literature data and data obtained from process suppliers provided
the basis for the evaluation.
When used alone, none of the processes considered would be economically applicable to
on-site mobile unit treatment of the variety of concentrated wastes encountered,
although reverse osmosis, ion exchange, and wet air oxidation meet many of the applica-
tion requirements and hence require less pretreatment and/or post-treatment. The
estimated capital costs for a unit suitable for trailer mounting vary from as low as
$35,000 for a 227,000 a/day (60,000 gpd) ultrafiltration unit to as high as $1.25 to
$1.5,million for a 54,000 fc/day (14,400 gpd) 2-trailer wet air oxidation unit. For
short-term operation, the operating cost of the mobile unit is determined largely by
non-process specific costs (for example, transportation, labor subsistence, analytical
support, etc.), which vary from situation to situation.
17
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b IDENTIFIERS/OPEN ENDED TERMS
c COSATl Field/Croup
Mobile Treatment
On-site
Reverse Osmosis
Ultrafiltration
Ion Exchange
Wet Air Oxidation
Biological Oxidation
Ozone/UV
Physical Chemical Treatmf
nt
18 DISTRIBUTION STATEMENT
19 SECURITY CLASS I This Report/
Unclassified
20 SECURITY CLASS (Thupoge>
Unclassified
22 PRIC:
EPA Form 2220-1 (R.v 4-77)
-------
DISCLAIMER
This report has been reviewed by the Municipal Environmental Research
Laboratory, U.S. Environmental Protection Agency, and approved for publica-
tion. Approval does not signify that the contents necessarily reflect the
views and policies of the U.S. Environmental Protection Agency, nor does
mention of trade names or commercial products constitute endorsement or re-
commendation for use.
ii
-------
I FOREWORD
!
I
The U.S. Environmental Protection Agency was created because of increasing
public and government concern about the dangers of pollution to the health and
i welfare of the American people. Noxious air, foul water, and spoiled land are
i tragic testimonies to the deterioration of our natural environment. The com-
plexity of that environment and the interplay of its components require a
i concentrated and integrated attack on the problem.
I Research and development is that necessary first step in problem solution;
it involves defining the problem, measuring its impact, and searching for
solutions. The Municipal Environmental Research Laboratory develops new and
improved technology and systems to prevent, treat, and manage wastewater and
| solid and hazardous waste pollutant discharges from municipal and community
i sources, to preserve and treat public drinking water supplies, and to mini-
l mize the adverse economic, social, health, and aesthetic effects of pollution.
I This publication is one of the products of that research and provides a most
I vital communications link between the researcher and the user community.
i
The suitability and economics of using reverse osmosis, ultrafiltration,
ion exchange, wet air oxidation, high purity oxygen activated sludge process,
i ultraviolet-ozone oxidation, and coagulation/precipitation for on-site t^eat-
ment of concentrated wastes were evaluated in this report.
Francis T. Mayo, Director
Municipal Environmental Research
Laboratory
m
-------
ABSTRACT
The suitability and economics of using reverse osmosis, ultrafiltration,
ion exchange, wet air oxidation, high purity oxygen activated sludge process,
ultraviolet-ozone oxidation, and coagulation/precipitation for on-site treat-
ment of concentrated wastes were evaluated. Published literature data and
data obtained from process suppliers provided the basis for the evaluation.
When used alone, none of the processes considered would be economically
applicable to on-site mobile unit treatment of the variety of concentrated
wastes encountered, although reverse osmosis, ion exchange, and wet air oxi-
dation meet many of the application requirements and hence require less pre-
treatment and/or post-treatment. The estimated capital costs for a unit
suitable for trailer mounting vary from as low as $35,000 for a 227,000 «,/day
(60,000 gpd) ultrafiltration unit to as high as $1.25 to $1.5 million for a
54,000 A/day (14,400 gpd) 2-trailer wet air oxidation unit. For short-term
operation, the operating cost of the mobile unit is determined largely by non-
process specific costs (for example, transportation, labor subsistence,
analytical support, etc.), which vary from situation to situation.
This report was submitted in fulfillment of Contract No. 68-03-2560, Work
Directive T5009, by TRW Environmental Engineering Division under the sponsor-
ship of the U.S. Environmental Protection Agency.
-------
CONTENTS
Foreword lii
Abstract iv
Figures vii
Tables vm
Acknowledgement x
1. Summary 1
2. Conclusions 2
3. Introduction 10
Background and study objectives 10
General requirements for use of a process in a mobile unit. . 11
Organization of the report 12
4. Reverse Osmosis 13
Process description 13
Process applications and related studies 16
Use of RO in mobile units for treatment of waste spills ... 22
References 27
5. Ultrafiltration 30
Process description 30
Process applications and related studies 32
Use of UF in mobile units for treatment of waste spills ... 35
References 38
6. Ion Exchange 40
Process description 40
Process applications and related studies 41
Use of ion exchange in mobile units for treatment of
waste spills ^3
References 48
7. Wet Air Oxidation 50
Process description 50
Process applications and related studies 54
Use in mobile units for the treatment of waste spills .... 56
References 59
-------
CONTENTS (Continued)
8. Biological Treatment 60
Process description 60
Process applications and related studies 64
Use of high purity oxygen activated sludge process in
mobile units for treatment of waste spills 71
References 73
9. Combined Ozonation and Ultraviolet Radiation 75
Process description 75
Process applications and related studies 80
Use of 03/UV in mobile units for treatment of waste spills. . 83
References 86
10. Chemical Treatment (Precipitation/Coagulation) 87
Process description 87
Process applications and related studies 95
Use in mobile units for the treatment of waste spills .... 96
References 96
11. Miscellaneous Processes 97
Gravity separation 97
Filtration 100
Carbon adsorption 103
Incineration 108
References 114
VI
-------
FIGURES
Number Page
1 Membrane module configurations .................. 17
2 4-Stage module arrangement for a 227,000 «./day (60,000 gal/day)
mobile RO treatment system ................... 26
3 Pore size vs. flux for filtration processes ........... 31
4 Effect of operating parameter on UF flux ............. 33
5 Schematic flow diagram for the WAO process ............ 51
6 Oxidation curves for concentrated simulated or actual wastes. . . 53
7 Time-temperature effect on the degree of oxidation ........ 53
8 Wet air oxidation mobile unit .................. 57
9 Schematic diagram of high purity oxygen system .......... 63
10 Schematic diagram of the UNOX mobile unit pilot plant and the
external clarifier ....................... 69
11 Comparison of Os/UV combination treatment with ozonation
(a = fraction of ozone used by the reaction). ... ...... 76
12 Schematics of an 03/UV reactor .................. 79
13 Solubility of metal hydroxides and sulfides ........... 88
14 Precipitation of pyrophosphate with Fe (III) at a 2:1 cation-
to-pyrophosphate equivalence ratio (initial pyrophosphate
concentration, 18 mg/2P) .................... 90
15 Residual orthophosphate in precipitation of orthophosphate
with Fe (III) at a 1:1 cation-to-orthophosphate molar ratio
(initial orthophosphate concentration, 12 mg/x,P) ........ 92
16 Orthophosphate removal in Fe (Ill)-orthophosphate reaction at
pH 4.0 (initial orthophosphate concentration, 12 mg/iP) .... 93
17 Schematic of coagulation-flocculation system ........... 94
18 Gravity separators ........................ 98
19 Typical filtration bed ...................... 101
20 Two- vessel granular carbon adsorption system ........... 105
21 Rotary kiln incineration unit .................. Ill
22 Mobile environmental restoration incinerator comolex ....... 113
vii
-------
TABLES
Number Page
1 Rejection of Various Salts and Organics by the RO Membrane. ... 15
2 Comparison of Reverse Osmosis Module Configurations 18
3 Designs and Operating Features for Some Commercial RO Plants
Treating High TDS Waters 19
4 Use of RO for the Demoralization of Acid Mine Drainage 20
5 Use of RO for the Treatment of an Oily Wastewater 21
6 Comparison of the Three RO Modules for Use in Spills Treatment
Mobile Units 25
7 Features of Commercial UF Module Designs 34
8 Results from UF Pilot/Laboratory Scale Studies 36
9 Pertinent Features of Ion Exchange Systems 42
10 Specific Examples of Full-scale Ion Exchange Application to
Industrial Wastewater Treatment 44
11 Heating Values for Some Comoounds and Waste Materials and Air
Requirements for Their Oxidation 55
12 WAO Efficiency for Ten Priority Pollutants (1-hr Detention
Time) 55
13 Comparison of Process Design Conditions for the High Purity
Oxygen System and for Conventional Air Aeration Systems for
Typical Municipal Wastewater 62
14 The High Purity Oxygen System Performance in Full-scale
Applications 65
15 Performance of the High Purity Oxygen System Pilot Plants .... 67
16 Typical Coal Conversion Wastewater Characteristics (Wastewater
Diluted 1+19 With River Water Prior to Treatment) 68
17 Diluted Coal Conversion Wastewater Treatment Results With the
High Purity Oxygen Process 68
18 List of Organics Determined to be Economically Treatable by
the Os/UV Process 78
19 Typical Os/UV Pilot Plant Test Results 81
-------
TABLES (Continued)
Number Page
20 Comparison of the Economics of 03/UV and Carbon Adsorption
for the Treatment of PCB Wastewaters 82
21 Oxidation of Metal Complexed Cyanide in a Staged Reactor 82
22 Ozone Exposure Levels Proposed by Various Regulatory and
Professional Organizations 85
23 General Features of Gravity Separators for Use in Mobile Units. . 99
24 General Features of Filtration for Application to Wastewater
Treatment in Mobile Units 102
25 Feature of a Dual-Media Filter Unit Used in the EPA's Mobile
Physical/Chemical Treatment System 103
26 Amenability of Typical Organic Compounds to Activated Carbon
Ads'orption -.-•'.•:>•.. . . .•;•/. . .- ". . . . 106
27 Some Features of Activated Carbon Adsorption Process for
Remo*va-l'oYfOrganicSafronrWa'StewaterS'.,. .'.<•.> •."".* .'. . 107
28 Toxic Compounds Removed from Water Using EPA's Mobile Physical/
Chemical^Treatment System- -.-..» -.•..<.. 109
29 Key Features of Major Types of Incinerators 110
30 Some Advantages and Disadvantages of Incineration Systems for
the Destruction of Toxic Organic Wastes 112
-------
ACKNOWLEDGMENTS
The authors wish to express their gratitude to the EPA Project Officer,
Mr. Frank Freestone, for his advice and guidance during the course of this
effort.
Special thanks are extended to the following individuals and process/
equipment suppliers who supplied technical data for use in this engineering
evaluation:
I. Abrams and R.Y. Lawrence, diamond Shamrock, Redwood City,
California 94064
Sam Akinbami, Linde Division of Union Carbide, Tonawanda,
New York 14150
Gerald Alexander, Permutit Co., Glendale, California 91071
Claude Ellis, Zimpro Inc., Rothschild, Wisconsin 54474
A.M. Fradkin, Dow Chemical, Pasadena, California 91101
Roy Langslet, Air Products and Chemicals, Inc., Allentown,
Pennsylvania 18105
R. Lawrence, Envirogenics Co., El Monte, California 91731
Craig Netwig, Fluid Systems Division, UOP, Inc.", San Dieao,
California 92101
H. William Prengle, Jr., Houston Research, Inc., Houston,
Texas 77043
B.W. Stevens, Rohm and Haas Company, Philadelphia, Pennsylvania
19105
Jack D. Zeff, Westgate Research Corporation, West Los Angeles,
California 90025
At TRW, Ms. Pam Painter and Mr. Robert Scofield assisted in the prepara-
tion of Chapter 8, Biological Treatment, and Mr. Kraig Scheyer assisted in the
compilation of data for Chapter 11, Miscellaneous Processes. The authors wish
to express their sincere appreciation to Ms. Monique Tholke for typing the
manuscript.
-------
SECTION 1
SUMMARY
Under a contract with EPA, an engineering evaluation has been made of the
suitability of several commercially available water and wastewater treatment
processes for userin mobile units for on-site treatment of highly contaminated
waters encountered in hazardous materials spills situations. The^processes
evaluated are^reverse osmosis, ultrafiltration, ion exchange, wet air oxida-
tion, biologicar treatment, combined ozonation/ultraviolet radiation treatment,
and coagulation/flocculation. Gravity separation, filtration, activated carbon
adsorption, and incineration* which have been used*or are under development
for use in«spill control situations, are also briefly reviewed.
The process evaluation is based on published literature and data obtained
from process and equipment suppliers and assumes the use of a single trailer,
227,000 A/day (60,'OOQ gpd) hydraulic capacity, and the use of a process alone
and not in combination with other processes in a treatment train. The evalua-
tion criteria have included weight and size factors, capability for self-
sustained operation, applicability to a diversity of waste types and contami-
nant concentrations, quick system startup and shutdown, minimum requirement
for skilled labor, minimal intermedia pollution transfer potential, commercial
availability or ease of fabrication of equipment, experience in mobile unit
applications, and capital and operating costs.
REVERSE OSMOSIS (RO)
RO uses high pressures to force solvent (for example, water) through a
membrane that is permeable to the solvent molecules but not to the solute
molecules. Several types of membranes and membrane designs are available. RO
is being used on a commercial scale to demineralize brackish waters and to
treat a variety of industrial wastewaters (for example, food processing waste-
waters, plating rinses, and cooling water blowdown). Numerous pilot plant
studies have also been conducted or are under way to assess process suitability
for the treatment of a range of industrial wastewaters, sea water, and biologi-
cally treated municipal sewage. RO is very effective in the removal of most
dissolved organics, inorganic salts, heavy metals, and emulsified oils. In-
dustrial wastewaters containing several thousand ppm of chemical oxygen demand
(COD) and total dissolved solids (TDS) have been successfully treated with RO,
as well as sea water that contains 3.5 percent TDS. One process developer has
used RO in a 189,000 A/day (50,000 gpd) unit for brackish water demineraliza-
tion; the U.S. Army has a 54,500 a/day mobile unit for field use to obtain
potable water from brackish waters.
-------
RO meets many of the requirements for process use in a mobile spill con-
trol unit. RO units are compact, are commercially available, can be started
and shut down relatively quickly, can be serviced conveniently, would produce
only a small volume of residue (10 to 25 percent of the feed), would not
require skilled field labor, and can be operated with the electric power pro-
duced on-site. The major shortcoming of the system is membrane susceptability
to fouling or degradation caused by the presence of suspended solids or strong
oxidizing chemicals in the wastewater or very low pH of the wastewater. Based
on a preliminary design developed for a system using the "spiral wound" mem-
brane design, the capital cost for a 227,000 i/day (60,000 gpd) mobile RO
system and support equipment has been estimated at 570,000. This cost does
not include the trailer cost and costs for pretreatment equipment (for example,
coagulation/filtration for suspended solids removal) that may be required. The
operating cost per unit waste volume treated depends on the specific applica-
tion (waste characteristics, site location, treatment chemicals required, etc.).
ULTRAFILTRATION (UF)
UF is a pressure-driven membrane separation process that operates at a
lower pressure than RO and is suitable for applications involving larger mole-
cules. Several types of membranes and membrane designs are offered commer-
cially. To date, all reported full-scale uses of UF have been in industrial
applications (for example, concentrating cheese whey, dye rinses, and emulsi-
fied oils). A number of pilot plant studies have also been conducted or are
under way to treat other industrial wastewaters including wastes from Kraft
pulp mills, laundries, and textile mills. In these applications, the process
has proved effective in concentrating total solids, including color. Feed
solids concentrations of as high as 46,300 ppm have been tested. When used in
applications to high concentration wastewaters, the effluent may require addi-
tional treatment. Although UF has not been deployed in mobile units, process
vendors have skid-mounted units that can be installed on a trailer; these units
consist of one or two modules with hydraulic capacities in the 18,900 to 39,800
A/day (5,000 to 10,000 gpd) range.
Sharing many characteristics of RO, UF meets nearly all key requirements
for mobile unit application. The primary limitation of the system is inapoli-
cability to wastes containing low molecular substances and for producing low
concentration permeates from highly concentrated wastes. The capital cost for
a 227,000 £/day (60,000 gpd) spiral wound UF mobile unit is estimated at
$35,000.
ION EXCHANGE
Ion exchange is a process whereby the toxic or undesirable ions in waste-
water are replaced with relatively harmless ions. The exhausted resins are
regenerated with acids, bases, or brine solutions. Exchange resins used in
commercial applications are primarily made of synthetic organic materials.
Recent advances in resins technology have involved development of "sorptive"
resins that remove organics via adsorption rather than an ion exchange
mechanism. Sorptive resins are reaenerated with organic solvents or inorganic
solutions of appropriate pH. Ion exchange resins are commonly employed in
-------
columnar beds through which the aqueous phase is oassed in an upflow or down-
flow direction.
Ion exchange has been widely used in full-scale applications for water
softening and boiler water treatment. Industrial wastewater treatment appli-
cations include treatment of electroplating wastewaters to recover chromium
and water for reuse, mercury removal from chlor-alkali plant effluent, acid
wastewater demineralization at an Army ammunition plant, color removal from
pulp and paper mill effluent, recovery of ammonium nitrate from fertilizer
production plant effluent, radioactive waste treatment at nuclear power plants,
and recovery of phenols from concentrated industrial brines. Ion exchange can
handle highly concentrated wastes; the process economics, however, may be un-
favorable because of the requirement for frequent regeneration. Numerous pilot
plant studies have.been conducted or are currently under way to evaluate the
applicability and economics of ion exchange for the treatment of various in-
dustrial wastewaters. Ion exchange process vendors have used a number of
mobile ion exchange units for process treatability studies. A 276 i/rnin (70
gpm) mobile unit has been used for tertiary treatment of domestic sewage.
Ion exchaYfge meets nearly'all "the-requirements for mobile unit spill con-
trol application. Ion exchange units are compact, are commercially available,
are easily automated,, can be.started-and shut-down.quickly, can1 handle a range
of organic-and inorganic concentrations (including hi'ghly acidic and alkaline
wastes), can be serviced conveniently, would produce a small volume^of residue
that may be Suitable for material recovery, and woul'd-not require skilled
operating labor. Ion exchange is not an energy-intensive process, and the
power required for pumping can be supplied by an on-board generator. Based on
the data provided by one process supplier, a preliminary design has been pre-
pared for a 227,000 a/day (60,000 gpd) ion exchange system for use on a trailer
flatbed. By loading the system with proper type(s) of resins (exchange resins
and/or sorptive resins), the six-bed system can be used for demineralization
of water low in inorganics, removal of both organic and inorganic substances,
and removal of organics only. The total capital cost for the system is
estimated at $140,000. The operating cost per unit volume of waste treated
would vary with the speficic application and would be affected largely by
waste characteristics that dictate the regeneration frequency. A total regene-
ration cost of $58 per cycle has been estimated for regeneration of one cation
and one anion exchange column in the system with 5 percent solutions of HC1
and NaOH. The corresponding cost for the regeneration of a sorotive resin
column is estimated at $42, $400, and $1,150 for regeneration with 5 percent
NaOH, methanol and acetone, respectively. No credit is taken for acetone or
methanol, which may be potentially recovered via distillation of the spent
regenerant.
WET AIR OXIDATION (WAO)
WAO involves aqueous phase oxidation of reduced inorganic and orqanic
substances with air at relatively high temperatures and pressures. The orocess
is especially suitable for the treatment of high strength or toxic organic
wastes or wastes containing nonbiodegradable organics. Because oxidation takes
olace in an aqueous medium, air pollution problems are not generally associated
-------
with WAO. By proper selection of operating conditions (detention time, pres-
sure, and temperature), any degree of oxidation (including complete destruc-
tion) of reduced substances can be achieved.
There are currently more than 150 WAO units in operation worldwide; about
90 percent of the units handle municipal sludges, whereas the rest process
industrial wastes. Other reported commercial applications of WAO are for
treatment of cyanide, photographic, acrylonitrile, and pulp mill wastes. On
a pilot plant scale WAO has been applied to the treatment of explosives, pesti-
cides, and plastic wastes. The process developer owns a 550 £/day (155 gpd)
WAO unit that it uses at the company headquarters for waste treatability
studies; a 54,000 A/day (14,000 gpd) mobile unit which will be mounted on two
trailers is currently under design and should be available in 1981.
WAO does not meet all the requirements for use in a mobile spill control
unit. The major system limitations are the small capacity of units suitable
for trailer mounting; requirements for special design, material of construc-
tion, and skilled operating labor; and relatively high capital cost. The
capital cost for the 54,000 a/day (14,400 gpd) titanium-constructed unit is
estimated at around $1.25 to $1.5 million. The operating costs are estimated
at $15/hr for fuel and 2.5i per gallon of waste that is detoxified.
BIOLOGICAL TREATMENT
Biological oxidation involves the use of microorganisms to convert the
dissolved and oxidizable organic matter to organic and inorganic end products
and to agglomerating microbial floes that can be removed by settling and/or
filtration. Biological oxidation is most efficient and results in the produc-
tion of stable end products when carried out aerobically, that is, in the pre-
sence of dissolved oxygen. The activated sludge process is one of the most
widely used biological wastewater treatment processes. In the conventional
design, air is provided to the wastewater in a reactor containing a high con-
centration of microbial floes. The treated waste is then clarified prior to
discharge; a portion of the settled sludge is recycled to the reactor, and the
excess sludge -is wasted. Use of high purity oxygen instead'of air is one of
the latest developments in activated sludge technology. Considerable saving
in capital and operating cost is realized because of higher oxygen utilization
efficiency that enables use of smaller equipment, reduces power consumption,
and improves sludge settleability. High purity oxygen activated sludge, which
is offered in this country under the trade names UNOX and OASES, has been
evaluated in this study for use in mobile spill control units.
The high purity activated sludge process has been used in a number of
large scale applications for the treatment of municipal and industrial waste-
waters. Pilot plant treatability studies have also been carried out on a
range of industrial wastewaters, including petrochemical, pulp mill, oharma-
ceutical, fruit canneries, brewery, and textile mill wastewaters. Tne most
concentrated chemical wastes that have been treated without prior dilution
have had COD values in the 1,000 to 3,000 range and have required long deten-
tion times (10 to 30 hours) to obtain high removal efficiencies. Union Carbide
Corporation, the developer of the UNOX process, has seven mobile pilot plants
-------
having hydraulic capacities ranging from 27,000 to 217,000 a/day (7,200 to
57,600 gpd). The company offers a total pilot plant program that includes the
pilot plant equipment and necessary engineering services and supervision for
plant operation.
The high purity activated sludge process does not meet some of the key
requirements for use in a mobile spill control unit. System limitations in-
clude those inherent in biological processes (primarily, unsuitability for
handling toxic or refractory wastes); long startup and stabilization period
required; requirement for elaborate and lengthy bench-scale testing to deter-
mine treatability of complex wastes and optimum operating conditions; suscep-
tibility to "shock" loads; unsuitability for treatment of wastewaters con-
taining volatile hydrocarbons (because of explosion hazard), requirement for
trained operating personnel; and very large reactor size required for handling
concentrated wastes. It is estimated that the largest reactor size that can
be accommodated on a single trailer is 47,300 £ (12,500 gal) that, at a deten-
tion time of 48 hours (which would be typically required for wastes containing
1,000-3,000 ppm of COD), would have a hydraulic capacity of only 23,650 A/day
(6,250 gpd). The purchase price for an existing UNOX mobile unit with a
reactor volume of 6,050 I (1,600 gal) is about $250,000.
COMBINED OZONATION AND ULTRAVIOLET RADIATION (Oa/UV)
03/UV is a newly-developed chemical oxidation process that has proved
effective for the destruction of many organics, including refractory,and toxic
chemicals, organometallic complexes, and reduced inorganic substances. The
oxidation takes place in a reactor where the waste is contacted with ozone and
UV radiation simultaneously. The use of UV radiation enhances the oxidation
power of ozone, thereby increasing oxidation rate and efficiency. Wastewater
characteristics that influence process efficiency and hence system design are
wastewater flow rate, nature and concentration of organics and suspended solids
(which interfere with passage of UV light), and temperature.
Only a very limited number of full-scale units are reportedly in use
today. These plants handle photographic and metal plating wastes from an Army
ammunition plant and cyanide waste from a tool company. On a pilot plant
scale, 03/UV has been investigated for the treatment of "pink water" from
ammunition plants, wastewater from field hospitals, PCB-containing wastewaters,
and secondary effluent. No mobile 03/UV unit is reportedly in operation today.
03/UV meets many of the requirements for applicability to the treatment
of waste spills in a mobile unit. The major shortcoming of the system relates
to the lack of extensive experience with the process and the limited amount of
ozone that can be generated on-site. Because of ozone generation capacity
limitations, highly concentrated wastes cannot be treated on-site without pre-
treatment. For spills containing 10 and 1,000 ppm TOC, the theoretical volumes
of wastewater that can be treated in a mobile unit equipped with a 91 kg/cay
03 generator are 1,130,000 a/day (300,000 gpd) and 11,300 a/day (3,000 gpd),
respectively. The caoital cost for a 227,000 a/day (60,000 gpd) mobile unit
is estimated at $285,000. Fuel for the power generator is the major operating
cost item. A 125 kw generator would consume about $6/hour of fuel.
-------
CHEMICAL TREATMENT (PRECIPITATION/COAGULATION)
Addition of chemicals to wastewaters to precipitate dissolved substances
and to coagulate colloidal particulates are widely used in the treatment of
raw water supplies and industrial wastewaters. The tieatment involves addi-
tion to and rapid mixing of the chemicals in the raw water, gentle mixing to
promote floe growth, and settling and/or filtration for floe removal.
Precipitation results when the solubility of a compound is exceeded by
addition of a "common ion". Many heavy metals can be precipitated as hydrox-
ides or sulfides by addition of chemicals such as lime, sodium hydroxide, and
sodium sulfide. The most widely used coagulants are aluminum and ferric salts.
Coagulation is believed to result from the neutralization of the negative
charge on the colloidal particles by the positively charged coagulant ions and
their hydrolysis products and/or chemical reaction between the coagulant ions
and the surface groups on the colloidal particles. Because of the many factors
that affect coagulation efficiency (pH, nature, and concentration of organic
and inorganic substances in water, coagulant dosage, temperature, etc.), no
scientific method has yet been developed (despite extensive research) that
would enable the determination of optimum coagulant dose based on some measu-
rable raw water characteristics. In practice, the optimum coagulation condi-
tions are determined by a "trial-and-error" approach using the "jar test".
At dosages that are commonly employed in the treatment of natural waters
and dilute wastewaters (for example, municipal sewage), aluminum, ferric, and
calcium salts have proved to be effective and economical for coagulation of
particulates and precipitation of certain organic and inorganic substances.
These coagulants, however, are not suitable for the treatment of concentrated
wastes because of the very high chemical dosages required and the production
of voluminous quantities of bulky sludges that must be separated, processed,
and disposed of. In addition, these coagulants are ineffective in removina a
broad spectrum of soluble organic and inorganic substances. From these stand-
points, chemical coagulation alone is not suitable for use in a mobile spill
control unit. The process, however, has been used in conjunction with filtra-
tion and carbon adsorption -in the EPA's Mobile Physjcal/Chemical Treatment
System.
MISCELLANEOUS PROCESSES
Miscellaneous processes that are reviewed are gravity separation, filtra-
tion, carbon adsorption, and thermal oxidation. The review of these processes
is ve>*y brief and is intended primarily to make the study more "complete".
These processes are commercially available and are widely used in municipal
and industrial waste treatment applications. With the exception of thermal
oxidation, these processes are also used in the EPA's Mobile Physical/Cnemical
Treatment System that has been employed in a number of spill control situa-
tions.
The EPA's Mobile Physical/Chemical Treatment System uses large rubber,
collapsable, portable tanks, which are set up next to the trailer, for cneiii-
cal coagulation and settling. The trailer houses three dual-media filters
-------
designed for a maximum filtration rate of 4.22 fc/sec (67 gpm) and three acti-
vated carbon columns containing 19.6 m3 (700 ft3) of carbon. The mobile unit
has been in operation for six years, and considerable operating experience
exists in actual applications to spill situations.
EPA has currently contracted out the work for design, construction, and
demonstration of a mobile incineration system (Environmental Restoration In-
cinerator Complex, ERIC) for the destruction of residuals from hazardous
material spill clean-up operations. The system, which will be mounted on three
trailers, will use a rotary kiln incinerator equipped with afterburner, quench
tower, and scrubber.
-------
SECTION 2
CONCLUSIONS
Based on the engineering evaluation described in this report, the follow-
ing general conclusions can be offered on the suitability of reverse osmosis
(RO), ultrafiltration (UF), ion exchange, wet air oxidation, biooxidation,
ozonation/ultraviolet radiation (03/UV), and precipitation/coagulation pro-
cesses for use in mobile units for on-site treatment and detoxification of
spill-impacted waters.
• RO and ion exchange processes appear to meet many of the requirements
for applicability of a process to the treatment of waste spills in a
mobile unit. These processes offer low cost compact units, are com-
mercially available, can be started and shut down relatively quickly,
can be serviced conveniently, would not require skilled operating
field labor, can be operated with electricity produced by on-board
generators, can handle a spectrum of wastes including those containing
high concentration of toxic substances and refractory organics, and
produce a relatively small volume of waste residue requiring disposal.
Both these processes have been used in mobile units for water and
wastewater treatability studies.
0 Wet air oxidation is particularly applicable to the destruction of
refractory and toxic organics in concentrated wastes encountered in
spill situations. The major limitations of the process are high
capital cost, small capacity of trailer mounted units and the require-
ment for skilled operating personnel.
t While meeting many of the mobile unit application requirements, UF
and 03/UV suffer from the limitations of inapplicability to waste
containing low molecular weight substances or highly concentrated
large volume wastes, respectively. There is apparently no experience
on the use of these processes in mobile units.
• High purity oxyaen activated sludge process appears to be the pre-
ferred biooxidation system for possible use in mobile units. Bio-
oxidation processes, in general, suffer from the following limitations
that would make them unsuitable to be used alone in spill control si-
tuations: inapplicability to wastes containing high concentration of
toxic substances and refractory organics; requirement for long deten-
tion time (and hence a large reactor size) for processing concentrated
wastes; and considerable time required for process start-up and sta-
bilization.
-------
• Aluminum, ferric, and calcium salts, at dosages that are commonly
employed in the chemical treatment of raw waters and dilute waste-
waters would not be effective in handling concentrated wastes. Use
of much higher dosages that would be needed for concentrated wastes,
results in the production of voluminous quantities of bulky sludges
that are difficult to process for disposal.
• In general, very limited data are available on the performance of
the candidate processes for use in mobile units in handling very con-
centrated wastes.
t Although some of the processes considered (for example, UF and 03/UV)
would be inapplicable to the treatment of certain spills (for example,
wastes containing low molecular weight substances or large-volume
concentrated wastes), they would be applicable to other types of
spills (for example, dilute wastewaters or wastes containing large
molecular weight substances).
• When used alone, none of the processes considered would be economical-
ly applicable to the treatment of the large variety of wastes en-
countered in spill situations. The applicability of these processes
would be enhanced and the treatment costs would be reduced, however,
if these processes are used in combination in a treatment train. The
optimum process combination would vary with the spill type. The
specific process combinations that would be applicable to the greater
variety of spills need to be evaluated for possible deployment in
mobile units.
-------
SECTION 3
INTRODUCTION
BACKGROUND AND STUDY OBJECTIVES
Even with the most comprehensive spill prevention programs, accidental
spills of hazardous substances cannot be totally eliminated in an industrial
society that handles and transports voluminous quantities of industrial chemi-
cals and process wastes every day. Since spills generally occur at "unexpected"
times and places, appropriate spill contaminant measures are seldom available
or seldom can be delivered to the spill site on time to prevent spilled material
from reaching nearby watercourses (ditches, ponds, streams, lakes, etc.). When
a watercourse becomes contaminated, countermeasures including containment and
treatment of the spill-impacted waters must be implemented as soon as practical
to prevent further spread of pollution and to restore the watercourse to its
pre-spill condition. A similar need for containment and on-site treatment of
wastes and wastewaters has recently come to light with the discovery of sites
where hazardous wastes have been improperly disposed of. The more publicized
examples of such incidents have been the Pollution Abatement Services operation
at Oswega, NY, the chemical disposal operation on the Albert Harris prooerty
at Ditmer, MO, and the former Hooker Chemical Corporation dump site at Love
Canal at Niagara Falls, NY.
The need for emergency response capabilities for the treatment of leachate
and wastes at hazardous waste dump sites and of spill-impacted waters at or
near spill and dump sites has led the U.S. Environmental Protection Agency to
carry out a number of programs aimed at the develooment of water and wastewater
treatment technologies and associated hardware for such applications. These
programs have led to the development and successful deployment of a Mobile
Physical/Chemical Treatment System (employing chemical coagulation/floccula-
tion, filtration, and activated carbon adsorption) and an on-going effort in-
volving design, construction, and demonstration of a Mobile Environmental
Restoration Incinerator Complex (ERIC).
The study described in this report, which has been carried out for the
EPA, has had as its objective the engineering evaluation of a number of com-
mercially available water and wastewater treatment orocesses for possible use
in mobile units for on-site treatment and detoxification of spill-imoactea
waters containing high concentration of contaminants. The processes evaluated
are as follows: reverse osmosis, ultrafiltration, ion exchange, wet air ox^da-
tion, biological treatment, combined ozonation/ultraviolet radiation, and
coagulation/flocculation. To make the study more "complete", gravity separa-
tion, filtration, activated carbon adsorotion, and incineration, which have
10
-------
been used in or are under development for spill control applications, have
also been briefly reviewed. The engineering evaluation of the processes con-
sidered has been based on the information available in the published litera-
ture and the data obtained from process and equipment developers and vendors.
GENERAL REQUIREMENTS FOR USE OF A PROCESS IN A MOBILE UNIT
The characteristics listed below have been used as the general guidelines
in the evaluation of the applicability of various processes to the treatment
of concentrated waste spills. The mobile wastewater treatment system must:
(1) Be relatively light weight and small in size*. There are strict
highway regulations that limit the size and weight of motor vehicles.
The regulations vary from state to state. In California, vehicle
dimensions are limited to the following: 12.2 m long, 2.4 m wide
and 4.1 m high (40 ft by 8 ft by 13.5 ft). The weight limitation
depends on the distance between the extreme axles and the number of
axles on the vehicle. For a five-axle vehicle with a distance of
11 m (36 ft) between the first and last axles, the gross weight is
limited to 29,900 Kg (66,000 Ib). Although certain components of
the system can be mounted on a separate vehicle, it would be pre-
'ferred to have the entire unit (including auxiliary equipment) con-
fined to a single unit to achieve greater operational flexibility
• • and reduce cost and-labor requirements. -
(2) Be capable of self-sustained operation; that is, the unit should have
an on-board power supply.
(3) Be capable of handling wastes having a range of physical and chemical
properties as would be the case for a spectrum of chemicals that are
encountered in different spill situations.
(4) Be capable of quick startup and shutdown and have minimum require-
ments for skilled labor.
(5) Be suitable for periodic operation and should not present special
storage requirements.
(6) Present no, or minimum of, intermedia oollution transfer; any residue
that is generated should be relatively small in volume and inoccuous.
(7) Be commercially available or be easily fabricated.
(8) Have reasonable capital and operating costs.
A treatment capacity of 227,000 £/day (60,000 gpd) has been assumed in this
study for the purpose of unit sizing and cost estimating.
-------
ORGANIZATION OF THE REPORT
A separate chapter has been devoted in this report to each of the
-------
SECTION 4
REVERSE OSMOSIS
PROCESS DESCRIPTION
Process Principle
Reverse osmosi-s (RO) is a unit,operation for producing waters low in dis-
solved solids from brackish waters and wastewaters. The process involves use
of a semi permeable membrane that allows the passage of solvent (water) mole-
cules but notj.the 'soluble molecules (both organic and inorganic compounds).
When a dilute and a concentrated solution are separated by a semi permeable
membrane, there is a spontaneous transport of solvent from the dilute to the
concentrated side of the membrane. The driving force for this phenomenon,
referred to as osmosis, is the difference in the solvent vapor pressure on the
two sides of the membrane. Equilibrium is reached between the two solutions
at a certain pressure called the osmotic pressure, at which point the amount
of solvent that passes in each direction throuah the semipermeable membrane is
equal. The magnitude of the osmotic pressure is given by the Van't Hoff equa-
ti on:
w = nRT (4.1)
where v = osmotic pressure, atm
n = solute concentration, moles/2*
R = gas constant, 0.083 atii.A/moles-°K
T = temperature, °K
By applying enough pressure to the concentrated solution to overcome the
osmotic pressure, the flow of the solvent molecules can be reversed (that is,
from the concentrated solution compartment to the dilute solution compartment).
In water/wastewater treatment applications, this would result in the production
of a procuct water (permeate) low in dissolved solids and a small volume ot
concentrated waste (reject) that can be sent to material recovery or further
treatment/disposal.
The performance of a reverse osmosis system is determined by two para-
meters, the solvent (for example, water) flux and solute (for example, salt)
*
At high solute concentrations, use of solute activity in place of solute con-
centration would yield a more accurate value of osmotic pressure.
13
-------
permeability. The solvent flux, which determines the amount of water recovered,
is given by the following equation:
J = K (AP - ATT) (4. 2)
where J = solvent flux, g/cm2.sec
K = water permeability coefficient, g/cm^-sec-atm (value dependent
on membrane characteristics and temperature)
AP = P] - ?2> P! = applied pressure, atm
?2 - pressure on downstream side of membrane, atm
ATT = ir-| - TTJ; if] = osmotic pressure of the solution at the membrane
interface, atm
iT2 = osmotic pressure of permeate, atm
In many cases, ir2 can De ignored and in tubular systems P? is atmospheric
pressure and can also be ignored. The flux decreases with the increase in AT,
which in turn decreases with the lowering of solute concentration.
Even with the "best" systems, some solutes pass through the membrane. The
amount of solute that passes through a unit area of membrane (that is, the
solute flux) is independent of the applied pressure as indicated by the follow-
ing equation:
F = 3 (CH - CL) (4.3)
where F = solute flux,
e = solute permeability coefficient, cm/sec
CH = concentration of solute on high pressure side of membrane, g
C|_ = concentration of solute on low pressure side of membrane, g/CTi3
The solute flux affects the quality of the product water. Table 1 is a summary
of the performance of RO for various chemical substances. In general, good
removal can be expected for high molecular weight organics (for example,
protein) and charged anions and cations. The degree of rejection (a) increases
with the degree of ionization and the size of- the charged ions and neutral
molecules, (b) decreases with the hydrogen bonding tendency with the membrane,
polarities of the molecules, and (c) is affected by molecular geometry and
membrane characteristics (2,3).
Equations 4.2 and 4.3 show that while the solvent flux (J) increases with
the increase in the applied pressure, the solute flux (F) is independent of
the pressure. Thus, the higher the pressure the better is the quality and
quantity of the water recovered. However, when too high a pressure is applied,
the porous structure of the membrane is compressed, causing a decrease in flux.
Optimum operating pressure for most membrane modules is around 2.5 to 4.2 MPa
(350 to 600 psi).
Types of Membrane Modules
Membranes are usually fabricated in flat sheets or tubular fons and are
assembled into modules. The most common materials used are cellulose acetate
H
-------
TABLE 1. PERCENT REJECTION OF SELECTED CATIONS, ANIONS, AND ORGANICS (1)*
Salts
Cationst
Name
Sodium
Calcium
Magnesium
Potassium
Iron (Fe+2)
Manganese (Mn+2)
Aluminum
Ammonium
Copper (Cu+2)
Nickel (Ni+2)
Strontium
Hardness
Cadmium
Silver
Percent
rejection
94-96
96-98
96-98
94-96
98-99
98-99
99
88-95
96-99
97-99
96-99
96-98
95-98
94-96
Anions^
Name
Chloride
Bicarbonate
Sulfate
Nitrate
Fluoride
Silicate
Phosphate
Bromide
Borate
Chromate
Cyanide
Sulfite
Thiosulfate
Ferrocyanide
Percent
rejection
94-95
95-96
89
93-96
94-96
95-97
99-
94-96
35-70
90-98
90-95
98-99
99
99
Organics5
Maine
Sucrose sugar
Lactose sugar
Protein (10,000+ M.W.)
Glucose
Phenol
Acetic acid
Lactic acid
Dyes (400 to 900 M.W.)
Biochemical oxygen
demand (BOD)
Chemical oxygen
demand (COD)
Bacteria and virus
Pyrogen
Percent
rejection
100
100
100
99.9
—
—
—
100
90-99
80-95
100
100
Letter from I. Nusbaum, Fluid Systems Division, UOP, San Diego, CA, January 22, 1980.
Aluminum, copper, nickel, and strontium may precipitate; in the presence of sulfate, calcium,
boriiim, and strontium may be limiting; ammonium rejection is pH-dependent.
^Fluoride, sulfate, sulfite, and silicate may precipitate; rejection of some anions (for example,
borate and cyanide) is pH-dependent.
Rejection of certain ionizable organics (for example, phenol, acetic acid, and lactic acid) is
extremely pH-dependent; the data for phenol is characteristic of cellulose acetate membrane and
not polyamide membrane; sucrose and lactose are not 100 percent rejected by conventional mem-
branes and, operating at high concentrations, require elevated temperatures since viscosity
would be a factor; COD and BOD rejections affected by the presence of low molecular weight
organics; periodic disinfection of system would be required to maintain high rejections of bac-
teria, viruses, and pyrogens.
-------
and other polymers such as polyamides. There are three basic module designs,
tubular, hollow fiber, and spiral wound (see Figure 1). Each type of membrane
module has its own advantages and limitations. The tubular module provides
the largest flow channel and allows for turbulent fluid flow reaime; thus, it
is least susceptible to plugging caused by suspended solids and has the highest
flux. However, because of its small area/volume ratio the total product
recovered per module is small. The cost of a tubular module is approximately
five times that for the other modules for an equivalent rate of water recovery
(4), and the total space requirement is about three to five times that for the
spiral wound system (5).
A hollow fiber membrane is constructed of polyamide polymers and cellulose
triacetate by Dupont and Dow, respectively*. Polyamide membrane permits a
wider operating pH range than cellulose acetate, which is commonly used for tKe
construction of spiral wound and tubular membranes (see Table 2). Each fiber
is about the size of a human hair. The flow channel and the flux are about an
order of magnitude lower than the other configurations. This small flux,
however, is compensated for by the large surface area/volume ratio, with the
total product water per module being close to that obtainable with spiral wound
modules. However, because of the small size of the channels (about 0.1 mm or
0.004 in.) and the laminar fluid flow regime within the channels, this modu1e
is most susceptible to plugging and may require extensive pretreatment to pro-
tect the membrane.
The spiral wound module consists of an envelope of flat sheet membranes
rolled around a permeate collector tube (see Figure 1). This configuration
provides for a higher flux (Table 1} and greater resistance to fouling than
the hollow fiber modules; it is also less expensive and occupies less space
than a tubular module.
PROCESS APPLICATIONS AND RELATED STUDIES
Commercial Scale Municipal/Industrial Applications
RO has played and continues to play an important role in desalination of
brackish waters for municipal and industrial uses. In 1975, there were nearly
300 large (>95,000 a/day) plants in operation around the world producing a
total of 170 million liter per day (45 mgd) of purified water (6). Some data
on these applications are summarized in Table 3.
RO has been used in full scale applications in food processing (7), in
industrial wastewater treatment (8,9) (for example, elating rinses, cooling
tower blowdown, pulp and paper spent liquor, acid mine drainage, petroleum
stripping water), and in the treatment of agricultural wastewaters (10). There
are more than 70 RO units in operation in the metal finishing and plating in-
dustry (4). There is little information on the performance of the RO units in
these and other industrial wastewater treatment applications.
Letter from I. Nusbaum, Fluid Systems Division, UOP, San Dieao, CA, January
22, 1980.
-------
CASING^
MEMBRANE
WATER
FLOW
a. TUBULAR MEMBRANE
ROLL TO
ASSEMBLE
FEED SIDE
SPACER
FEED FLOW
X
PERMEATE FLOW
(AFTER PASSAGE
THROUGH MEMBRANE)
PERMEATE OUT
PERMEATE SIDE BACKING
MATERIAL WITH MEMBRANE ON "^
EACH SIDE AND GLUED AROUND
EDGES AND TO CENTER TUBE
b. SPIRAL WOUND MODULE
O RING
SEAL
CONCENTRATE FLQW
OUTLET SCREEN
OPEN END
OF FIBERS EPOXY
TUBE SHEET
POROUS
BACK UP DISC
END PLATE
\
FIBER
.SNAP RING
PERMEATE
SHELL POROUS FEED v ,
DISTRIBUTOR 0 RING END PLATE
TUBE SEAL
c HOLLOW FIBER MODULE
Figure 1. Membrane module configurations (4),
17
-------
TABLE 2. COMPARISON OF REVERSE OSMOSIS MODULE CONFIGURATIONS (1)*
oo
Membrane surface area per
volume, ft2/ft3
Product water flux, gpd/ft^
Typical module factors:
Brine velocity, ft/sec
Brine channel diameter, in
Spiral wrap
„ 100 - 300
8 - 251-
f
0.031
Tubular Hollow fine fiber
40 - 100 5,000 - 10,000
8-25 0.1-2
1.5 0.04
0.5 0.004
Method of membrane replacement
Membrane replacement labor
High pressure limitation
Pressure drop, product water side
Pressure drop, feed to brine exit
Concentration polarization problem
Membrane cleaning - mechanical
- chemical
Particulate in feed
Permissible feed ranges, pll**
Permissible temperature, °F**
As a membrane module
assembly - on site
Low
Membrane compaction
Medium
Medium
Medium
No
Yes - pH and solvent
limited
- Some filtration
required
As tubes - on site
High
Membrane compaction
Low
High
High
Yes
Yes - pH and solvent
limited
No problem
As entire pressure mo-
dule - on site, module
returned to factory
Medium - requires
equipment
Fiber collapse
High
Low
High
No
Yes - less restricted
Filtration required
Letter from I. Nusbaum, Fluid Systems Division, UOP, San Diego, CA, January 22, 1980.
Product flux varies with the net driving pressure and temperature; a flux of 10-25 gpd/ft^ is typical
at a pressure of about 400 psi.
'It is difficult to define velocity in a spiral element since the space between membrane is filled with
a polypropylene screen which acts as a spacer and turbulent promoter.
Height of brine channel (not diameter).
A* i
Permissible pll and temperature ranges dependent primarily on membrane type and not on module configura-
tion; for example, polyannde hollow fine fiber is pll limited from 4 to 11, cellulose acetate from 3 to
7.5, thin film composite (TFC) spirals have been operated and cleaned at pH levels ranging from 1 to 12.
-------
TABLE 3. DESIGNS AND OPERATING FEATURES FOR SOME COMMERCIAL RO PLANTS TREATING HIGH TDS WATERS
Application
Municipal
(Florida)
Computer
manufacturer
(Colorado)
Boiler
feedwater
(Illinois)
Condensate
makeup
Capacity
106 SL/d (gpd)
1.9
(500,000)
0.19
(50,000)
1.8
(475,000)
0.22
(57,600)
Module
type
Hollow
fiber
Spiral
wound
Spiral
wound
TDS, ppm Comment
Raw water Product water
8,000 • . <500 System designed for
50 percent water
conversion, membrane
is cleaned after
every 870-1,700
hours of operation.
800 ' 25
550 50 Cost per 1,000 & of
product water is
$0.44 ($1.68/1,000
gal)
1,125 ' 217
Reference
11
12
13
14
-------
Pilot Laboratory Scale Studies
In addition to the actual process/wastewater treatment applications in
the industry, a number of engineering/pilot-plant studies have been conducted
or are currently in progress to assess the applicability of RO to the treat-
ment of municipal and other industrial wastewaters and in connection with the
development of improved membranes and RO systems. Some of the most pertinent
of these studies are briefly described below.
A pilot-plant evaluation of the desalinization of sea water (IDS of
35,000 ppm) at the San Diego "Sea World" demonstrated that at an applied pres-
sure of 7 mpa (1,000 psi) a product water with less than 300 ppm IDS can be
obtained (15). Projected costs for this application ranged from $0.43/1,000 a
($1.62/1,000 gal) for an 18.9 million £/day (5 mpg) plant to $1.53/1,000 a
($5.79/1,000 gal) for a 94,600 Ji/day (24,000 gpd) plant.
The use of RO for the demineralization of acid mine drainage to produce
low IDS waters suitable for domestic and industrial purposes has been demon-
strated by Fluid Systems (previously Gulf) (16,17) (see Table 4).
TABLE 4. USE OF RO FOR THE DEMINERALIZATION OF ACID MINE DRAINAGE (17)*
Parameters/constituents Value
Raw water Treated water
PH
Fe (total)
so4
Ca
Al
3.6
100
1,580
185
27
3.9
2
21
1.5
1
*
Except for pH, all values are in mg/1.
Laboratory and pilot studies on the removal of nutrients from raw and
secondary treated wastewaters have indicated the following removal efficien-
cies (18): phosphorus - 100 percent, ammonia - 85 percent, and nitrate and
nitrite - 56 percent*. A spiral wound module tested at the Los Angeles County
Sanitation District demonstrated the feasibility of using RO for tertiary
treatment of municipal wastewaters. In this demonstration effort, it was
*
By proper pH adjustment, ammonia removal of better than 95 percent can readi-
ly be obtained; new thin film composite (TFC) membranes attain N03 removal of
98 percent (letter from I. Nusbaum, Fluid Systems Division, UOP, Ssn Diego,
CA, January 22, 1980).
20
-------
found that membrane deposits contained large amounts of polyhydroxy aromatic
compounds and substances containing carboxylic acid groups. The flux could be
restored by periodic treatment with detergent and relaxation of the apolied
pressure (19).
The feasibility of the application of RO for concentrating soluble oil
coolants in a wastewater was demonstrated in a 2,650 fc/day (700 gpd) pilot
study (20). The data on the system performance after 3,000 hours of on-line
operation are summarized in Table 5. Since RO approaches the absolute filter,
the rejection of oil and suspended material is near perfect but fouling pro-
blems would essentially negate this application for anything except possibly
tubular system*.
TABLE-5. USE OF RO FOR,THE TREATMENT OF AN OILY WASTEWATER (20)
Constituents
TDS
TOC' *'
COD
BOD5 "- *'
Phenol
Oil
Concentrations, pom
Feed
1,404
2,000'
6,900
9,000 -
1.32
2,400
Treated water
39
50° *. '
150
82
0.72
6
Percent
removal
97.2
97.5
97.8
90.9
45.5
99.7
Very few studies appear to have been conducted to evaluate the effective-
ness of RO for the removal of specific organic compounds. In a laboratory
study, RO was shown to be capable of removing better than 99 percent of 15
major pesticides (7 chlorinated hydrocarbons, 4 organophosphorus, and 4 other
types of pesticides) for wastewaters containing pesticide concentrations ran-
ging from 40 ppb to 1.5 ppm (21). Part of the removal was attributed to the
adsorption of pesticides on the membrane.
In another laboratory study, an average rejection of 12 to 80 percent was
found for 13 low molecular weight polar organic compounds, includina methanol,
ethanol, isopropanol, acetic acid, formaldehyde, acetone, ethyl ether, urea,
glycerol, phenol, hydroquinone, aniline, and methyl acetate (22). Synthetic
solutions having a TOC of 1,000 ppm were used as feed in these experiments.
The reported data on the effectiveness of RO for the removal of ohenols
are inconsistent. A 45 percent removal of phenol was observed in one stjdy
*
Letter from I. Nusbaum, Fluid Systems Division, UOP, San Diego, CA, January
22, 1980.
21
-------
(20) whereas a "negative" removal was reported in another (6). Phenol rejec
tion is a function of pH and membrane type.
Mobile Unit Applications
Use of RO in mobile systems for the treatment of brackish waters has been
demonstrated in a number of applications. Some of these applications are
briefly discussed below.
In 1969, Envirogenics (El Monte, CA) used a 189,000 Vday (50,000 gpd)
trailer-mounted "plate-and-frame"* RO unit for the treatment of brackish
waters (23). In 1978 the company built another trailer-mounted RO unit usina
the spiral wound design. The unit could be operated on either a 50 or 60 cycle
external power source or from a diesel generator (24). Recently, the U.S. Army
has constructed a 54,500 fc/day (14,400 gpd) mobile unit to provide potable
water in the field from any fresh, brackish, or sea water source (25).
The U.S. Army has also tested a lightweight, hand-operated tubular RO
unit with the following characteristics (26): size - 0.0085 m3 (0.3 ft3), dry
weight - 4.7 kg (10.3 lb), wet weight - 5.7 kg (12.6 Ib), design capacity - 500
fc/day (135 gpd), and design performance - 70 percent product water recovery and
90 percent salt rejection.
USE OF RO IN MOBILE UNITS FOR TREATMENT OF WASTE SPILLS
This section evaluates the feasibility of using RO in a mobile unit to
treat concentrated waste spills.
Pertinent Features of RO for Use in Mobile Units
In general, nearly all the requirements discussed in Section 3 for appli-
cability of a systems to the treatment of waste spills in a mobile unit would
be met by RO. RO units are compact, are commercially available, can be started
and shut down relatively quickly, can be serviced conveniently, would generally
produce only a small volume of residue, would not require skilled field labor
for operation, and can be operated with the electric power produced on-site by
a diesel generator. The major shortcomings of the RO unit relate to possible
membrane fouling by suspended solids; membrane degradation due to very low or
high pH levels (depending on membrane type); presence of strong oxidizing
chemicals, high levels of phenol and phenolic substances, and organic solvents;
and generation of a reject requiring ultimate disposal.
As discussed earlier in this section, RO is a commercial process, is
offered by a number of manufacturers"*", and has been used in mobile water treat-
*
The "plate-and-frame" units represent the early stage of RO development and
have been abandoned in favor of newer membrane designs (spiral wound, etc.).
Partial list of RO manufacturers and suppliers: Aqua Media, Sunnyvale, CA;
Dow, Walnut Creek, CA; DuPont, Wilmington, DE; Envirogenics, El Monte. CA;
Hydronautics, Santa Barbara, CA; Permutit, Paramus, NO; Polymetrics, San
Jose, CA; UOP, San Diego, CA.
22
-------
ment applications. The unit is fairly compact; a 5,680 £/day (1,500 gpd)
product water capacity unit using the spiral wound module design would weight
about 36 kg (8 Ib) and would be approximately 10 cm (4 in) in diameter and
100 cm (40 in) in length. The startup and shutdown of RO consist primarily
of turning on-and-off of the high pressure pump and adjusting the pressure and
flow control valves. RO modules are connected in series or in parallel with
provisions for flow by-pass so that a single module can be by-passed for
maintenance service. Operation of RO units does not require highly skilled
labor; the operating effort would be limited to turning valves, reading gauges,
and collecting samples for chemical analysis. In mobile unit applications, the
system would be equipped with on-line monitors for continuous measurement of
pH, temperature, pressure, etc. Ordinarily no more than 6 to 12 operator-hours
would be required for a 24-hour operation of an RO unit. For safety reasons,
however, operation of a mobile unit in the field may require a minimum of 2
persons. The power for the operation of an RO unit can be provided by a diesel
electric generator that can be housed in the same mobile unit. Approximately
2.6 kw of electricity would be required to produce 1,000 a of product water
(10 kw/1,000 gal).
As with any wastewater treatment system, RO produces a residue (reject)
that requires disposal. The volume of this reject (waste concentrate) is
•usually about"10 t'o-25 percenteof the volume of feed water. In field apolica-
tions, the reject may be collected in drums and subsequently transported to
hazardous waste management facilities or municipal sewage treatment plants for
further treatment/disposal.
As with all membrane processes, RO units are susceptible to membrane
fouling and deterioration. Waste spills would generally contain suspended
solids/debris and may also contain chemicals that can attack the membrane
materials. Suspended solids in the waste can be removed by pretreatment using
mixed-media filters followed by cartridge filtration (for example, through 10-
25 \i filters). The growth of slimy biomass on the membrane surface or the pre-
sence of organic macromolecules in the waste may also foul up the membrane.
This organic fouling can be minimized by pre-chlorination, addition of bio-
cides, and/or pretreatment with carbon adsorption. Membrane fouling can also
result from precipitate (scale) formation as the waste becomes concentrated.
The precipitates that would be most commonly encountered are those of calcium
salts (primarily carbonate and sulfate). Scale formation can be minimized by
adjusting the pH and/or the use of "scale inhibitors" such as polyphosphates.
Choice of Membrane Configuration
There are three basic membrane module configurations (designs). Although
the general performance (with respect to salt rejection) is similar for the
three membrane designs, there are certain design features and membrane proper-
ties that make one design superior to others under certain conditions (for
example, pH or presence of suspended solids) and inferior under other condi-
tions (see Table 2 for comparison of RO module configurations).
In genera1, the hollow fiber modules with polyamide membranes can be
operated over a wider pH range than the tubular or the spiral wound modules
that commonly use cellulose acetate membranes. At relatively low pH levels,
23
-------
the cellulose acetate is susceptible to degradation via hydrolysis, l/aste-
water having too high or too low pH values would require pH adjustment by
addition of acid or base prior to treatment in RO units employing spiral wound
or tubular module designs.
Hollow fiber modules with polyamide membranes are more susceptible to
chlorine attack than the soiral wound and the tubular modules. A chlorine
concentration of as low as 0.1 ppm can cause degradation of the hollow fiber
membranes. Spiral wound and tubular membranes can tolerate up to 1 ppm of
free chlorine (28).
Having much narrower passages for wastewater flow, the hollow fiber con-
figuration is considerably more susceptible to fouling and hence, less de-
sirable for waste spill treatment application than tubular or spiral wound
designs. Membrane fouling can result from the presence of suspended solids in
the waste. The solids may be initially present in the wastewater or may be
produced during treatment as a result of chemical precipitation (for example,
of calcium sulfate or carbonate) brought about as the waste becomes more con-
centrated. Although suspended solids can be removed by prior filtration and
the precipitate formation can be minimized by use of chemical additives, the
requirement for filtration or chemical addition would be less for the tubular
and spiral wound designs than for the hollow fiber design.
Because of its relatively smaller membrane surface area to module volume
ratio, a tubular module would occupy a larger space (approximately 4 times as
much) than the hollow fiber or the spiral wound modules for the same water
throughput. This would constitute a disadvantage for the tubular module for
use in a mobile unit that would require a small equipment volume. Another
disadvantage of the tubular module is its relatively high cost. For the same
throughput capacity, the capital cost for a tubular design is about five times
as much as that for a spiral wound design.
Table 6 summarizes the relative advantages and disadvantages of the three
modules for use in spills treatment mobile units. Even though the tubular
design would be less susceptible to fouling, it would probably be less suitable
for use in mobile units because of its larger size (and higher cost) than the
spiral wound or the hollow fiber designs. The spiral wound appears to be more
desirable than the hollow fiber design primarily because of its lower suscep-
tibility to fouling that significantly reduces the pretreatment requirements.
Accordingly, in the following section, only the spiral wound design has been
evaluated and costed for use in mobile treatment units.
Engineering Evaluation and Cost Estimates
Based on data obtained from one RO manufacturer*, a preliminary design
has been developed for a 227,000 A/day (60,000 gpd) spiral wound RO unit for
use on a trailer flatbed. Figure 2 presents the flow diagram for the system
*
Telephone communication with Frank Trabert, Envirogen cs Company, El Monte,
CA, January 19, 1979.
24
-------
TABLE 6. COMPARISON OF THE THREE RO MODULES FOR USE
IN SPILLS TREATMENT MOBILE UNITS
Parameter Hollow fiber Tubular Spiral wound
Broad pH range of applica-
bility*
Relative susceptibility to
fouling by suspended solids
Relative unit size (4)
Relative cost (5)
Yes
High
1
1
No
Low
4
5
No
Medium
1
1
As noted previously, pH range of applicability is dependent on membrane type
and not on module design; spiral wound TFC would have the broadest pH range
of applicability.
based on design recommendations by an RO manufacturer*. The system, which
consists of ten pressure vessels arranged in four stages, can be mounted on
two heavy skids with five vessels on each skid. Each vessel contains six mo-
dules. Excluding the equipment for pretreatment, the system would have dimen-
sions of 8 m long, 1 m wide, and 2.6 m high (24 ft x 3 ft x 8 ft), would weigh
about 2,720 kg (6,000 Ib) and would occupy 20.8 m3 (576 ft3) of space. The
water recovery is estimated at 90 percent. Other components of the system
(not all of which are shown in the figure) include a high pressure pump, pre-
treatment equipment (cartridge filters, carbon adsorption, chemical feed sys-
tem), valves and piping, pressure valves, instrument panel, and the power
supply source. Power can be supplied to the unit using a diesel electric
generator. The generator, a 30 kw unit, would have the approximate dimensions
of 2 m (1) x 0.8 m (w) x 1.1 m (h), 6 ft x 2 ft x 3.5 ft, would weight aporoxi-
mately 900 kg (2,000 Ib), and would occupy a volume of 1.8 m3 (42 ft3). A
storage tank having a volume of about 1.9 m3 (500 gal) would suoply the fuel
to the unit.
The major logistics support required in field applications relate to
possible requirements for membrane replacement and sufficient supply of treat-
ment chemicals (acid, base, scale inhibitors, etc.). Under normal operation,
it is unlikely that membrane replacement would be required during short-tern
field operation. In application to brackish water treatment, a membrane life
of 2 to 3 years has been reported for continuous operation. Lower membrane
life would be expected in application to concentrated waste spills. Assuming
a requirement for replacement of 50 percent of the modules during a field
application, the volume of the spare modules which may have to be carried on-
board is about 0.24 m3 (8.7 ft3).
*
Telephone communication with Frank Trabert, Envirogenics Company, El Monte,
CA, January 19, 1979.
25
-------
STAGE I
STAGE II
STAGE III
STAGE IV
ro
cr>
WASTE STREAM
(10% OF RAW
WATER FLOW)
PRODUCT WATER STREAM
(90% OF RAW WATER FLOW)
Figure 2. 4-stage module arrangement for a 227,000 £/day (60,000 gal/
day) mobile RO treatment system.
-------
Chemical requirements are determined by the characteristics of a specific
waste spill. The quantity of the chemicals which may have to be carried on-
board for short-term field operation would be relatively small*. For extended
field operation, the mobile unit may have to be resupplied with chemicals on
an as necessary basis.
The estimated total capital costs for the RO unit1" and the diesel electric
generator^ are $60,000 and $10,000, respectively. The cost for the RO unit
does not include costs for pretreatment equipment. It does, however, include
costs for RO modules, piping, valves, high pressure Rump, chemical storage tank,
chemical feed pumps, and the instrumentation panel. The cost also covers the
engineering and installation fee.
The operating cost for the mobile RO unit per unit volume of waste treated
should vary depending on the specific application (for example, waste charac-
teristics, waste volume chemicals requirement, location of the spill, etc.).
Assuming that in most field applications the-mobile unit will have to be
powered by a diesel electric generator, the fuel for the generator would con-
stitute one of the major items of operating cost. A 30 kw diesel generator
would consume about 11 Ji/hr (3 gal/hr) of fuel (about $2/hr)=J:.
REFERENCES-
t
1. Bureau of Reclamation and Office of Saline Water, Desalting Handbook for
Planners, May 1972.
2. Duvel, W.A., Jr. & T. Helfgott. Removal of Wastewater Organics by Reverse
Osmosis, Journal of Water Pollution Control Federation, Vol. 47, No. 1,
p. 57, Jan. 1975.
3. Weber, W.J., Jr. Physicochemical Processes for Water Quality Control,
Wiley-Interscience, New York, 1972.
4. McNulty, K.J., D.6. Grant, et al. Treatment of Metal Finishing Wastes by
Reverse Osmosis, AIChE Symposium Series, No. 166, Vol. 73; Water-1976:
I. Physical, Chemical Wastewater Treatment; p. 176, 1977.
5. Mason, D.G. Engineering Evaluation of Reverse Osmosis and Ion Exchange:
Demineralization for use in a Self-Sustained, Air-Transportable Waste-
water Renovation Unit. Interim Technical Report, AFSC Report AD-840
198/6ST, 1968.
Assuming that a dose of 100 ppm HC1 would be used for pH adjustment, the acid
requirement would only be 100 g/1,000 a (0.8 lb/1,000 gal). Ordinarily, much
lower quantities of other chemicals (for example, scale inhibitors) would be
required because of the much lower dosages used.
telephone communication with Frank Trabert, Envirogenics Company, El Monte,
CA, January 19, 1979.
O.
Telephone communication with L. Anderson, Anderson-Belviar Company, Long
Beach, CA, January 31, 1979.
27
-------
6. Arthur D. Little. Physical, Chemical, and Biological Treatment Techniques
for Industrial Wastes. PB-275-287, EPA Office of Solid Waste Management
Programs, 1976.
7. Water Purification Associates. Innovative Technologies for Water Pollu-
tion Abatement, PB-247-390, 1976.
8. Witmer, F.E. Reusing Wastewater by Desalination, Environmental Science
and Technology, Vol. 7, No. 314, 1973.
9. Kremen, S.S. Reverse Osmosis Makes High Quality Water Now, Environmental
Science and Technology, Vol. 9, No. 314, 1975.
10. Blanton, M. California's Water Reclamation and Desalination Projects,
Water and Sewage Works, Vol. 124, No. 6, 1977, p. 60.
11. Mclntosh, A.L. Florida Reverse Osmosis Plant Turns Out Potable Water,
Water and Sewage Works, Vol. 124, No. 12, 1977, p. 30.
12. Cruver, J.E. Reverse Osmosis - Where It Stands Today, Water and Sewage
Works, Vol. 120, No. 10, 1973, p. 74.
13. Loeb, M.8. and R. Schuler. RO Units Cut Cost of Feedwater Treatment,
Power Engineering, December 1978, p. 57.
14. Wadlington, M. Chemical Regenerant Savings Can Pay for a Reverse Osmosis
Unit, Industrial Water Engineering, Vol. 13, No. 3, 1976, p. 17.
15. Cruver, J.E. and J.H. Sleigh. Reverse Osmosis - The Emerging Answer to
Seawater Desalination, Industrial Water Engineering, Vol. 13, No. 3, 1976,
p. 8.
16. Kremer, S.S., A.B. Reidinger, et al. Reverse Osmosis Field Testing on
Acid Mine Waters at Norton, West Virginia, Research and Development Pro-
gress 'Report No. 586, U.S. Department of Interior, 1970.
17. Gulf Environmental Systems Company, Acid Mine Waste Treatment Using
Reverse Osmosis, Water Pollution Control Research Series 140/ODYG, August
1971.
18. Lim, H.S. and H.K. Johnston. Reverse Osmosis as an Advanced Treatment
Process, Journal of Water Pollution Control Federation, Vol. 48, No. 7,
p. 1804, July 1976.
19. Cruver, J.E. and I. Nusbaum. Application of Reverse Osmosis to Wastewater
Treatment, Journal of Water Pollution Control Federation, Vol. ^6, '\o. 2,
Feb. 1974, p. 301.
20. Markind, J., J.S. Neri, et al. Use of Reverse Osmosis for Concent^a-
ting Waste-Soluble 0.1 Coolants, Water - 1975, AIChE Symposium Series,
No. 151, Vol. 71, p. 70, 1976.
28
-------
21. Chian, E.S.K., W.N. Bruce, et al. Removal of Pesticides by Reverse Osmosis,
Environmental Science & Technology, Vol. 9, No. T, p. 52, Jan. 1975.
22. Fang, H.H.P. and E.S.K. Chian. Reverse Osmosis Separation of Polar
Organic Compounds in Aqueous Solution, Environmental Science and Technolo-
gy, Vol. 10, No. 4, p. 364, April 1976.
23. Watson, Z.K. Design and Fabrication of a 50,000 GPD Portable Reverse
Osmosis Pilot Plant, Research and Develooment Progress Report 431, Depart-
ment of Interior, 1969.
24. Envirogenics Systems Company, Newsletter, Vol. 2, No. 1, May 1978.
25. Schmitt, R.P. and R.P. Carnahan. U.S. Army Mobility Equipment Research
and Development Command, Environmental Pollution Abatement Program, In-
dustrial Water Engineering, Vol. 13, No. 1, p. 25, 1976.
26. Williams, R.H., D.R. Fort, et al. Lightweight, Hand-Operated Brackish
Water Purifier, AD-782119, 1974.
29
-------
SECTION 5
ULTRAFILTRATION
PROCESS DESCRIPTION
Process Principle
Ultrafiltration (UF) is a pressure-driven separation process that employs
semi permeable membranes operating under dynamic flow conditions. The process
is especially suitable for concentrating dilute products or recovering water
or certain chemicals from waste streams. Figure 3 compares the particle sepa-
ration capability of UF with those of reverse osmosis, microporous filters,
and conventional filters. At the small-molecule/low flux end of the scale lies
the commercial cellulose acetate reverse osmosis membranes with the capability
of retaining hydrated sodium and chloride ions. Next on the scale is UF mem-
branes with pores that cover a size range of about 10-3 to 10-2 microns (10-
100A) with filtration fluxes of about 0.5 to 10 gallons/sq ft/day/psi of
pressure-driving force. Compared to the reverse osmosis process that operates
at a pressure of 2.6 to 4.2 mpa (350 to 600 psi), UF operates at much lower
pressure, usually around 0.18 to 1.1 mpa (25 to 1,250 psi). UF can operate at
such low pressures because unlike RO, which is affected by the osmotic pressure
of the solutes, the osmotic pressure of the materials retained by UF membranes
is so low that it does not significantly influence the solvent flux, and RO
membranes are nonporous diffusion barriers, while UF membranes are microporous
structures.
In actual UF operation, the degree of separation or the amount of water
recovered (flux) depends on the physical properties of the membranes such as
porosity and thickness, and system variables such as pressure, temperature,
feed velocity, and waste composition. Semipermeable membranes used in the UF
process are porous structures made of organic polymers such as cellulose ace-
tate or inorganic chemicals such as zirconium oxide. The membranes have an
extremely thin (0.1 to 1 y) with an effective pore size of about 0.002 y (20
A) or larger (2). Solute molecules or particles larger than the membrane
pore size would be retained by the membrane while the smaller molecules or
particles would penetrate the membrane with f"e solvent (water) and are col-
lected as permeates. In general, dissolved salts and small organic molecules
are not retained by UF membranes. The lower and upper molecular weight cut-
off limits are around 500 and 500,000, respectively (3). The solute molecules
or particles to be separated by the UF process should be at least one to two
orders of magnitude larger than the solvent molecules (2). With the advance-
ment in membrane technology, the membrane porosity can be "custom maae" for
specific molecular size separations.
30
-------
(0 004)
Red Blood Cells
Light Microscope
Small Bactaena
(i
c
1
I
«
Influenza |
102
10
1 0
0
5
5
3 10'
Tobacco £
Starch Molecule -
102
•2 v
Electron o ™ ,
Microscope | « 10
Glucose
Chloride Ion
Molecule
— 10^
Reverse
Osmosis
Membranes
X
€>
'
/ s
&
•X"
Microporous
Filters
/ /
///
y
Jllrafiltration
Membranes
Conver
Part
Fill
ft
itional /
icle /
ers /
/ s
/ s
>c
>y
/
105
10'
i
0)
101 S
CO
-------
Figure 4 presents the effects of operating parameters on the UF flux for
concentrating cheese whey. The solvent fux across the UF membrane surface is
given by the following equation:
J = KP/t (5.1)
dhere J = solvent flux, g/cm^ sec
K = membrane permeability coefficient, g/cm/sec/atm
P = pressure difference across the membrane, atm
t = membrane thickness, cm
As expected, in the case of pure water (Figure 4a) the flux varies nearly in
proportion to the applied pressure. For the "cheese whey", however, the flux
increases at low pressures and reaches a plateau at a pressure of about 1.5
kg/cm2. This phenomenon is caused by "concentration polarization" involving
formation of a solute "gel layer" at the membrane surface as the solution be-
comes more concentrated. This "gel layer" acts as a secondary membrane that
offers additional resistance to the solvent flux of the membrane system.
The effect of waste concentration, feed velocity, and temperature on the
ultrafiltration flux is illustrated in Figures 4b, 4c, and 4d, respectively.
As shown in Figure 4b, the flux declines initially with an increase in feed
concentration and levels off at higher solute concentrations. The initial
decline in the flux is caused by the concentration polarization effect men-
tioned above; the flux reaches a constant value when the gel layer reaches the
maximum width. Figure 4c indicates that the flux increases with the increase
in linear feed velocity, presumably because of the reduced polarization at
higher velocity. Figure 4d shows the effect of temperature on flux rates. An
increase in temperature generally increases the permeability coefficient (see
equation 5.1) and thus the flux.
Types of Membrane Modules
Membranes that can be designed for specific applications are usually
fabricated in flat sheets, plates, or tubular forms and are assembled into 110-
dules. As with the reverse osmosis process, common module configurations are
tubular, spiral wound, and hollow fibers. The advantages and disadvantaoes of
these modules configurations are discussed in Section A. Key features of some
commercial module designs are presented in Table 7.
PROCESS APPLICATIONS AND RELATED STUDIES
Commercial Scale Municipal/Industrial Applications
To date, nearly all commercial uses of the UF process have been in seoa-
rating and concentrating macromolecules from dilute solutions in industrial
applications. There are full scale UF operations in the food processing in-
dustry (to concentrate cheese whey), textile industry (to concentrate dye
rinses), and in metal cutting and related industries (to concentrate emulsified
oils). Recently, UF has also been used in the electronic industry for treating
plating rinses and as a post-treatment to ion exchange for raw water treatment.
32
-------
Average Pressure psi
10 20 30 40 50 60 70
1234
Average Pressure kg/cm'
a Presture Effect on Ultrattltration Flux
Whole Cheese Whey
Tubular UPON"
4 kg/cm' (57 psO
37m/sec(12ft;sec)
25 °C
30
20
10
0 1 2 3 4 5 6 7 8 9 10 11 12 13
°'o ProtPin n Feed
b Effect of Concentration on Ultraflltration Flux
40
e30
V
x
,? 20
10
Linear Velocity ft/sec
8 8 10 12 14
16
Whole Cheese Whey
Tubular UFCN
4 0 kg/cm' (57 psi)
25 -C
20
15 TS
10
2345
Linear Velocity mi/sec
c. Effect of Feed Velocity on Ultraflltration Flux
100
90
80
I 7°
i 60
X
Z. 50
40
30
20
Whole Cheese Whey
Tuouiar UPON-
4 0 kg/cm' (57 psi)
3 7 m;sec (12lt/sec)
60
35
C0
45 >.
ra
5
40 r
^
35 &
X
30 |
2^
20
15
10 15 20 25 30 35 40 45
Temperature "C
d Effect of Temperature on Ultraflltration Flux
Figure 4. Effect of operating parameter on UF flux (4).
33
-------
TABLE 7. FEATURES OF COMMERCIAL UF MODULE DESIGNS
Company
Geometry
Membrane type
Abcor
Door Oliver
Romi con"1"
Union Carbide
Millipore
1" diameter tubular membranes
and spiral wound flat membranes
Multiple flat plate
Narrow channel tubular (flat
sheet membrane); also hollow
fibers
1/4" tubes
Multiple plate and screen
cartridge
Originally cellulose acetate; now, other
materials that are pH and temperature resis-
tant
Polyelectrolyte complex (developed by Amicon)
Dynel^ (acrylic), cellulose acetate, Nomex^
(nylon) polyelectrolyte complexes, others
Zirconium oxide on carbon
Cellulose acetate and other more solvent re-
sistant materials
Osmonics
Selas
UOP
Envirogenics
Spiral wound modules, 2" and 4"
diameter
Tubular
Tubular
Spiral wound
Cellulose acetate and
Zirconium oxide deposi
ceramic
Cellulose acetate
Cellulose acetate
polyarylsulfone
ted in-situ on porous
I-
Based on data from Reference 1.
Joint venture of Rohm & Haas and Amicon.
-------
There is little operating data on the UF units used in these and other indus-
trial/municipal applications.
Pilot/Laboratory Scale Studies
In addition to the actual industrial process/wastewater applications men-
tioned above, a number of engineering/pilot plant studies have been conducted
or are currently in progress to assess the applicability of UF to the treatment
of other industrial process streams and wastewaters. Results of soire of the
most pertinent of these studies are summarized in Table 8. The data indicate
that the process is very effective in concentrating protein, color, oil, and
COD in wastewaters. In treating cheese whey, lactose and salts are essentially
unaffected by the process because of their small sizes. Based on the data in
Table 8, when used for the treatment of very concentrated wastes, UF permeates
(treated effluents) may not meet certain effluent discharge requirements witn-
out additional treatment (for example, biological treatment or RO).
Mobile Unit Applications
Most UF manufacturing companies have skid-mounted pilot units that can be
installed on a trailer for field pilot studies. These units usually consist
of one or two modules having capabilities from about 18,900 £/day to 39,800
i/day (or 5,000 to 10,000 gpd)*.
USE OF UF IN MOBILE UNITS FOR TREATMENT OF WASTE SPILLS
This section evaluates the feasibility of using UF in a mobile unit to
treat concentrated waste spills, based on the general requirements'discussed
in Section 3.
Pertinent Features of UF for Use in Mobile Units
Essentially all key requirements for mobile unit application (see Section
3) would be met by UF. UF units are compact, are commercially available, can
be started up and shut down relatively quickly, can be serviced conveniently,
would generally produce only a small volume of residue, would not require
skilled labor for operation, and can be operated with the electric power pro-
duced by a diesel generator. The UF process would be suitable for the treat-
ment of emulsified oil and humic acids and can be used as a pretreatment to
RO or other processes for the removal of colloidal substances and high mole-
cular weight organics. Based on the above discussion, the primary limitations
of the UF are inapplicability to wastes containing low molecular weight sub-
stances and for producing low concentration permeates from highly concentrated
wastes.
As discussed earlier, UF is a commercial process, is offered by a number
of manufacturers (see Table 7), and has been used in mobile water treatment
*
Telephone communication with Gerald Alexander, Permutit Company, Glenaale,
CA, February 20, 1979.
35
-------
TABLE 8. RESULTS FROM UF PILOT/LABORATORY SCALE STUDIES
CJ
en
Stream
Kraft black liquid
Oil coolant
Screen room effluent
Cheese whey
Laundry wastewater
Textile sizing (PVS)
Characteristics
Total solids, ppm
Color, ppm
Total solids, ppm
Freon extractable oil,
ppm
Total 'solids, ppm
Color, ppm
Chemical oxygen demand
Percent protein
Percent lactose
Percent salts
COD, ppm
Percent solids
Feed
25,000
9,500
46,300
36,900
2,900
2,000
2,300
0.65
5.0
0.70
—
_ ——
Concentrate
250,000
94,200
331,000
320,000
30,100
25,000
41,300
13.1
4.92
0.79
1,950
4.00
Permeate
5,000
7,000
5,400
300
1,600
150
860
0.2
5.03
0.69
193
0.68
Reference
5
5
5
4
4
4
wastewater
-------
applications. The unit is fairly compact; a 18,900 a/day (5,000 gpd) product
water capacity unit using the spiral wound module design would weight about
3.6 kg (8 Ib) and would be approximately 10 cm (4 in) in diameter and 100 cm
(40 in) in length. The start-up and shut-down of UF consist primarily of
turning on-and-off of the high pressure pump and adjusting the pressure gauges.
UF modules are connected in series or in parallel with provisions for flow by-
pass so that a single module can be by-passed for maintenance service. Opera-
tion of UF units does not require highly skilled labor; the operating effort
would be limited to turning valves, reading gauges, and collecting samples for
chemical analysis. In mobile unit applications, the system would be equipped
with on-line monitors for continuous measurement of pH, temperature, pressure,
etc. Ordinarily no more than 6 to 12 operator-hours would be required for a
24-hour operation of a UF unit. For safety reasons, however, operation of a
mobile unit in the field may require a minimum of 2 persons. The power for
the operation can be provided by a diesel electric generator that can be housed
in the same mobile unit.
As with any wastewater treatment system, UF produces a residue (reject)
that requires disposal. The volume of this reject (waste concentrate) is
usually about 10 to 25 percent of the volume of feed water. In field applica-
tions, the reject may be collected in drums and subsequently transported to
hazardous waste management facilities or municipal sewage treatment plants for
further treatment/disposal.
As with all membrane processes, UF units are susceptible to membrane
fouling and deterioration. However, the magnitude of the fouling problem is
not nearly as big as with RO. When the membrane is fouled to an "unacceptable"
level, the unit is taken off service and is cleaned by flushing with deter-
gents or water (this is similar to ordinary sand filtration where the filter
is backwashed periodically to remove suspended solids trapped in the filter
media). Fouling by the growth of biomass on the membrane surface can be mini-
mized by prechlorination of the waste or periodic flushing of the system with
1 percent formaldehyde.
Engineering Evaluation and Cost Estimates
Based on data obtained from one UF manufacturer*, a preliminary design
has been developed for a 227,000 a/day (60,000 gpd) spiral wound UF unit for
use on a trailer flatbed. The system, which consists of nine pressure vessels,
can be mounted on two heavy skids. Each vessel contains 2 modules. Excluding
the pretreattrent equipment, the system would have dimensions of 5 m long, 1 m
wide, and 1.7 m high (15 ft x 3 ft x 5 ft), would weigh about 680 kg (1.200
Ib), and would occupy 8.5 m3 (225 ft3) of space. The water recovery is
estimated at close to 100 percent. Other components of the system include a
high pressure puma, valves and piping, instrument panel, and the power supply
source. Power can be supplied to the unit using a diesel electric generator.
The generator, a 5 kw unit, would have the approximate dimensions or 2 m (1)
*
Telephone communication with R. Gibbons, Envirogenics Coirpany, El lonte, CA,
April 17, 1979.
37
-------
x 0.8 m (w) x 1.1 m (h) (6 ft x 2 ft x 3.5 ft), would weigh approximately 900
kg (2,000 Ib), and would occupy a volume of 1.8 m^ (42 ft^). A storage tank
having a volume of about 1.9 nH (500 gal) would supply the fuel to the unit.
The major logistics support required in field applications relate to
possible requirements for membrane replacement and sufficient supply of treat-
ment chemicals (acid, base, scale inhibitors, etc.). Under normal operation,
it is unlikely that membrane replacement would be required during short-term
field operation. In industrial applications, a membrane life of 2 to 3 years
has been reported for continuous operation. Assuming a requirement for re-
placement of 50 percent of the modules during a field application, the volume
of the spare modules that may have to be carried on-board is about 0.24 m3
(8.7 ft3).
Chemical requirements are determined by the characteristics of a soecific
waste spill. The quantity of the chemicals which may have to be carried on-
board for short-term field operation would be relatively small. For extended
field operation, the mobile unit may have to be resupplied with chemicals on
an as necessary basis.
The estimated total capital costs for the 227,000 Ji/day (60,000 god) UF
unit* and the accompanying diesel electric generatort are estimated at $30,000
and $5,000, respectively. The cost for the UF unit includes cost for UF mo-
dules, piping, valves, pressure pump, chemical storage tank, chemical feed
pumps, and the instrumentation panel. The cost also covers the engineering
and installation fees.
The operating cost for the mobile UF unit per unit volume of waste treated
should vary depending on the specific application (for example, waste charac-
teristics, waste volume, chemicals requirement, location of the spill, etc.).
Assuming that in most field applications the mobile unit will have to be
powered by a diesel electric generator, the fuel for the generator would con-
stitute one of the major items of operating cost. A 5 kw diesel generator
would consume about 3 i/hr (1 gal/hr) of fuelt (about $0.70/hr).
REFERENCES
1. Arthur D. Little Co. Physical, Chemical, and Biological Treatment Tech-
niques for Industrial Wastes, EPA Office of Solid Waste Management Pro-
grams, PB-275-287, 1976.
2. Bansal, I.K. Progress in Developing Membrane Systems for Treatnent of
Forest Products and Food Processing Effluents, Water-1976: I. Physical,
Chemical Wastewater Treatment, AIChE Symposium Series, No. 166, Vol. 73,
1977.
*
Telephone communication with R. Gibbons, Envirogenics Company, El Monte, CA,
April 17, 1979.
Telephone communication with L. Anderson, Anderson-Belviar Company, Long
Beach, CA, January 31, 1979.
38
-------
3. Weber, W.J., Jr. Physicochemical Processes for Water Quality Control,
Wiley-Interscience, New York, 1972.
4. Envirogenics Systems Company. Membrane Ultrafiltration for Water Treat-
ment.
5. Bansal, I.K. Reverse Osmosis and Ultrafiltration of Oily and Pulping
Effluents, Industrial Waste, May/June, 1977, p. 32.
39
-------
SECTION 6
ION EXCHANGE
PROCESS DESCRIPTION
Process Principle
Ion exchange is a process whereby the toxic or undesirable ions in a
wastewater are exchanged with relatively harmless ions, such as H+, OH", Na ,
HC03~, Cl", held by the ion exchange material (usually an organic matrix).
The process is used for water softening (removal of Ca++ and Mg++ hardness),
treatment of wastewaters, and for concentration and recovery of ionic sub-
stances. The exchange reaction is reversible and concentration dependent,
thus making it possible to regenerate the exchange resins for reuse. The ex-
change reactions can be represented by the following equations:
exchange »(»V + A"* "T (IT)/* * ** (6.1)
excnansa regeneration "
excSance n(R+)D" + c"" • <*>" + nD" ^^
exchange regeneration n
where R = ion exchange resin matrix
An+, Cn~ = cation and am'on to be removed from solution
B+, D- = exchangeable ions on the resin matrix
Certain natural substances (for example, natural zeolite, clay, protein)
exhibit ion exchange properties, and some (for example, natural zeolite) have
been used in water purification. Modern ion exchange resins, however, are
primarily synthetic organic materials containing ionic functional groups
(-COO", -NHj+j -S03~, etc.) to which exchangeable ions are "attached". These
synthetic resins are structurally stable (that is, can tolerate a range of
temperatures and pH conditions), exhibit a high exchange capacity, and can be
"tailored" to show selectivity toward specific ions.
Recent advances in synthetic resin technology include the development of
"sorptive" (or macroporous) resins capable of removing organics from aqueous
systems. The removal mechanism is generally considered to be one of sorotion
rather than ion exchange; the resin regeneration consists of "elution" of the
adsorbed material with an organic solvent or an inorganic salt solution.
40
-------
Sorptive resins have been used for the removal of phenols (1), alkylbenzensul-
fonate (ABS) (2), and viruses (3) from waters, and for the treatment or
municipal and industrial wastewaters (4-7).
Process Design
Ion exchange resins used for water and wastewater treatment are commonly
employed in columnar beds through which the aqueous phase is passed in an up-
flow or downflow direction. The operation consists of four steps: exchange
(service), backwashing, regeneration, and rinsing. In the fixed-bed downflow
design, raw water is introduced through the top distributor, and the treated
water is collected at the bottom. When the bed is exhausted ("breakthrough"
occurs), the feed water flow is stopped, and the bed is backwashea (to remove
particulates trapped in the bed). The exhausted resin is then regenerated
using an appropriate regenerant. Regenerants most commonly used include brine,
hydrochloric acid, sulfuric acid, sodium hydroxide, sodium carbonate, and
ammonium hydroxide. In general, one to four bed volumes of regenerants (5 to
10 percent concentration) are used in the regeneration step. The regeneration
is followed by a rinse step whereby the excess regenerant is removed from the
bed before the bed,is returned to service. The spent regenerant is concen-
trated in substances-removed'from the wastewa'ter and requires disposal.
%, To insure-'ltfng -service
-------
TABLE 9. PERTINENT FEATURES OF ION EXCHANGE SYSTEMS (8)
Types
Cocurrent Fixed Bed
REGENERATION
Countercurrent Fixed Bed
SERVICE
REGENERATION
Continuous Countercurrent
SERVICE
REGENERATION
Description
of Process
Indications
for Use
Advantages
Disadvantages
Downflow of raw fluid to be treated
(loading phase) Upflow backwash
Downflow regeneration Downflow rinse
Batch loading and regeneration
Low loads (200 PPM in softening 250
TDS In demin) Lower thruput (about
tOUOGPM) Where regeneration chemi
cal cost is not critical disposal of waste
volume in large single batch not critical,
and dilution of feed no problem Manual
operation acceptable
Low capital cost Automatic controls in
strumentation optional Simple basic
type of unit Easy maintenance
High regenorant cost Fluctuating efflu
ent quality Large single batch waste
disposal High water consumption thru
dilution and waste Requires substantial
floor space
Regeneration flows opposite in direction
to Influent Backwash (in regeneration)
does not occur on every cycle to pre
serve resin stage heights Resin bed Is
locked In place during regeneration
Handles high loads at moderate thruput
or low loads at high thruput (GPM x TDS
or GPM x PPM removal = 40000 or
more) Where effluent quality must be
relatively constant regeneration cost is
relatively critical disposal of single
batch waste volume no problem
Moderate capital cost Can be operated
with periodic attention Moderate
regeneration cost Lesser volume of
waste due to less frequent backwash
Consistent effluent quality
Increased controls and instrumentation
higher cost Requires mechanism to lock
resin bed Large single batches of waste
disposal Moderate water consmptlon
thru dilution and waste Requires sub
slanlial floor space
Multl stage Countercurrent movement of
resin in closed loop providing simul
taneous treatment, regeneration, back
wash and rinse Operation Is only Inter
rupted for momentary resin pulse
Highloads with high thruputs (GPM x
TDS or GPM x PPM removal = 40 000 or
more) Where constant effluent quality is
essential regeneration costs critical
total waste volume requires small con
centrated stream to be controllable
Where loss of product thru dilution and
waste must be minimized Where avail
able floor space is limited
Lowest regeneration cost Lowest resin
Inventory Consistent effluent quality
Highest thruput to floor space Large
capacity units factory preassembled
Concentrated low volume waste stream
Can handle strong chemical solutions
and slurry Fully automatic operation
Requires automatic controls and Instru
mentation, higher capital cost More
headroom required
-------
deimneralization of acid mine drainage, treated municipal wastewaters, and a
range of industrial wastewaters have been the subject of a number or -investi-
gations (9). Some specific examples of full-scale ion exchange application to
the treatment of industrial wastewaters are listed in Table 10.
Pilot and Laboratory Scale Studies
Almost every large-scale application of ion exchange to the treatment of
industrial wastewaters has been preceded by laooratory and pilot plant studies
aimed at selection of the most suitable resin and development of design cri-
teria (resin quantities and regeneration requirements, optimum hydraulic
loading, etc.) and basis for cost estimates. Many of the laboratory and pilot
scale studies that are currently in progress (largely by resins manufacturers)
are aimed at the development of ion-specific and longer life resins (,vith
minimum susceptability to organic fouling). Tne investigation of the use of
ion exchange resins in new applications (for example, treatment of miscella-
neous industrial wastewaters) is a continuous effort that is carried out ooth
by resin manufacturers and environmental engineers in the industry.
Mobile Unit Application
Trailer-mounted ion exchange resin systems have been used for field pilot
plant studies. A 276 Vmin (70 gpm) mobile unit was used for evaluation of
the ion exchange as a tertiary treatment for the polishing of secondary erfluent
from sewage treatment plants in Ricnland, Washington; Pomona, California; South
Lake Tahoe, California; and Washington, D.C. (9). Rohm and Haas Company
(Philadelphia, PA) operated two mobile ion exchange units, one for the evalua-
tion of the treatability of TNT production wastewaters at an Army ammunition
olant and another for the evaluation of the treatability of a process waste-
water at a naval installation*.
USE OF ION EXCHANGE IN MOBILE UNITS FOR TREATMENT OF WASTE SPILLS
This section evaluates the feasibility of using ion exchance in a mobile
unit to treat concentrated waste spills, based on the general requirements
discussed earlier.
Pertinent.Features of Ion Exchange for Use in Mobile Units
In aeneral, nearly all the requirements discussed for applicability of a
system to the treatment of waste spills in a mobile unit would be met by rhe
ion exchange process. Ion exchange units are compact, are commercially avail-
able, are easily automated, can be started and shut down relatively quickly,
can handle a range of organic and inorganic concentrations (including highly
acidic and alkaline wastes), can be serviced conveniently, would produce a
snail volume of residue that may be suitable for processing for mater'di
recovery, would not require skilled fiela labor for operation, c"d Can be
operated with the electric power produced on-site by a diesel generator.
*
Information provided by Rohm and Haas Company, Philadelphia, DA.
-------
TABLE 10. SPLCIFIC EXAMPLES Of FULL-SCALE ION EXCHANGE APPLICATION TO INDUSTRIAL UASTEWATCR TREATMENT
Application
System description
Performance/specification Reference
Acid wastewater
demineralization
at an Army ammu-
ni tion plant
Ammonium nitrate
recovery from
fertilizer plant
wastewater
Organic color
removal from dye-
stuff production
plant
Organics removal
from pulp and
paper mil 1
effluent
Two trains fixed bed ion exchange with
gravity filter, countercurrent regenera-
tion with IIN03 or l^SO^j of cation units;
NII^OII regeneration of anion units.
Recovery of NII4N03 as by-product and
water for boiler feed. 2 cation units
10 ft dia., 12 ft high, 2 anion units
9 ft dia., 10 ft high.
Several continuous ion exchange resin
systems in operation at fertilizer
manufacturing plants; ammonium nitrate
is recovered from wastewater from
reuse.
Two large resin sorption systems in
operation in Sweden and one system in
operation in Japan. The systems in
Sweden consist of three resin columns,
each containing 20 cubic meters of
resins; Ihe system in Japan consists of
four columns, each with about 30 cubic
meters of resins. The resins are re-
generated with a caustic wash followed
by a reactivation with an acid stream
(e.g., H2S04).
310 gpm capacity; reduce
TDS from 2,800 ppm to less
than 20 ppm.
Capacities vary from 40
gpm to 695 gpm; reduce
ammonia from as high as
2,500 ppm to less than
25 ppm.
300,000 gpd waste stream;
color reduced from 75,000
to 500 APIIA units; COD,
copper and chromium also
reduced.
92 to 96 percent color
removal achieved (from an
initial level of 30,000
to 40,000 units); 80 to
90 percent COD removal and
40 to 60 percent BOD remo-
val also achieved.
12
13
13
Ir
Information provided by Chemical Separations Corporation, Oak Ridge, TN.
-Continued-
-------
TABLE 10. (Continued)
Application
System description
Performance/specification Reference
-P.
en
Para-nitrophenol
(PNP) recovery
from parathion
manufacturing
wastewater
Recovery of
phenol from
phenolic resin
manufacturing
waste
Brine
purification
Wastewater (pH=2) is passed through 50
cu ft bed of cross-linked polystyrene
adsorbent resin. PNP recovered by re-
generation with 2 bed volume of 4 percent
NaOH followed by water rinse; recovered
sodium PfJP recycled to the parathion pro-
cess; two resin beds used; one in service
removing phenol; the other in regenera-
tion or standby; since plant operates
only part of the year, resin beds are re-
generated and kept wet while idle.
A weakly basic macroporous ion exchange
resin with a cross-linked polystyrene
matrix used for phenol recovery from an
aqueous process waste. Methanol (9.4 to
22.4 gal/cu ft of resin) is used for re-
generation with the phenol/water/inethanol
recycled to production process.
Acidified brine solution contaminated
with phenol is passed through two beds of
adsorbent beds in series; lead adsorbent
bed is periodically regenerated with 4
percent NaOH at 70°C; after one year of
operation, resin condition is essentially
unchanged.
Phenol concentration re-
duced from 1,000-1,800 ppm
to 1-5 ppm.
14
Phenol concentration re-
duced from 1.5-2 percent
to less than 120 ppm;
higher phenol recovery
achieved with higher
methanol usage.
Phenol in 20 percent
waste brine solution is
reduced from 10-150 ppm to
less than 0.5 ppm.
14
14
-------
Ion exchange is a commercial process and is offered by a number of ven-
dors*. The system is fairly compact; a 227,000 «./day (60,000 gpd) system
would have two columns, each approximately 2.1 m (7 ft) high and 0.91 m (3 ft)
in diameter, and when loaded, weighing approximately 2,500 kg (5,500 Ib). The
system can be designed for manual or automatic operation. Process startup and
shutdown would consist primarily of turning on-and-off the feed pump. For
manual operation, the operator must have some familiarity with the process;
he will decide when to stop one cycle (for example, service) and start the
next (for example, backwash). Manual operation would be more suitable for
mobile unit application because of the diversity of wastes that are encountered
in spill situations. Ordinarily no more than six to twelve operator-hours
would be required for a 24-hour operation of an ion exchange unit. For safety
reasons, however, operation of a mobile unit in the field may require a minimum
of two persons.
The ion exchange system is not an energy intensive unit. The only energy
required is for pumping; this energy can easily be provided by a small diesel
electric generator that can be housed in the same trailer.
As with any wastewater treatment system, ion exchange produces a residue
(spent regenerant) that requires disposal. The volume of residue as a percent
of the feed is dependent on factors such as resin capacity, solution concen-
tration, and regeneration parameters, and is seldom higher than 10 percent.
In field applications, the regenerant may be collected in drums and subsequent-
ly transported to hazardous waste management facilities for further treatment/
disposal, or to a recovery facility if the waste can be economically processed
for material recovery.
Engineering Evaluation and Cost Estimates
Based on data obtained from one ion exchange process supplier, a prelim-
inary design has been developed for a 227,000 a/day (60,000 gpd) cocurrent
ion exchange system for use on a flatbed trailer. The system consists of six
steel tanks each containing 0.69 m3 (24 ft3) of resin and measuring 21.1 m
(7 ft) in height, 0.91 m (3 ft) in diameter, and weighing 2,500 kg (5,500 Ib).
Other components of the system include a feed water pump, valves and piping,
a tank for storage of processed water for backwashing, and the power supply
source. Power required for the system can be supplied by a 5 kw diesel elec-
tric generator that would have the approximate dimensions of 2 m (1) x 0.8 m
(w) x 1.1 m (h) (6 ft x 2 ft x 3.5 ft), would weight approximately 900 kg
(2,000 Ib), and would occupy a volume of 1.8 m3 (42 ft3).
The proposed ion exchange system can be "prepared" to be operated in any
one of the following three ways: (a) as a complete demineralization system
for application to waters containing little or no organics; (b) as an adsorp-
Major suppliers of ion exchange systems are: Chemical Separations Corp.,
Oak Ridge, TN; Crane Co., King of Prussia, PA; Ecodyne, Union, NJ; Illinois
Water Treatment, Rockford, II; Infilco, Richmond, VA; LA Water Treatment Co.,
Industry, CA; and Permutit, Paramus, NJ.
46
-------
tion-demineralization combination system for the removal of both inorganic
ions and organic substances; and (c) as a sorptive system for the removal of
organics (for application to wastewaters containing high organic loadina).
In the first type of application, three columns will carry cation exchange
resins on the H+ cycle whereas three columns will carry anion exchange resins
on the OH" cycle. The cation and anion exchange columns can be operated in
series (to increase service life between regeneration of the lead beds) or in
parallel (for maximum hydraulic capacity). The piping arrangement would allow
for one or more beds to be taken out for regeneration while the remaining
columns would remain in service.
In the adsprption-demineralization combination arrangement, the two lead
columns would carry the adsorptive resins and would act as organic scavengers.
The other four columns would contain the anion and cation exchanae resins and
would be operated similar to the "demineralization only" system described
above. In the all sorptive resin system, all columns would be filled with
adsorptive resins and would be operated in series or in parallel. By carrying
two or more different types of adsorptive resins (for example, "polar" and
"non-polar" resins) in the columns, the system can provide capability for the
removal of a broad spectrum of organics.
• . In field-applications, the bed regeneration requirement is dictated by
the concentration of the substances-in the raw wastewater and the selected
mode of operation. For an organic adsorption application where five beds may
be in service-at any one time and a wastewater containing one percent organics
such as phenols is to be processed, a service life of about 30 minutes would
be expected for the lead column before requiring regeneration.
The major logistics support required for field operation relate to chemi-
cals needed for resin regeneration and possible requirements for resin replace-
ment. Under normal operation, it is unlikely that substantial resin replacement
would be required during short-term field operation. In various demineraliza-
tion applications, a resin life of longer than two years has been reported.
For continuous operation, resin loss resulting from regeneration is reported
at about one percent per year under ideal conditions and at about 10 percent
under normal conditions; specific installations which operate under harsh
conditions have reported greater than 50 percent losses*.
Chemical requirements for regeneration are determined by the types of
resins used and the characteristics of a specific waste spill. In general,
H2S04 is used to regenerate the cationic exchangers and NaOH is used to re-
generate the anionic exchangers. Sodium hydroxide or organic solvents such
as acetone and methanol are used to regenerate sorptive resins. One to three
bed volumes of organic solvents or acid or alkali (5 to 10 percent solutions)
are required in each regeneration cycle. For extended field operation, the
mobile unit may have to be resupplied with chemicals on an as necessary basis.
*
Letter from Ronald Y. Lawrence, Diamond Shamrock Corp., Redwood City, CA,
February 14, 1980.
47
-------
The estimated total capital costs for the ion exchange system described
above is $140,000*, which includes $5,000* for the diesel aenerator*. The
operating cost for the mobile unit per unit volume of spill treated snould vary
depending on the specific application (for example, waste characteristics,
waste volume, regenerant requirement, location of the soil!, etc.). Tne re-
generant cost would likely constitute a major element of the operating cost.
A total regeneration cost of $58 per cycle has been estimated for the regenera-
t^'on of one cation exchange and one am on exchange bed with four bed volumes
of five percent solutions of HC1 and NaOH. This estimated cost assumes
chemical costs of S39/MT and $155/MT for HC1 (32 percent solution) and NaOH
(50 percent solution), respectively (15):
The estimated cost for the regeneration of an adsorotion res^n witn four
bed volumes of an eluent is estimated at $42 when the bed is regenerated with
five percent NaOH solution, $400 when the bed is reaenerated with methanol,
and $1,150 when the bed is regenerated with acetone. Tnese estimates assume
42c/2. (S1.60/gal) for acetone and 15c/£ (55c/gal) for methanol (15) and taKe
no credit for acetone or methanol that could be potentially recovered /ia dis-
tillation of the spent eluents.
REFERENCES
1. Pollio, F.X. and R. Kunin. Sorotion of Phenols by Anion Exchange Resins,
Environmenta1 Science and Technology, Vol. 1, No. 2, p. 160, Feb. 1967.
2. Hinricns, R.L. and V.L. Snoeyink. Sorption of Benzenesulfonates by ]ieak
Base Anion Exchange Resins, Water Research, ^0, pp 79-87, 1976.
3. Kim, B.R., et al. Adsorption of Oraanic Compounds by Synthetic Resins,
Journal of Water Pollution Control Federation, £8, 120, 1976.
4. Rook, J.J. and S. Evans. Removal of Trihalorrethane Precursors From
Surface Waters Using Weak Base Resins, Journal of Arrencan Water Works
Association, 71_(9), 520, 1979.
5. Chudyk, W.A., et al. Activated Carbon Versus Resin Adsorption of 2-
Methylisoborneol and Chloroform, Journal of American Water '/orks Associa-
tion, 71(9), 529, 1979.
6. Brenner, !_., et al. Evaluation of the Caoacitv of Granular Activated
Carbon and SAD-2 Resin to Remove Trace Organics from Treated Drinking
Water, Environmental Science and Technology, V2_(2), 1315, 1973.
Telephone communication with Gerald Alexander, Permutit Company, Glendale,
^CA, February 20, 1979.
•"Telephone communication with L. Anderson, Anderson-Sevier Co., Long Beach,
CA, January 31, 1979.
This cost is for a 6-column system design presented here. A 2 , 3-, or 4-bed
system, which would cost about 1/3 to 2/3 as much, can also be jsed in mooile
units; such a system, however, dould require more freauent reaeneration and
does not provide as much operational flexibility as a 6-bea aesian.
48
-------
7. Anon, Decolorization of Kraft Pulp Bleaching Effluents Using Amberlite
SAO-8 Polymeric Adsorbent, Rohm and Haas Company, Philadelphia, PA, 1971.
8. Ion Exchange Systems, Chemical Separations Corporation, Oak Ridge, TN.
9. Water Purification Associates, Innovative Technologies for Water Pollu-
tion Abatement, NTIS PB-247-390, 1975.
10. AKZO IMAZ TMR Process for the Removal of Mercury in Water, AKZO ZOOT
Chemie, Netherland.
11. Pollio, F.X., et al. Ion Exchange Resins Treats Sour Water, Oil and Gas
Journal, 67_, 126, 1967.
12. Brennah, J.F. The Chem-Seps Nitrogen Recovery Process: A Pollution
Solution that Works, paper presented at the Fertilizer Institute Environ-
mental Symposium, New Orleans, LA. January 14-16, 1976.
13." Arthur D. Little. Physical",'Chemical, and Biological Treatment Techniques
for Industrial Wastes, Vol. I, NTIS PB-275-054, 1977.
14. Fox,'C.R. ' Plant Uses'Prove-Phenol Recoverywith Resins, Hydrocarbon
Processing, November 1978, pp 269-273.
15. Chemical Marketing Reporter, October 8, 1979.
49
-------
SECTION 7
WET AIR OXIDATION
PROCESS DESCRIPTION
Process Principle
Wet air oxidation (WAO), also referred to as the Zimpro process, wet oxi-
dation, wet combustion, and wet incineration, is a process especially suitable
for the treatment of high strength or toxic organic wastes or wastes containing
an appreciable concentration of non-biodegradable organic matter (as reflected
by a high COD/BOD ratio). Zimpro, Inc. (Rothschild, Wisconsin) holds a patent
on this process. WAO operates at relatively high temperatures, 200 to 320°C
(392 to 608°F) and pressures, 1 to 29 mpa (150 to 4,000 psi), and results in
the oxidation of most reduced inorganic and organic compounds to inoccuous
inorganic end-products (for example, C02, H20, S04=, etc.).
Compared with the conventional combustion processes that generate air
pollutants (for example, NOX, SOX, etc.) requiring control, no such pollution
problems are associated with WAO (1). Also, since oxidation is carried out in
a water environment, waste pretreatment such as sludge dewatering or drying
that would be required in conventional combustion would not be necessary.
Figure 5 presents a schematic flow diagram for WAO. The raw waste is
pressurized.to^the system pressure by means of a high pressure positive dis-
placement pump . Air is then injected into the waste by an air compressor and
the mixture is heated in the process heat exchanger by the hot reactor effluent.
Depending on -influent COD'concentration and" 1'eve.l ."of'treatment desired, addi-
tional heating may be necessary. This supplementary heating is provided by a
heater that can be either an.electrical or an oil-fired oil heating unit.
In the reactor, the organic or reduced inorganic constituents are oxi-
dized. Being an exothermic reaction, the reaction results in a temperature
rise in the reactor. The hot effluent is cooled in the process heat exchanger
as previously mentioned and in a final cooler to a discharge temperature of 38
to 60°C (100 to 140°F). The final cooler uses cooling water as a cooling
medium. The final step in the process is to reduce the pressure of the oxi-
dized waste by means of the pressure control valve. The pressure control valve
If the waste contains very large sized (>6 mm or 0.25 in) particles, such
particles must be removed or reduced in size (by use of a "grinder") before
pumping.
50
-------
PROCESS
HEAT
EXCHANGER
RAW WASTE.
PUMP
REACTOR
AIR COMPRESSOR
ELECTRIC
HOT OIL
HEATER
OXIDIZED
WASTE
COOLER
Figure 5. Schematic flow diagram" for the WAO process.
-------
maintains the system pressure at a pre-set point. When waste contains appre-
ciable quantities of undesirable inorganic constituents (for example, fluorides,
heavy metals, etc.),, further treatment of the WAO effluent may be necessary
prior to final discharge. In certain appl'icat-ions ffor example, treatment of
organic sludges to improve dewaterability) where UAO is operated at relatively
low temperatures, little destruction of organics is affected. In large-scale
applications, the reactor effluent is passed through a gas-liquid "separator"
where hot exhaust gases are recovered and used to drive the air compressors.
The exhaust gas, which may contain volatile organic matter, would be treated
before final discharge. In mobile spill waste treatment applications, where
complete oxidation of organics is desired, WAO reactor would be operated at
very high temperatures and the exhaust, which would consist primarily of steam,
C02, and N2 (from the air supply), would not require separation and treatment.
For such small scale application it would also not be economical to use the
effluent steam to power the air compressor.
Process Control Parameters
By proper selection of operating parameters, any degree of oxidation (in-
cluding complete destruction) of reduced organic and inorganic waste consti-
tuents can be achieved. Operating parameters that affect oxidation efficiency
are temperature, detention time, and pressure. The process air input and
supplementary heating requirements are determined by the waste characteristics
and the selected operating conditions.
The effect of temperature on the oxidation of five concentrated simulated
or actual wastes is shown in Figure 6. As noted in the figure, the oxidation
efficiency, as indicated by the reduction in COD, generally increases with the
rise in temperature. Figure 7 presents typical data on the effect of detention
time on oxidation efficiency for several temperatures. The data in this figure
indicate that (a) at low temperatures, only partial oxidation of the waste is
affected even with very long detention tfmes and (b) at temperatures greater
than 200°C (392°F), maximum achievable oxidation is attained in about~30
minutes. Although the oxidation efficiency increases with the rise in tempera-
ture, the operating temperature cannot exceed the critical temperature of water
(374°C or 905°F) because continuous presence of a liquid phase is essential
for oxidation. Operations at high operating temperatures require that the
process be run at high pressures in order to avoid complete vaporization of
water. Ordinarily, the operating temperature is selected to achieve the
desired degree of oxidation for a selected reaction time. If the reaction time
is to be short and near complete oxidation is to be achieved (as for mobile
spill treatment applications), the system would have to be operated at high
temperature and pressure.
The air-to-steam ratio at saturation in the reactor vapor space is given
by the following equation (9):
. 144 (P - P ) v
n _ T. 5 S I-, -, \
- (7J)
52
-------
CHEMICAL
OXYGEN DEMAND, gll
0 *• o g 8 § g g
SUGAR
SOLUTION
.
N
\
.-
s
s
SEWAGE
S. SLUDGE
B
* *
\\
\\
\
I
\
\
\
'
SULFUR
SUSPENSION
%
™
•
\
\
\
•
\
CARBON
SUSPENSION
—W
•\
\
1
I
\
\
—r
SODA BASE
WASTE
SULFITE
LIQUOR
i
\
\
T
\
\
V
\
100 200 300 100 200 300 100 200 300 100 200 300 100 200 300
TEMPERATURE «C < - • • .
Figure 6. Oxidation curves for concentrated simulated or actual wastes (2)
05
.10 15 20
Time at temperature, hr
25
30
Figure 7. Time-temperature effect on the degree of oxidation (2).
__ _ - 53
-------
where S = steam, Ibm
A = air, Ibm
T = temperature, PF
PJ. = total pressure, psia
Ps = saturated steam pressure, psia
Vs = specific volume of saturated steam,.ftfyIbm
The quantity of air added to the wastewater should be sufficient to achieve
oxidation and should not exceed the a. ratio corresponding to the operating tem-
perature and pressure of the reactor? Higher A ratios would result in complete
evaporation of the water. If the quantity of'^air required for a desired
level of oxidation exceeds that indicated-tiy the A ratio for the selected
reactor operating temperature and pressure,' the S oxidation level that actual-
ly would be achieved would be less than that desired. To'achieve the desired
level, the reactor operating conditions would be adjusted.
Depending on the selected operating temperature and the heating value of
the waste, supplementary preheating of the waste may not be required (except
for startup). The heating value of the waste is determined by the nature and
concentration of the oxidizable waste components. Heating values for selected
compounds and waste materials and the theoretical amount of air required for
their complete oxidation c.re presented in Table 11.
PROCESS APPLICATIONS AND RELATED STUDIES
Commercial Scale Municipal/Industrial Applications'
There are currently more than 150 UAO units in operation worldwide; about
90 percent of the units handle municipal sludges whereas the remaining 10 per-
cent process industrial wastes (4). In application to municipal sludges, where
the objective is to improve sludge characteristics and-not to achieve destruc-
tion of organics, the units are operated at relatively.low temperatures (150
to 200°C or 302 to 392°F). A total" sludge processing cost of $30'to $40 per
tonne of dry solid has been reported for these applications(3).
WAO has been used for the treatment of'cyanide wastes, pulp and paper
waste liquor (5,6), glue manufacturing (4), and photographic wastes (7). There
are at least four plants treating acrylonitrile and coke production effluents
for the destruction of cyanides in these wastes (8). In one such application,
the cyanide content is reportedly reduced from 200 ppm to 6 ppm, and the COD
is reduced from. 3-2. g/i to-8-.g/x, (4.)-.- •«. . - • •
Pilot and Laboratory Scale Studies
Based on laboratory test data, WAO is effective in the treatment of a wide
variety of wastes. Explosive materials such as nitrocellulose can be safely
destroyed by WAO (4). As a pretreatment method, WAO can render certain toxic
or non-biodegradable materials (for example, pesticides, plastics, etc.)
amenable to biological oxidation (3). Table 12 presents data on WAO efficien-
cy for ten "priority pollutants". As noted in the table, destruction efficien-
cies of greater than 99.8 percent and 81 percent are achieved at temperatures
54
-------
TABLE 11. HEATING VALUES FOR SOME COMPOUNDS AND WASTE MATERIALS
AND AIR REQUIREMENTS FOR THEIR OXIDATiON (2)
Material
Hydrogen
Ethyl ene
Carbon
Acetic acid
Oxalic acid
Pyridine
Fuel oil
Lactose
Casein
Waste materials: •
Waste sulfite
liquor solids
o Semi -chemical
• ",pulp solrds "••i*,>V
Sewage sludge
. primary .,
Sewage sludge
activated
Btu/lb
61,000
21 ,460
14,093
6,270
1.203
14,950
19,376
7,100
10,500
7,900.
„' " » *
5,812
.•*••*
7,820
" - «
6,540
Lb of Oe/lb Lb
of material of
7.937
3.42
2.66
1.07
0.178
2.53
3.26
1.13
1.75
1.32._ ,^
0.955
.". :„' «*-'::•
1,334'
• -/•;•' '.
1.191
of air/lb
material
34.34
14.8
11.53
4.6
0.77
10.9
14.0
4.87
7.55
5.70, . .
•» **
4.13
"* "' ' ' <
5.75
^ ,
5.14
Btu/lb
of air
1,780
1,450
1,220
1,365
l,56t>
1,370
1,380
1,455
1,395
1,335
1,410
1,365
1,270
TABLE 12. WAO EFFICIENCY FOR TEN PRIORITY POLLUTANTS (1-HR DETENTION TIME)(9)
Compound
Acenaphthene
Acrolein
Acrylonitrile
2-Chlorophenol
2, 4-Dimethyl phenol
2,4-Dinitrotoluene
1 ,2-Diphenylhydrazine
4-Nitrophenol
Pentachlorophenol
Phenol
Starting con-
centration (g/1)
7.0
8.41
8.06
12.41
8.22
10.0
5.0
10.0
5.0
10.0
% Starting material
320°C
99.96
>99.96*
99.91
99.86
99.99
99.88
99.98
99.%
99.88
99.97
destroyed
275°C
99.99
99.05
99.00t
94.96t
99.99
99.74
00.08
99.60
81.96-
99.77
The concentration remaining was less than the detection limit of 3 mg/1.
TThe % destruction for acrylonitrile, 2-chlorophenol, and oentachlorophenol
at 275°C were increased to 99.50, 99.88, and 97.3 by addition of cupnc
sulfate (catalyst).
55
-------
of 320°C (608°F) and 275°C (527°F), respectively. Bioassay tests conducted
on the raw and treated wastes indicated a 15 to 4,000 fold reduction in toxici-
ty (9). - '
As noted in Table 12, use Of a catalyst can'increase .the efficiency of
WAO for certain substances (thus permitting operation at lower temperatures).
Similar results have been reported in laboratory batch studies using 0.001 to
0.005 mole/2, of catalyst.s such as CuSQ^, Fe2(S04)3,'and H202(10).
Mobile Unit Applications
Currently there is a WAO mobile unit available far waste treatment. This
unit, which is owned and operated by Zimpro Inc., the developer of the process,
has a capacity of 23 &/hr (6 gal/hr) and is used for waste treatability demon-
stration studies (11). Although the trailer-mounted unit can be moved from
site-to-site, for economic reasons the unit has been stationed at the company's
headquarters in Rothschild, Wisconsin* and wastes for continuous treatability
tests are brought to the site in 1,890 a (500 gal) batches. Zimpro has a
2,300 Ji/hr (10 gpm) mobile unit currently under design which-will.be mounted
on two trailers. An isometric drawing of the unit is presented in Figure 8.
USE IN MOBILE UNITS FOR THE TREATMENT OF WASTE SPILLS
This section evaluates the feasibility of using WAO in a mobile unit to
treat concentrated waste spills, based on the general requirements discussed
previously.
Pertinent Features of WAO for Use in Mobile Units
The major limitations of the WAO for mobile unit applications are (a) the
relatively small capacity of the units suitable for trailer mounting, (b)
special design and material of construction requirements, (c) requirements for
skilled labor, and" (d) relatively high capi-tal costs.
As noted previously, Zimpro has a 54,000 i/day--(14,400 gal/day)-unit
currently under design. Th-i-s"unrtMs considered1 probably theJlargest capacity
unit that can be trailer truck-mounted. The unit, which will be mounted on
two trailers, thus will not-meet.-the mobile.unit size requirement ,of a single
trailer unit'with a capacity of 227,000 a/day"(60,000 gal/day); neither of the
two Zirapro trailers, however, would exceed the criteria for trailer weight
limit.
The WAO unit shown in Figure 8 is expected to be thermally self-sustaining
for wastewaters that have a minimum COD of 3,500 mg/i. More dilute wastewaters
require supplemental heating from the hot oil system. The WAO trailer unit has
been primarily designed ^or on-site use at a waste generator's manufacturing
*
Most of the technical data presented in this section are those supplied by
Zimpro (letter from L.A. Schaefer, ?impro Inc., Rothschild, WI, March 25,
1980).
56
-------
i INSTRUMENT PANEL
HIGH PRESSURE PUMPS
MOTOR CONTROL CENTER
FUEL OIL TANK
HOT OIL HEATER
REACTORS
PCVS
SOLVENT PUMP
SOLVENT TANK
DOOR CONTROL
COOLER
DC BLOWER
HEAT EXCHANGER
i
AIR COMPRESSOR |
1
Figure 8. Wet air oxidation mobile unit.*
*Provided by L.A. Schaefer, Zimpro Inc., Rothschild, WI, March 25, 1980.
-------
facilities. As a result, an electric generator (200 kw) and cooling water
supply (40 gpm) have not been incorporated into the design:' Including'a gener-
ator, however, is not considered to pose a problem. The cooling water could
be obtained in an emergency, short term situation'from a stream or lake.
Failing that, a small self-contained package cooling tower, or refrigeration
unit could be incorporated into'the design./ "' "" . '"' • •"
Although ineffective in handling mineral wastes, WAO .is most suitable for
treatment and detoxification of organic and oxidizable inorganic wastes. The
process does not have'the toxicity sensitivity of biological treatment or the
very large fuel cost of disposal by incineration. In contrast to ot^er techno-
logies which concentrate the waste to smaller volumes for subsequent disposal
to landfills or by incineration, WAO renders the organic wastes innocuous.
To be compatible with the spectrum of wastes (including highly corrosive
wastes) anticipated in spill situations, a WAO mobile unit must be desioned to
enable operation at very high temperatures and pressures and the reactor, heat
exchanger, and fluid conduits should be constructed of titanium. Even when
titanium is used as the material of construction, potential problems could
exist if the pH is less than 2 and chlorides were simultaneously greater than
15 to 20 mg/i. Problems could also exist with fluorides at a pH greater than
7 and highly caustic situations where the pH is greater than 10. Because of
its titanium construction, the capital cost of WAO is considerably higher than
those of alternative processes (for example, RO).
A mobile WAO unit could be set up and started in twelve hours; once appro-
priate conversions are made, actual start-up would take about four hours. Even
though WAO designs would incorporate a number of safeguards against possible
explosions, because of the high temperature and pressure nature of the opera-
tion, it is very desirable to have the unit operated by trained personnel. The
only requirement for periodic-operations and-subsequent storage i-s'the need
to drain the water from the unit for freeze protection purposes.
Engineering-Evaluation and Cost'Estimates* •""•-<*' •' «"•••".. *
Zimpro,has,,had its 2-3 £/hr0(6 gal/hr) •mopile^unit avail able'for'the last
five years, during~which time°a wide varietyVf'actual industr»ial"wastewaterss.
have been treated. The mobile unit shown in Figure 8 is under design and
should be available-in 1981. .I.f.tne'"waste'Sp.tVi rnauket oroves to be a viable
one, Zimpro does not anticipate any problems adapting its mobi'l-e unit experi-
ence to ma.ke the minor modifications necessary to be consistent with the guide-
lines for spill treatment equipment presented in this study. After building
150 WAO units over the last 20 years, Zimpro considers the small mobile units
relatively easy to build.
Though the capital cost for a WAO unit is high, the operatina cost is
low. The capital cost for the mobile unit shown in Figure 8 is around $1.25
*
Based on technical data provided by L.A. Schaefer, Zimpro Inc., Rothschild,
WI (letter dated March 25, 1980).
58
-------
to $1.5 million. This cost includes the costs of a diesel electric power
generator and all system components (high pressure displacement pump, air com-
pressor, process heat exchanger, hot oil heater, reactor and final cooler).
The operating costs for a mobile WAO unit would vary somewhat depending
on the COD of the wastewater. Assuming that in most field applications, sup-
plementary heating of the waste would be necessary to raise the waste temoera-
ture to the desired reactor temperature and that the mobile unit will have to
be powered by diesel electric generator, the fuel consumption for the generator
would be essentially the only operating cost. A 200 lew generator would consume
about 56 Vhr (15 gal/hr) of fuel. This would result in a cost of S15/hr, not
including labor, which is approximately 2.5<£ per gallon of waste that is de-
toxified.
REFERENCES
1. Process Design Manual for Sludge Treatment and Disposal, EPA Report No.
625/1-74-006.
2. Teletzke, G.H. Wet Air Oxidation, Chemical Engineering Progress, Vol. 60,
No. 1, p. 33, 1964.
3. Liptak, E.G. Environmental Engineers1'Handbook, Vol. 1. Water Pollution,
Chilton Book,Co., Radnor, PA, 1974.
4. Pradt, L.A. Developments in Wet Air Oxidation, Chemical Engineering
Progress, VoK 68, No. 12, p. 72, 1972.
5. Hoeft, J.E. and C.L. Soukup. Wet Air Oxidation at Ontonagon, paper pre-
sented at TAPPI Akaline Pulping Conference, Williamsburg, Virginia, 1975.
6. Morgan, J.E. Innovation in Soda Black Liquor Recovery, presented at TAPPI
Non-Wood Plant Fiber Conference, Atlanta, GA, October 1973.
7. Water Purification Associates, Innovative Technologies for Water Pollution
Abatement, NTIS Report No. PB-247-390, 1975.
8. PAT Report, Wet Air Oxidation Comes of Ages, Environmental Science and
Technology, Vol. 9, No. 4, p. 300, 1975.
9. Randall, T.L. and P.V. Knopp. Detoxification of Specific Organic Substan-
ces by Wet Oxidation, paper presented at the 51st Water Pollution Control
Federation, Anaheim, Calif., September 1978.
10. Chowdhury, A.K. and L.W. Ross. Catalytic Wet Oxidation of Strong Waste
Waters, Water-1975, AIChE Symposium Series, No. 151, Vol. 71, p. 46, 1975.
11. Wilhelmi, A.R. and R.B. Ely. The Treatment of Toxic Industrial Waste-
waters by a Two-step Process, presented at 30th Annual Purdue Industrial
Waste Conference.
59
-------
SECTION 8
BIOLOGICAL TREATMENT
PROCESS DESCRIPTION
Process Principle
Biological oxidation involves the use of microorganisms to convert the
dissolved and oxidizable organic matter to organic or inorganic end products
and to agglomerating microbial floes (biological sludge) that can be removed
subsequently by settling and/or filtration. Biological oxidation can be
carried out aerobically (in the presence of dissolved oxygen) or anaerobically
(in the absence of dissolved oxygen). The end products of aerobic oxidation
are stable gaseous and dissolved inorganics such as carbon dioxide, water,
sulfate, and nitrate. The end products of anaerobic decomposition are pri-
marily methane ana C0£ with sulfur-bearing compounds also yielding reduced
sulfur species such as mercaptans and hydrogen sulfide. Anaerobic oxiaation
is considerably slower and less efficient and requires a greater degree of
process control than aerobic oxidation. It is primarily used for the stabi-
lization of organic sludges and concentrated organic wastes.
Being biological processes, both aerobic and anaerobic oxidation are
unsuitable for the treatment of wastes containing high concentration of toxic
substances (for example, heavy metals, phenols) and refractory organics (for
example, tertiary butyl alcohol or pyridine), wastes deficient in certain
nutrients (for example, primarily nitrogen and phosphorus compounds), wastes
having very Tow or high pH, and wastes that fluctuate widely in chemical com-
position. Some of these wastes, however, can be made amenable to biological
oxidation by proper pretreatment such as nutrient addition, pH adjustment,
dilution, extended acclimation, or removal of toxic elements (for example,
precipitation of heavy metals).
Process Design Considerations
Aerobic biological treatment systems must be designed to provide for (a)
an adequate level of dissolved oxygen in the wastewater, (b) retention of a
proper concentration of an active biomass in the biological "reactor" (in
relation to the concentration of organics in the feed wastewater (that is,
food-to-microorganism ratio), (c) sufficient mixing and contact time in the
reactor, and (d) subsequent settling and removal of solids to provide a
clarified effluent.
60
-------
The most widely used biological treatment systems are activated sludge,
trickling filters, lagoons (waste stabilization ponds), and oxidation towers.
From the standpoint of potential for use in mobile units, activated sludge is
probably the most relevant and is the only process considered in this assess-
ment.
The conventional activated sludge process consists of a biological reactor
unit containing a high concentration of microorganisms. Air is supplied either
by mechanical aeration or by a diffused air system. The treated waste is sent
to a clarifier for solids/liquid separation. A portion of the settled sludge
is recycled to the biological reactor to "seed" the raw wastewater; the excass
sludge is sent to disposal.
A recent advance in the activated sludge process is the development of
the high purity oxygen activated sludge process where high purity oxygen is
used instead of air for biological oxidation. A considerable saving in capital
and operating cost is realized because of higher oxygen utilization efficiency
that enables the use of smaller reactors, reduces power consumption, and im-
proves the settleability of the biological floes. The high purity activated
sludge process, developed and marketed by Union Carbide Corporation (New York,
NY) as the "UNOX" process and by Air Products and Chemicals, Inc. (Allentown,
PA), is currently used in a number of large scale applications involving
biological treatment of municipal and industrial wastewaters. Some distin-
guishing process performance'and operating features of the high purity oxygen
process include the following:
• Economical operation at high mixed-liquor DO levels (-10 mg/£ DO).
• Multistage or plug flow operation at high organic loadings and high
MLSS levels without oxygen limitation.
• High volumetric oxygenation capacity per unit of gas-liquid contacting
power input.
• Operation under high rate, high MLSS levels with good sludge settle-
ability, compactability, and low sludge recycle rations.
• Low sludge production under low retention time, high organic loading
conditions.
Table 13 compares the process design conditions for the high purity
oxygen system and for the conventional air activated sludge process for typi-
cal municipal wastewater.
The high purity oxygen system, which appears to be a suitable bioloaical
oxidation process for use in mobile units, is briefly reviewed in the following
section.
The High Purity Oxygen Activated Sludge Process Description
Figure 9 is a schematic diagram of the high purity system. The reactor
tank is divided into several essentially identical sections or stages by means
61
-------
TABLE 13. COMPARISON OF PROCESS DESIGN CONDITIONS FOR THE HIGH
PURITY OXYGEN SYSTEM AND FOR CONVENTIONAL AIR
AERATION SYSTEMS FOR TYPICAL MUNICIPAL WASTEWATER (1)
Parameter
UNOX -
oxygenation system
Conventional air
aeration systems
Mixed-liquor DO level -
Aeration detention time (raw
flow only) - hours
MLSS concentration - mg/Ji
MLVSS concentration - mg/£
Volumetric organic loading -
Ibs BOD/day/1,000 cu ft
Food biomass ratio - Ibs BOO/
Ib MLVSS
Recycle sludge ratio - Ibs
recycle/1b feed
Recycle sludge concentration -
mg/a
Sludge volume index (Mohlman)
Sludge production - Ibs VSS/
Ib BOD removed
6-10
1 - 2
6,000 - 10,000
3,900 - 6,500
150 - 200
0.4 - 0.8
0.2 - 0.4
20,000 - 40,000
30'- 50
0.3 - 0.45
1 - 2
3-6
1,500 - 4,000
900 - 2,600
30 - 60
0.3 - 0.6
0.3 - 1.0
5,000 - 15,000
100 - 150
0.5 - 0.75
of baffles and is covered with an air-tight lid. High purity (90-100 percent by
volume) oxygen gas is fed into the first stage at a pressure of about 2.5 to
10 cm (4 in) of water column above ambient. Feed wastewater and recycled
sludge are also introduced into the first stage. The liquid and gas phases
then flow concurrently through the system.
The successive aeration stages are connected to each other with only a
slight pressure drop between stages. This allows the gas to flow freely from
one stage to the next, but prevents backmixing or interstage mixing of the
aeration gas. Within a given stage, gas is recirculated at a rate usually
higher than the rate of gas flow from one stage to another. As a higher pro-
portion of the oxygen demand is met in the initial stages, the volume of gas
required to maintain the desired dissolved oxygen level in the mixed-liquor
will be lower in the second and third stages than in the first stage.
Mass transfer and mixing within each stage is accomplished either with
surface aerators or with a submerged-turbine rotating-sparge system. In the
62
-------
AERATION
CONTROL TANK COVER
VALVE
AGITATOR
OXYGEN.
FEED GAS
WASTE
LIQUOR •
FEED
RECYCLE.
SLUDGE
EXHAUST
GAS
MIXED LIQUOR
' EFFLUENT TO
CLARIFIER
STAGE BAFFLE
figure 9. Schematic diagram of high purity oxygen system (1),
-------
first case, mass transfer occurs in the gas space; in the latter, gas is
sparged into the mixed-liquor where mass transfer occurs from the gas bubbles
to the bulk liquid. In both cases, the mass-transfer process is enhanced by
the high oxygen partial pressure maintained under the tank covers in each
stage.
Since the oxygen gas fed to the system is devoid of nearly all nitrogen,
and since approximately 90 percent of the oxygen gas normally is used, the
total gas venting from the system is relatively small. Since the reactors
are covered, gas is vented at a single point, thus allowing for effective
control of odor and biological aerosols.
Process control for this system is relatively simple. A small positive
pressure is maintained by the feed gas flow controller. As the organic load
and respiration (oxygen demand) of the biomass increases, the pressure tends
to decrease and feed oxygen flow into the system increases to re-establish
the pressure set point of the controller. Feed oxygen to the multiples
system can be controlled on this pressure demand basis by a simple regulator,
or differential pressure controller, automatic valve compilation.
Sludge separation and oxygen supply can be achieved by conventional
means. Oxygen can be generated on-site, although for a mobile treatment unit,
purchased liquid oxygen would probably be more economical. For large apolica-
tions, oxygen is lenerated on-site from air by the cryogenic air separation
process or by the pressure-swing adsorption process. In mobile unit appli-
cations, oxygen can be supplied from a liquid oxygen supply tank. Sludge
separation and effluent clarification can be achieved using conventional
gravity settling and/or filtration.
PROCESS APPLICATIONS AND RELATED STUDIES
Commercial Scale Municipal/Industrial Applications
The high purity oxygen process was first given a full-scale demonstration
in 1968 at Batavia, NY, by Union Carbide (2). Since that time there have been
numerous full-scale municipal and industrial applications of the process. The
process has partially or completely replaced conventional air activated sludge
at a number of municipal wastewater treatment plants (3). It has also been
successfully used to treat industrial wastes including brewery waste, citrus
waste, and wastewater from a chemical plant producing silicone based products.
Table 14 summarizes some data on these applications. Full-scale process de-
signs have also been proposed for Kraft mill effluent (4) as a result of suc-
cessful pilot plant programs. A 5.67-million liter per day (1.5 mgd) plant
treating high strength citrus waste has been in operation in California since
1976*. The plant handles average raw wastewater COD and suspended solids con-
centrations of 9,500 ppm and 4,000 ppm, respectively, and produces an effluent
with a COD of 300 to 400 ppm and a suspended solids concentration of 50 to
*
Letter from Keith H. Conarroe, Sunkist Growers, Inc., Corona, CA, November
1979.
64
-------
en
en
TABLE 14. THE HIGH PURITY OXYGEN SYSTEM PERFORMANCE
IN FULL-SCALE APPLICATIONS
Wastewater
Brewery
waste
Municipal
sewage
Silicon
chemical
plant
Citrus
waste
Influent characteristics, mg/A . Percent Removal
BOD COD TSS TOC BOD COD TSS TOC
e
1,632 2,844 536 ' 97 94 88
136-240 290-365 101-159 90-110 81-94 72-83 79-92 50-81
f
>
425 595 - , 95
t +
9,500 4,000 - - 96 98 -
Reference
5
6
7
*
*
Letter from Keith H. Conarroe, Sunkist Growers, Inc., Corona, CA, November 1979.
-------
100 ppm. Because of the low pH and nutrient deficiency of the wastewater,
ammonium hydroxide and phosphoric acid are added to the raw wastewater. The
reactor carries a MLVSS concentration of 5,100 ppm and is operated at a deten-
tion time of 37 hours and a DO level of 6 to 10 ppm. When first started and
proper "seed" was unavailable, it took about two weeks to achieve full, stable
operation.
Pilot and Laboratory Scale Studies
By 1973 over 40 municipal and industrial plants had used mobile high
purity oxygen pilot plants to test the feasibility of using an oxygen activa-
ted sludge system to treat their wastewater. Representative .industries whose
wastewaters were treated include petrochemical, Kraft mill pulp and paper,
pharmaceutical, brewery and food processing, meat packing, grain, cereal,
syrup processing, textile, poultry processing, fruit canneries, and mixed in-
dustrial operations (8). Typical results obtained in some of these pilot
plant operations are shown in Table 15.
High purity oxygen activated sludge has been tested on industrial waste-
waters containing as high as 9,500 ppm of COD (see Table 14). Application to
high concentration wastes, however, has required a very long detention time
to achieve high levels of organics removal. Air Products and Chemicals re-
ports having treated an industrial wastewater containing 2,500 ppm BODs with
a detention time of 20 to 25 hours*. The Union Carbide UNOX system has been
pilot tested on a chemical plant wastewater containing 3,000 ppm BODs; a BODs
removal efficiency of 90 to 95 percent was obtained when the system was
operated with a detention time of 10 to 12 hourst. UNOX process has been
pilot tested for the treatment of diluted coal conversion wastewater. The
wastewater (see Table 16 for typical composition) was diluted 1:19 with river
water and treated at an organic loading of 0.54 kg BODs/day/kg of MLVSS using
a detention time of 19.6 hours and a MLVSS of 3,800 mg/a in the aeration
tank. The results obtained are summarized in Table 17.
Mobile Unit Applications
Air Products and Chemicals, Inc. and Union Carbide Corporation operate
mobile, high purity oxygen, activated sludge process pilot plants that are
used for wastewater treatability studies. These pilot plants have a hydraulic
capacity of 15 to 23 £/min (4 to 6 gpm). The majority of pilot plant studies
using the UNOX process have been carried out'in mobile pilot plant units
supplied by Union'Carbide Corporation. Union Carbide presently has seven
mobile pilot plants available for UNOX wastewater treatment studies7. These
pilot plants, which have a hydraulic capacity of approximately 19 to 151
£/min (5 to 40 gpm), are contained within 12 m (40 ft) van trailers and in-
clude an external clarifier (see Figure 10). The van includes a 6,048 I
Letter from Roy Lagslet, Air Products and Chenicals, Allentown, PA, March 14,
1980.
TTelephone communications with S.O. Akinbami, Union Carbide Corporation, Linde
Division, Tonawanda, NY, November 1979.
66
-------
TABLE 15. PERFORMANCE OF THE HIGH PURITY OXYGEN SYSTEM PILOT PLANTS
Influent concentration, mg/«, Percent removal Reference
Petrochemical
complex
Petrochemical
complex
Pulp and paper
mill
Kraft mill
effluent
BOD
885
2,670-4,040
277-445
215-291
COD TSS
1,905 85
4,030-7,150
874-1,383
814-1,040 86-109
TOC BOD
97
1,310-2,350 77-90
91-93
77-91
COD TSS TOC
75
54-76 - 60-76
52-60 -
35-45 -
7
8
8
4
Municipal plus
chemical indusl
iron processing
and textile waste
-------
TABLE 16. TYPICAL COAL CONVERSION WASTEWATER CHARACTERISTICS (7)
(WASTEWATER DILUTED 1+19 WITH RIVER WATER PRIOR
TO TREATMENT)
Constituent/parameter mg/2
BOD5
COD
TOC
TC
Phenols
Acetone
Methyl ethyl ketone
Pyrrole
Pyridine
Cresols
Xylenols
Oil and grease
Ammonia, as N
TKN, as N
Nitrate, as N
Total sulfur, as S
PH
30,000
50,000
12,000
13,000
8,500
150
100
150
no
2,800
1,300
500
4,600
4,700
200
3,000
9.5 units
TABLE 17. DILUTED COAL CONVERSION WASTEWATER TREATMENT RESULTS
WITH THE HIGH PURITY OXYGEN PROCESS (7)
Characteristic Influent Effluent Percent
removal
BOD5, mg/i, 1,700 21 99
COD, mg/£ 2,800 177 9*
TOC, mg/i 690 33 95
TSS, mg/£ 140 64 55
Oils, mg/s. 30 10 70
Phenols, mg/i 700 1.2 99+
68
-------
en
10
47 VS' \. LABORATORY
Figure 10. Schematic diagram of the UHQX Mobile Unit Pilot Plant and the external clarifier.*
* Drawing provided by S.O. Akmbanri, Union Carbide Corporation, Linde Division, Tonawanda, NY, October
23, 1979.
-------
(1,600 gal) four-stage UNOX reactor. Each stage (1,512 t, or 400 gal in capa-
city) contains a submerged turbine rotating sparger device, which'includes a
mechanical mixer and a gas recirculation compressor. Rotameters are provided
on each stage to monitor and control the recirculation flow. Influent, return
sludge, and sludge wasting pumps are provided. These are variable speed,
positive displacement Moyno pumps. Influent and return sludge can be fed to
any of the four reactor stages. An in-line oxygen gas composition system is
provided (Servomex analyzer), as is a hydrocarbon monitoring system (LEL
meter). A dry gas meter is used to monitor the oxygen feed flow and a wet
drum meter is used to monitor the vent'gas flow. Portable pH, dissolved
oxygen, and sludge blanket meters are provided to obtain daily operational
data. The mobile unit also houses a laboratory for wa.stewater analysis and
process control work.
The secondary clarifier associated with the UNOX pilot plant is 2.1 m
(7 ft) in diameter and 3.35 m (11 ft) deep. This unit is externally mounted
and requires a hole, approximately 1.5m (5 ft) deep. In field applications,
the trailer itself is positioned on a firm, flat surface, such as a roadway.
Ideally, this would be adjacent to a non-paved area for clarifier installa-
tion.
When handling wastes containing volatile flammable organics, the buildup
of volatile hydrocarbons in the oxygen rich space of the UNOX system could
constitute a hazard. The system, however, does include features for the eli-
mination of combustible gas buildup in the reactor. To eliminate ignition
sources, no electrical components are installed under the tank cover and no
metal-to-metal contact of moving parts is present. To-eliminate the collec-
tion of combustible vapors, combustible gas analyzers are employed to conti-
nuously monitor for the presence of combustible materials. Should an approach
to a lower explosive limit (LEL) occur, the analyzer activates the necessary
controls that cause the gas space to be purged with air until the gas and
liquid have carried the combustible material out of the system.
Union Carbide Corporation, which markets the'UNOX process, offers a total
pilot plant program for the evaluation of the applicability of the UNOX process
to the treatment of specific wastewater and for the development of criteria
for the design of full-scale units. The program includes the pilot plant
equipment described above (that is, van and clarifier), a full-time on-site
technician (40 hour work week and 24 hour on-call),and the services of a pilot
plant engineer who-directs'the program from-Tonawanda (NY) and makes periodic
site visits. The oxygen required in these programs is also provided by Union
Carbide. The customer is expected to provide installation (teardown)'labor,
operating manpower, analytical support, and utilities. A typical installation
requires three to four days each'for an electrician and a pipefitter or mill-
wright. A "cherry picker" or small crane is also required for the clarifier
installation. The operating manpower consists of 24-hour coverage to take
samples and routine measurements. The utilities include 240-volt, single-
phase, 200-amp electrical service* and a supply of potable water (for example,
*
Although the existing UNOX pilot plant requires an external power supply
source, there is sufficient room on the van for insta^ation of a diesel
electric power generator.
70
-------
a garden hose supply is adequate).
The mobile UNOX system can be either rented or purchased from the Union
Carbide Corporation. The estimated rental costs are as follows*:
(a) $5,000 for the checkout and refurbishment of equipment to make
it operational
(b) $420/day for on-site service including engineering consultation
on program plan and execution
(c) $70/day rental of equipment
(d) transportation charges to get the equipment from Tonawanda, to
the site of operation and back again to Tonawanda
The purchase price for-tnVUN'dX mobile unit is-between $200,0'00 and
$250,000.
U'S'E OF HIGH'PURITY "OXYGEN ACTUATED -SLUDGE PROCESS IN'MOBILE'UNITS FOR TREAT-
MENT OF WASTE SPILLS
This 'section evaluates the feasibility-of using a high purity oxygen
activated sludge process .in a mobile unit -to treat concentrated waste spills.
Pertinent Features of High Purity Oxygen Activated Sludge Process for Use in
Mobile Units
Although high purity activated sludge process is commercially available
in mobile field units, being a biological process it suffers from a nunber of
limitations that may restrict its applicability or make extensive wastewater
pretreatment necessary. These limitations primarily relate to the requirement
for very long aeration time (and hence large reactor size) for treating con-
centrated wastes, inapplicability to wastes containing very high concentrations
of toxic substances or refractory organics, and the considerable time reauired
for process startup and waste acclimation. Since many chemical spill incidents
involve toxic or refractory chemicals and highly concentrated wastes, the use
of activated sludge process may not be applicable to many such situations.
Relatively dilute wastes (200-500 mg/2, 8005) that do not contain concentrations
of toxic substances* however, can be treated by high purity oxygen activated
sludge process in mobile units such as those that are currently available con-
mercially.
Another shortcoming of the biological processes that may limit their
application to short-term emergency response situations is the often lengthy
start-up time required and the requirement for elaborate and lengthy bench-
scale or small pilot plant scale testing to determine waste treatability and
The costs shown are 1979 estimates and do not include taxes and any perform-
ance bonds.
• 71
-------
optimum treatment conditions. Generally, it takes several weeks for activated
sludge systems to stabilize. This time period will vary depending on the type
of waste being treated. Start-up times can be decreased if the wastewater is
"seeded" with an optimal mix of microorganisms. With a mobile unit that
treats wastewaters of highly variable types, it may be extremely difficult to
maintain the variety of cultures needed to seed a specific wastewater.
As a biological process, high purity oxygen activated sludge process is
susceptible to "shock" loads because of fluctuations in the characteristics of
the raw wastewater. If such fluctuations are anticipated in field applica-
tions, use of an equalization basin would be necessary to "smooth" variations
in the quality of feed to the mobile unit.
High purity oxygen activated sludge process would not be suitable for
treatment of wastewaters containing high concentration of volatile hydro-
carbons and other organics (for example, when spills of ether, benzene, pen-
tane, etc., are involved). The presence of such low flash-point compounds
can present a potential fire hazard and, as noted earlier, the UNOX system is
equipped with hydrocarbon analyzers and control systems that deactivate the
system when dangerously high concentrations of volatile hydrocarbons are
detected in the oxygen-rich space of the reactor.
The operation of a mobile high purity activated sludge unit would require
trained operators. As with most other processes considered, the process
generates a residue (waste sludge) that may require further processing and
disposal. The quantity of sludge varies with the type of wastewater and sys-
tem design. For application to municipal sewage, about 0.35-0.45 kg of sludge
(VSS) is produced per kg of BODs removed. In field application, the waste
sludge may have to be collected and transported to hazardous waste management
facilities for disposal. As noted previously, oxygen for the process can be
provided from a liquid oxygen source that would be delivered to this site in
special containers*. The oxygen requirement varies with the process design
and organic loading and would typically be in the ,0.6 to 1.5-kg/kg of BODs
removed (10). Most waste spills would be deficient in nutrients (for examole,
nitrogen and phosphorus compounds)-or have a pH too, high or too. low for
effective'biooxidation. Such wastewaters would require addition of supplemen-
tary nutrients or acids or alkali for pH adjustment.
Engineering Evaluation and Cost Estimates
The largest existing high purity activated sludae process mobile unit
has a reactor capacity of 6,000 SL (1,600 gal). When handling municipal waste-
waters or dilute industrial wastewaters, the system can be operated with a
detention time of as low as one hour. At one-hour detention time, the system
*
Large quantities of liquid oxygen can be delivered with only a few hours
notice to virtually any place in the Continental United States. This capa-
bility currently provides oxygen supply backup to thousands of steel manu-
facturers, metal fabricators, hospitals, and chemical users throughout the
United States.
72
-------
would have a hydraulic capacity of 144,000 Vday (38,400 gpd). For treating
higher concentration wastewaters, a significantly higher detention time would
be required that would subsequently reduce the hydraulic capacity of the unit.
At a 24-hour detention time, which is commonly required for the treatment of
industrial wastes having BOOs concentrations in the 1,000-3,000 ppm range, a
unit with a reactor size of 6,000 i (1,600 gal) can handle a flow of only
6,000 Vday (1,600 gpd). The hydraulic capacity of the system would be re-
duced even further when higher detention times are required. It is estimated
that the largest reactor size that can be accommodated on a single trailer is
47,300 £ (12,500 gal); at a detention time of 48 hours, this reactor would
have a hydraulic capacity of only 23,650 £/day (6,250 gpd) that would be too
small for mobile unit application to the management of large spills.
The existing UNOX mobile units require an external power source. Tnese
units, however, can also carry a generator for on-board power generation. The
overall power required for liquid mixing and gas recirculation will vary with
specific system configuration, but will generally lie between 0.02 and 0.03 kw
per 1,000 a (0.08 and 0.14 HP per 1,000 gal) of mixed liquor under aeration.
For a 46,300 I (12,500 gal) reactor capacity, the power requirement would be
in the 0.74 to 1.3 kw range that can be provided with a 5 kw diesel engine.
As discussed previously, the oxygen required in the process can be pur-
chased in liquid form and would generally be delivered to the site by the
suppliers. .Typically, between 0.6 and 1.5 kg of oxygen would be required per
kg of BOD5 removed. Liquid oxygen is delivered in bulk or in tanks. The
price varies considerably; one supplier on the west coast quotes unit ourchase
prices for bulk and tank load liquid oxygen of $42/metric ton ($38/ton) and
$57/metric ton ($52/ton), respectively. The requirements for other chemicals
(nutrients and pH adjustment chemicals) would be dependent on the wastewater
characteristics and would generally be very small (typically, a BOD:N:P ratio
of 100:5:1 would be required for optimum biooxidation).
Typical rental and purchase costs and operating labor requirements for
the existing UNOX mobile units were presented earlier.
REFERENCES
1. Union Carbide UNOX System Wastewater Treatment, Union Carbide Corporation,
Linde Division, Bulletin F-3424.
2. Albertsson, J.G., et al. "Batavia I, EPA-PNI 17050 DN W, May 1970.
3. Chapman, T.D., et al. Effect of High Dissolved Oxygen Concentration in
Activated Sludge Systems, Journal of Water Pollution Control Federation
Vol. 48, No. 11, pp. 2486-2510, Nov. 1976.
4. Peterson, R.R. Design Criteria for High Purity Oxygen Treatment of
Kraft Mill Effluent, Journal of Water Pollution Control Federation
Vol. 47, No. 9, pp. 2317-2329, Sept. 1975.
73
-------
5. Bell, B.A. and J.M. Welday. Comparison of Completely Mixed Activated
Sludge and UNOX Treatment of Brewery Wastes, in Water-1977, AIChE
Symposium Series, No. 178, Vol. 4, pp. 29-36, 1978.
i *
6. Nash, N., et al. Oxygen Aeration at Newton Creek, Journal of Water
Pollution Control Federation 49(3), pp. 388-400, March 1977.
7. Hardisty, O.M. and H.E. Bishop, Jr. Wastewater Treatment Experience
at Organic Chemical Plants Using a Pure Oxygen System, in Water-!976:
II. Biological Wastewater Treatment, AIChE Symposium Series, No. 167,
Vol. 73, pp. 140-144, 1977.
8. Matsch, L.C. and W.C. Oedeke. Use of Pure Oxygen in the Secondary
Treatment of Wastewater, The Petroleum/Petrochemical Industry and the
Ecological Challenge, AIChE Symposium Series, No. 135, Vol. 69, pp. 175-
178, 1973.
9. Vaseleski, R.C. The UNOX Process: Effective Wastewater Treatment Practice,
in Water-1977, AIChE Symposium Series, No. 178, Vol. 4, pp. 23-28, 1978.
10. EPA Technology Transfer Seminar Publication, Oxygen Activated Sludge
Wastewater Treatment Systems - Design Criteria and Operating Experience,
August 1973.
74
-------
SECTION 9
COMBINED OZONATION AND ULTRAVIOLET RADIATION
PROCESS DESCRIPTION
Process Principle
Ozonation in combination with ultraviolet radiation (03/UV) is a newly-
developed chemical oxidation process that has proved effective for the des-
truction of many organic compounds including refractory chemicals and organo-
metallic complexes, and reduced inorganic substances in aqueous systems. The
oxidation takes place in a reactor where the waste is contacted with ozone
and UV radiation simultaneously. The use of ultraviolet radiation enhances
the oxidation power of ozones thereby increasing the reaction rate and oxida-
tion efficiency.
By itself, ozone is a very strong oxidizing agent with an oxidation power
higher than those of chemicals such as chlorine compounds, hydrogen peroxide,
and potassium permanganate which are commonly used for water disinfection and
oxidative treatment of waters and wastewaters. Ozone is extensively used in
Europe for water disinfection, for the treatment of waters containing iron
and manganese, and as a pretreatment method to convert toxic and refractory
substances in a wastewater into biodegradable materials. Ozone exists as a
gas at ordinary temperatures and pressures (bailing point of ozone is - 112°C
at atmospheric pressure)(l). It is only slightly soluble in water (about 20
times as soluble as oxygen). Because of its low solubility, supplying ozone
at a sufficiently fast rate to the reactor becomes a major mass transfer
problem in the treatment of concentrated wastes (1), especially those con-
taining substances that are rapidly oxidizable with ozone (for examole,
sulfides, nitrites, bacteria, phenols, and unsaturated organics). Ozone
transfer to the wastewater, however, would not be a major problem for sub-
stances that are oxidized very slowly with ozone (for example, acetic acid,
oxalic acid, urea, and unsaturated aliphatic alcohols).
In actual application where ozone is used alone, complex organic sub-
stances are only partially oxidized. These substances are generally converted
to intermediate products such as acetic and oxalic acids and to other products
that are either refractory to ozone oxidation or that undergo oxidation at
such a low rate that they are not removed from the system during the treatrent
period. The effectiveness of ozone for wastewater treatment, however, can be
enhanced by simultaneous application of ultraviolet radiation. Figure 11 is
a comparison of 03/UV combination treatment with the application of ozone
75
-------
03 OZONOLYSIS
REGION 1
OXIDATION OF INITIAL SPECIES
REGION 2
OXIDATION OF INTERMEDIATES
0,/UV
GOES TO
COMPLETE
OXIDATION
REGION 3
UNOXIDZABLE
REFRACTORY
PRODUCTS
n DIMENSIONLESS TIME
Figure 11. Comparison of O^/UV combination treatment with ozonation (2)
(a = fraction of ozone used by the reaction).
76
-------
alone. The figure indicates that 03/UV is superior to ozonation both in
terms of rates and efficiency of TOC removed.
While the specific role of the ultraviolet radiation in enhancing the
efficiency of ozonation is still under investigation, it is currently hypothe-
tized that the major effect of radiation is to bring about a photodecomposi-
tion of the substances undergoing oxidation, thereby converting them to more
reactive "free radical" species.
Table 18 is a list of organic compounds that have been determined to be
economically treated by 03/UV process. The substances listed range from
simple polar aliphatic compounds such as acetaldehyde and acetic acid (which
cannot generally be removed from wastewaters by carbon adsorption) to highly
nonbiodegradable complex and toxic compounds such as PCB and kepone. Although
not listed in the table, metal complexes of cyanides that cannot be oxidized
with chlorine or ozone alone are completely destroyed by 03/UV treatment (2).
Process Flow Diagram and Other Considerations in 03/UV Treatment
Figure 12 presents one commercial design for a 03/UV reactor for waste-
water treatment*. Since the optimizations of the 03/UV contacts with the
wastewater and the 03 mass transfer are important design parameters, the
reactor is divided into compartments (created by use of baffels) with cylin-
drical UV lights (similar to commercial fluorescent lights) placed vertically
at equal distances along the flow path in each compartment. Ozone, which is
generated on-site, is introduced as a gas into the reactor through diffusers
that are located at the bottom of each compartment. The excess ozone in the
reactor offgas is discharged to the atmosphere. In general, the specifics of
a reactor design such as reactor size, number of compartments, and spacing of
ozone diffusers and UV lights are determined by the characteristics of the
wastewater to be treated and the desired process efficiency. For most aopli-
cations, the optimum design can best be determined using pilot plant tests.
Wastewater characteristics which would impact process efficiency and
hence reactor design are wastewater flow, nature and concentration of sub-
stances to be oxidized and concentration of suspended solids which would inter-
fere with the passage of UV light through the wastewater, and wastewater
temperature. Even though the reaction rate increases with the rise in tem-
perature, increased temperature can reduce the overall efficiency by reducing
the half-life of ozone and its solubility in wastewater. Although pH of the
wastewater is not considered a critical factor in determining process effi-
ciency, for certain wastes (for example, cyanides) maintenance of a specific
pH level is necessary to avoid generation of toxic gases (for example, hydrogen
cyanide). Pilot plant studies should be conducted to determine the optimum
reactor design and operating conditions for specific wastes.
*
Another commercial design, the "multi-stage reactor", uses several reactors
in series instead of one reactor with multiple compartments. Each reactor
"Stage" is a "completely" mixed system with the mixing provided by a cen-
trally-located mechanical mixer.
" ' 77 •
-------
TABLE 18. LIST OF ORGANICS DETERMINED TO BE ECONOMICALLY
TREATABLE BY THE 0;/UV PROCESS (4)
J * v *
Acetaldehyde
Acetic acid
Alcohols ' "
Aldrin
Amines
Anisole
Benzoic acid
Chelating compounds
Chlorinated phenols
Chlorobenzene
Detergents
Dieldrin
Dioctylpthalate ' •
Endrin
Ethyl-ene dichloride
Formaldehyde
Formic acid
Glycerols
Glycine
,G1yco'ls
Hydroquinone
Kepone
Methyler>e chloride
.nitrobenzene'
Nitrophenol
Organic phosphates
Organosulfur compounds
Organo-tin compounds
PCB's
Phenol
•Phthalic acid
'RDX '"
Sodium,acetate
Styrerve
Sugars
TNT
Vinyl chloride
Xylenol
78
-------
GRAVITY OR PUMP
WATER FEED
OZONE IN
UV LAMPS
POWER
CONDUIT
SPENT OZONE
OUT
PURIFIED
WATER
OUT
(a) Reactor Module
OZONE DIFFUSERS
CONTAMINATED
PURIFIED V/A7£fl OUT
^£V&&l5wte«
(b) Cross Section
of Reactor
Figure 12. Schematics of an 0,/UV reactor.
79
-------
PROCESS APPLICATIONS AND RELATED STUDIES
Commercial Scale Municipal /Industr4a1 Applications •
Because of'the relatively new nature of -the 0-j/UV process, there are only
a few full-scale plants currently 1n operation. These plants, for which little
design and operating" data have been published, reportedly 'handle cyanide and
organic wastewaters from a tool company, ''photographic 'and metal plating waste
from a U.S. Army ammunition plant, arvd a' combination of organic nitrogen and
cyanide waste from a company in France (2). There are also a number of plants
in the various design phases including a plant to treat PCB from a capacitor
production facility and "pink water" ,from a U.S. Army ammunition plant (4).
Pilot and Laboratory Scale Studies
A number of pilot plant studies 'have been conducted or are currently in
progress to assess the applicability of 03/UV to the treatment of various
wastewaters including wastewaters from army field, hospitals, a contaminated
groundwater from an Army arsenal, "pink water" from an army ammunition plant,
organic contaminated brine -and elutriate containing PCB's, and wastewater
from a paint-stripping operation ,.(4), Typtpa.1 results obtained in some of
these applications are showrMn, Table 19.
A pilot plant was set up at General Electric Company's Capacitor Product
Department facilities in Hudson Falls, New York, to demonstrate the efficiency
and cost effectiveness of the Os/UV system for destruction of PCB's (5 to ^0
ppb) from an industrial effluent (5). Results of this study indicated that it
is both feasible and economical to reduce PCB's to below one ppb. The capital
costs for a 151,400 a/day (40,000 gpd) and 567,800 «./day (150,000 gpd) plant
were estimated to be $124,500 and $300,000, respectively; the operating costs
were estimated to be $1.15/1,000 SL ($4,35/1,000 gal) and $0.58/1,000 i ($2.21/
1,000 gal), respectively. In another study that compared the cost of 03/UV
and carbon adsorption to remove PCB's to one ppb level, it was found that
is actually, more economical ,tha.n the ..carbon, absorption (see Table 20) (6).
Pilot studies have shown that for some wastewaters it is advantageous to
use several reactors in stages instead of a single-stage or a single-conoart-
ment reactor (6). As discussed earlier, ozone mass transfer may be the rate
limiting step for oxidizing certain chemicals. With staged reactors/compart-
ments the ozone and UV dosages can be adjusted in each stage/comoartment to
the actual "demand" that varies as the treatment progresses. Table 21 shows
the treatment results for a six-stage reactor treating a wastewater containing
metal -complexed cyanides. The data show that ozone alone in the first few
stages can reduce the complexed cyanide to very low levels, with the UV light
used primarily as the "polishing" treatment in the last stages. Final concen-
tration of less than 0.3 ppm of complexed Fe cyanide, and less than 0.1 ppm of
Cu- or Ni-complexed cyanides can thus be obtained.
The effectiveness of 03/UV and ozonation alone have been compared in a
number of laboratory and pilot plant studies using a spectrum of wastewaters
including secondary effluent, TNT and hospital wastes j(4), and synthetic
80
-------
TABLE 19. TYPICAL 03/UV PILOT PLANT TEST RESULTS (3)
Wastewater
Pink water
Ground water
Dredge spoil
elutriate
Influent
analysis, nig/2.
70 TOC
76 TNT
2.6 DIMP* •
2.6 TOC
0.030 PCB's .
Effluent
analysis, mg/£
5 TOC
1 TNT
0.012 DIMP
2.0 TOC
<0.00> PCB's
Residence
time, min
177
95
83
03 mass
flow, rng/min
721
400
410
No. of 40w
UV lamps
29
29
18
*
Disporopylmethyl phosphonate.
-------
TABLE 20. COMPARISON OF THE ECONOMICS OF Oj/UV AND CARBON ADSORPTION
FOR THE TREATMENT OF PCB WA5TEWATERS (6)
Flow (gpm)
Capital cost ($1,000)
0,/UV
O
Carbon adsorption
Operating cost ($/l,000 gal)
Including amortization
o3/uv
Carbon adsorption
'|0
152
174
3.49
4.17
160
422
570
1.72
1.73
1,700
1,380
3/07
1.09
1.11
TABLE 21. "OXIDATION OF'METAL COM.PL-EXED CYANIDE IN A STAGED-REACTOR (6)
Cu-complexed
CN, mg/«,
Cyanate, mg/£
Temp., °C
PH
UV watts/ i
Ni-complexed
CN, rag/x,
Cyanate, mg/ji
Temp., "C
PH
UV watts/ i
Fe-complexed
CN, mg/x.
Cyanate, mg/£
Temp., °C
PH
UV watts/ t
Influent
4,000
0
20
11.5
-
4,000
0
20
11.8
-
4,000
0
20
-
-
1
17
-
"20
-
0
no
-
20
-
0
2,680
-
66
-
0
Effluent
2
0.5
-
20-
-
0
0.6
-
20
-
0
1,630
-
66
-
0
from reactor stage
3 45 6
<0.1
no
66 . -
7.7
1.2
<0.1
470
66
8.5
1.2
710 105 13 <0.3
47
66 66 66 66
8.9
0 0 0 1.2
82
-------
wastewaters using model chemicals such as pentachlorophenol, malathion, and
iron-organic complexes (6). The results from these studies provide the basis
for the generalized plot in Figure 11. When used for disinfection, 03/UV has
been shown to require only 10 percent as much ozone as ozonation for achieving
the same degree of disinfection (6).
Mobile Unit Applications
At the present time, there is no report as to the existence of any 03/UV
mobile treatment units.
USE OF 03/UV IN MOBILE UNITS FOR TREATMENT OF WASTE SPILLS
Pertinent Features of O-^/UV Treatment for Use in Mobile Units
Based on discussions with the process developers, nearly all the require-
ments for the applicability of a process to the treatment of waste spills in
a mobile unit would be met by the 03/UV process. The 03/UV reactors are
modular, portable, and commercially available. The system can-be started and
shut down relatively quickly, can be serviced conveniently, would not require
skilled labor for field operation, and can be operated with the electric power
produced on-site by.a diesel generator. The system can handle troublesome
wastes (such as cyanide) and convert them to harmless end products (such as
water and €03). The major shortcomings of the 03/UV unit relate to the limited
amount of ozone that can be generated on-site and the discharge of residual
ozone to the atmosphere. The amount of ozone that can be generated on-site in
a mobile unit may be less than the quantity required in an application. Os/UV
is a process that has been commercialized only recently, and little operating
and design data are available for the process. 03/UV process is currently
being offered by the following two companies: Houston Research, Inc., Houston,
Texas and Westgate Research Corp., West Los Angeles, California. Houston
Research offers a "multi-stage" reactor design, whereas Westgate Research offers
a single-stage, m'ultiple-compartment design. Both designs offer units that are
modular, portable, and can be skid mounted. The modular design allows flexi-
bility that the treatment capacity can be altered via use of additional modules,
number and combination, of,.UV -lamps-, or adjustments of ozone flow to each stage.
The 03/UV units require minimal labor for operation and maintenance. The
process start-up merely involves turning on the wastewater pump, the UV lights,
and the ozone-generator. .Periodic cleaning of the reactor by opening the
drain valves at the bottom may be required to prevent excessive build-up of
sludge at the bottom of the reactor. The "burned" or "weak" UV lamps would
require replacement. On the average a UV lamp would have a life expectancy of
about one year.
The Os/UV process is especially suitable for handling hard-to-treat
wastes (for example, complexed cyanides and refractory organics). When
treating these wastewaters, the end products of 03/UV treatment are onmarily
harmless gases. The power for the operation of an 03/UV unit can be supplied
by a diesel electric generator that can be housed in the same or a separate
mobile unit.
83
-------
Some of the major shortcomings of the 03/UV process for mobile unit
application relate to the heavy weight and Targe size of the ozone generator
and the discharge of residual ozone to the atmosphere. The largest ozone
generator that can be installed on a trailer would generate 91 kg 03/day
(200 lb/day)*. As discussed before, theoretically, eight parts of ozone is
required to remove one part of TOC in wastewater. Thus a 91 kg/day 03 genera-
tor can theoretically handle a wastewater containing 11.3 kg/day of TOC. For
spills containing 10 ppm and 1,000 ppm TOC, the theoretical amounts of waste-
water that can be treated by the unit would be 1.13 x 10^ x,/day (3 x 105 gpd)
and 1.13 x 10^ i/day (3 x 10^ gpd), respectively. For large volume, concen-
trated waste spills, the 03/UV can be used as a polishing step after bulk of
the organics are removed by other treatment methods (for example, carbon
adsorption).
Although the commercial designs feature the use of a catalytic converter
for destruction of unused ozone in the reactor off-gas, the destruction effi-
ciency is usually less than 100 percent and some residual ozone (about 5 per-
cent of the ozone fed to the wastewater) would escape to the atmosphere. Good
ventilation is required for safety of the operating personnel. The maximum
ozone exposure levels proposed by various regulatory and pro/essional organi-
zations are listed in Table-22,
Engineering Evaluation and'Cost•Estimates -• -
. < ,
Based on data obtained from one 03/UV process developer, a preliminary
design has been developed for a 227,000 a/day (€0,000 gpd) 03/UV single-stage
multiple-compartment unit for use on a trailer flatbed. The unit would be
constructed of stainless steel, sized for one hour detention time, and would
have dimensions of 3.7 m long, 2.1 m wide, and 1.6 m high (21 ft x 7 ft x 5
ft). The UV light would be supplied by 336 40w UV lamps and the ozone by a
91 kg/d (200 Ib/d) generator that uses compressed air from an air compressor.
The entire system, including wastewater pump, air compressor, ozone generator,
reactor, and catalytic converter would weigh about 9,070 kg (20,000 Ib).
Power can be supplied by a 125 kw diesel generator. The power requirements
for the operation of the UV light and the ozone generator are estimated at
22 kw and 83 kw, respectively. The diesel generator would have the approximate
dimensions of 2.5 m (1) x 0.9 m (w) x 1.3 m (h) or 8.3 ft x 2.8 ft x 4.3 ft
and would weight about 2,270 kg (1,000 Ib).
The major logistics support required in field application relates to
possible requirements for UV lamp replacement. Under normal operation, it is
unlikely that lamp replacement would be required in the field during short-
term operations. Based on experience with pilot plant operation, a UV lamp
life of one year is estimated for continuous operation.
The estimated total capital cost for the 03/UV unit and the diesel elec-
tric generator are $265,000 and $20,000*, respectively. These costs include
*
Information supplied by Jack D. Zeff, Westgate Research Corporation, West Los
Angeles, CA, in a meeting on February 23, 1979.
84
-------
TABLE 22. OZONE EXPOSURE LEVELS PROPOSED BY VARIOUS REGULATORY
AND PROFESSIONAL ORGANIZATIONS
Agency/organi zati on
Exposure level
Occupational Safety and
Health Administration (OSHA)
American National Standards
Institute/American Society
for Testing and Materials
(ANSI/ASTM)
American Conference of
Government Industrial
Hygienists (ACGIH)
American Industrial Hygiene
Association
Maximum permissible exposure to airborne
concentrations of ozone not in excess of
0.1 mg/£ (by volume) averaged over an
8-hour work shift
Control occupational exposure such that
workers will not be exposed to ozone con-
centrations in excess of a tine weighted
average of 0.1 mg/Ji (by volume) for eight
hours or more per workday, and that no
worker be-exposed to a ceiling concentra-
tion of ozone in excess of 0.3 mg/a (by
volume) for^more than ten minutes
Maximum ozone level of 0.1 mg/x, (by volune)
for a normal 8-hour work day or 40-hour
work week, and a maximum concentration of
0.3 mg/£ (by volume) for exposure of up to
15 minutes.
Maximum concentration for 8-hour exposure
of 0.1 mg/a (by volume)
the equipment costs (air compressor, ozone generator, reactor with UV lamps,
catalytic converter, wastewater pump, piping, and valves) and the engineering
and installation fees.
The major operating and maintenance cost elements for the 03/UV unit are
electrical power for ozone generating and operation and replacement of UV
lamps. For field applications, the operating cost per unit volume of waste-
water treated would vary depending on the specific characteristics of a spill
situation (for example, waste characteristics, waste volume, location of
spill, etc.). Assuming that in most field applications the mobile unit will
have to be powered by a diesel electric generator, the fuel for the generator
would constitute one of the major items of operating expense. A 125 kw diesel
generator would consume about 76 fc/hr (10 gal/hr) of fuel ($6/hr).
-------
REFERENCES
1. Miller, G.W., R.G. Rice, et al. An Assessment of Ozone and Chlorine
Dioxide Technologies for Treatment of Municipal Water Supplies, EPA-
600/2-78-147.
2. Prengle, H.W., Jr. Recent Experience - Os/UV Photo-Oxidation of Indus-
trial Wastewater and Source Water Components., presented at 101 Ozone
Symposium and Exposition, Los Angeles, California. May 23-25, 1978.
3. Leitis, E. An Investigation into the Chemistry of 'UV-Ozone Purifica-
tion Process , annual report on research supported by NSF under Grant
Number ENV 76-24652, 1979. ' -
4. Zeff, J.D. Ultrox Process Treatment of Organic Wastewater, presented
at the 3rd Annual Conference on Treatment and Disposal of Industrial
Wastewaters and Residues, Houston, Texas. April 18-20, 1978.
5. Arisman, R.K., T.C. Crase, et al. Destruction of PCB's in Industrial
and Sanitary Waste Effluents by the UltroxR (UV-Ozone) Process,
presented at the AIChE 86th National Meeting, Houston, Texas. April 1-5,
1979.
6. Prengle, H.W., Jr. and C.E. Mauk. New Technology: Ozone/UV Chemical
Oxidation Wastewater Process for Metal Complexes, Organic Species and
Disinfection, Water-1977, AIChE Symposium Series 178, Vol. 76, p. 228,
1978.
86
-------
SECTION 10
CHEMICAL TREATMENT
(Precipitation/Coagulation)
PROCESS DESCRIPTION
Process Principle
Removal of substances from aqueous systems. by addition of chemicals that
form insoluble precipitates with such substances is widely used for the treat-
ment of municipal • and industrial waters and wastewaters. Examples of such
'applications include hydroxide and carbonate precipitation of Ca++ and Ma"1"*"
(water softening), sulfide and hydroxide precipitation of Zn++, Cd++, Ni^+,
Pb++, etc., and Ca(II), Al(III), and Fe(UI) precipitation of phosphates.
Concentrations of cations and anions in aqueous systems are governed by the
"solubility product constant" (Ksp) for the insoluble precipitate. For
example, the solubility of Cu in water is governed by Ksp for Cu(OH)2 (or
Ksp for CuS when sulfide is also present):
= 10"19 at 20°C
where [] denote concentrations in mole/A. " •
At a pH of 5.0 ([OH"] = 10"9 mole/A), Cu++ concentration in v/ater cannot
theoretically exceed 0.1 mole/i or 6,500 ppm. By increasing the pH to 9
(through addition of sodium hydroxide or lime), the solubility of Cu++ would
reduce to 10~9 mole/2, or 0.065 ppb. Thus, near complete removal of Cu"1"1" from
solution should be achieved by hydroxide precipitation.
Most metal -hydroxide precipitation systems have amphoteric properties in
that they are capable of exhibiting either acidic or basic characteristics
depending on the pH level. In these systems, the metal cation becomes
"resolubalized" as the pH is raised beyond the level of minimum solubility.
As deoicted in Figure 13, the pH of minimum solubility for Cu++ is close to 9.
As the pH is raised beyond this level, the concentration of Cu++ in solution
increases because of the formation of anionic complexes such as Cu(OH)~3.
Fiqure 13 also indicates that the sulfides of the cations shown are considera-
- 87
-------
o>
•a
o
a
3
i
102
10°
10-2
10-4
10"8
10'*
10-12 -
Pb(OH),
CRfOH),
2n(OH),
Ag(OH)
Cu(OH),
Ni(OH),
Cd(OH),
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
PH
Figure 13. Solubility of metal hydroxides and sulfides.
88
-------
bly more insoluble than their hydroxides and hence more complete precipitation
of the cations can be achieved at lower pH,levels via addition of sulfide ion.
Hydrolyzing cations such as Al(III) and Fe(III) have been widely used in
water and wastewater treatment for the removal of colloidal organic and in-
organic particles. Although the exact mechanism of removal (coagulation) is
not completely understood (despite extensive research), two theories have been
forwarded to explain the observed phenomenon. The physical theory considers
colloidal particles to be negatively charged and the coagulation to result
from neutralization of the charge with the positively charged coagulant ions
(Fe+++ and Al+++) and their hydroxide complexes (for example, A1(OH)++,
Fe(OH)+2, etc.). The chemical theory considers coagulation to result from
specific chemical reactions between the coagulant ions and their hydroxide
complexes and the surface groups on the colloidal particles. In most coagu-
lation systems a combination of both charge neutralization and chemical reac-
tions is probably operative.
Determination of the Optimum Conditions for Precipitation and Coagulation
• Solubility-considerations such as those illustrated in the above example
for the precipitation of Cu++ coupled with chemical analysis information on
wastewater constituents can provide a reasonably accurate basis for defining
the optimum.conditions and chemical dosage requirements for precipitation of
metal ions by addition of appropriate anions (for example, hydroxide, sulfide,
carbonate or phosphate). Practical applications emphasize identification of
operating conditions that are most conducive to the formation of readily
settleable and/or filterable precipitates.
Many factors affect efficiency of coagulation/precipitation with Al(III)
and Fe(III) salts. These factors, which include coagulant dose, pH, tempera-
ture, and ionic composition of^the water, -determine the nature and charge of
the coagulant species and colloidal particles and anionic species that parti-
cipate in the coagulation or precipitation reactions. Because of the extremely
complex aqueous chemistry of Al(III) and Fe(III) salts and the coagulation/
precipitation reactions, it is not possible to estimate the optimum conditions
for coagulation/precipitation based on raw water analysis data. In practice
the optimum conditions (pH and chemical dosage) are usually determined by a
"trial-and-error" approach using the conventional "jar test".
The optimum pH of coagulation for the removal of turbidity from most
natural waters is between 6 and 8 for Al(III) and Fe(III) salts. Removal of
natural coloring matter and phosphates from waters and wastewaters can best
be accomplished at a pH near 5 for A!(Ill) and close'to 4 for Fe(III). For
some systems such as polyphosphates, the pH range for optimum removal is
extremely narrow with both Al(III) and Fe(III) salts (see Figure 14), and
unless the pH is maintained within this narrow pH range (something which is
impractical in actual water and wastewater treatment application), little re-
moval can be expected. In the case of.phosphates (ortho- and polyphosphates)
it has been demonstrated that the decrease in the removal efficiency in the
vicinity of the pH range for optimum removal and at coagulant doses exceeding
the optimum dosage is because of the dispersion of the colloidal pre^pitates.
89
-------
s
-3
"•3
crt
OC
O)UJ
Eo
—55
ZLU
Q5
IS
UJ £
o£
§2
a. a
ujO
g
en
IT
D
001
SOLUTION pH
Figure 14. Precipitation of pyrophosphate with Fe(III) at a 2:1 cation-
to-pyrophosphate equivalence ratio (1) (Initial
pyrophosphate concentration, 18 mg/£P).
90
-------
As shown in Figures 15 and 16, the optimum pH and dosage range for ortho-
phosphate removal can be extended when the water is passed through very fine
filters subsequent to chemical reaction. In actual water and wastewater
treatment application, the removal of coagulated or precipitated products in
any subsequent settling and/or filtration step is improved by addition of
coagulant aids such as organic polyelectrolytes to the water following addi-
tion of the coagulant salt.
Process Configuration and Considerations
Figure 17 is a flow diagram for the precipitation/coagulation and floc-
culation system. For industrial wastewaters that fluctuate in characteristics
the raw wastewaters are fed to a retention/equalization basin where they are
mixed to produce more constant feedwater quantity and quality. The equalized
wastewater is pumped to the coagulation-flocculation unit that usually con-
sists of three chambers: a mixing chamber, a flocculation compartment, and a
sedimentation chamber. In the mixing chamber, the wastewater is flash-mixed
with chemical coagulant*, coagulant aids and pH adjustment chemicals by means
of vertical or horizontal mechanical paddles. The wastewater passes from the
mixing chamber.to,the flocculation comoartment where it is agitated.by slowly
moving paddles. It then flows into the sedimentation chamber by means of
an inlet device that distributes the waste uniformly throughout the cross-
sectional ^area of the chamber.,. Clarified water,leaves the sedimentation
chamber over a weir and is usually filtered prior to discharge/use. Residual
sludge is scraped from the bottom of the sedimentation chamber and discharged.
Two basic types of systems are in commercial use: the sludge-blanket
(reactor/clarifier) unit that combines mixing,, flocculation', and settling in
a single unit and the conventional system using a rapid mix tank, followed by
a flocculation tank containing longitud-inal paddles that provide slow mixing,
and a conventional settling tank. -
The sludges produced with A!(Ill) and Fe(III) salts in the treatment of
certain industrial wastewaters (for example, pulp mill effluent which contains
a high concentration of organics) are usually very bulky, voluminous, and
difficult to dewater for ultimate disposal. Accordingly, coagulation with
these hydrolyzing coagulants has not been employed in full-scale units for
the treatment of concentrated waste streams. For certain such applications
lime has proved more effective and has been used in full-scale applications
where lime addition can be integrated into the production process (for example,
in the removal of organic color from Kraft pulp mill waste where lime precipi-
tation can be integrated with the chemical recovery step in the pulping
process). Lanthanide salts, which hydrolyze to a much lesser extent and can
effect removals over a*broader, pH .range than Al(III) and Fe(III) salts, have
been investigated for-application to wastewater treatment for phosphate
removal (3).
In smaller applications, chemicals are often added directly to the feed pump
inlet with the mixing taking effect in transit.
91
-------
LU
O
LU
12
11
2 10
a>
E
E
CO
§ 5
a.
O
i «
o
< 3
g
to
ai „
DC 2
O WHATMAN NO 42 FILTRATES
A 450 mfj MEMBRANE FILTRATES
D 100 m/j MEMBRANE FILTRATES
10
SOLUTION pH
Figure 15. Residual orthophosphate in precipitation of orthophosphate
with Fe(III) at a 1:1 cation-to-orthophosphate molar
ratio (2) (Initial orthophosphate concentration, 12 mg/aP).
-------
100
to
co
O
ai
OL
I
0.
O
I
cc
O
O
I i CC
I I UJ
o.
40 -
20 -
O
o
O WHATMAN NO 42 FILTRATES
V 450 mu MEMBRANE FILTRATES
D 100 m/J MEMBRANE FILTRATES
r\_
04 08 12 16 20 • 24 28 32 36
' : ' Fe (III)/PC^ MOLAR RATIO
Figure 16. Orthophosphate removal in Fe(III)-orthophosphate reaction at pH 4.0
(initial orthophosphate concentration, 12 mg/AP) (2).
-------
CHEMICAL FLOCCULANTS,
FLOCCULATION AIDS
AND pH ADJUSTERS
MECHANICAL
PADDLES
INFLUENT
WASTE RETEr
BA
.•
WON t N
ZATION ' *\ J
SIN • E=3>
PUMP
-
1°
2 ^
4>
k 1
rn;
-------
PROCESS APPLICATIONS AND RELATED STUDIES
Commercial Scale Municipal/Industrial Applications
Chemical precipitation and coagulation is widely u§ed in the treatment
of municipal and industrial water supplies for the removal of particulates,
hardness, and naturally occurring organic color ("humic" substances). Hydrox-
ide and sulfide precipitation have been used in full-scale processing of in-
dustrial wastewaters for the removal of heavy metals (for example, hydroxide
precipitation of dissolved copper and nickel from the effluent from copper
sulfate production, and of lead, chromium, and zinc from effluent from chrome
pigment production (4)).
Lime, ferric, aluminum, and ferrous salts have been used in combination
with biological or physical wastewater treatment processes to precipitate
phosphates, improve particulate/floe settling and filtering characteristics,
and to increase treatment plant throughput rate under overloaded conditions.
These applications generally involve dilute systems and the chemical dosages
are relatively small (100-300 ppm range). Where high concentration organic
wastewaters are encountered, aluminum and iron salts have proved ineffective
because of the very high chemical dosage requirement and the production of a
large quantity of a "bulky" sludge that cannot be economically processed and
disposed of. "Massive" lime treatment has been employed in large-scale
application for the treatment of highly colored bleach waste in Kraft pulp
mills. In this application the entire lime requirement of the mill is added
to the bleach waste to effect substantial color removal; the organic-laden
lime is then returned to the chemical recovery section of the mill where, as
a process chemical, lime is used to regenerate the pulping liquor.
Pilot and Laboratory Scale Studies
Numerous-laboratory and pilot plant studies have been reported in the
literature on-chemical treatment of actual and simulated industrial waste-
waters and landfill leachates with aluminum and iron salts and lime. The
most relevant of these studies from,the standpoint of possible application
to chemical treatment of concentrated waste spills and spill-impacted waters
are those pertaining to landfill leachates.
Laboratory studies of the chemical treatment of landfill leachate with
alum, ferric chloride, and lime have been reported (5). Treatment of a
leachate containing 9,100 mg/£ COD with art alum dosage of 1,000 mg/& yielded
a COD removal of only 5.1 percent and a "substantial" amount of sludge.
Treatment of the same leachate with 1,000 mg/Ji of ferric chloride and at an
optimum pH of 7.0 resulted in a COD removal efficiency of only 16 percent.
Treatment of a leachate containing 10,650 mg/Ji COD with a lime dosage of 1,840
mg/2. yielded a COD removal of only 3.5 percent,. Color removal with lime,
however, was excellent.
Chian and DeWalle (6) have summarized'the results reported in the lite^a-
ture on the effectiveness of various physical and chemical processes for the
-------
treatment of landfill leachate. Their data indicate reported COD removal
efficiencies with iron, alum, and lime ranging from 0. percent to as high as
31 percent. . • - - • • " •
USE IN MOBILE UNITS FOR THE TREATMENT OF WASTE SPILLS . .
Precipitation/coagulation -per severely Involves addition to and rapid
mixing of the chemicals in-the rainwater. For-continuous, operation, metering
devices and chemical pumps .are.used for continuous deeding''of- the chemicals
in dry, slurry, or solution form.-. Chemica-1 addition is followed by flo.ccjl.a-
tion and settling and/or filtration-. The'equipment used for these subsequent
operations and their suitability for use 1'n mobile units are discussed later
in connection with the EPA's Mobile Phystcal/Chemical Treatment System and in^
connection with biological oxidation, ••-."'
The cost for chemical feed system for use in-mobile units-would be less
than $3,000. Chemical costs would vary with the type of chemical used and
the dosage required. The current costs for alum (17 percent A1203), ferric
chloride, and lime are reportedly $161/mti •$110/mt, and $36/mt, respectively
(7). -• . .. ...
REFERENCES ' -
1. Ghassemi, M. and H.L. Recht. Precipitation, of Polyphosphate with Alumi-
num and Ferric Salts, paper..presented at.-the-25th Purdue Industrial Waste
Conference, Lafayette, Indiana, May 5-7, 1970.
2. Recht, H.L. and M. Ghassemi. Kinetics and Mechanism of Precioitation and
Nature of the Precipitate Obtained in Phosphate Removal from Wastewater
Using Aluminum (III) and Iron (III) Salts, Federal Water Quality Adminis-
tration, Water Pollution Control Research Series 170/OEKI04/70, Aoril
1970.
n
3. Recht, H.L., M. Ghassemi, and E.V. Kleber. Precipitation-of'Phosphates
from Water and Wastewater Using Lanthanum Salts, Proceedings of the 5th
Internati.onal Water Pollution.ResearchuCflnfe^e/tice, San.Francisco, CA, ,„
July-August*1970. * ' .•'•.'
4. Development-Document for-Interim Final Effluent Limitations Guidelines
and Proposed New Source Performance Standards for the Significant In-
organic Products Segment of the Inoraanic Chemicals Manufacturing Point
Source Category, EPA Report EPA-440/1-75/037, 1975.
5. Ho, S., W.C. Boyle, and R.K. Ham. Chemical Treatment of Leachates from
Sanitary Landfills, Journal of Water Pollution Control Federation, Vol.
46, No. 7, pp. 1776-1791, July 1974.
6. Chian, E. and F.B. Dewalle. Sanitary Landfill Leachates and Their Treat-
ment, Journal of the Environmental Engineering Division, ASCE, 102, No.
EE2, 411-431, April 1976.
7. Chemical Marketing Reporter, October 8, 1979.
96
-------
SECTION 11
MISCELLANEOUS PROCESSES
Miscellaneous processes reviewed in this section include gravity separa-
tion, filtration, carbon adsorption, and thermal oxidation. These processes
are commercially available and have been widely used for the treatment of
municipal and industrial wastewaters and, except for the carbon adsorption,
for concentration and disposal of sludges. With the exception of the thermal
oxidation, these processes have also been used in mobile applications for
waste treatability studies and for spill control. The review of the processes
in this section is very brief and is primarily intended to make the study more
complete.
Discussion of mobile unit applications of the various processes relies
heavily on previous and current U.S. EPA experience in spills control, in-
cluding the work on development and application of the Mobile Physical Chemical
Treatment System and the Environmental Restoration Incinerator Complex (ERIC).
GRAVITY SEPARATION
Process Description
Gravity separation is a solids-liquid unit operation whereby particles
denser than water are settled and collected as sludge, and particles lighter
than water (for example, oil) are allowed to rise to the surface and are
collected by skimming equipment. The efficiency of separation can be enhanced
by addition of chemical coagulants' and/or-flotation agents.
Three of the most common types of gravity separators are the API separa-
tors, conventional settling tanks, and parallel plate/tube separators. In
the API separator (Figure 18) the wastewater enters the basin and passes under
the oil retention bafflej then over the diffusion baffle (to minimize turbu-
lence). As the wastewater travels the length of the channel, the oil globules
move toward the surface and the heavy particles settle downward. Flignt
scrappers push the oil that has reached the surface towards one end and into
the slotted pipe for removal. At-the same time the flight scrappers oush
sludge deposits on the bottom of the basin to sludge hoppers. Clarified
water passes under the oil retention baffle and leaves the unit.
Conventional settling tanks a"re circular or rectangular in design. In
the circular design the influent enters at a central location and effluent is
collected around the tank periphery. The sludge is collected in a central
hopper.
97
-------
Olt RETENTION BAFFLE
DIFFUSION BAFFLE
\j
IT I T I I I I I II 1 I I 1 I •*
(a). API Separator
DIFFUSION BAFFLE
LEGEND
1 — Influent
2 — Clarified Effluent
3 -r- Collected Oil
4 — Sludge
OIL RETENTION BAFFLE
•~£
(b). Parallel Plate Separator
Figure 18. Gravity separators.
98
-------
In parallel plate (or tube) separators (Figure 18), the wastewater enters
the separator and flows over a weir and through the parallel plates (or tubes).
The plates can either be corrogated or flat. The oil particles coalesce on the
underside of the plates and rise up to the surface where they are removed.
Solids collect on the bottom of the plates and slide downward toward the sludge
hopper for removal.
Gravity separators are widely used for the treatment of municipal and
industrial wastewaters and they often constitute the first module in waste-
water treatment trains. Because of differences in the characteristics of
various wastewaters (for example, nature and concentration of suspended solids,
temperature of feed stream, presence or absence of emulsions), design and
operation of separators must be "tailored" to the specific wastewaters to be
treated. Depending on the wastewater characteristics and the unit loading, a
suspended solids removal efficiency of as high as 70-80 percent can be achieved
by gravity separation. Some general features of gravity separators from the
standpoint of possible use in mobile units are listed in Table 23.
TABLE 23. GENERAL FEATURES OF GRAVITY SEPARATORS
FOR USE IN MOBILE UNITS
Advantages Disadvantages
Equipment very simple and easy to operate Incomplete removal of suspended
Units suitable for trailer mounting com- s°]^ ;"< "° removal if dls~
mercially available or can be readily S0lvea sonas
fabricated - Settled solids require removal
Low energy consumption an lsP°sa
Can provide effective pretreatment for
other processes
Mobile Unit Applications
Although trailer-mountable gravity separation tanks are commercially
available*, for logistic reasons, less "sophisticated" light-weight portable
*
For example, AFL Industries (Chicago, IL) markets a 10 to 500 gpm "Vertical
Tube Coalescing Separator" (VTC) for the removal of oil and settleable solids
from wastewaters (AFL Industries Bulletin No. 2-13.B.I and supplementary in-
formation provided by Mr. Don Summer, Los Angeles Reoresentative of AFL In-
dustries). The unit is constructed of fiberglass with PVC fittings and has
overall dimensions of 4.8 m length, 2.3 m width, and 2.1 m height (14 ft x
7.7 ft x 6.7 ft). The unit weighs 2,050 kg (4,500 Ib) when empty and 6,840
kg (15,050 Ib) when loaded.
- - 99
-------
units are considered more suitable for use in connection with other spill
treatment systems for field application. The EPA's Mobile Physical Treatment
System uses large rubber, collapsable* portable tanks that are-set up next to
the trailer for pretreatment of wastewater by -flocculation/sedimentation (V).
The unit consists of two concentric tanks: a 11,350 i (3,000 gal) reaction/
flocculation center tank and a 56,775'fc (15,DOO gal) sedimentation periphery
tank. At a design flow rate of 12.6 £/sec (200 gpm) the unit pr.o.vides for a
reaction/flocculation time of 15 minutes and a sedimentation time of 60
minutes. The raw waste is introduced at the bottom of the fl.occulajtion tank.
Flocculation is achieved by hydraulic mixing via the use of two gj'ectar-s ...
placed opposite to each other. The flocculated wastewater flows out' of the
center reaction tank into the 45;'500 a (12,QOO-gaT)'annular space'of the
sedimentation tank through a series p-f submerged orifi-ces-. located'around 'the >
periphery of the-reaction tank. -Both these tanks-are cylindrical-open top -
tanks and are supported Jby staves anchored into the ground. The open top
tanks also permit easy acdessibility'for manual-removal of-sludge and floating
materials. . -.-•''
A IT.9 A/sec ,(50 gpm) "sludge pump and special suction fittings are pro-
vided for removal of settled.and floating'contaminants and a 11,350 a (3,000
gal) rubber stave tank is provided for storage. The sludge can'subsequently
be pumped out of the stave tank by the sludge pump into a ta,nk truck or other
container for further treatment and/or final ih'sppsal.-
FILTRATION
Process Description ' .
Filtration is a unit operation employed to remove suspended solids from
water and wastewaters*. A flow diagram of a typical filtration system for the
removal of suspended solids from liquids is shown in Figure 19. Wastewaters
are slowly percolated through a bed of filter media (for example, sand, char-
coal, diatomaceous earth, anthracite, etc.). Suspended solids are collected
on the surface and in the interstices of the filter. The accumulation of the
solids on and in the filter media results in an increase in pressure drop (or
head loss) across the filter bed. When the pressure drop becomes "excessive"
(usually 1.5 to 2.4 m of water column), the filter is backflushed to remove
the suspended solids from the surface and interstices of the filter. Back-
wash water contains a very high concentration of suspended solids and is
usually treated by sedimentation (for example, in primary sedimentation unit
that precedes filtration).
Depending on the design and operating conditions, a water very low in
turbidity can be obtained by conventional sand filtration. Chemical coagulant
may be added to the filter influent to increase particle removal efficiency
and/or extend the filter "run". Table 24 lists some general features of
Filtration is also used for solid concentration (e.g., in dewatering of
sludges); this type of filtration is not considered in this discussion.
100
-------
BACKWASH
DRAIN
AIR
DIFFUSES •
HIGH HEAD
BACKWASH
TROUGH
UNDER DRAIN
RAW FEED
SINGLE OR - -
MULTIPLE LAYER
FILTER MEDIUM
•4-AIR
BACKWASH
EFFLUENT
Figure 19. Typical filtration bed.
101
-------
conventional filtration from the standpoint of use in mobile units for waste-
water treatment. , ....
TABLE 24. GENERAL FEATURES OF FILTRATION -FOR'APPLICATION -TO "
WASTEWATER TREATMENT IN MOBILE UNITS
Advantages
Disadvantages
Equipment reliable, simple, and commer-
cially available
Ability to handle wide fluctuations in
influent suspended solids concentrations
Ease of operation and control • •
Can provide effective pretreatment for
other processes
No secondary air pollution problem
Considerable experience exists in mobile
unit application ""• " ^ ".
Ineffective in removal of dis-
solved substances
-Produces backwash wastewater
.containing high solids concen-
tration
Frequent backwashing of filter
'required when treating influent
high in suspended solids
Mobile Unit Applications
The EPA's Mobile Physical /Chemical Treatment System, which has been in
operation for more than six years, incorporates three "dual" media (sand-
anthracite) filters in its treatment train. The three filter units are con-
nected in parallel and may be taken off line individually for backwashing.
The filters consist of 61 cm (24 in) anthracite over 45 cm (18 in) fine sand
and are designed for a maximum hydraulic loading of 4.8 £/sec/m2 (7.0 gpm/
Additional features of the filters are described in Table 25.
The supernatant from the sedimentation tank is drawn off by the filter
pump through a submerged orifice header ring at the outside tank wall. A
pneumatic level sensor in the sedimentation rank controls filter pump flow to
match raw flow. After the addition of a filtration conditioner, the settled
effluent is pumped through the three dual media filters in parallel for re-
moval of residual suspended solids. Two in-line turbidimeters monitor the
turbidity of the total filter influent and effluent from each tank. A
differential pressure gauge indicates the degree of filter cloaging. Tne
filters are backwashed with air and clean system effluent stored in a 11,350
-------
TABLE 25. FEATURE OF A DUAL-MEDIA FILTER UNIT USED IN THE
EPA'S MOBILE PHYSICAL/CHEMICAL TREATMENT SYSTEM (1)
Filter diameter
Filter area
Design filtration rate
Maximum filtration rate
Design backwash rate
Maximum allowable differential pressure
Maximum tank pressure
Depth of sand
Effective size
Uniformity coefficient
Quantity of sand
Depth of coal
Effective size
Uniformity coefficient
Quantity of coal
1.17 m (3.5 ft)
0.893 m3 (9.62 ft2)
2.11 £/sec (33.5 gpm)
4.22 a/sec (67 gpm)
6.3 A/sec (100 gpm)
1.06 kg/m2 (15 psi)
p
4.93 kg cm/cm (70 psi)
45 cm (18 in)
0.5 mm
1.5
860 kg (1,900 Ibs)
61 cm (24 in)
0.85 - 9.85
1.7
0.425 m3 (15 ft3) 0.39 ton
CARBON ADSORPTION
Process Description
Carbon adsorption is a unit operation for the removal of soluble organics
from wastewaters. Removal of the soluble organics 'is affected by contacting
the wastewater with activated carbon. Two methods of contact are commonly
used: passing wastewater through a column or bed of granular activated car-
bon and adding powdered activated carbon directly to a treatment system.
Spent granular activated carbon containing adsorbed organics is usually re-
generated by thermal treatment. Powdered activated carbon is in most cases
not recovered and is disposed of with other wastewater treatment sludges.
Granular activated carbon systems commonly employ two or more beds and
provide for continuous treatment with periodic removal of one of the adsorbers
from service for backwashing and removal of spent carbon*. Backwashing serves
Fixed beds may be arranged in series or parallel with either uoflow or down-
flow design. "Pulsed" columns with countercurrent flow of carbon and waste-
water have also been used.
103
-------
to remove particulate buildup on the column surface that increases bed pressure
drop, thereby increasing energy consumption. A schematic diagram of a two-
vessel carbon adsorption system is shown in "Figure 20.
Regeneration of the spent carbon is performed by thermal regeneration,in
a furnace. The regeneration,process produces offgases that must be treated
for particulate, odor, carbon monoxide, residual organics, and -hydrocarbon
removal. Wastewaters resulting from bed backwashing, reactivated carbon1
quenching, and air emission controls are usually returned to upstream treat-
ment systems.
Use of powdered activated carbon for wastewater treatment has several
drawbacks, including unavailability of-commercial regeneration systems, diffi-
culty in handling large dosage requirement, and production of sludge requiring
disposal.
Removal of an organic compound by activated carbon is affected by the
characteristics of the specific carbon"used (adsorbent), the specific sub-
stances to be removed (adsorbate), and the solution properties (pH, tempera-
ture, nature, and characteristics .pf other dissolved solids in the wastewatar).
In general, activated carbon preferentially adsorbs high molecular weight and
less polar organic compounds. Table 26 shows the relative adsorbability of
several representative compounds differing in structure and molecular weight._
In actual wastewater applications"," a wide.range of substances would be en-
countered and the actual carbon performance would have to be determined by
laboratory and pilot testing. Critical design criteria are contact time and
organic loading. Contact time defines the adsorber size and number of units
required. The amount of organic material removed per weight of carbon deter-
mines the carbon usage rate and the regeneration requirements.
Activated carbon systems for wastewater treatment are employed in indus-
tries such as coke production, oil refining, petrochemical production, and
pesticide manufacture. Carbon systems are also used for trace organics, taste
and odor removal from potable water supplies, sugar decolorization, and puri-
fication of fats, oils, foods, beverages, and Pharmaceuticals. In commercial
refinery applications, from 59 percent to 83 percent COD removal has been
obtained with granular carbon systems used without prior biological treatment.
Studies of an activated carbon system for treatment of a coke plant effluent
after clarification and filtration reported 80 percent COD removal, 91 percent
'TOC removal, and 99 percent phenol removal. Even though these removal effi-
ciencies are very high, the effluent from activated carbon adsorption can
contain substantial concentration of organics {for example,'in the'ease of
concentrated waste spills). Some general features of activated carbon adsorp-
tion as a treatment process for the removal of oroanics from wastewaters are
listed in Table 27.
It is estimated that there are approximately 100 large-scale carbon ad-
sorption systems in use for industrial/municipal waste treatment (5). A
number of carbon manufacturers own and operate trailer-mounted carbon adsorp-
tion units that are used to assess treatability of industrial wastes and/or
to develop criteria for the design of large-scale carbon adsorption units for
industrial application.
104
-------
o
in
FEED WATER
REGENERATED/MAKEUP
ACTIVATED CARBON '
BACK WASH EFFLUENT
BACK WASH FEED
ADSORBER 1
ADSORBER 2
T
REGENERATED/MAKEUP
ACTIVATED CARBON
BACK WASH EFFLUENT
BACK WASH FEED
TREATED EFFLUENT
I
SPENT CARBON
VALVE CLOSED
VALVE OPEN
figure 20. Two-vessel granular carbon adsorption system (2).
-------
TABLE 26. AMENABILITY OF TYPICAL ORGANIC COMPOUNDS
TO ACTIVATED CARBON ADSORPTION* (3)
Compound ' •" ' . - Adsorbability
(grams 'compound/grajns carbon.)
Ethanol '_ ' 0.02 • ' '
2-ethyl butanol ' 0.170.
Acetaldehyde , O.Q22 ' .
Benzaldehyde •• ." Q.T88'
Di-n-butylamine " 0.174'
Monoethanolaraine ' 0.150
2-methyl 5-ethyl pyri.djne " 0.179
Benzene / '" . "" " 0.080
Hydroquinone , , 0.167 .
Ethyl acetate • 0,100
Butyl acetate 0.193
Isx)propy1- ether - - 0.162
Ethylene glycol 0.0136
Tetraethylene glycol 0.116
Acetone 0.054
Acetophenone 0.194
Formic acid 0.047
Valeric acid 0.159
Benzoic acid 0.183
*
Westvaco Nuchar WV-G (12 x 40 mesh, coal based) carbon.
4.
'5 g carbon added to 1 liter of solution containing 100 mq/i
of compound.
106
-------
TABLE 27. SOME FEATURES OF ACTIVATED CARBON ADSORPTION PROCESS
FOR REMOVAL OF ORGANICS FROM WASTEWATER5
Advantages
Disadvantaaes
High efficiency of organics removal
Apolicable to a wide range of
orgamcs, including toxic orgamcs
not amenable to biological treatment
Low sensitivity to variations in
feed concentrations
Relatively small space requirements
Commercially available
Considerable operating.experience
available, including use of trailer-
nounted units
Generally limited to wastes with less
than 1 percent organic content
High capital and operatina cost
Pretreatment of waste to reduce sus-
pended solids concentration generally
necessary
Does not remove most inorganic SUD
stances
In mobile unit application, on-site
regeneration may not be practical and
spent carbon may require ultimate
disposal.
Mobile Un^'t Applications
The U.S. EPA's Mobile Physical/Chemical Treatment System (1), which has
been in operation for more than six years, uses three activated carbon colurrns
that handle the effluent from the dual-media filters. Tne filtered effluent
flows through three pressure carbon columns that may be used in parallel or in
series. Altogether, they contain 19.6 m3 (700 ft3) of carbon. This volume
represents a dry carbon weight of 8,172 kg (18,000 Ib) of carbon, which is the
maximum possible weight that can be accommodated on the trailer because of tne
overall weight constraints on the mobile system. The caroon columns are de-
signed for a hydraulic loading rate of 3.4 2/sec/m2 (5 gpm/ft2). Three 2.1 m
(7 ft} diameter carbon columns with carbon bed deoth of 1.8 m (6 ft) are oro-
vided on the mobile treatment system. The selected carbon volume of 19.6 cu m
(700 cu ft) on the trailer provides a maximum contact time or 27 minutes for
the three columns at a flow rate of 12.6 a/sec (200 gpm). This carbon contact
time has been found to be suitable for many of the hazardous materials eval-
uated. However, when h^'gh contact times are reauired, these may be orovided
by reducing proportionately the hydraulic flow rates througn the carbon columns.
The caroon tanks, which occupy the back ore-half of the trailer, are com-
pletely plumoed so that only valve ad]ustments are necessary to control t*e
various nodes of operation. The t^aile*1 is designed so that it nay be trans-
ported with the carbon columns full of the wet, drained carbon. However,
whether done on the site or back at the homeoase, the carbon will ne°d to be
recharged either because of exhaustion, to prevent undesiraole effects of
mixing contaminants, or because storage with contaminants in the carbon could
be hazardous. Spent caroon nay be removed using the backwash pump ana ore
107
-------
cessed clean water stored in one of the rubber tanks for this purpose. The
clear effluent water is pumped through the underdrain system of the carbon
column to fluidize the bed and cause the slurry to drain out of the tank drain
fitting. . '
When the adsorptive capacity of the carbon for the processed hazardous
materials is depleted, new carbon may be installed into the tanks in the field
by a slurry pumping system. 'Depletion time, or breakthrough, occurs simul-
taneously for parallel operation and sequentially for series operation. Thus
all of the carbon must be.replaced at once for oarallel operation as opposed
to one tank at a time for series-opera tion%. .The tank plumbing permits a rota-
tion of the flow sequence through^the tanks for series operation. Thus, the
leading tank eventually becomes trie second tank in line and finally assumes
the third position until breakthrough and carbon replacement, whereon it again
assumes the first position. The slurry pumping .system for carbon replacement
utilizes a dry carbon hopper feeding an eductor through which clean effluent
from the effluent storage tank is pumped to form a slurry. The slurry is
pumped into the carbon column where it is dewatered by the carbon column under-
drains. The water is then returned to the effluent storage tank completing
the closed loop slurry pumping system. A manual sampling valve in the carbon
column inlet line and in each tank effluent line permits analysis of process
water to measure removal effectiveness and to sense carbon column breakthrough.
Table 28 summarizes some operating results for the treatment of a number ,
of spill-inpacted waters by the EPA's Mobile Physical/Chemical Treatment
System.
INCINERATION
Process Description
Incineration is the thermal destruction of wastes in a controlled envi-
ronment. It has been used for the destruction of toxic organic wastes and
disposal of municipal and industrial sludges.
Several types of incinerators that are in commercial use include the
following: fluidized bed, rotary kiln, multiple hearth, and liquid injection.
Some key features of these incinerators are summarized in Table 29. Inciner-
ators for the destruction of hazardous wastes should provide a sufficiently
high temperature and residence time in the combustion chamber to assure near
complete destruction of the organics.
Figure 21 depicts the basic components and input/output streams asso-
ciated with the rotary kiln incinerator. The basic components consist of the
following:
• a waste storage and waste feeding mechanism
t fuel-fired burner and oxidation chamber (may include afterburner,
not shown)
• flue gas purification system (usually a wet scrubber with liquid
collection tank)
108 \
-------
TABLE 28. TOXIC COMPOUNDS REMOVED FROM WATER USING ERA'S MOBILE PHYSICAL/CHEMICAL TREATMENT SYSTEM (6)
o
10
Compound
DNBP
PCB
Toxaphene
Chlordane
lleptachlor
Aldrin
Dieldrin
Kepone
Pentachlorophenol
Methylene chloride
Carbon tetrachloride
Benzene
Toluene
Xylenes
Trichloroelhane
Trichloroethylene
Phenol
Cresol
Dimethyl phenol
Trimethyl phenol
Butyl phenol
Dioctyladipate
Dimethylanil ine
PCB
Location
of incident
Clarksburg, N.J.
Seattle, Wash.
The Plains, Va.
Strongstown, Pa.
Strongstown, Pa.
Strongstown, Pa.
'
Strongstown, Pa.
Hopewell, Va.
Haverford, Pa.
Oswego, N.Y.
Oswego, N.Y.
Oswego, N.Y.
Oswego, N.Y.
Oswego, N.Y.
Oswego, N.Y.
Oswego, N.Y.
Oswego, N.Y.
Oswego, N.Y.
Oswego, N.Y.
Oswego, N.Y.
Oswego, N.Y.
Oswego, N.Y.
Oswego, N.Y.
Dittmer, Mo.
Quantity
treated
U)
7,600,000
2,270,000
45,000,000'
950,000
380,000
11,400
380,000
11,400
380,000
11,400
380,000
11,400
850,000
820,000
950,000
950,000
950,000
950,000
950,000
950,000
950,000
950,000
950,000
950,000
950,000
950,000
950,000
950,000
180,000
Contact
time
(min)
26
30-40
8.5
26
17
240
17
240
17
240
17
240
45.5
26
8.5
8.5
8.5
8.5
8.5
8.5
8.5
8.5
8.5
8.5
8.5
8.5
8.5
8.5
60
Influent
cone.
(mg/s.)
8
400
1
36
13
1,430
6.1
80
8.5
60.5
11
60.5
4,000
10,000
190
1.1
1
120
140
12
21
140
230
1,220
130
300
360
380
19
Effluent
cone.
(mg/fc)
<0.002
<0.075
<0.075
1
0.35
0.43
0.06
0.1
0.19
0.15
<0.01
<0.01
<1
<1
51
<0.1
0.1
0.3
<0.1
<0.1
0.3
<0.1
8.1
5.4
10
15
320
23
<0.1
Removal
(%)
99.98
99.98+
92.5+
97.22
97.3
99.99
99.02
99.87
97.76
99.75
99.99+
99.99+
99.98
99.99
73.15
90.91+
90
99.75
99.92+
99.17+
98.57
99.92+
96.47
99.56
92.3
95
11
93.95
99.47+
-------
TABLE 29. KEY FEATURES OF MAJOR TYPES OF INCINERATORS (7)
Type
Process principle
1 Application
Combustion temp.
Residence time
Rotary Slowly rotating cylinder
kiln mounted at slight incline
to horizontal. Tumbling
action improves efficiency
of combustion.
Multiple Solid feed slowly moves
hearth through vertically stacked
hearths; gases and liquids
fed through side ports and
nozzles.
Liquid Vertical or horizontal
injection vessels; wastes atomized
through nozzles to increase
rate of vaporization.
Fluidized Wastes are injected into a
bed hot agitated bed of inert
granular particles; heat
is transferred between the
bed material and the waste
during combustion.
Most organic wastes;
well suited for
solids and sludges;
liquids and gases.
Most organic wastes,
largely in sewage
sludge; well suited
for solids and
sludges; also handles
liquids and gases.
Limited to pumpable
liquids and.slurries
(750 SSU or less for
proper atomization)k
Most organic wastes;
ideal for liquids,
also handles solids
and gases.
810-1,640°C
(1,500-3,000°F)
• 760-980°C •
(1,400-1,800°F)
650-1,650°C
(1>200-3,000°D
750-870°C
(1,400-1,600°F)
Several seconds
to several hours
Up to several
hours
0.1 to V second
Seconds for gases
and liquids;
longer for solids
-------
WASTE
STORAGE
HOPPER
FLUE GAS
OXIDATION SCRUBBER
CHAMBER BURNER
,
1#
_**
A«
ASH
REMOVAL
MECHANISM
LIQUID
HOLDING
TANK
©
STACK
*«5
i i
— N-^-^^x^
LEGEND.
1 INFLUENT WASTE
2. COMBUSTION AIR
3 FLUE GAS
4 RESIDUALS
5 SCRUBBER WATER
6 FUEL
figure 21. Rotary kiln incineration unit.
-------
t ash removal mechanism
• flue gas stack or vent
All incineration processes require some auxiliary fuel to initiate combustion
and bring the oxidation chamber up to temperature. Depending on the heat
content of the waste, auxiliary fuel may or may not be required during incin-
eration. Some incineration systems (especially those used for hazardous waste
incineration) may require an afterburner to insure complete destruction or
organics.
The effluent streams from the incineration process are ash and the
scrubber water. The ash is generally inorqanic inerts and has a small volume.
Scrubber waters have high inorganic solids content and are-usually Created
before disposal. Some advantages and disadvantages or incineration as a pro-
cess for the disposal of hazardous organic wastes are summarized in Taole 30.
TABLE 30. SOME ADVANTAGES AND DISADVANTAGES OF INCINERATION SYSTEMS
FOR THE DESTRUCTION OF TOXIC ORGANIC WASTES
Advantages
Disadvantages
Effective for destruction of solid,
liquid and gaseous toxic organic
wastes
Reduces waste to a small volume of
inerts (ash) which can be disposed
of safely
Incineration technology well de-
veloped and considerable experience
exists in connection with hazardous
waste destruction
High degree of process control re-
quired to insure efficient operation,
especially when highly halogenated
organic wastes are to be destroyed
Use of afterburners and scrubbing
systems required to minimize air
pollution
Ash and scrubber water require dis-
posal
Auxiliary fuel required for wastes
low in Btu content
Mobile Unit Applications
M.B. Associates (San Ramon, CA) is currently under an EPA contract to
design, construct, and demonstrate a mobile incineration system for aest^uc-
tion of residuals from hazardous material spi1! clean-uo operations. The
system, referred to as the Environmental Restoration Incinerator Corcnlex
(ERIC), will be mounted and transoorted on three heavy duty truck trailers.
As shown in Figure 22, Trailer No. 1 houses the control room, ram *eeae^,
loading bin, shredder, and rotary kiln; Trailer No. 2 houses the afte^ourrer
and the quench unit, whereas the scrubber, blowers and diesel generators are
loaded on Trailer No. 3. Each trailer unit will be 13.7 m (^5 rt) and when
pulled by a tractor will meet most interstate requirements for both wiatn and
112
-------
CHEAP PARTICLE SCRUBBER
MASS TRANSFER SCRUBBER
CONNECTING
DUCT
2 STAGE BLOWER
EMERGENCY STACK
ROUND LEVEL SUMP
SHREDDER
SOUND SUPPRESSOR
BLOWER DRIVE
DIESEL GENERATOR
TRAILER 03
CONTROL BOOTH
RAM FEED
LOADING BIN
KILN
TRAILER #2
TRAILER HI
BREECH UNIT
Figure 22. Mobile environmental restoration incinerator complex.
-------
length. Some key design concepts incorporated in the Mobile ERIC system in-
clude the following:
• All components must be suitable and compatible with both over-the-
road and off-the-road shock and vibration loads.
• The system must be capable of processing solids, liquids, and sludges
with or without Btu content.
• Kiln temperatures of 1,800°F to insure volatilization of HM's with
one hour dwell for solids in order to assure clean ash residuals.
• Afterburner minimum temperatures of 2,012°F per Federal Regulations
of PCB's.
• Afterburner two second dwells to insure thermal decomposition of
pesticides and poly-chlorinated biphenyls per Federal Regulations.
• Excess 02 above 3 percent per Federal Regulations.
• Particle scrubbing sufficient to effectively meet a 10 percent perma-
nent plume opacity when decomposing organic phosphates that will
generate
• An absorber scrubber to effectively meet S02 and HC1 emission
constraints for incinerators at various locations.
Since the Mobile ERIC has not yet been field tested, no operating data
are available on its application to spill clean-up wastes.
REFERENCES
1. Gupta, M.K. Development of a Mobile Treatment System for Handling
Hazardous Material Spills, EPA Report No. 600/2-76-109. July 1976.
2. U.S. EPA, Process Design Manual for Carbon Adsorption, EPA 62-5/1 -7] -002a,"
October 1973.
3. Giusti, D.M., et al. Activated Carbon Adsorption of Petrochemicals,
Journal of Water Pollution Control Federation, Vol. 46, No. 5, May 1974.
4. Van Stone, G.R. Treatment of Coke Plant Waste Effluent, Iron and Steel
Engineer, April 1972, pp 63-66.
5. Arthur D. Little. Physical, Chemical, and Biological Treatment Tech-
niques for Industrial Wastes, NTIS PB-275-287, November 1976.
6. La Fornora, J. Cleanup After Spills of Toxic Substances, Journal of
Water Pollution Control Federation, April 1978.
7. Shen, T., M. Chen, and J. Lauber. Incineration of Toxic Chemical Wastes,
Pollution Engineering, October 1978, pp 45-50.
114
------- |