United States
Environmental Protection
Agency
Industrial Environmental Research EPA-600/2-78-176
Laboratory August 1978
Cincinnati OH 45268
Research and Development
&EPA
Treatment
of Wastewaters
From Adhesives
and Sealants
Manufacture
by Ultrafiltration
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the ENVIRONMENTAL PROTECTION TECH-
NOLOGY series. This series describes research performed to develop and dem-
onstrate instrumentation, equipment, and methodology to repair or prevent en-
vironmental degradation from point and non-point sources of pollution. This work
provides the new or improved technology required for the control and treatment
of pollution sources to meet environmental quality standards.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/2-78-176
August 1978
TREATMENT OF WASTEWATERS FROM
ADHESIVES AND SEALANTS MANUFACTURE BY ULTRAFILTRATION
by
Myles H, Kleper
Robert L. Goldsmith
Tarn Van Tran
Walden Division of Abcor, Inc.
Wilmington, Massachusetts 01887
and
David H. Steiner
John Pecevich
Michael A. Sakillaris
Dewey and Almy Chemical Division
W.R. Grace and Company
Lexington, Massachusetts 02173
Grant No. S804350010
Project Officer
Ronald Turner
Industrial Pollution Control Division
Industrial Environmental Research Laboratory
Cincinnati, Ohio 45268
INDUSTRIAL ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
-------
DISCLAIMER
This report has been reviewed by the Industrial Environmental Research
Laboratory, U.S. Environmental Protection Agency, and approved for
publication. 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 recommendation for use.
-------
FOREWORD
When energy and material resources are extracted, processed, converted,
and used, the related pollutional impacts on our environment and even on our
health often require that new and increasingly efficient pollution control
methods be used. The Industrial Environmental Research Laboratory-
Cincinnati (lERL-Ci) assists in developing and demonstrating new and
improved methodologies that will meet these needs both efficiently and
economically.
A field demonstration of ultrafiltration for the treatment of waste-
waters generated during the manufacture of adhesives and sealants is dis-
cussed in this report. The technical merits of various post-treatment unit
processes are considered based on pilot-scale tests, bdndh-scale tests, and
literature reviews. A full-scale system design for treatment of this
industry's wastewater is conceptualized, and purchased equipment and oper-
ating cost projections are made- These costs are compared with those which
were included in the preliminary contractor recommendations presented in the
draft development document for this industry as Best Practical Control
Technology Current Achievable, Best Available Demonstrated Control Technology
and Best Available Technology Economically Achievable costs for similar waste
streams. It is hoped that the results of this study will increase the
knowledge of both the public and industry in this complex area and will
promote further activity in the review of waste treatment problems in the
adhesives and sealants industry.
The Organic Chemicals and Products Branch of the Industrial Pollution
Control Division should be contacted for further information on this
subject.
David 6. Stephtn
Director
Industrial Environmental Research Laboratory
Cincinnati 45268
m
-------
ABSTRACT
The overall goal of this program was to demonstrate technology for the
treatment of effluents from adhesives and sealants manufacture to produce
water of a quality suitable for discharge to municipal sewers. A secondary
goal was to collect information on the nature and variability of these
wastes to better characterize effluents generated in adhesives and sealants
manufacture. The principal element of the program was a several-month field
demonstration performed at the San Leandro, California, manufacturing
facility of the Dewey and Almy Chemical Division of W.R. Grace and Company.
Additionally, the program consisted of preliminary studies of UF permeate
post-treatment alternatives, documentation of full-scale ultrafiltration
system performance at the Dewey and Almy Chicago, Jllionis plant, and develop-
ment of full-scale treatment system designs and economic projections for
plants with wastewater flows ranging from 3.8 m3/day (1,000 gpd) to 75.8 m3/
day (20,000 gpd).
Ultrafiltration was proven to be a viable unit process for separating
adhesives and sealants manufacturing wastewaters into a low-volume con-
centrate stream and a high-volume permeate stream. The UF permeate was
characterized by the following average contaminant loadings: 100 mg/£ total
freon extractives, <7.4 mg/£ nonpolar extractives, <27 mg/£ (typically
<5 mg/£) suspended solids, 0.43 mg/£ free cyanide, 3.6 mg/£ total cyanide,
8,900 mg/£ BOD, 36,600 mg/£ COD, 44.6 mg/£ phenolic compounds, and 1.5 mg/£
zinc. A treated effluent of this quality is acceptable for discharge under
the San Leandro Municipal Discharge Limitations, with the exception of the
phenolic compound and total cyanide loadings. Surcharges would be imposed,
however, based on the suspended solids and BOD loadings.
If significant levels of phenolic compounds and cyanide are not present
in a particular plant's wastewater discharge, ultrafiltration is judged
capable of meeting local municipal discharge standards. This claim has been
verified over the past two years by the operation of a full-scale UF system
at the Dewey and Almy Chicago Plant. When phenolic compounds and cyanide
are present at significant levels, either ozonation or reverse osmosis are
considered the preferred post-treatment processes. A UF/ozonation or UF/
reverse osmosis treatment system is projected to meet all municipal dis-
charge standards. For either treatment system option, ultrafiltration, or
IV
-------
ultra-filtration followed by a post-treatment process; equalization of
the plant wastes before ultrafiltration is recommended. This treatment step
dampens flow and composition variations and provides gravity settling and
flotation of suspended solids and oils and grease. The latter reduces the
loading on the UF system and therefore lowers the membrane area require-
ment.
The preliminary contractor recommendations as presented in the draft
development document for this industry proposed double-effect liquid
evaporation as the model technology to meet the proposed standards for two of
the six manufacturing subcategories. This treatment technology has not, as
yet, been demonstrated. Projected capital investments for double-effect
liquid evaporation are conservatively from 4 to 6 times the cost of the
equalization/ultrafiltration and post-treatment costs. Projected annual
operating costs for the evaporation system are in the range of 3 to 6 times
the operating costs for the UF system options.
This report was submitted in fulfillment of Grant No. S804350010 by
the Wai den Division of Abcor, Inc., and the Dewey and Almy Chemical Division
of W.R. Grace and Company under the sponsorship of the U.S. Environmental
Protection Agency. This report covers the period from March 1, 1976, to
July 30, 1977, and work was completed as of November 30, 1977.
-------
CONTENTS
Foreword i i i
Abstract i'v
Figures viii
Tables ix
English-Metric Conversion Table xii
Acknowl edgments xi i i
1. Introduction 1
2. Conclusions 14
3. Recommendations 20
4. Program Overview 22
5. Discussion of Unit Processes 24
6. Test Systems, Procedures, and Analyses < 36
7. Experimental Results and Discussion 48
8. Summary of Ultrafiltration System Operation at the Dewey and
Almy Chicago, Illinois, Plant 87
9. Full-scale System Design and Economics 93
Appendices
A. Additional Field Demonstration Test Data 108
B. Chicago Plant UF Data 119
vn
-------
FIGURES
Number Page
1 Cut-Away View of a Tubular Ultrafiltration Assembly 29
2 Various System Designs for Modular Membrane
Equipment 30
3 Flow Schematic of San Leandro Plant Field Demonstration
Test System 37
4. Ultrafiltration System Flow Schematic 38
5 Simplified Flow Schematic for Reverse Osmosis
Test System 43
6 Cut-Away View of a Permasep Hollow-Fine-Fiber Reverse
Osmosis Module 45
7 UF Permeate Flux vs. Time for San Leandro Plant Waste-
water Concentration Runs #2 and #6 52
8 UF Permeate Flux vs. Time for Surfactant Addition Test
at San Leandro 54
9 UF Permeate Flux vs. Time for Total Effluent Less CPD
Stream Test at San Leandro 55
10 UF Permeate Flux vs. Time for Total Effluent Less "Z"
Stream Test at San Leandro 57
11 Productivity vs. Feed Volumetric Concentration for DuPont
B-9 Processing of Dewey and Almy Chicago Plant
Ultrafi1trate 74
12 Equilibrium Adsorption Isotherm at 20°C for BOD Removal
from Dewey and Almy Chicago Plant Ultrafiltrate
(Second Sample) 78
13 Equilibrium Adsorption Isotherm at 20°C for COD Removal
from Dewey and Almy Chicago Plant Ultrafiltrate
(Second Sample) 79
14 UF Permeate Flux vs. Time for Electrocoagulation Process
Test at San Leandro 84
15 Proposed Treatment Systems for Adhesives and Sealants
Manufacturing Wastes 94
vm
-------
TABLES
Number Page
1 Standard Industrial Classification of Products from
the Adhesives and Sealants Industry 2
2 Principal Raw Materials Used in Adhesive and Sealants
Manufacture 3
3 Summary of the Raw Wastewater Loading for Major
Pollutants in the Adhesives and Sealants Industry 5
4 Summary of EPA Proposed Effluent Limitation Guidelines
for the Adhesives and Sealants Industry 1
5 Municipal Discharge Limitations 8
6 Hexane Extractible Levels Following Conventional Treat-
ment of Dewey and Almy Chicago Plant Waste Streams 10
7 Principal Raw Materials Used in Manufacture of
San Leandro Product Line 13
8 Differences Between Reverse Osmosis and Ultrafiltration .. 25
9 Comparison of Commercially Available Ultrafiltration
Module Configurations 27
10 Chemical Reactions which Occur During Oxidation of
Cyanide and Phenolic Compounds 35
11 Characteristics of Ultrafiltration Membranes Employed
During Field Tests 40
12 Assays and Methods Employed During Experimental Program .. 47
13 Summary of San Leandro Plant Effluent Composition 49
14, Summary of Ultrafiltration Membrane Flux During
San Leandro Field Tests 51
15 Flux Recovery and Accumulated Operating Times for UF
Membranes Operated on the San Leandro Plant Effluent 58
16 UF Membrane Flux Recovery Data Subsequent to
Latex Foul ing 60
17 Summary of Analytical Data for Ultrafiltration Feed
After Settling at San Leandro Plant Through Test #12 62
IX
-------
Number paqe
=>—
18 Summary of Analytical Data for Ultrafiltration Con-
centrate at San Leandro Plant Through Test #12 63
19 Summary of HFM Permeate Qua! i ty 54
20 Effect of Settling Pretreatment on San Leandro
Plant Effluent 66
21 Average UF Membrane Removal Efficiency Data During
Processing of the Total Plant Effluent 67
22 Average UF Permeate Quality During Processing of the
Total Plant Effluent with Surfactant Addition 69
23 Average UF Permeate Quality During Processing of the
Total Plant Effluent Less the CPD Stream 70
24 Average UF Permeate Quality During Processing of the
Total Plant Effluent Less the "Z" Stream 70
25 Comparison of HFM Permeate Quality with Local and
Federal Discharge Standards During Treatment of the
Total PI ant Ef fl uent 72
26 Analytical Data and RO Module Removal Efficiencies During
Processing of the Grace Chicago Plant Ultrafiltrate 75
27 Standard Salt Rejection Test Data for DuPont B-9 Hollow-
Fiber Module During Processing of Dewey and A!my
Chicago Plant Ultrafiltrate 77
28 Analytical Data for Ozonation of San Leandro Plant
Ultrafi1trate 81
29 Average UF System Removal Efficiencies Following
Electrocoagulation Pretreatment 85
30 Average UF Permeate Quality During Processing of
Chicago PI ant Effluent 89
31 Average UF System Operating Data During 1976 90
32 Annual Operating Costs for Chicago Plant Waste
Treatment System 92
33 Design Bases for Projections of Unit Process Purchased
Equipment and Operating Costs 97
34, Estimated Purchased Equipment Costs for Selected Unit
Processes of Various Capacities (Thousands of Dollars) ... 98
35 Estimated Purchased Equipment Costs for the Three
Treatment System Options (Thousands of Dollars) 100
36 Estimated Annual Operating Costs for Unit Processes
of Various Capacities (Thousands of Dollars) 101
-------
Number
37 Estimated Annual Operating Costs for the Three
Treatment System Options (Thousands of Dollars)
38 Summary of Development Document for Effluent
Limitations Guidelines Projected Costs for Treat-
ment of Subcategory B and C Adhesives and
Sealants Wastewaters 104,
39 Comparison of Option 1, 2 and 3 System Costs with
Double-Effect Liquid Evaporation Costs 105
xi
-------
ENGLISH-METRIC CONVERSION TABLE
To Convert from
To
Multiply by
Inch
Feet
Square inch
Square feet
Cubic feet
Gallon
Pound
Pound per sq. inch
Horsepower
Gallon per day
Gallon per minute
Gallon per sq. ft-day
Gallon per minute per sq. ft.
Meter
Meter
Square meter
Square meter
Cubic meter
Cubic meter
Kilogram
Atmosphere
Watt
Cubic meter per day
Cubic meter per day
Cubic meter per sq. meter-day
Cubic meter per sq. meter-day
2.54xlO"2
6.45xlO'4'
9.29xlO'2
2. 83x1 0"2
3.79xlO"3
4.54X10"1
6. 80x1 0~2
7.46x10 2
3.79xlO"3
5.45
4.10xlO'2
5.87x10 ]
xn
-------
ACKNOWLEDGMENTS
The authors wish to thank Robert Holt, plant manager, and Laughlin Nines,
plant engineer at the San Leandro field demonstration site for their support
and cooperation throughout the field test program. The efforts of the test
system operator, Don Benge, in handling the day-to-day operation of the test
system are gratefully acknowledged.
The authors express their appreciation to Ronald Turner, EPA Project
Officer on the demonstration grant, for his technical and administrative
guidance during the course of this program.
Acknowledgment is also made of the skill and patience of Sharon Collins
in preparing this manuscript.
xm
-------
SECTION 1
INTRODUCTION
BACKGROUND
The adhesives and sealants industry consists of 745 companies which
operate 1114 manufacturing plants (1). Within this highly competitive
industry, manufacturing facilities range from single-man operations to large
industrial complexes (2), with the 50 leading adhesives and sealants formu-
lators accounting for less than 33% of industrywide sales (1). Manufacturing
facilities for this industry are concentrated mainly within the states of
Illinois, New Jersey, California, Massachusetts, and New York.
For the purpose of developing effluent limitation guidelines, only those
commodities within Standard Industrial Classifications 2891 and 2899 in the
adhesives and sealants category were included. These commodities are listed
in Table 1 (2). The diversity of products manufactured by the adhesives and
sealants industry is clearly evident. Because of this diversity the industry
was subcategorized into six groups in the preliminary contractor recommen-
dations.
A. Water-Based Animal Glues and Gelatins
B. Water-Based Adhesive Solutions Containing Synthetic and Natural
Materials
C. Solvent Solution Adhesives and Cements Generating Contaminated
Wastewaters
D. Solvent Solution Adhesives and Cements generating Noncontact
Cooling Water Only
E. Solid and Semi-Solid Hot Melt Thermoplastic Adhesives
F. Dry Blended Adhesive Materials
Detailed descriptions of the manufacturing processes for each subcategory
are presented in Reference (2). The major similarity between the six
groups is that they are all manufactured in batch processes by compounding
raw materials in mix tanks or jacketed kettles.
*
San Leandro and Chicago plants are in subcategories B and C.
-------
TABLE 1 . STANDARD INDUSTRIAL CLASSIFICATION OF PRODUCTS
FROM THE ADHESIVES AND SEALANTS INDUSTRY (2).
SIC 2891
Adhesives
Adhesives, plastic
Calking compounds
Cement (cellulose nitrate
base)
Cement, linoleum
Cement, mending
Cement, rubber
Epoxy adhesives
Glue, except dental:
animal, vegetable, fish
casein, and synthetic
resin
SIC 2899
Iron cement, household
Laminating Compounds
Mucilage
Paste, adhesive
Porcelain cement, household
Rubber cement
Sealing compounds for pipe
threads and joints
Sealing compounds, synthetic
rubber and plastic
Wax, sealing
Gelatin: edible, technical, photographic, and pharmaceutical
WASTEWATER CHARACTERIZATION
Numerous raw materials are used throughout this industry and therefore
complex and varied wastewater streams are developed. The principal raw
materials in adhesives and sealants manufactured are listed in Table 2 (2)
indicating the wide variety of solvents, latices, surfactants, fillers and
preservatives commonly employed. The main source of wastewater from adhe-
sive manufacturing processes is the washing of the process vessels and lines
while a small portion of wastewater is generated from area housekeeping and
laboratories. The ratio of washwater volume to the other waste volumes is
not well-defined.
The wastewater flow from adhesives plants varies markedly between
subcategories. Typical averages are 249 m3/kkg (30,000 gal/1000 Ibs) for
subcategory A, 0.94,m3/kkg (113 gal/1000 Ibs) for subcategory B, 0.34,
m3/kkg (41 gal/1000 Ibs) for subcategory C, and zero discharge for sub-
categories D, E, and F. Plants in subcategory D discharge noncontact
cooling water only. Process vessel cleaning in subcategory E is performed
with a hot wax solution which is recycled back into production. For dry
blends (group F) neither water nor solvents are compatible with the
manufacturing process.
A summary of the raw wastewater loadings, by subcategory, is presented
in Table 3 for several major pollutants. The oxygen demand is significant
in the wastewater from subcategories A, B, and C. The suspended solids
loading in groups A and B is >3,000-4,000 mg/l, while oil and grease levels
-------
TABLE 2. PRINCIPAL RAW MATERIALS USED IN
ADHESIVE AND SEALANTS MANUFACTURE (2).
1. Water based solutions containing natural and synthetic materials.
Starches Polyvinyl acetate - homo-polymers,
Dextrins co-polymers
Sugars Acrylic polymers and co-polymers
Syrups Synthetic and natural elastomeric
Animal & Fish Glue lattices
Gelatins Polyvinyl chloride
Caseins Polyvinyl alcohol
Cellulose Rosin and rosin derivatives
Marine Colloids Bituminous
Lignin Hydrocarbon resins
Phenolic resins
2. Solvent solution adhesives and cements (Non-water based solutions).
Synthetic and natural Solvents
elastomers Aliphatic hydrocarbons
Synthetic and natural Ketones and mixed ketones
resins Aromatic hydrocarbons
Synthetic and natural Nitrated halogenated hydrocarbons
rosins, and modified Alcohols
rosins Esters
Plasticizers Ethers
Anti-oxidants Amines
Peptizing Agents
3. Solid and Semi-Solid and Thermoplastic Thermosetting Compounds
Synthetic polymers and Synthetics and natural waxes
copolymers Synthetics and natural oils
Synthetic and natural rosins Plasticizers
and modified rosins Synthetic and natural resins
4, Dry-Blended Adhesive Materials
Silicas Assorted Cements
Clays Fillers
Plaster
5. The following materials can be found, as additives, in the preceeding
groups.
Fillers Solvents and Plasticizers
Clays Aliphatic hydrocarbons
Calcium carbonates Ketones and mixed ketones
Calcium sulphates Aromatic hydrocarbons
Talcs Nitrated halogenated hydrocarbons
Pigments, dyes, oxides Alcohols
Asbestos Esters
Sand Ethers
Fly Ash, etc. Amines
-------
TABLE 2 (CONTINUED) PRINCIPAL RAW MATERIALS USED IN
ADHESIVE AND SEALANTS MANUFACTURE (2)
Surface Active Agents Preservatives
Surfactants Fungicides
Soaps Mildewicides
Defoamers Bactericides
Penetrating Agents
Miscellaneous Components
Organic Salts
Inorganic Salts
Acids
Bases
Humectants
Metals
Thickeners - Polymeric or
cellulosic, etc.
Anti-oxidants
-------
TABLE 3. SUMMARY OF THE RAW WASTEWATER LOADING FOR MAJOR
POLLUTANTS IN THE ADHESIVES AND SEALANTS INDUSTRY (2)*
Assay
Total Suspended Sol
Total Dissolved Sol
BOD5,
COD,
TOC,
Oil and Grease,
Average Wastewater
Flowrate,
ids,
kg/kkg
mg/£
ids,
kg/kkg
mg/£
kg/kkg
mg/£
kg/kkg
mg/&
kg/kkg
mg/£
kg/kkg
mg/£
m3/kkg
A
1,140
4,560
800
3,200
942
3,770
2,520
10,100
669
2,680
254
1,020
249
Subcategory
B
3.1
3,300
13.9
14,800
3.4
3,610
15.5
16,500
5.0
5,310
1.28
1,360
0.94
C
,i
0.01
29
0.4
1,180
4.4
12,900
7.3
21,500
1.4
4,100
0.0003
0.9
0,34
*No process water discharged in Subcategories D, E and F.
-------
in these two groups are >1,000 mg/£. Phenolic compounds may be present in
significant quantities in wastewaters generated from phenol formaldehyde
resins in subcategories B or C (2). Also, these wastewaters contain highly
stable chemical emulsions.
Both the flow and composition of wastewaters from adhesives plants are
highly variable, over both short term (minutes to hours) and long term (days
to-months) due to the batch nature of the manufacturing processes.
EFFLUENT LIMITATION GUIDELINES
In February, 1975, the Environmental Protection Agency included the
adhesives and sealants industry in a draft report entitled, "Development
Document for Effluent Limitations Guidelines and Standards of Performance,
Miscellaneous Chemicals Industry" (2)*. Best Practical Control Technology
Currently Available (BPCTCA) Standards and Best Available Technology
Economically Achievable (BATEA) Standards were developed for existing plants.
Best Available Demonstrated Control Technology (BADCT) Standards of Perfor-
mance were drafted for new sources. The preliminary contractor recommen-
dations are summarized in Table 4 for subsequent comparison with the results
of the field demonstration.
Local Municipal Treatment Authorities often have effluent limitation
guidelines which are more inclusive, and at times more difficult to satisfy,
than the Federal Standards. The discharge limits for two cities; Chicago,
Illinois and San Leandro, California and the ranges of sewer discharge
standards for approximately 20 municipalities (3) are given in Table 5.
Limitations on wastewater BOD and total suspended solids content are
generally written such that once a specified level is exceeded, surcharges
are imposed. In San Leandro surcharges are assessed on the entire BOD and
suspended solids loading of the wastewater, and the wastewater volume per
the following schedule (4):
Volume: $0.721 per m3
BOD: $1.229 x BOD (mg/A)/1000 per m3
Suspended Solids: $1.244.x SS (mg/£)/1000 per m3
These rates became effective 1 July 1977 and, in total, represent a 230%
increase over previous surcharges.
CURRENT TREATMENT METHODS
The majority of adhesive plants discharge their wastewaters directly
to municipal treatment facilities and, therefore, end-of-pipe treatment and
control technology are not practiced extensively within this industry (2).
In a survey of the adhesives industry conducted during the Effluent
Limitation Guidelines development program, "no adhesive plants which conduct
*
The miscellaneous chemicals industry classification also includes:
Pharmaceuticals, gum and wood chemicals, pesticides and agricultural
chemicals, explosives, carbon black, photographic processing and hospitals.
-------
TABLE 4. SUMMARY OF EPA PRELIMINARY CONTRACTOR RECOMMENDED EFFLUENT LIMITATION
GUIDELINES FOR THE ADHESrVES AND SEALANTS INDUSTRY
Subcategory
Wastewater
Flow, m3/kkg
Waste
Parameter
Raw. Wastewater
Kg/kkg mg/H
BPCTCA
Kg/kkg mg/t
BADCT
Kg/kkg mg/4
BATEA
Kg/kkg mg/d
B.
C.
glues and gelatins 249
Water-based ashesive
solutions containing
synthetic and natural
materials 0.94
Solvent solution
adhesives and cements
generating contaminated
wastewaters 0.34
BOD
COO
TSS
BOD
COD
TSS
BOD
COD
TSS
942
2,520
1,140
3,
15.
3,
4.
7.
,4
,5
,1
.4
,3
0.01
3,770
10,100
4,560
3,610
16,500
3,300
12,900
21 ,500
29
65.9
176
—
0.24
1.09
--
0.31
0.51
265
710
50
255
1,300
20
910
1 ,530
20
65
170
—
0.24
1.09
—
0.31
0.51
260
680
20
255
1,160
20
910
1,500
20
25
66.9
—
0.24
1.09
--
0.31
0.51
"
100
267
20
255
1,160
20
910
1,500
20
tNo process water discharge In subcategories D, E, and F.
-------
TABLE 5. MUNICIPAL DISCHARGE LIMITATIONS
Parameter
Arsenic, mg/£
Cadmium, mg/£
Chromium (total), mg/£
Cyanide (free), mgA
Cyanide (total ), mg/&
Lead, mg/£
Mercury, mg/&
Phenolic Compounds, mg/£
Zinc, mg/&
Oil and Grease, mg/£
pH, units
* Hexane extract! bles
Chicago,
Illinois (3)
-
2.0
25.0
2.0
10.0
0.5
0.0005
0.1
15.0
<100*
4,5-10.0
San Leandro,
California
0.1
0.2
0.5
-
1.0
1.0
0.01
1.0
3.0
300/1 OO1"
>6.0
Ranges for
%20 Cities (3)
<0.05-5
<0.02-5
_
-
0-10
<0.1-5
<0. 0005- 1.5
<0. 02-10
<2-15
<50-120tf
4.5-10.5
tt
300 mg/H oil and grease of animal or vegetable origin,
100 mg/£ oil and grease of mineral or petroleum origin.
Listed as grease, no details of analysis provided.
-------
complete treatment of their wastewaterdischarges were found. The only
type of treatment observed was physical treatment for suspended solids
removal" (2).
Pollution abatement measures which are typically exercised by adhesives
and sealants manufacturers include:
1. Controlled rinsing of process vessels and lines with rinse water
recycling as opposed to use of a simple filling and draining
technique;
2. Scrape down of tank residue where controlled rinsing is not
practical;
3. Blowing out process lines and pumping systems where rinse water
would not be reusable;
4, Cleaning of accidental spills before they enter the plant
effluent; and,
5. Use of a holding tank or pond for flow equalization and sus-
pended solids removal.
Recently, a full-scale ultrafiltration (UF) system was installed at the
Chicago Plant of the Dewey and Almy Chemical Division of W.R. Grace, Inc.
This plant has an average wastewater discharge of 76 m^/day (20,000 gal/day).
Since the UF system was installed this plant has not been cited by the
Metropolitan Sanitary District (MSD) of greater Chicago for any discharge
violations.
In contrast, an extensive examination of the application of coagulation
and flocculation, followed by dissolved air flotation showed only limited
success on Dewey and Almy's Chicago plant waste stream. A tabulation of the
best effluent hexane extractive levels attained with this conventional treat-
ment process is given in Table 6, while a summary of a full year's operation
of the Chicago Plant's ultrafiltration system is presented in Section 8.
BPCTCA, BADCT AND BATEA TREATMENT METHODS
The preliminary contractor report recommended biological treatment
(activated sludge) for subcategory A to provide BPCTCA technology level.
For this subcategory dual-media depth filtration (DMDF) of the secondary
treated effluent is suggested to meet BADCT requirements and a two-stage
activated sludge system with DMDF of the effluent is recommended for BATEA
treatment.
The recommended technology to provide an efflueat consistent with all
three technology levels for both subcategories B and C was double effect
liquid evaporation. Two days of detention time in a well-mixed equalization
basin were planned prior to evaporation.
-------
TABLE 6. HEXANE EXTRACTIBLE LEVELS FOLLOWING CONVENTIONAL TREATMENT OF
DEWEY AND ALMY CHICAGO PLANT WASTE STREAMS
Stream
Number
1.
2.
3.
4.
5.
6.
7.
8.
Common
Constituents*
Saponified oils, acids and resins
Calcium lignin sulfonate* phenol
Paraffin wax, PVC resin, oils
Resins, SBR latex, mineral oil
Latices.wax, resin, oils,
surfactants
Latex, rubbers, oils, surfactants
Dioctyl sebacate and phthalate,
zinc resinate
Resin, wax, oils, acids
Volume
m3, wk
37.9
2.3
3.8
76-95
26.5
17.0
3.9
5.7
Untreated
PH
11.0
10.7
12.0
8.9
9.8
9.2
12.3
12.0
Untreated Hexane
Extract! bles , mg/£
8,270
1 ,330
2,090
14,000
10,600
3,880
11,600
2,910
Treated Hexane
Extractive, mg/A'
330
330
172
243
77
108
623
147
*Partial listing.
Best results with conventional treatment consisting of coagulation and dissolved air flotation.
Note: Chicago'Municipal Discharge Limitation for Hexane Extractives is 100 mg/&.
-------
It is the authors' belief that no full-scale biological treatment or
double effect liquid evaporation systems are currently in use in the
adhesive and sealants industry.
TREATMENT APPROACH SELECTED FOR EVALUATION
The principal unit operation selected for evaluation was the membrane
separation process of ultrafiltration. This unit process had demonstrated
the ability to successfully treat one adhesive and sealants plant wastewater
discharge and its application to other subcategory B and C waste streams
deserved investigation.
Ultrafiltration is particularly effective in removing suspended solids,
oil and grease and complexed heavy metals. Advantages of ultrafiltration
over conventional treatment processes and evaporation are:
1. Ultrafiltration requires no heat addition (cooling may be required);
2. Ultrafiltration is economical at both small and large sizes,
because of its modular nature;
3. Ultrafiltration systems are very simple to operate since they
involve, primarily, the pumping of liquids;
4. Energy requirements for ultrafiltration are low since operation
proceeds through the pumping of liquids;
5. Ultrafiltration performs best with chemically-stabilized
emulsions (as are found in adhesives manufacturing wastewaters);
6. Ultrafiltration is generally insensitive to shock loading. During
processing of streams containing latex, latex instability may
cause membrane fouling;
7. The ultrafiltrate will be essentially free of suspended solids
and will have a low oil and grease content;
8. Heavy metals should be efficiently removed by ultrafiltration.
This is because the metals will be insolubilized by reaction
with negatively charged colloids and will not pass through the
ultrafiltration membranes; and,
9. The concentrate volume will be significantly less than that
produced in a coagulation-solids separation process.
Because of these characteristics of ultrafiltration this process has the
potential to be a viable alternative to double effect evaporation recom-
mended for subcategory B and C wastes with BPCTCA and BATEA standards.
In conjunction with the ultrafiltration process three methods of post-
treatment were studied: reverse osmosis (RO), activated carbon adsorption
(ACA) and chemical oxidation. These processes are all capable of removing
dissolved organic species from wastewater, thus lowering the ultrafiltrate's
biological and chemical oxygen demands. Also, with proper post-treatment
water reuse within an adhesives plant is considered feasible.
11
-------
FIELD DEMONSTRATION SITE
The site of the field demonstration tests was the manufacturing facility
of the Dewey and Almy Chemical Division of W.R. Grace & Company located at
San Leandro, California. At this plant a variety of adhesives, sealants and
construction products additives are produced. Although over twenty "custom
tailored" products are manufactured at the San Leandro Plant, the product
line can be summarized into five basic categories. These categories are:
Product Identification Use
Water Base Sealant for jars, cans, caps, etc.
Cover and Drum Sealant for drums, large containers
Solvent Base Sealant (fast drying) for cans and caps
"Z" Internally used latex product; added
to solvent base sealants
Construction Products
Additives (CPD) Various uses include: reducing water
content in concrete; reducing setting
time of concrete; improving pump-
ability of concrete, etc.
All products are manufactured in batchwise operations.
Wastewater is generated from the manufacturing of all products with the
exception of solvent base sealants. For these products, the cleaning
operations are "closed-loop". Table 7 details the principal raw materials
used in the manufacture of the other four product types. The most prevalent
raw material is styrene-butadiene latex.
Several other waste stream Sources are present at the San Leandro
Plant. These include: periodic washdown of the crystal flux apparatus
(contains zinc ammonium chloride), CPD tank truck rinsing, cooling water
overflow and boiler blowdown.
All plant wastewater flows into a sump [estimated capacity 1.9-3.8 m^
(500-1000 gal) ] prior to discharge to the sewer. Approximately 19 nr
(5000 gal) of wastewater are generated per day .
12
-------
TABLE 7. PRINCIPAL RAW MATERIALS USED IN MANUFACTURE
OF SAN LEANDRO PRODUCT LINE
Product
Raw Materials
Water Base Sealant
Cover and Drum Seal ant
Construction Products
Additives
Styrene-Butadiene (SBR) latex, alcohol, small quantities of
xylene and toluene
SB latex, some neoprene latex
SB latex with clay fillers and antioxidants, isopropanol, sodium
nitrate, pinene resins
Sodium lignum sulfonate, phenols, wetting agents
-------
SECTION 2
CONCLUSIONS
This demonstration of adhesives and sealants manufacturing wastewater
treatment by ultrafnitration has been augmented by a review of a number of
post-treatment alternatives for the UF permeate. From this work three treat-
ment system options for adhesives and sealants wastes have been conceptual-
ized and full-scale treatment costs have been projected for system
capacities of interest to this industry. In general, a high-quality product
water is produced by the UF system; however, it may contain (depending on the
manufacturing site) residual quantities of phenolic compounds and cyanide
which are in excess of local Municipal Discharge Limitations. These
contaminants, which are incompatible with subsequent biological treatment in
publicly-owned treatment plants can be removed from the UF permeate by
either ozonation or reverse osmosis.* The results of the field demonstration
were compared to the preliminary contractor recommendations as presented in
the draft development document for this industry. The contractor's recommended
values for BOD and COD were not achieved by any combination of waste equali-
zation, ultrafiltration and post-treatment evaluated. The large capital
investments associated with double-effect liquid evaporation (the treatment
model presented in the Draft Development Document) may, however, severely limit
the use of such technology.
These general conclusions are supplemented by the following specific
findings:
1. Plant Effluent Characteristics
- The flow of wastewaters from the San Leandro Plant did not
follow any set pattern. This is because of the batchwise
nature of all manufacturing operations. Equalization of the
plant wastewater flow before UF processing is therefore
advisable.
- The plant effluent exhibited wide ranges in contaminant loadings.
For example, the effluent total suspended solids ranged from
59 nog/a to 70,800 mg/£ and its zinc content varied from 2.4,mg/£
to 740 mg/£. Again, wastewater equalization prior to ultra-
filtration would be advisable.
- The suspended solids loading in the total plant effluent
averaged 10,600 mg/£ during the field demonstration tests.
Gravity settling and/or flotation of a portion of these
* As indicated by laboratory-scale feasibility tests. These post-treatment
processes remain to be demonstrated.
14
-------
suspended solids would reduce the loading on the ultra-
filtration system and therefore lower the membrane area
requirement. This, in turn, would reduce the UF system
capital cost and would lower the UF system operating
expense.
2. Ultrafiltration System Performance Characteristics
- During treatment of the total effluent from the San Leandro
Plant the UF permeate averaged 100 mg/£ total freon extractives,
< 7.4lmg/£ non-polar freon extractives, < 27 mgA suspended
solids, 0.43 mg/£ free cyanide, 3.6 mg/i total cyanide,
8,900 mg/& BOD, 36,600 mg/£ COD, 44,6 mg/£ phenolic compounds
and 1.5 mg/£,zinc. The average removal efficiencies for total
freon extractives, non-polar freon extractives and suspended
solids were 92.2%, 94,7%, and 99.6%, respectively. The UF
system removal efficiency for BOD averaged 24,0% and for COD
averaged 38.2%. The mean.zinc removal was 90.8%. Accurate
removal efficiencies for free and total cyanide could not be
calculated due to interference in the assays for these
contaminants. A treated effluent of the above quality is
acceptable for discharge under the San Leandro Municipal Dis-
charge Limitations with the exception of the phenolic compound
and total cyanide loadings. Surcharges would be imposed,
however, based on the suspended solids and BOD loadings. An
effluent of this quality does not conform to recommended BPCTCA
or BATEA standards for BOD and COD, subcategories B and C.
- The suspended solids loading in the UF permeate was typically
below 5 mg/£. Periodically, however, secondary precipitation
occurred which increased the permeate suspended solids loading
(during processing of the total plant effluent) to as high as
160 mg/£. The mechanism by which this secondary precipitation
occurs was not identified during this program. In Chicago it
was found that this phenomenon occurred with high bacteria
counts and low pH.
- The average plant effluent concentrations of cadmium, total
chromium, lead and mercury were below the San Leandro
Municipal Discharge Limitations before Ultrafiltration. The
concentration of .arsenic in the wastewater was below the
detection limit of the assay method employed (0.2 mg/a). Thus
discharge of the plant wastewater is not limited by any of these
constituents.
- Average flux for the Abcor, Inc. type HFM Ultrafiltration
membranes during processing of the total plant effluent was
1.38 m3/m2-day (33.8 gfd). This flux was averaged throughout
six individual tests with a total operating time of 1021 hours
and is an economically acceptable design flux.
- Typically, a short duration (1/2-hour) detergent wash cycle was
capable of recovering membrane flux to acceptable levels. In
two instances, severe latex fouling occurred. The use of
mechanical cleaning in one case, and solvent cleaning with
mechanical cleaning in the second case were necessary to affect
15
-------
membrane flux recovery. Membranes which are solvent cleanable
and membrane modules which can be mechanically cleaned (if
necessary) are therefore required for ultrafiltration of
adhesives and sealant manufacturing wastewaters.
- Continuous addition of surfactant to the UF tank was exercised
on two occasions during the field demonstration program. No
significant gain or loss in UF membrane flux or rejection char-
acteristics were observed. Therefore, surfactant addition (used
at the Chicago plant) is not viewed as a required process step.
If, however, fairly unstable latices are present in a plant
wastewater surfactant addition may be necessary to maintain
economic flux levels.
- Treatment of the total plant effluent less the CPD waste stream
during one field test and less the "Z" dispersion waste stream
in another test was performed in an effort to identify a "bad
actor" stream in terms of either membrane flux or rejection.
While no significant benefits in UF system performance were
gained by segregation of either stream, it was observed that
segregation of the CPD stream lowered the remaining plant
effluent phenolic compounds loading to an average of 4,8 mg/£.
Even with this lessening of the phenolic compound loading,
treatment of the total plant effluent was preferred over
stream segregation.
3. Post-treatment Process Evaluation
- Pilot-scale tests performed with the Grace Chicago Plant UF
permeate demonstrated the technical feasibility of reverse
osmosis post-treatment. However, membrane degradation during
one of the two experiments conducted limits the generalization
of these results. While phenolic compound and cyanide concen-
trations in an RO product water are predicted to be below
municipal discharge limitations, BOD and COD loadings are
expected to be ~3X the values recommended by the contractor
as the draft BPCTCA, BADCT and BATEA standards.
- Carbon isotherm experiments with the Grace Chicago Plant UF
permeate indicated poor adsorption capacity for BOD and COD.
Therefore carbon adsorption is not considered a viable post-
treatment unit operation.
- Bench-scale ozonation tests on a sample of the San Leandro
plant UF permeate indicated the feasibility of this process
for phenolic compound and cyanide destruction. BOD and COD
content in the ozonated product water were each an order-of-
magnitude above the values recommended by the contractor as
the draft BPCTCA, BADCT and BATEA standards.
- Limited literature reviews on the processes of chlorination,
hydrogen peroxide oxidation and potassium permanganate oxidation
for phenolic compound and cyanide destruction were performed.
Each of these processes was eliminated from consideration as a
viable posttreatment alternative on either a technical or
economic basis.
16
-------
4, Conceptual Treatment System Design
- Three treatment system design options have been conceptualized.
These are:
Option 1. Equalization •*• Ultrafiltration"*Discharge
Option 2. Equalization •* Ultrafiltration"5" Ozonation"*
Discharge
Option 3. Equalization •* Ultrafiltration•* Reverse Osmosis "*
Reuse and/or Discharge
The Option 1 system is capable of producing an effluent
meeting the local municipal discharge standards if significant
levels of phenolic compounds and cyanide are not present in
the plant wastewater discharge. The Option 2 system is capable
of producing an effluent meeting local municipal discharge
standards when phenolic compounds and cyanide are present in
plant wastewater. System Option 3 is capable of concentrating
phenolic compounds and cyanide producing an effluent suitable
for discharge. It may also produce a product water reusable
within the manufacturing plant. None of the three treatment
options described is considered capable of reducing the
adhesives and sealants manufacturing plant wastewater BOD and
COD loadings to the BPCTCA, BADCT and BATEA levels recommended
in the draft development document.
5. Estimated Process Costs
- Purchased equipment costs were projected for each treatment
system option for each of three manufacturing plant waste-
water flow rates: 3.8 m3/day (1,000 gpd), 18.95 m3/day
(5,000 gpd) and 75.8 m3/day (20,000 gpd). One shift operation,
two shift operation and three shift operation was assumed for
the three system capacities, respectively. A matrix of
estimated purchased equipment costs (in thousands of dollars )
follows:
Plant Wastewater Flowrate (m-Vday)
System -
Option 3.8 18.95 75.8
1 $23.5 $52.9 $90.1
2 37.5 66.9 104,1
3 38.5 77.9 125.1
These cost estimates exclude installation costs which are
highly site specific.
- Annual operating costs for three capacities of each system
option have been developed. The annual operating cost
projections (in thousands of dollars) are summarized below.
17
-------
System Plant Wastewater FlowratP Qn3/day)
Option 3.8 18.95 75.8
1 $8.84 $20.8 $52.8
2 11.0 23.7 56.8
3 13.2 29.9 74,3
- Sludge disposal costs vary from 15% to 31% of the total
operating expense. If an effective means of sludge de-
watering is identified operating costs could be reduced
significantly.
6. Comparison of System Options, 1, 2, and 3 Costs with the prelimin-
ary contractor report's BPCTCA, BADCT and BATEA Technology Costs
- The recommended BPCTCA, BADCT and BATEA technology presented in
the Draft Effluent Limitations Guidelines Development Document
for the Adhesives and Sealants Industry is double effect liquid
evaporation for similar types of wastewater to those under
consideration in this report. The following costs (in
August, 1972 dollars) were developed:
Plant Wastewater Flow Rate (m3/day)
22.7 37.9
Capital Costs
($000) 324, 412
Annual Operating
Costs ($000) 139 189
While these costs cannot be directly compared with the treat-
ment system Options 1, 2 and 3 costs, it is clear that
double effect liquid evaporation would require a capital
investment of from 4,to 6 times (using 1972 dollars) the
cost of the equalization/ultrafiltration and posttreatment
system costs (using 1977 dollars). Annual operating costs
would also be significantly increased if the contractor
recommended BPCTCA, BADCT, BATEA treatment model is employed.
7. Evaluation of a Full-scale Equalization/Ultrafiltration System
Treating Adhesives and Sealants Manufacturing Wastewaters
- During the early 1970's a dissolved air flotation system was
installed at the Chicago, Illinois plant of the Dewey and Almy
Chemical Division of W.R. Grace and Company. The system
improved the quality of the plant's effluent in terms of oil and
grease, however, due to the effluent's variability from hour to
hour, there were frequent periods during which the pollutant
levels were excessive. Even with modifications and improvements
to this treatment system it was projected that the effluent
would be out of specification (> 100 mg/£ oils and grease) from
5% to 10% of the time. This treatment method was therefore
unacceptable.
18
-------
Pilot-scale ultra-filtration tests were performed and a full-
scale UF system (design capacity 81.9 m-Vday) was installed in
1974,at a total cost of $180,000. This cost includes all
accessory equipment, tanks and piping and installation expenses.
The UF permeate oil and grease (hexane extractibles) loading
has averaged 35 mg/£ compared to a Metropolitan Sanitary
District (MSD) of Greater Chicago specification of 100 mg/&.
The suspended solids concentration has averaged 24,mg/&. Iron
and,zinc concentrations in the permeate have averaged 0.8 mg/£
and 1.25 mg/Jl, respectively. Similar to the San Leandro results,
modest reductions in BOD are achieved by the UF membranes. The
UF permeate has maintained a 99% compliance level with MSD
specifications. Thus, UF treatment provides an effluent meeting
local municipal discharge standards but the effluent will not
achieve the values recommended for the draft Federal BOD and
COD discharge limitations.
The total operating costs for the Chicago Plant pollution control
system were $3.34/m3 ($12.66/1000 gal) in 1975 and $3.51/m3
(13.23/1000 gal) in 1976. These costs are in-line with System
Option 1 cost projections.
19
-------
SECTION 3
RECOMMENDATIONS
On the basis of the knowledge gained during this program, the following
recommendations are offered:
1. Field demonstration programs should be conducted for the post-
treatment alternatives of ozonation and reverse osmosis. Such
demonstrations would provide sufficient data for optimization of
each system's operating parameters and would develop a data
base for design calculations, and economic analyses. An accurate
profile of product water quality from each process should be
obtained and the reuse potential of each product water should be
assessed.
2. A comprehensive investigation should be conducted to identify
the preferred dewatering technique (s) for sludges generated
during the treatment of adhesive and sealants manufacturing
wastewaters. The percentage of the total operating costs
contributed by the sludge disposal operation for treatment
systems comprised of equalization, ultrafiltration and reverse
osmosis (Option 3) is 27% for a system processing 3.8 m-Vday
(1000 gpd); 29% for a system processing 18.95 m3/day (5000 gpd);
and, 31% for a system processing 75.8 m3/day (20,000 gpd). These
percentages and their related costs are quite significant. Thus
substantial savings in operating costs can be realized and overall
attractiveness of the treatment system can be enhanced if an
acceptable sludge dewatering method is demonstrated.
3. The contractor's draft development document and the recommended
treatment model (double effect liquid evaporation) for the ad-
hesive and sealants industry should be reconsidered. None of
the treatment schemes investigated during this program were
capable of meeting the contractor's recommended BPCTCA, BADCT
and BATEA standards for BOD and COD. However, they were
capable of meeting local municipal treatment discharge
standards and of removing pollutants incompatible with publicly-
owned treatment works. The Draft Development Document for this
industry has identified BOD and COD as pollution parameters
which are compatible with publicly-owned treatment works.*
*
However, COD may not be totally compatible with publicly-owned treatment
works in all cases.
20
-------
Considering the relatively small quantities of wastewater being
generated by subcategory B and C manufacturing sites, the capital
investment required for the proposed treatment model and the
energy-intensive operation of double effect liquid evaporators
it does not appear economically feasible to meet the contractor's
recommended discharge goals. Rather, "pretreatment" of adhesives
and sealants wastes by equalization, ultrafiltration and ozonation
or reverse osmosis (or other proven treatment schemes) to remove
oils and grease, suspended solids and pollutants incompatible
with municipal biological treatment systems should be required.
If the preliminary contractor's recommendations are subsequently
endorsed by the U.S. EPA, the proposed model treatment of double
effect liquid evaporation should be demonstrated at a represen-
tative adhesives and sealants manufacturing site.
21
-------
SECTION 4
PROGRAM OVERVIEW
The overall goal of this program was to demonstrate acceptable
technology for the treatment of effluents from adhesives and sealants
manufacture to produce water of a quality suitable for discharge to munici-
pal sewers. A secondary goal was to collect information on the nature and
variability of these wastes to better characterize effluents generated in
adhesives and sealants manufacture. Toward these ends a pilot-scale
ultrafiltration system, incorporating feed pretreatment via gravity
settling and flotation, was installed at the San Leandro, California plant
of the Dewey and Almy Chemical Division of W.R. Grace and Company. The
test program conducted at the San Leandro plant consisted of several tasks
including:
a. Determination of UF membrane flux and rejection characteristics
with the total plant effluent.
b. Evaluation of the effect of segregating various waste streams on
UF membrane performance characteristics.
c. Evaluation of the effect of surfactant addition on UF membrane
performance characteristics.
d. Determination of the maximum concentration achievable by ultra-
filtration.
e. Evaluation of membrane flux recovery with cleaning.
f. Sampling and analysis of the plant effluent, the effluent after
settling, the UF concentrate and the ultrafiltrate.
g. Determination of variability in flow of the plant effluent.
Also at the field demonstration site, samples of the plant effluent and the
UF concentrate were treated by several dewatering methods. Although in-depth
experiments were not performed a measure of effectiveness for the various
dewatering techniques was derived.
As a check of prior experimental work a one week evaluation of a
dissolved air flotation process was performed at the San Leandro Plant. The
process employed was LECTRO-CLEAR TM treatment developed by Swift Environ-
mental Systems Company. This is an electrolytic process involving an
electrocoagulation cell followed by an electroflotation basin. Periodically,
the LECTRO-CLEAR ™ effluent was further treated by the UF pilot system.
22
-------
The test program described above was supplemented by feasibility
experiments on the treatment of the product waters from ultrafiltration by
reverse osmosis and carbon adsorption. These experiments consisted of two
reverse osmosis batch concentrations and two carbon isotherm tests with
ultrafiltrate from the Dewey and Almy Chicago Plant. Additional post-
treatment work involved a brief literature survey of cyanide and phenolic
compound removal by chemical oxidation processes employing chlorine,
hydrogen peroxide, potassium permanganate and ozone. A single sample of
ultrafiltrate from San Leandro was ozonated by U.S. Ozonair Corp. to provide
an order-of-magnitude estimate of the economics for this process.
Effluent quality and membrane flux performance during one year's
operation of the Chicago Plant's ultrafiltration system were analyzed to
provide input into an evaluation of the technical feasibility of full-scale
treatment systems. Data from this plant also provided accurate operating
costs, especially with regard to membrane life, operating labor requirements,
and concentrate disposal.
Based on the results of the field demonstration experiments, the post-
treatment feasibility studies and the Chicago Plant's UF system operating
experience a full-scale treatment for the San Leandro Plant was designed.
This design included P & I drawings and sizing of all major process
components. The economics of full-scale system operation, including
estimates of capital and operating costs, were also developed.
23
-------
SECTION 5
DISCUSSION OF UNIT PROCESSES
The purpose of this section is to set forth certain principles and
definitions which will be used in subsequent sections. Many general
references are available which describe the relevant unit processes in
more detail .
ULTRAFILTRATION
Ultrafiltration and reverse osmosis are similar processes in that both
employ a semipermeable membrane as the separating agent and pressure as the
driving force to achieve separation. There are important differences,
however, which lead to different applications, process conditions and
equipment for each of the two processes. The approach in this program is
based on ultrafiltration as the principal unit process with reverse osmosis
as a possible posttreatment step. The differences between ultrafiltration
and reverse osmosis arc summarized in Table 8 while each process is
discussed separately below.
In an ultrafiltration process a feed solution/suspension is introduced
into a membrane unit, where water and certain solutes pass through the
membrane under an applied hydrostatic pressure. Solutes whose sizes are
greater than the pore size of the membrane and all suspended solids are re-
tained and concentrated. The pore structure of this molecular filter is
such that it does not become plugged because suspended solids are rejected
at the surface and do not penetrate the membrane.
For solutions which have no rejected species, such as water, the flux
through the membrane is given by:
- AP
where ,
J0 = Flux rate (m3/m2 - day)
AP = Pressure drop across the membrane (pressure driving force)
(atm)
Rm = Resistance of clean membrane (m2-day-atfn/m3)
R. = Resistance of fouling layer (m2-day-atm/m3)
No material from the process stream builds up on the membrane surface and,
therefore, for water the flux is pressure dependent and flow independent.
24
-------
TABLE 8. DIFFERENCES BETWEEN REVERSE OSMOSIS AND ULTRAFILTRATION
Item
Reverse Osmosis
Ultrafiltration
ro
Size of solute retained
Osmotic pressures of feed
solutions
Operating pressures
Nature of membrane retention
Chemical nature of membrane
Typical membrane flux levels
Molecular weights generally
less than 500
High salt retention
Important, can range to over
69 bar
Greater than 28 bar, up to
138 bar
Diffusive transport barrier;
possibly molecular screening
Important in affecting trans-
port properties
0.08 to 0.61 m3/m2-day
Molecular weights generally
over 1000
Nil salt retention
Negligible
0.7 to 6.9 bar
Molecular screening
Unimportant in affecting transport
properties so long as proper pore
size and pore size distribution are
obtained.
0.82 to 8.2 m3/m2-day
-------
When ultrafiltering solutions having high concentrations of rejected
material, the observed flux levels are much lower than the water flux of the
clean membrane. A gel layer develops and the following equation applies:
J] = AQX An Cb (2)
where,
J
A
C.
1
Flux rate
A constant which is a function of feed channel dimensions
and fluid properties
Concentration of rejected species in the gel layer
y
C, = Concentration of rejected species in the bulk solution
Q = Circulation rate of fluid through the membrane modules
(m3/min)
X = Empirical constant (generally 1
-------
TABLE 9. COMPARISON OF COMMERCIALLY AVAILABLE ULTRAFILTRATION
MODULE CONFIGURATIONS
Commercially-
Available
Configurations
Advantages
Disadvantages
Tubular
Spiral-Wound
Hollow Fiber
(Tubules),
Tubeside Feed
1. easily cleaned chem-
ically or mechanically
if membranes become
fouled
2. can process dirty feeds
with minimal pretreat-
ment
3. good hydrodynamic control
4, individual tubes can be
replaced
1. compact-good membrane
surface to volume ratio
2. less expensive than
tubular modules
1. compact-very good
membrane surface to
volume ratio
2. economical
PIate-and-Frame 1.
2.
good membrane surface
to volume ratio
well-developed equip-
ment
relatively high volume
required per unit
membrane area
relatively expensive
at present
susceptible to plugging
by particulates
badly fouled membranes
difficult to clean
very susceptible to
plugging by
particulates
potentially difficult
to clean
susceptible to plugging
at flow stagnation
points
potentially difficult
to clean
presently very
expensive
27
-------
become fouled. Spiral-wound and hollow fiber modules are less expensive
then tubular modules in dollars/m^ of membrane area, generally have lower
power requirements, and are more compact. Both spiral-wound and hollow-fiber
modules are, however, more susceptible to plugging and may be difficult to
clean. Plate-and-frame modules are quite expensive and, in case of failure,
the entire membrane module must be replaced.
For treatment of latex containing wastewaters from adhesives and
sealants manufacture the tubular geometry is judged to be most suitable in
terms of both process reliability and ease of membrane cleaning. A tubular
membrane, as shown in Figure 1, consists of a porous tubular support with
the membrane either cast in place, or inserted into the support tube. The
feed solution is pumped through the tube; the concentrate is removed down-
stream; and the permeate passes through the membrane/porous support
composite.
SYSTEM DESIGNS FOR ULTRAFILTRATION EQUIPMENT
Three common ultrafiltration system designs are shown in Figure 2.
In the batch concentration mode of operation (Figure2a), the feed tank is
charged with waste only at the beginning of each concentration cycle.
During operation the permeate is continuously withdrawn while the concentrate
is recycled to the feed tank. As the run proceeds the volume of waste in
the feed tank decreases, and its concentration increases. When the volume
of waste is sufficiently low, it is discharged and a fresh batch of waste
is charged to the feed tank. The degree of volumetric concentration is
given by
Cv - ^ (4,
where V0 and Vp are'the initial batch volume and the collected permeate
volume, respectively. The degree of volumetric concentration, Cv, is
related to the overall water recovery, by the relationship
x 100 (5)
Corresponding values of the volumetric feed concentration and the system
water recovery are shown below.
c Equivalent Water
_v Recovery (%)
lx (Vp=0) 0
2x 50
lOx 90
20x 95
50x 98
lOOx 99
28
-------
IV5
PERMEATE
CONCENTRATE
FEED
MEMBRANE
PVC END FITTING
DOUBLE O RING
SEAL
EPOXY RE-1NPORCED
FIBERGLASS BACKING
FIGURE 1. CUT-AWAY VIEW OF A TUBULAR ULTRAFILTRATION ASSEMBLY.
-------
Intermittent
Addition of
Waste
Feed
Tank
Intermittent
Slowdown of
Residue
(a) Batch Concentration
Feed
Concentrate
Permeate ^
UF Membrane
Module
Pump
Concentrate
Concentrate
Waste v
Recycl e
t Feed ^
,,- *"
— — — y
Permeate v
(b) Continuous Feed and Bleed
Permeate
Permeate
(c) Continuous Staged: Once-Through
FIGURE 2. VARIOUS SYSTEM DESIGNS FOR MODULAR MEMBRANE EQUIPMENT.
30
-------
There are three advantages to the batch concentration mode of operation
1. Feed circulation rate within the modules can be adjusted
to control membrane fouling and/or concentration
polarization.
2. High system conversions can be obtained by concentrating
to a very low residual volume.
3. The average feed concentration over the batch concentration
is minimized (compared to other modes of operation)
resulting in a maximum time-averaged module flux and
rejection over the concentration cycle.
The disadvantages of this mode relate to its intermittent nature of
operation. Since it is not continuous, it requires large holding tank
capacity and somewhat more operator time than the other modes.
The continuous feed and bleed mode of operation is shown in Figure 2b.
The advantages of this mode are:
1. It is continuous.
2. Feed circulation can be adjusted to control concentration
polarization.
3. High system conversions can be obtained.
The disadvantage of this mode is that the system is operated at the
concentration level of the concentrate stream. Thus, the average flux and
rejection will be low relative to that of the batch concentration mode.
For sufficiently large systems continuous once-through operation, shown
in Figure 2c, is preferred. This mode combines the advantages of both the
batch and the feed-and-bleed modes of operation. The feed passes through
each module in a single-pass which minimizes the average feed concentration
and achieves maximum utilization of the modules in terms of flux and re-
jection. In this mode, operation is continuous and a high overall system
conversion can be obtained.
The preferred mode of operation for any given application may be a
modified form of one of these three more common modes. The operating mode
selection depends upon UF feed flow conditions, membrane flux, water
recovery desired, membrane cleaning frequency, etc.
POSTTREATMENT UNIT PROCESSES
Reverse Osmosis
Reverse osmosis, a second membrane separation process, is based on the
unique property of semipermeable membranes to selectively pass water while
retaining dissolved solutes. If dilute and concentrated solutions are
31
-------
separated by a semi permeable membrane, water spontaneously passes by direct
osmosis from the dilute solution to the concentrated one, in order to
establish thermodynamic equilibrium (equal chemical potential on both sides
of the membrane). By imposing a hydrostatic pressure on the concentrated
solution, exceeding its "osmotic pressure", water can be forced from the
concentrated side to the dilute side by reverse osmosis. Separation by
reverse osmosis will continue until equilbrium is reestablished, at which
point the difference in applied hydrostatic pressure across the membrane will
be equal to the difference in osmotic pressure.
Reverse osmosis membranes are characterized by high rejection for
dissolved inorganics and poor to high rejection of dissolved organics,
depending on the specific characteristics of the organic solutes. For
example, the rejection for some low molecular weight uncharged organics is
rather poor, and is a complex function of the membrane polymer material and
solute diffusivity and solubility in the membrane. Ionic species are highly
rejected by interaction with fixed charges on the membrane surface. In
general, ionic species and large organics will be substantially rejected by
RO membranes; small hydrogen-bonding organics and non-ionized acids and
bases will be poorly rejected.
There are a number of reverse osmosis membranes materials presently
under development, but only two are in commercial use. The most widely
applied is cellulose acetate. It exhibits excellent water permeation rates
and high rejection of ionic species. Unfortunately, it is limited to a
fairly narrow pH range (2.5-7). The other commercial membrane material is
aromatic polyamide. It is available in a spiral-wound configuration from
UOP, Inc. or in a hollow-fiber configuration from DuPont, Inc.*. The
polyamide membranes have a broader pH range (5-11) but are sensitive to low
levels of free-chlorine or other oxidants.
Reverse osmosis systems may be operated in the same manner as discussed
for ultrafiltration systems.
Carbon Adsorption
Adsorption by activated carbon is a surface phenomenon in which
dissolved organics are removed from wastewater and concentrated at the
carbon-liquid interface. The degree of adsorption which occurs is a
combination of solute solubility in the wastewater and the strength of the
attractive forces between the solute and the carbon. The more hydrophilic
the organic, the less likely it is to move toward the carbon-water inter-
face. Thus, highly soluble organics tend to be poorly adsorbed by carbon;
whereas less soluble organics are more highly adsorbed.
Activated carbon is a highly porous material which is characterized by
a typical surface area yield of 1000 m2/gram. Since adsorption is a surface
*
UOP's PA-300 aromatic polyamide spiral-wound modules were not available
commercially until after the posttreatment experiments were completed.
32
-------
phenomenon, activated carbon has the potential (depending on the nature of
the dissolved organics) to be a highly-effective, economical unit process
for improving water quality.
The amount of organic adsorbed at equilibrium is usually expressed by
an "adsorption isotherm". The isotherm is a plot of the weight of organic
adsorbed per unit weight of carbon (X/M) versus the organic concentration in
the waste (C) when equilibrium is established at a constant temperature. A
number of mathematical expressions have been proposed (5) to describe the
shape of the isotherm. The most generally applicable expression is the
Freundlich adsorption equation:
^=kC1/n (6)
where,
X = amount of organic adsorbed
M = weight of carbon
k = constant
n = constant
C = concentration of unadsorbed organic in surrounding solution
at equilibrium
Restating this equation in logarithmic form,
^) = log k + (1) log C (7)
A plot of X/M vs. C on logarithmic paper will yield a straight line with
slope 1/n if the Freundlich isotherm is followed.
Little progress has been made for liquid systems in predicting the
isotherm from the properties of the carbon and organic. Therefore, isotherms
must be determined experimentally for each waste-carbon combination. The
Freundlich expression given by Equations (6) and (7) is useful for
correlating the experimental data.
Chemical Oxidation
Dissolved organics or toxic species can be removed from wastewater by
the application of strong chemical oxidants which breakdown the dissolved
species into other less harmful species, water and gases. Those oxidants
which are currently used to treat industrial wastes include ozone, chlorine,
hydrogen peroxide and potassium permangante. The selection of the preferred
chemical oxidant depends on several factors including: the type and nature
of the organics, the concentration of the organics, the wastewater pH, the
degree of organic destruction required (i.e., partial or total destruction),
pretreatment requirements, maintenance requirements, and economics.
33
-------
Two families of compounds which may be present in wastewater generated
from adhesives and sealants production are cyanides and phenolic compounds.
Discharge limitations on both of these groups are in effect in most
municipalities and residual amounts of these compounds are to be expected in
an ultrafiltrate stream. Various chemical reactions take place when cyanide
or phenolic compounds are exposed to the different chemical oxidants.
Table 10 summarizes these chemical reactions and provides relevant comments
on the processes, as appropriate.
DEWATERING TECHNIQUES
There are five major types of dewatering methods used for concentrating
liquid wastes. These methods are described briefly below.
Gravi ty Sedimentation
The effluent may be both thickened and clarified using sedimentation
which removes suspended solids from the liquid by gravity settling. Examples
of gravity settlers are a simple trench and the over-under sump.
Gravity Filters
The separation of solids in this filtering system is the result of the
hydrostatic pressure of the effluent on a static, vibrating or rotating
screen. Examples of this are the Sweco vibrating separator and the Bauer
hydrasieve.
Pressure Filters
Pressure filters are those which operate under forced positive or
negative pressure on a filter medium. Examples are the filter press and
rotary vacuum drum filter.
Centrigugal Filters
Through centrifugal force and a difference in density, an effective
liquid-solids separation can be accomplished. Examples are the basket
centrifuge and Bauer's liquid cyclone.
Solids Drying
The removal of a liquid by evaporation can be used when mechanical
methods are not feasible. Because of the high energy usage this method
may be prohibitive in dewatering sludge. Examples of this are evaporators
and spray dryers.
34
-------
TABLE 10. CHEMICAL REACTIONS WHICH OCCUR DURING OXIDATION OF CYANIDE AND
PHENOLIC COMPOUNDS (6, 7, 8, 9, 10)
Chemical
Oxidant
Cyanide Destruction
Corner) ts
Phenolic Compound
Destruction
Comments
CO
en
a) NaCN«-Cl2-».CNC1+NaC1
b) CNCHZNaOH -vNaCHO*
NaCl + HZ0
c) 2NaCNO+4NaOH+3Cl2-»
6NaCI+2C02+N2+2H20
Chlorine
Hydrogen d) CN" +H202-«-CN
Peroxide fij CNO"+2H;,0->-CO,+NH,+
OH- ' i 3
Potassium f) 2KMn04+NaCN+2KOH-»-
Permanganate 2KzMn04+NaCNO+«20
Ozone g) CN'+03->CNO-+02
h) CN
1) NH3+403+ HN03+402+
Product of reactions a + b 1s
cyanate. Cyanate can be further
oxidized as shown in reaction c.
Ratio of chlorine to cyanide for
reduction to cyanate is 2; for
complete destruction Is 8.5.
Process requires 3-4 parts H202
and 2-3 parts formaldehyde per
part cyanide in the presence of
a copper salt catalyst.
Process requires 12 parts KMn04
and 3 parts caustic per part
cyanide. Complete destruction
is not possible using
permanganate.
Ozone demand Is not well defined.
Complete destruction estimated
to require 3 to 9 parts ozone
per part cyanide.
5H20-*18C02+28KOH+
28Mn02
m) CeHcOH+03 •* various
products + 02
Partial oxidation achieved
through reaction j, however
chlorinated phenols are also
toxic. Complete oxidation
may occur in an excess of
chlorine (i.e. 25 parts
chlorine to 1 part phenol).
Iron salts as catalyst.
Initial reaction products
are hydroquinone and catechol.
These are further oxidized to
dibasic acids. Destruction of
phenols to dibasic acid state
requires 2 parts H202 per part
phenol. .
Process requires 6-7 parts
KHn04 per part of phenol.
For removal of phenol and its
aromatic oxidation products
the stoichpmetric ratio of 0^
to phenol is 3 to 1. Competing
reactions may make It advisable
to use a 6 to 1 ratio.
-------
SECTION 6
TEST SYSTEMS, PROCEDURES, AND ANALYSES
FIELD DEMONSTRATION TESTS
Held Test System and Procedures
A flow schematic of the complete San Leandro Plant Test System is pre-
sented in Figure 3. Plant wastewater was collected in an existing sump
[approximate capacity 3.8 m3 (1000 gal)] in which high and low level switches
were installed. During normal operation a 109 m3/day (20 gpm) constant
volume pump (Moyno Model IL6 Type SSF) transferred the plant effluent to
the first stage of a four stage 10.6 m3 (2800 gal) settling/flotation tank.
The Moyno pump was equipped with a timer/recorder to monitor its on/off
cycle. If a surge in the plant effluent occurred, a second pump (Gorman-
Rupp Model 12C-2B, also equipped with a timer/recorder combination was
activated and discharged the effluent to the sewer (by-passing the settling
tanks).
The second and fourth stages of the 10.6 m3 tank acted as flotation
tanks while the first and third sections provided settling area. A transfer
pump (Gorman-Rupp Model 81-1/2D3) delivered the effluent from the last stage
of this tank to a 3.8 m3 (1000 gal) holding tank. This tank served to
further clarify the plant wastewater. Overflow from this tank was piped
to the sewer. A solenoid valve regulated by a high/low level switch in the
UF system feed tank controlled the flow from the 3.8 m3 tank. Surfactant
could be fed into the UF feed tank by a metering pump (FMI Model RP2G).
Figure 4,presents a detailed flow schematic of the Ultrafiltration Pilot
System, The plant effluent, after settling and flotation, entered the
0.57 m3 (150 gal) UF feed tank and was passed through an in-line strainer
for coarse solids removal and into the suction of the centrifugal circulation
pump (Worthington Model D12). This pump pressurized the feed and delivered
it to the membrane inlet manifold. Pressures before and after the tubular
membranes and the feed temperature were measured. A low pressure switch on
the membrane outlet shut down the circulation pump when the concentrate
pressure fell below 0.3 atm (4,psig). A heat exchanger was provided on the
concentrate line to maintain the processing temperature below 52°C (125°F).
The concentrated waste returned to the UF feed tank and was combined with
fresh feed.
36
-------
Plant
Effluent
To Sewer
TM R
Gorman-Rupp
Pump
Existing Sump
3> H/LLC ™ R :i°9 m3/day
Constant
"Vol ume
— Pump (Moyno)
Overflow to
Sewer
n H/LLC
.--1
10.6 m3 SETTLING/FLOTATION TANK
co
Overflow to
Sewer
3.8 m3 Tank
LEGEND
BV - Ball Valve
DV - Diaphragm Valve
FT - Flow Totalizer
H/LLC - High/Low Level Control
TM - Timer
R - Recorder
SV - Solenoid Valve
1,2,3,4 - Sampling Locations
Surfactant
(Triton-X 100)
Metering pump
Concentrate
!T^h If
A L—^
H/LLC i
v
0.58 m3
UF
Circulation
Pump
Transfer
Pump
DV
Ultrafiltration
Membranes
Permeate
FIGURE 3. FLOW SCHEMATIC OF SAN LEANDRO PLANT FIELD DEMONSTRATION TEST SYSTEM.
-------
OJ
CO
Legend:
T - Temperature gauge
FI- Flow Indicator
FT- Flow Totalizer
3 - Pressure
/- Valve
gauge
Water I
In
t
1
-|X3
i
?
1
I
|
1
\
\
\
i
1
i
vJ
'
c
(r
$.
\
) "
lff
i
^
b. VI 3
Ultraf iltrate
-^ sewer
f Strainer f~\
Ik
Drain
Circulation
Pump
FIGURE 4. ULTRAFILTRATION SYSTEM FLOW SCHEMATIC.
-------
The ultra-filtrate flowrate was measured and the ultrafiltrate was
discharged to drain. The ultrafiltrate flow meter was equipped with an alarm
to signal low flow (i.e. membrane fouling) and a by-pass loop to allow
stop-watch and graduated cylinder flowrate readings should the ultrafiltrate
flow be off-scale. A flow totalizer (Kent PSM 190) recorded the cumulative
volume of ultrafiltrate produced.
The UF pilot system was operated in a modified batch mode. The concen-
trate was recycled to the 0.57 m3 UF feed tank, while the ultrafiltrate was
continuously withdrawn.
Each field demonstration test was generally conducted continuously,
Monday through Friday, for two consecutive weeks. Following this 10 day
concentration cycle the UF feed tank was drained, the membranes were
detergent cleaned and the membrane flux recovery was measured.
Two types of ultrafiltration membranes were utilized during this
program. Initially, the UF system was equipped only with Abcor, Inc. type
HFM (non-cellulosic) tubular membranes. Midway through the program a
second Abcor membrane, type HFD (non-cellulosic) was tested in parallel with
the HFM membranes. The characteristics of each membrane type are given in
Table 11. Both the HFM and HFD membranes were supplied in tubular assemblies,
0.025 m (1 inch) in diameter x 1.52 (5 feet) long. Typically, 21 tubular
assemblies (three parallel banks of seven tubes in series) were in place
on the ultrafiltration system.
The standard operating conditions for the UF system during the tests at
San Leandro were:
Feed circulation rate: 163.5 m3/day (30 gpm)
Membrane inlet pressure: 2.8-3.4 atm (40-50 psig)
Process temperature: Typical - 32.2°C (9QQF)
Range - 18.3 to 40.5<>C (65 to 105°F)
Feed pH: natural
Samples were collected of the plant effluent from the sump, the plant
effluent after the 3.8 m3 settling tank, the UF concentrate and the ultra-
filtrate. These sampling locations are identified in Figure 3, while the
assays performed on these samples are detailed at the end of this section.
Daily composite samples were formed by collecting equal volume grab samples
throughout an 8-hour day. Weekly samples were composited from the daily
samples for determination of certain assays. All samples were refrigerated
on-site.
UF System Cleaning Procedures
Detergent Cleaning--
The ultrafiltration membranes were cleaned with a detergent solution
between each experiment. Three different detergents were employed at
various points in the test program: Ultraclean (Abcor, Inc.), Dishmate
39
-------
TABLE 11. CHARACTERISTICS OF ULTRAFILTRATION
MEMBRANE EMPLOYED DURING FIELD TESTS
Resistance to
Solvents and Oils
Membrane Type
Parameter
pH range @ 38°C
Maximum Temperature
(°C) @ pH = 7
Maximum Operating
Pressure (atm)
0
Average Pore Size (A)
Equivalent Molecular
Weight Cut-Off
Tolerance to Free
Chlorine
HFD
3-12
85
5.2
50+25
20,000
None
HFM
0.5 - 12.5
90
5.2
50+25
20,000
Chemically iner
Good
to concentrations
to 50 ppm
Excellent
40
-------
(Calgon, Corp.) and trisodium phosphate. Dishmate was used only with HFM
membranes since it contains free-available-chlorine. The generalized
procedure for detergent washing was as follows:
1. The concentrated waste was drained from the system.
2. Clean water was passed through the system at a low
flowrate 109 m3 (20 gpm) to flush out the residual
concentrate.
3. A 1/2% by weight, detergent solution was recirculated
through the system for 30 minutes under the following
operating conditions:
Recirculation Flowrate: 109-136 m3/day (20-25 gpm)
Inlet Pressure: 1.4-1.7 atm (20-25 psig)
Temperature: 46-49°C
4. Clean water was passed through the system for 20-30
minutes at low flow and low pressure to flush out
the detergent.
5. The water flux of the clean membranes was determined.
Solvent Cleaning—
Twice during the field demonstration tests an irregular, grey rubber
(latex) coating deposited on the membrane surface. This latex coating is
difficult to remove with detergents, thus cleaning with methyl ethyl ketone
(MEK) became necessary. Solvent cleaning was performed both in the field
and at Maiden's pilot facilities as described below.
1. The tubular assembly was removed from the test unit.
Washers were removed and one end of the tube was plugged,
at the double o-ring seal (see Figure 1), with a rubber
stopper.
2. One liter of MEK was poured into the tube. Care was
taken to avoid contact of the MEK with any PVC fittings.
3. After a specified time period (between 5 and 35 minutes),
the MEK was drained and the membrane flushed with water.
4, (Field cleaning only). A laboratory test tube brush soaked
in MEK was used to remove the softened, swelled rubber
coating.
5. Spongeballs (see below) were passed through the tube until
the outlet water was clean.
6. The tubular assembly was reinstalled on the test unit and
its water flux was measured.
7. (Walden cleaning only.) Steps 1, 2, 3, 5, and 6 were
repeated with a Step 3 MEK exposure of 20 minutes.
41
-------
Mechanical Cleaning—
Mechanical cleaning of the tubular membrane assemblies was periodically
performed as an additional step to detergent cleaning and was always used
during solvent cleaning procedures. The mechanical cleaning procedures
involves the use of "spongeballs", cylindically shaped pieces of polyurethane
foam (0.025 m in diameter x 0.050 m long), and proceeds as follows:
1. The inlet of the first tubular membrane and outlet of the last
tubular membrane, in series, are disconnected from the rest of
the system. Two spongeballs are carefully inserted into the
first membrane. Water pressure from a hose is used to force
the spongeballs through the membranes.
2. Step 1 is repeated until two spongeballs have passed through
the membranes three times.
3. The membrane inlet and outlet are reconnected with the system
and the membrane flux recovery is measured.
Measurement of Flux Recovery—
The measurement of the flux of tap water through UF membranes, under
standardized conditions, indicates the water transport properties of the
membranes and is one means of detecting membrane degradation due to
compaction, plugging, biological attack and/or chemical attack. This measure-
ment was routinely made after any membrane cleaning operation.
During flux recovery (i.e. water flux) measurements the system was
operated with water at 163.5 m3/day (30 gpm) recirculation, 3.4 atm (50 psig)
inlet pressure, and ambient temperature. The observed ultrafiltrate flux
and temperature readings were used to obtain a flux reading corrected to
32.2°C (90°F) by use of the following relationship:
(Flux)32 20c = (Flux)TOcX (Viscosity of Water) T°C
(Viscosity of Water) 32.2°C
Reduction of flux data to a standard temperature simplifies data
analysis. A standard temperature of 32.2°C was selected since this was the
average process temperature during the first field demonstration test.
POSTTREATMENT EXPERIMENTATION
Reverse Osmosis Tests
A simplified flow schematic for the reverse osmosis test system is
presented in Figure 5. Ultrafiltrate from the Dewey and Almy's Chicago
Plant, was transferred to a 5.68 m3 (1500 gal) feed tank. A booster pump
(Dayton, Model 6F507) was used to pass the feed through two string-wound
42
-------
30y
Cartridge Filter FI
Concentrate
bleed
Booster 5y In
Pump String Wound
Cartridge Filters
High
Pressure
Pumo
Concentrate
UKiend
AC - Accumulator
BPR - Back Pressure Regulator
DV - Drain Valve
FI - Flow Indicator
FT - Flow Totalizer
IPS - Low Pressure Switch
P - Pressure Indicator
PS - Pressure Switch
SV - Sample Valve
TIC - Temperature Indicator/Controller
RO Permeate
to Collecting
Tank
FIGURE 5. SIMPLIFIED FLOW SCHEMATIC FOR REVERSE OSMOSIS TEST SYSTEM.
-------
cartridge filters, in series, to the suction of the high-pressure positive-
displacement pump. This pump (Gaulin Model 75E) increased the feed pressure
to 27-54,atm (400-800 psig). The feed pressure was controlled by the back
pressure regulator (BPR) and the flow rate through the module was controlled
by the concentrate throttle valve (V-3). An accumulator (AC) was used to
dampen pressure pulsations from the pump. The reverse osmosis module was
protected against overpressurization by a high pressure switch (PS), and
the pump was protected against running dry by a low pressure switch (IPS).
The feed temperature was measured and controlled by an indicating temperature
controller (United Electric, Type 1200). The flow rates of the permeate and
concentrate were measured, and the feed flow rate was calculated (sum of
permeate and concentrate flow rates). The feed pressure and pressure drop
across the module were also measured.
The reverse osmosis test system was operated in the batch operating
mode (i.e., concentrate returned to the feed tank, permeate continuously
withdrawn). The membrane module employed during the RO tests was a B-9
polyamide membrane in a hollow-fine-fiber configuration manufactured by
DuPont, Inc., Permasep Products Division (see Figure 6). This module has a
pH range of 4-11, a maximum operating pressure of 27 aim (400 psig) and an
operating temperature limitation of 35°C.
Typical values of the RO system operating parameters during tests with
the adhesives manufacturing plant ultrafiltrate were:
Feed circulation rate: 27.3 m3/day (5 gpm)
Module inlet pressure: 27 atm (400 psig)
Feed temperature: 27-30°C
Following each reverse osmosis experiment a standard test was performed
to determine if any decline in membrane flux or rejection had occurred.
After the system was drained and flushed with dechlorinated water, it was
operated in the total recycle mode on a standard NaCl solution (5000 ppm).
System operation proceeded under the set of conditions given above. At
steady state the feed and permeate flows and concentrations were measured
and the measured rejection was corrected to a module conversion of 0%.
Carbon Adsorption Isotherms
The carbon adsorption isotherm tests were conducted using the procedure
outlined below. Filtrasorb 400 (Calgon Corp.) a general-purpose carbon was
used.
1. Filtrasorb 400 granular activated carbon was ground with
a mortar and pestle and screened to 45y (335 mesh) size.
2. Seven samples of dried carbon were weighed out: 2 mg, 5 mg,
10 mg, 20 mg, 50 mg, 100 mg, and 500 mg.
3. Each sample of dried carbon was placed in a separate
erlynmeyer flask.
4. 100 (*I) ml ultrafiltrate was added to each flask.
44
-------
CONCENTRATE
SNAP RING OUTLET
•0' RING SEAL
SNAP RING
FEED
PERMEATE
END PLATE
FIBER
SHELL
•0' RING SEAL
FEED END PLATE
DISTRIBUTOR TUBE
FIGURE 6. CUTAWAY VIEW OF A PERMASEP HOLLOW-FINE-FIBER REVERSE OSMOSIS
MODULE.
45
-------
5. The flasks were stoppered and placed on a Burrel Wrist
Action Shaker for 24-48 hours.
6. The flask contents were filtered through a 0.22y Mi Hi pore
Filter, and the center portion of filtrate was collected
for analysis.
7. The seven carbon treated samples, an original feed sample
taken through all procedures except for carbon addition,
an original feed sample not taken through the isotherm
procedures, and a high purity water sample were analyzed
for TOC, BOD, and/or COD.
8. The data were fit to a Freundlich Isotherm Expression by
plotting, on log-log paper, X/M, the amount of pollutant
adsorbed per unit weight of carbon versus C, the residual
pollutant concentration and drawing the best straight line
through the points.
SAMPLE ANALYSES
Table 12 lists the sample analyses routinely performed during the field
demonstration tests and the methods employed for each analysis. During the
first six field tests the samples (both daily and weekly composites of grab
samples) were air-freighted to the Walden Analytical Laboratory for analysis.
All assays were performed by the Walden Analytical Laboratory except total
cyanide and mercury which were performed by Environmental Research and
Technology, Concord, Massachusetts. Following the sixth field test,
analytical work was performed in California by Engineering Science, Inc.
Throughout the field tests pH and total solids measurements were made by the
Dewey and Almy San Leandro Plant Analytical Laboratory.
During the posttreatment experiments BOD, COD and TOC assays were
conducted by Wai den's laboratory.
46
-------
TABLE 12. ASSAYS AND METHODS EMPLOYED DURING EXPERIMENTAL PROGRAM.
Constituent
Assay Method
Reference
Arsenic
BOD
Cadmi urn
Chromium (total)
COD
Cyanide (free)
Cyanide (total)
Freon Extract!bles (total)
Freon Extractives (non-polar)
Lead
Mercury
PH
Phenolic Compounds
TOC
Total Solids
Total Suspended Solids
Zinc
Dithiocarbamate Colorimetric
5 Day Incubation, Electrode
Atomic Adsorption
Atomic Adsorption
Dichromate Reflux
Selective Ion Electrode
Distillation, Titration
Separtory Funnel Extraction
Extraction, Gravimetric
Atomic Adsorption
Atomic Adsorption
Meter Reading
Distillation, AAP Colorimetric
Combustion-Methane Detection
Gravimetric
Glass Fiber Filtration
Atomic Adsorption
SM 404A*
SM 507, 422F, 422B
SM 301A
SM 301A
SM 508; EPA, p. 21**
Orion Manual
SM 413B, 413C
SM 502A
SM 502A, EMSL***
SM 301A
SM 301A VI
Manufacturer's Manual
SM 510, 510A, 510C
EPA, p. 236
SM 208A
SM 2080
SM 301A
* SM 404A (etc) refers to procedure number In "Standard Methods for Examination of Water and
Wastewater11, 14th Edition, APHA, 1975.
** EPA refers to "Manual of Methods for Chemical Analysis of Water and Wastes", U.S. EPA, 1974.
*** EMSL refers to method developed by Environmental Monitoring and Support Laboratory, U.S.
EPA, Cincinnati, Ohio (July 1975), Referenced in AQC Newsletter 22.
-------
SECTION 7
EXPERIMENTAL RESULTS AND DISCUSSION
SURVEY OF PLANT EFFLUENT CHARACTERISTICS
The batchwise nature of the manufacturing operation at San Leandro, and
throughout the industry, results in intermittant wastewater discharges.
Information was collected on the variability in flow and composition of the
San Leandro Plant effluent to better characterize effluents generated in the
manufacture of water-based adhesive solutions containing synthetic and
natural materials (Subcategory B), solvent solution adhesives and cements
generated contaminated wastewaters (Subcategory C) and solvent solution
adhesives and cements generating non-contract cooling water only (Sub-
category D).
Wastewater Flow Patterns
All the wastewater generated at the San Leandro Plant flowed into the
3.8 m3 sump (see Figure 3). In general, the constant volume Moyno pump had
the necessary capacity to transfer all wastewater to the first settling tank.
When the Moyno pump's capacity was exceeded a gear pump was activated by a
high level control in the sump.
The initiation and duration of each pumps activities were noted by
strip chart recorders during the first five field demonstration tests.
Throughout 56 days (including weekends) no repetitive wastewater discharge
cycles developed. This result, in conjunction into the fact that the batch
production processes are used exclusively, accentuates the high variability
of effluent flow within this industry and identifies the need for flow
equalization in any wastewater treatment system.
Hastewater Composition
A summary of the composition of the San Leandro Plant effluent is shown
in Table 13. Low, high and average values are presented for both daily and
weekly composite samples and overall mean values are given. For all assays
where the detection limit of the analysis was not reached, excepting
mercury, there is at least an order-of-magnitude difference between the high
and low contaminant levels. For example, the effluent total suspended solids
ranges from 59 mg/H to 70,800 mg/H and its. zinc content varies from 2A.mg/H
to 740 mgA• These data clearly point out the high variability of the
San Leandro Plant effluent and re-enforce the requirement for a wastewater
treatment system which is insensitive to shock loading.
48
-------
TABLE 13. SUMMARY OF SAN LEANDRO PLANT EFFLUENT COMPOSITION
10
..,,.. Number of Average Number of Average
y Daily Value Weekly Composite Value
Composite Samples (m
-------
FIELD DEMONSTRATION TESTS
An extensive evaluation of ultrafiltration for adhesives and sealants
wastewater treatment was conducted by performing twelve field demonstration
tests at the San Leandro Plant. During nine of these experiments the total
plant effluent was processed. The remaining three tests were performed with
the following wastewaters: total effluent less CPD stream, total effluent
less "Z" stream, and, electrocoagulation process effluent. Twice during the
tests with the total plant effluent surfactant was continuously added to the
UF feed stream to assess changes in UF membrane performance associated with
increased latex stability.
The results of these field demonstration tests are discussed, in detail,
below.
Ultrafiltration Membrane Flux Performance
The rate of ultrafiltration production per unit membrane area is termed
the membrane flux and is expressed in cubic meters per square meter per day
(m3/m2-day) (gallons per square foot per day (gfd)). The time-averaged flux
for a given concentration experiment is determined by dividing the volume of
ultrafiltrate produced during the run by the elapsed time and the membrane
area. A summary of time-averaged UF membrane flux data is shown in Table 14,
for the entire field program. This table provides an overview of UF flux
performance and will be referred to in the following discussions.
Flux with Total Plant Effluent-
Figure 7 presents the UF permeate flux vs. time plots for two typical
runs with the total plant effluent as the feed stream. The data shown are
for runs #2 and 6. In both instances, and generally throughout the test
program, a very irregular flux pattern is observed. This non-linearity in
the flux vs. time curves is most likely associated with changes in the
waste stream composition. During the night no wastewater is generated by
the plant and thus the contents of the 3.8 m3 (1000 gal) holding tank (see
Figure 3) feeding the UF feed tank are continuously reduced. As normal
plant operations resume in the morning the 3.8 m3 tank is refilled with an
equalized, but not identical, wastewater for processing by the UF system.
During run #2 only type HFM membranes were employed and an average flux
of 1.84,m3/m2-day (44.9 gfd) was recorded over a 208 hr processing period.
In Run #6 both type HFM and HFD membranes were used. The HFM tubular
assemblies averaged 2.09 m3/m2-day (51.0 gfd) during a 172 hr concentration.
The HFD membranes averaged 1.97 m3/m2-day (48.0 gfd), however no flux
readings were recorded after 66 hours operating time.
Overall, throughout 6 individual tests with a total operating time of
1021 hours, the HFM membranes averaged 1.38 in3/m2-day (33.8 gfd) while
processing the total plant effluent. The total solids concentration
achieved by the end of each of these six runs varied from 2.3% to 12.8%
50
-------
TABLE 14; SUMMARY OF ULTRAFILTRATION MEMBRANE FLUX DURING SAN LEANDRO FIELD TESTS
tn
Test 1
1
2
3
4
5
6
7
8
9
10
11
12
Feed Stream Description
Total effluent with surfactant addition
Total effluent
Total effluent
Total effluent
Total effluent
Total effluent
Swift effluent
Total effluent with surfactant addition
Total effluent less CPD stream
Total effluent less "Z" stream
Total effluent, maximum concentration test
Total effluent, maximum concentration test
Average HFM membrane
Total Operating
Time (hrs) Membrane Type
198
208
99
220
152
172
66
17.6
137
141
162
28
170
flux during
HFM
HFM
HFM
HFM
HFM
HFD
HFM
HFD
HFM
HFD
HFM
HFM
HFM (orlg.)
HFM (new)
HFM (oHg.
and new)
Flux'c»V32*c! Fi"t
O . O . t f t\ lO V
m-Vm'-day (gfd)
1.47 (35.8)
1.84 (44.9)
0.94 (23.0)
0.84 (20.6)
1.43 (35.0)
1.60 (39.0)
2.09 (51.0)
1.97 (48.0)
1.99 (48.7)
2.17 (53.0)
1.92 (46.9)
1.71 (41.7)
1.35 (33.0)
1.88 (46.0)
0.43 (10.6)
HFM (oHg.) 1.15 (28.0)
HFM (new) 1.39 (34.0)
3 2
processing of total effluent • 1.38 m /m -day
1 Concentrate
al Sol Ids, %
9.7
4.8
2.3
3.1
7.6
7.6
9.4
6.9
16.4
16.4
0.73
12.8
12.8
(33.8 gfd)
Comments
Test terminated to
conduct dye test.
Surfactant added at
167 hours operating time
HFD flux readings not
recorded after 66 hours
Test aborted due to
latex fouling of UF
membranes
-------
en
ro
4.0
3.0
CO
-a
CO
E
X
3
2.0
-------
(see Table 13), however no correlation was observed between average
membrane flux levels and final concentrate total solids levels. The reason
that no flux vs. final solids concentration relationship developed is that
only during the final hours of some runs did the solids concentration
increase significantly. This short time at a high concentration becomes
relatively insignificant when averaged over a two-week operating period.
Flux with Surfactant Addition--
The addition of surfactant to the plant effluent prior to ultrafiltration
was given attention based on experience gained during operation of the UF
system at the Grace Chicago Plant. The surfactant acts to stabilize both
latex particles and oil emulsions, thereby reducing UF membrane fouling. The
addition of surfactant is a major operating expense at Chicago (seepage 92,
Section 8), therefore, its elimination as a processing steps at San Leandro
would be preferrable.
During two experiments (run f] and 8) surfactant (Rohm and Haas,
Triton-X-100) was continuously added to the UF feed tank during processing of
the total plant effluent. The flux vs time plot for run #8 is presented in
Figure 8. As was the case with the total effluent without surfactant
addition, erratic flux behavior is observed. The average flux for run #8
during 137 hours operation was 1.92 m3/m2-day (46.9 gfd). For run #1 the
flux averaged 1.47 m3/m2-day (35.8 gfd) throughout 198 hours processing.
These data are not significantly different from data obtained without
surfactant addition. Therefore, considering the additional costs associated
with surfactant addition, this procedure is not considered desirable for the
treatment of the San Leandro Plant effluent.
Flux with Segregation of Specific Streams from Total Plant Effluent--
During two of the field demonstration tests a selected process stream
was pumped directly to the sewer while the remaining waste streams entered
the plant sump and were processed by the pilot system per standard test
procedures. The objective of these tests was to determine if either the
CPD waste stream or the "Z" waste stream were "bad actors" in terms of UF
membrane flux or contaminant rejection.
The UF permeate flux vs. time plot during the processing of the plant
effluent, exclusive of the CPD stream, is given in Figure 9. Again, the
flux curve is fairly erratic. The average flux for the entire experiment
(141 hours) was 1.71 m3/m2-day (41.7 gfd). This is somewhat misleading,
however, because of a high initial flux level which was not maintained.
For design purposes an average flux of 1.43 np/m^-day (35 gfd) would be
appropriate since nearly 50% of the time the flux stabilized at this level.
An average flux of 1.43 m3/m2-day (35 gfd) is similar to total plant
effluent test results and does not indicate any benefit in membrane flux
performance from segregating the CPD stream.
53
-------
en
4.0
3.0
to
T3
x 2.0
O)
+j
fO
-------
5.0
en
CJl
Abcor Type HFM Membranes
Circulation Rate -164,m3/day
Inlet Pressure -2.7 atm
Permeate Flux Temperature
Corrected to32.2°C
0 20 40 60 80 100 120 140 160 180 200
Time (hours)
FIGURE 9. UF PERMEATE FLUX VS. TIME FOR TOTAL EFFLUENT LESS CPD STREAM TEST AT SAN LEANDRO.
-------
Prior to the processing of the total plant effluent less the "Z"
dispersion wastewater stream seven new tubular membranes were placed on the
UF system (see details under UF Flux Recovery discussion, below). As
observed in the data plot of Figure 10, the flux for the new membranes
exceeded the original membrane flux throughout the test period. This flux
differential, slightly more pronounced at the lower concentrations, is to be
expected since the surface of the original membranes following 1345 hours of
exposure to the San Leandro Plant wastes was not as clean as the surface of
the new membranes.
For both the original and the new membranes, permeate flux followed the
same pattern. An initial sharp increase in flux was followed by a gradual
flux decline. The average flux rates for the entire 162 hour experiment
were 1.35 m3/m2-day (33 gfd) and 1.88 m3/m2-day (46 gfd) for the original and
new membranes, respectively. Since the original membranes averaged
1.38 m3/m2-day (33.8 gfd) while processing the total plant effluent, no
benefit in membrane flux performance is indicated from segregation of the
"Z" wastewater stream.
Flux During Maximum Concentration--
Due to latex instability, the first experiment designed to assess the
maximum solids concentration achievable by ultrafiltration (run #11) was
aborted. A second attempt at defining the solids concentration achievable
before uneconomical flux levels are encountered (run #12) was interrupted
after 170 hours because of pump failure. Up to this point the concentrate
stream had reached a 12.8% total solids level with average flux values of
1.15 m3/m2-day (28 gfd) and 1.39 m3/m2-day (34,gfd) recorded for the original
and new membranes, respectively. Had pump failure not occurred the UF
concentration could have been continued with the contents of the UF feed
tank processed without fresh feed addition. The maximum concentration
achievable by ultrafiltration was, therefore, not assessed.
Ultrafiltration Membrane Flux Recovery and Cleaning
Overview—
Table 15 presents the flux recovery and accumulated operating time for
the UF membranes on the San Leandro Plant wastewater. One set of seven HFM
membranes remained in use throughout the entire test program (1704,hrs).
These membranes are termed "original" HFM. Following run #9, severe
membrane fouling problems occurred (see below) and seven "new" HFM
membranes were installed. Between runs #5 and 7 type HFD membranes were
operated in series with the HFM membranes.
On the basis of the data presented in Table 15, no irreversible mem-
brane fouling occurred during processing of the adhesives and sealants
manufacturing wastewater. While final water flux data were not available
because of pump failure at the end of the test program, the water flux for
56
-------
en
-•4
IO
-o
X
3
ra
<
low
fr
—^
Time (hours)
T2T
\W
T7T
"ZOO
FIGURE 10. UF PERMEATE FLUX VS. TIME FOR TOTAL EFFLUENT LESS "Z" STREAM TEST AT SAN LEANDRO.
-------
TABLE 15. FLUX RECOVERY AND ACCUMULATED OPERATING TIMES FOR UF MEMBRANES
OPERATED ON THE SAN LEANDRO PLANT EFFLUENT
en
00
Accumulated Water Flux
Membrane Exposure Time, at 32.2°C,
Type hrs m3/m2-day (gfd)
HFM (original) 0
198
406
505
725
— -
877
1049
—
1067
1204,
1345
1507
1534
1704
HFM (new) 0
162
189
359
HFD 0
152
324
342
8.2
8.5
6.5
6.5
7.1
9.7
5.0
6.3
8.3
10.3
—
4.7
6.6
6.2
—
6.3
9.0
6.2
—
—
5.5
—
—
(200)
(208)
(158)
(159)
(173)
(237)
(123)
(153)
(203)
(250)
—
(115)
(161)
(151)
—
(153)
(220)
(151)
—
—
(134)
—
—
Comments
After "spongeball"
1 .5 months later
Water flux not recorded
After solvent cleaning
After aborted run #1 1 .
Detergent washed and "spongeballed"
several times
Final water flux not recorded due to
pump malfunction
Installed after run #9
After aborted run #11.
Detergent washed and "spongeballed"
several times
Final water flux not recorded due to
pump malfunction
New membranes, flux not recorded
Flux not recorded
Membranes removed, flux not recorded
-------
the original HFM membranes was 6.2 m3/m2-day (151 gfd) after 1534,hrs
accumulated operating time. A water flux of this magnitude is quite
acceptable. In most instances, detergent cleaning cycles of only 1/2 hour
duration were able to fully recover the membrane water flux.
Flux Recovery Following Latex Destabilization--
On two separate occasions during the test program severe UF membrane
fouling occurred. The first instance was between runs #9 and 10*, while
the second occurrance was during run #11. In the first case, the UF
permeate flux decreased to 0.62 m3/m2-day (15 gfd) during normal operation.
The system was operated for an additional week with no improvement in flux
observed. A series of detergent wash cycles employing respectively,
"Ultra-Clean", "Dishmate" and trisodium phosphate were followed by hot
water (6QOC) rinsing and spongeball cleaning. The average water flux
increased to only 1.9 m3/m2-day (48 gfd). Inspection of the UF membranes
showed an irregular grey rubber coating had deposited on the membrane surface,
The nature of the coating and a review of the plant production schedule
indicated that the effluent stream from an extremely heat sensitive neo-
prene rubber based product had entered the plant sump prior to the fouling
problem. Unstable latex in this stream had previously coagulated within the
plant sump, however up until this instance a latex skin had not developed on
the membrane surface.
Once the fouling was identified as latexs standard solvent cleaning
procedures for latex removal were employed (see page41). Flux recovery
data following MEK cleaning are given in Table 16. A standard 20 minute
solvent exposure time was used at San Leandro with resultant water flux
measurements ranging from 2.9 to 5.6 m3/m2-day (72-136 gfd). These values
correspond to flux increases of 26%-231%. Initial water flux measurements
on tubular assemblies returned to Abcor were substantially above the 1.6-
2.5 m3/m2-day (40-60 gfd) readings observed in San Leandro. Variations in
MEK soaking times of from 5 to 35 minutes for these tubes produced little
variation in resultant water flux and only modest improvement over initial
values. An additional 20 minute exposure to MEK followed by spongeball
cleaning gave flux levels of 7.6-11.5 m3/m2-day (185-280 gfd). It thus
appears advantageous to expose the membranes to MEK twice. The first
exposure removes the thinner latex deposits and begins to attack the more
heavily coated areas. The second exposure swells the residual latex skin.
An initial exposure of 5-10 minutes appears adequate. No optimization of
the second exposure time was attempted.
The second occurrance of a latex skin formation on the membrane surface
took place during run #11, causing the permeate flux to fall below 0.21
m3/m2-day (5 gfd). It is believed that a few gallons of concentrated
hydrochloric acid were spilled into the effluent stream, lowering the waste-
water pH and thus destabilizing the latex. Detergent cleaning was not
successful in removing the latex skin, however spongeball cleaning was able
to recover the membrane water flux (for both the new and original membranes)
to 6.2 m3/m2-day (151 gfd). Solvent cleaning was not required.
No run number was assigned to this aborted test.
59
-------
TABLE 16. UF MEMBRANE FLUX RECOVERY DATA SUBSEQUENT TO LATEX FOULING*
Tubular
Assembly
Number
1
2
3
4
5
6
7
Mean Value
for Tubes
1-7
8
9
10
11
12
13
14,
Mean Value
for Tubes
8-14
Initial
Flux
(m3/m2-day)
1.93
1.68
1.76
2.34
1.97
2.09
1.93
1.97
3.12
3.97
3.12
2.55
4.67
4.39
7.18
4.14
Time Soaked
in MEK (min)
20
20
20
20
20
20
20
20
5
10
15
20
25
30
35
—
Resultant
Flux
(m3/nr-day)
4.80
5.58
4.80
2.95
4.71
3.20
5.58
4.51
4,26
4.96
3.97
4.26
3.97
5.66
5.66
4.67
Additional
Time in
MEK (min)
_
-
-
-
-
-
-
20
20
20
20
20
20
20
-
Resultant
Flux
_
-
-
-
-
-
-
-
7.63
8.90
7.91
7.18
7.91
9.35
11.5
8.61
Overall
148
231
172
26
139
53
189
137
144
124
153
181
69
112
60
120
NOTES: 1) Tubes 1 to 7 were cleaned at San Leandro.
2) Tubes 8 to 14 were cleaned at Wai den.
3) Tubes 1 to 7 were cleaned with a test tube brush soaked
in MEK before spongeball cleaning
-------
Summary—
Throughout most of the field demonstration program routine detergent
cleaning was satisfactory to recovery membrane water flux to acceptable
levels. It is evident, however, that periodically severe membrane fouling
due to latex destabilization can occur during processing of adhesives and
sealants manufacturing wastewaters. No fail-safe solution to the problem of
latex destabilization is available.
Prior to design of a full-scale system for a particular site it should
be determined by plant personnel (if possible) whether a latex fouling
problem may be a reoccurring circumstance or only the result of an
accidental spill. If latex stabilization is envisioned to occur several
times a year one or more of the following options should be considered if
full-scale UF treatment is to be employed:
1. Divert the effluent from a particular product's
wash cycle and treat it separately.
2. Stabilize the latex in the wastewater with
surfactant prior to UF treatment.
3. Stabilize the latex in the wastewater by main-
taining the effluent pH above 8.
4. Provide a solvent soak tank so that fouled membranes
can be removed from the UF system, soaked in MEK,
returned to the system and then spongeballed in place.
5. Construct the UF system with carbon steel membrane
shells and piping, and explosion proof pumps and
controls to allow solvent cleaning in situ.
Product Hater Quality
Introduction--
A summary of the plant effluent composition throughout the field
demonstration program was given in Table 13. Samples were also collected
of the feed after settling, the ultrafiltration concentrate and the ultra-
filtration permeate. The analytical data sets for these sampling stations
are summarized, respectively, in Tables 17, 18, and 19. Because of the
unique nature of the electrocoagulation test (Run #7) analytical data from
this experiment were not included in Tables 17-19. The analytical data
from individual tests are given in Appendix A.
Some analytical chemistry problems developed during the course of this
program. In particular, difficulties were encountered in the assays for
total cyanide, phenolic compounds and total freon extractives. In many
instances, the free cyanide levels in the wastewater samples were reported
to be in excess of the total cyanide levels. Also, the level of total
cyanides were often higher after treatment than in the raw waste. To
61
-------
no
TABLE 17. SUMMARY OF ANALYTICAL DATA FOR ULTRAFILTRATION FEED AFTER
SETTLING AT SAN LEANDRQ PLANT THROUGH TEST #12
Assays
Total Freon Extractive *
Non-Polar Freon Extractlble
Total Solids
Suspended Solids
BOD
Soluble BOD
COD
Soluble COD
Arsenic
Cadmium
Total Chromium
Free Cyanide
Total Cyanide *
Lead
Mercury
Phenolic Compounds *
Z1nc
Number of
Dally Samples
12
6
10
16
6
6
6
6
-
-
-
-
2
-
-
12
12
Average
Value
mg/1
544
117
7580
2200
7730
6870
20400
16200
-
-
-
-
4
-
-
81
38.7
Number of
Weekly
Composite
Samples
7
2
6
15
7
2
7
2
-
-
-
-
1
-
-
7
7
Average
Value
mg/1
486
133
8120
2320
5750
7710
29600
16500
-
-
-
-
5.5
,
-
89.3
66
Low
Value
mg/1
50
11
4320
242
1880
2220
13100
11000
-
-
-
-
3
-
-
1
1.1
High
Value
mg/1
1230
248
14300
4820
13000
13200
77300
26800
-
-
-
-
5.5
-
-
295
200
Overall
Average
Value
mq/1
522
121
7780
2260
6670
7080
25300
16300
-
-
-
-
4.5
-
-
84.0
48.7
Interference 1n assay suspected. See page 61.
-------
CO
TABLE T8. SUMMARY OF ANALYTICAL DATA FOR ULTRAFILTRATION CONCENTRATE
AT SAN LEANDRO PLANT THROUGH TEST #12
Assays
Total Freon Extract! ble*
Non-Polar Freon Extractlble
Total Solids
Suspended Sol Ids
BOO
Soluble BOD
COO
Soluble COD
Arsenic
Cadmium
Total Chromium
Free Cyanide
Total Cyanide*
Lead
Mercury
Phenolic Compounds*
Zinc
Number of
Dally Samples
29
23
10
31
7
6
6
6
1
1
1
1
2
1
-
13
13
Average
Value
mg/1
5130
1280
68600
34000
11500
10000
130000
93700
<0.2
<0.2
<0.5
2.6
4.0
<1
-
54.7
924
Number of
Weekly
Composites
Samples
16
10
6
17
16
2
7
2
9
6
6
8
9
6
6
16
15
Average
Value
mg/1
4720
1310
55300
27500
13600
11700
143000
111000
<0.22
<0,2
<0.5
4.1
3.58
<1.23
.0045
66
427
Low
Value
mg/1
100
28
4440
680
3600
3680
17100
14800
<0.2
<0.2
<0.5
0.53
0.1
<1
0.002
0.25
24
High
Value
mg/1
22700
7600
160000
140000
32000
17200
340000
183000
<0.35
<0.2
<0.5
11
n
<2.4
0.007
600
2800
Overall
Average
Value
mg/1
4980
1291
63600
31700
13000
10400
137000
98100
<0.21
<0.2
<0.5
3.94
3.66
<1.2
.0045
60.3
658
Interference 1n assay suspected. See page 61.
-------
TABLE 19. SUMMARY OF HFM PERMEATE QUALITY
Number of Average Number of Average
Assay Daily Value Weekly Composite Value
Composite Samples (mg/i) Samples (mg/i)
Total Freon Extractives'}"
Non-Polar Freon Extractives
Total Solids
Suspended Solids
BOD
Soluble BOD
COO
Soluble COD
Arsenic
Cadmium
Total Chromium
Free Cyanide
Total Cyanidef
Lead
Mercury
Phenolic Compounds "f
Zinc
33
23
14
35
7
6
6
6
1
1
1
1
2
1
—
17
17
130
9
4,340
51.4
5,390
6,640
16,500
15,900
< 0.2
< 0.2
< 0.5
0.26
4.26
< 1
—
60.9
73.7
19
10
10
21
20
2
7
2
9
6
6
8
9
6
6
20
19
218
7
5.020
58
7.660
7.715
27 ,000
16,700
< 0.2
< 0.2
< 0.5
0.41
5.18
< 1
0.0017
52
9.0
Low
Value
(mg/t)
9
2
1.460
5
590
3,990
6.570
10.500
< 0.2
< 0.2
< 0.5
0.26
0.5
1
i,
0.001
0.5
0.34
Overall
High Average
Value Value
(mg/d) {mg/J.5
953
43
9,700
192
*
19.000
10,400
53.200
22,800
< 0.2
< 0.2
< 0.5
0.65*
12*
< 1
*
0.003
388
1,100
162
8.4
4,620
53.9
7,070
6,910
22.200
16,100
< 0.2
< 0.2
< 0.5
0.39
5.01
< 1
0.0017
56.1
39.5
* Indicates low or high value observed in weekly composite sample.
t Interference in assay suspected. See page 61.
-------
investigate these inconsistencies, selected waste samples were "spiked"
with known amounts of cyanide and analyzed. For feed and concentrate
samples it was found that when 500 ml of sample were spiked with cyanide only
50% of the added cyanide was picked up by the assay. When a 50 ml sample was
used, cyanide recovery was 80%. It is known that grease, upon sample acidi-
fication, becomes a fatty acid and distills over along with the cyanide.
By employing a smaller sample volume, less interferring species distill over
and a more accurate cyanide reading is obtained. Permeate samples spiked
after the distillation step showed nearly complete cyanide recovery,
however a yellow organic compound was present in the distillate. No
further investigation of the total cyanide assay was performed.
Interferences from surfactants present in the waste were suspected in
the colorimetric assay for phenolic compounds. To test for interferences
permeate samples were spiked with Triton-X-100. The results of the phenols
assays are shown below:
HFD Permeate Sample Phenolic Compounds (mg/&)
As Received 203
With 211 ppm Surfactant Added 220
With 787 ppm Surfactant Added 214
No interference from surfactants is indicated from these data. Thus, in-
consistencies observed in the phenolic compound assays are unexplained.
Two problems arose in the assays for total freon extractives. First,
all of the total freon extract!ble assays on Runs fl, 2 and 3 samples and
some of the assays on Runs #4,and 5 samples were performed with a container
of freon which was later found to be contaminated. Therefore, average
values for this assay, and many individual data points represent the high-
est possible freon extractibles content rather than precise values.
The second difficulty with the total freon extractibles analysis was
that surfactant in the permeate was partially extracted by the freon. This
resulted in observed oil and grease levels higher than actually present in
the product water. This phenomenon was verified by analyzing two samples
of D.I. water, one spiked with Triton-X-100 and one control sample for
total freon extractibles. The results were as follows:
Concentration of Total Freon
Triton-X-100 in Extractibles
Sample No. Sample (nig/A) (nig/A)
1 0 < 5
2 1,190 296
Recovery of Triton-X-100 by the freon was -25% indicating a positive
interference exists.
65
-------
Effect of Settling on Plant Effluent Composition--
The scheme of employing a settling/flotation tank ahead of the ultra-
filtration system was adapted from the Grace Chicago Plant's treatment
system. By reducing the suspended solids loading on the UF system, higher
flux levels are achieved and less membrane area is required for a full-scale
treatment system. In addition, to suspended solids reduction, reduced
loadings were observed for several other contaminants. These data are
summarized in Table 20.
TABLE 20. EFFECT OF SETTLING PRETREATMENT ON
SAN LEANDRO PLANT EFFLUENT
. ' " — •• •' - ' — "
Assay
Total Freon Extractives*
Non-Polar Freon Extractives
Total Solids
Suspended Solids
BOD
Soluble BOD
COD
Soluble COD
Total Cyanide*
Phenolic Compounds *
Zinc
Overall
Plant
Effluent
2,220
296
14,100
10,600
8,740
6,760
27,100
17,300
1.88
154,
98.9
Average Value
Effluent
after
Settling
522
121
7,780
2,260
6,670
7,080
25,300
16,300
4,5
84,
48.7
(mg/fc)
Removal
Efficiency, %
76.5
59.1
44.8
78.7
23.7
6.6
5.8
45.5
50.8
* Interference in assay suspected. See page 61.
The most significant reductions (75% - 80%) occurred in the effluent
suspended solids and total freon extractive loadings. However, the average
loadings of these contaminants; 2,260 mg/£ and 522 mg/£, respectively, are
still significant and exceed discharge standards. In fact, in the final
settling tank overflow, none of the contaminants assayed met the San Leandro
Municipal Sewer Discharge Standards.
Ultrafiltration Membrane Removal Efficiency—
The ultra-filtration membrane removal efficiency data discussed below
was obtained with type HFM membranes unless otherwise stated. The membrane
removal efficiency data includes the effect of the settling/flotation tanks
66
-------
since it is based on the plant effluent as the feed stream. The analytical
data summary for the HFM permeate during all field demonstration tests was
given in Table 19.
Total Plant Effluent—Average UF membrane removal efficiencies during
tests with the total plant effluent (runs #2, 3, 4, 5, 6 and 12) are
summarized in Table 21.
TABLE 21. AVERAGE UF MEMBRANE REMOVAL EFFICIENCY DATA
DURING PROCESSING OF THE TOTAL PLANT EFFLUENT
Assay
Total Freon
Extract! bles*
Non-Polar Freon
Extractibles
Total Suspended
Solids
BOD5
COD
Free Cyanide
Phenolic
Compounds #
Zinc
Average
Concentration
in Plant Effluent
After Settling
(mg/£)
3,250
415
13,400
11,300
56,100
< 2.6f
244,
104,
Average
Concentration
in UF
Permeate
(mg/£)
100.0
< 7.4**
< 27.0**
8,890
36,600
0.43
44.6
1.5ft
Typical
Range
of
Removal
Efficiency
(%)
65- 99.7
90 -> 99.9
98.7-> 99.9
5- 50
25- 50
32- > 85
0- 70
60- 99
Average
Removal
Efficiency##
(X)
92.2
94,7
99.6
24.0
38.2
< 77.5
44,0
90.8
* Interference in analysis suspected. Maximum possible oil and grease re-
ported. See page 61.
** Most readings were <5 mg/fc.
t Samples diluted 1:10 to minimize interference; detection limit increased
to 2.6 mg/£.
tt Excludes the 1 reading.put of 11 which was ,>5.4,.mg/£.
# Interference in analysis suspected. See page 61.
## Includes effect of settling. Absolute UF removal efficiency may be lower,
The UF system exhibited excellent removal for total freon extractives.
The average oil and grease level in the permeate was 100 mg/£; the average
removal efficiency was 92.2%. In actuality, improvements in these figures
are to be expected since the difficulties previously mentioned with this
assay often gave false positive readings. The non-polar freon extractives
assay is performed by passing the extract from the total freon extractive
67
-------
assay through a silica gel column. The silica gel adsorbs surfactants and
animal and vegetable fats.* Thus, this assay in effect represents the oils
and grease of mineral and petroleum origins. The removal efficiency for
non-polar freon extractives averaged 94.7% with a mean UF permeate concen-
tration of <7.4,mg/£.
The suspended solids level in the permeate averaged <27.0 mg/£. While
most readings were below the detection limit of the assay (5 mg/&) three
readings of -100 mg/£ were recorded and one reading of 160 mg/£ was noted.
It is believed that secondary precipitation resulting from permeate insta-
bility occurred. The mechanism by which the secondary precipitation took
place was not investigated during this program. It was determined, however,
that the higher readings of suspended solids were a result of this chemical
reaction and not a function of membrane degradation or failure. As stated,
in most instances, the UF system reduced the suspended solids loading in
the plant effluent to <5 mg/£ .
BOD and COD reductions in the total plant effluent averaged 24% and
38%, respectively. These data indicate that a significant portion of the
organic pollutant loading in this wastewater is present as soluble matter.
An exact measure of the reduction in free cyanide by the treatment
process was not obtainable because the feed samples had to be diluted 1:10
to minimize interferences. This dilution increased the assay detection to
2.6 mg/£ which was above the actual feed free cyanide level. Interferences
with the total cyanide assays were discussed previously.
Phenolic compound removal was moderate, ranging from 0% to 70% and
averaging 44%. The average phenolic compound concentration in the UF
permeate was 44.6 mg/fc representing a range of values from 0.62 to 220 mg/£.
The detection limit of the analysis was reached for the feed, concen-
trate and permeate samples for cadmium, total chromium and lead. The same
is true for all arsenic assays except one feed and one concentrate sample.
Therefore, no removal efficiency data could be calculated for these
assays.
The degree of mercury of removal was essentially nil, however this is
of little practical significance since the level of mercury in the plant
effluent averaged only 0.0028 mg/£. Zinc removal of 91% was averaged
throughout the tests with the total plant effluent. The overall average
permeate loading was 1.5 mg/£ when a single, extreme reading (47 mg/£) was
eliminated from the data analysis. It is believed that this high reading
was the result of poor "housekeeping" measures in the crystal flux area,
where zinc ammonium chloride is present, and that it was therefore not a
representative data point.
*Based on unpublished information from EPA, Cincinnati; animal and
vegetable fats are adsorbed up to the capacity of the silica gel-about
500 mg oil and grease/15 grams silica gel. Also, it is expected that
< 7% of non-polar oil and grease might be adsorbed.
68
-------
Effect of Surfactant Addition On UF Permeate Quailty--The pertinent
analytical data from the two surfactant addition tests are summarized in
Table 22. Removal efficiency data are not presented because several incon-
sistencies in the data make these numbers suspect. For instance, the total
freon extractives loading in the feed during runs #1 and 8 was only 15%
of the average loading in the feed during the other runs with the total plant
effluent. Also, in many instances (see Table A8» Appendix A) contaminant
loadings in the treated wastewater were higher than in the feed.
TABLE 22. AVERAGE UF PERMEATE QUALITY DURING PROCESSING OF
THE TOTAL PLANT EFFLUENT WITH SURFACTANT ADDITION
Average Average
Concentration Concentration
in Plant in UF Permeate
Assay
Total Freon Extract! bl es *
Non-Polar Freon Extract! bles
Total Suspended Solids
BO 05
COD
Phenol ic Compounds *
Zinc
Effluent (mg/A)
478
117
4,230
8,700
23,000
148
116
(mg/£)
184,
< 7.8
61.3
8,570
16,900
102
9.3**
* Interference in assay suspected. See page 61.
** Excludes one reading of 1,100 mg/£.
The permeate total freon extractives were higher than observed for the
runs without surfactant addition (1841mg/Jl vs. 100 mg/£). As mentioned
previously these higher readings reflect the extraction of surfactant in
the permeate by the freon. The added surfactant neither prevented, nor
reduced, secondary precipitation in the UF permeate. In fact, the suspended
solids loading in the permeate during the surfactant addition tests was
greater than twice the loading when no latex stabilization was attempted.
Surfactant addition did not result in improved permeate quality in
terms of phenolic compounds or.zinc. BOD and COD levels were lower_in both
feed and permeate streams during the surfactant addition tests, again
indicating no improvement in treated effluent quality.
In summary, surfactant addition lessens, rather than improves,
permeate quality.
69
-------
Effect of CPD Stream Segregation on Permeate Quality—Analytical data are
presented in Table 23 for the field demonstration experiment performed with
the exclusion of the CPD stream. UF membrane removal efficiencies for most
assays were similar to the data obtained with the total plant effluent. In
terms of actual permeate quality, the oil and grease, suspended solids and
zinc loadings were lower in the total plant effluent than in the effluent
less the CPD stream.
TABLE 23. AVERAGE UF PERMEATE QUALITY DURING PROCESSING OF THE
TOTAL PLANT EFFLUENT LESS THE CPD STREAM
,
Assay
Total Freon Extractives
Non-Polar Freon Extract! bles
Total Suspended Solids
BODq
COD
Phenolic Compounds *
Zinc
Average
Concentration
in Plant
Effluent (mg/£)
2,490
288
16,200
5,330
20,800
5.7
117
Average
Concentration
in UF
Permeate (mg/&)
178
22.8
18.4
4,640
11,630
2.6
3.1
Average
Removal
Efficiency
(X)
t
61.2
98.6
15.6
40.6
42.0
96.2
fData are highly inconsistent, see Table A9.
interference in assay suspected. See page 61.
The only significant improvement in water quality from segregation of
the CPD stream resulted from a 2 order-of-magnitude decrease in the feed
phenolic compound loading. This decrease alone is not deemed sufficient to
warrant segregation of the CPD stream during full-scale treatment operations,
Effect of "Z" Stream Segregation on Permeate Quality—Analytical data
for the "Z" stream segregation test are shown below (Table 24).
TABLE 24. AVERAGE UF PERMEATE QUALITY DURING PROCESSING OF
THE TOTAL PLANT EFFLUENT LESS THE "Z" STREAM
Total
Total
BOD5
COD
Phenol
Zinc
Assay
Freon Extractives
Suspended Solids
ic Compounds *
Average
Concentration
in Plant
Effluent (mg/£)
1,020
6,940
1,620
19,800
82.9
51.5
Average
Concentration
in UF
Permeate (mg/£)
192
155
2,000
10,000
78.5
16.2
Average
Removal
Efficiency
(*)
79.4,
95.2
39.3
33.5
68.3
Interference in assay suspected. See page 61.
70
-------
Assays performed on samples collected from the plant sump show considerable
variation from the average total plant effluent values (see Table 21).
Major deviations from the total effluent analytical data are observed in the
lower loadings for suspended solids, BOD and total freon extractives. The
average UF removal efficiencies are lower with the "Z" stream segregated for
all assays except COD. Permeate quality is significantly worsened by
segregating the "Z" stream in terms of oil and grease, suspended solids,
phenolic compounds and,zinc. Permeate quality is improved in terms of BOD
and COD, resulting from reduced organics loadings in the feed stream.
Overall, the permeate quality is judged to degrade when the "Z" stream is
segregated.
Comparison of Effluent Quality with Local and Preliminary Federal
Discharge Standards--
Table 25 summarizes type HFM membrane permeate quality during processing
of the total plant effluent, San Leandro and MSD of Chicago Municipal
Discharge Standards and the recommended BPCTCA, BADCT and BATEA standards.
For oil and grease, free cyanide and zinc both the San Leandro and MSD dis-
charge limitations are met. The plant effluent, without treatment already
met the standards or reached the assay detection limit for arsenic, cadmium,
total chromium, lead and pH. The plant effluent also met the San Leandro
discharge limit for mercury. The metropolitan Chicago mercury limit of
0.05 mg/£ was not reached even after UB treatment.
The total cyanide content of the UF permeate (~5 mg/A) was acceptable
for discharge in the Chicago area but not San Leandro. The phenolic compounds
level in the permeate exceeded San Leandro limits while no discharge
limitation was published for Chicago.
For total solids, suspended solids, BOD and COD no municipal discharge
limitations are set, however, surcharges are imposed on the basis of BOD
and suspended solids loadings. The UF permeate suspended solids level is
quite low, however, the ~9,OQO mg/£ BOD level could incur a several thousand
dollar sewer surcharge.
The EPA Draft Development Document is written in terms of BOD, COD
$nd suspended solids loadings. While the UF permeate is essentially equal
to the draft BPCTCA, BADCT and BATEA discharge limits for suspended solids,
it is greatly in excess of the preliminary contractor recommendations for
BOD and COD.
Summarizing, ultrafiltration of the San Leandro Plant effluent cannot
produce a product water meeting all local and preliminary Federal discharge
standards. Some form of post-treatment will be required to make this
adhesives and sealants wastewater compatible with the discharge regulations.
71
-------
IN)
TABLE 25. COMPARISON OF HFM PERMEATE QUALITY WITH LOCAL AND FEDERAL
DISCHARGE STANDARDS DURING TREATMENT OF THE TOTAL PLANT EFFLUENT
Parameter
Total Freon Extractibles (mg/JlStt
Non-Polar Freon Extractives (mg/i)
Total Solids (mg/fc)
Total Suspended Solids (mg/t)
BOO (mg/i)
COD {mg/Jlj
Arsenic (mg/fc)
Cadmium (mg/£)
Total Chromium (mg/J.)
Free Cyanide (mg/ 6.0
Metropolitan
Chicago Sewer
Discharge Standards
<100
—
—
—
—
—
—
2.0
25.0
2.0
10.0
0.5
0.0005
—
15.0
4.5-10.0
Range for Subcategorles A, B and C
BPCTCA BADCT BATEA
—
-
„
20-50 20 20
255-910 255-910 100-910
710-1,530 680-1,500 267-1,500
--
—
..
—
._
..
—
—
—
**"* ~~ ~~
*Most readings were< 5 mg/fc.
tOetectlon limit of assay.
**300 mg/l oil and grease of animal or vegetable origin, 100 mg/t oil and grease of mineral or petroleum origin.
ttlnterferences suspected in some assays. See page 61.
-------
POST TREATMENT TESTS
Reverse Osmosis
Reverse osmosis experimentation was conducted in Wai den's Pilot Labor-
atory on two different ultrafiltrate samples obtained from the Grace
Chicago Plant. A DuPont B-9 (polyamide) hollow-fine-fiber permeator was used
throughout both tests.
RO Module Product!vity--
The B-9 permeate flow rate as a function of volumetric feed concen-
tration is plotted in Figure 11 for both pilot plant runs. The productivity
of the B-9 module decreases with increasing volumetric concentration in both
cases. This trend is typical of reverse osmosis operation since dissolved
solids build-up within the feed solution raises the feed osmotic pressure
and reduces the overall driving force across the membrane.
The average module productivity (m3 processed T processing time) in the
first test was 10.4,m3/day (1.91 gpm) during concentration to 13.7X (92.7%
conversion). Processing beyond this point was precluded due to the dead
volume holdup of the test system. For the RO experiment with the second
ultrafiltrate sample the initial module productivity of 8.2 m3/day (1.5 gpm)
was 25% lower than the IX productivity achieved during the first test. The
reduced permeate flow rate throughout the run is partially attributed to
the higher conductivity of the ultrafiltrate; 3000 ymhos/cm vs. 980 ymhos/cm
for the first sample.
Another factor influencing the lower module productivity was the
presence of suspended matter in the second ultrafiltrate sample. Since
ultrafiltrate is typically free of all suspended solids, the particulate
matter in the sample was believed to be a biological (bacteria) related floe,
which may have developed during a delay in sample shipment. The suspended
matter was allowed to settle overnight and the RO feed solution was passed
(as always) through 5y and ly string-wound cartridge filters, in series,
before processing. The average module productivity during this test was
4,85 m3/day (0.89 gpm) during concentration to 7.9X (87% conversion).
The ultrafiltrate sample used in the second experiment was aged longer
than the first sample and is thus believed to have degraded. Therefore, the
flux vs. volumetric feed concentration data developed in the first experiment
are considered more valid for design purposes.
RO Module Removal Efficiency--
Assays performed on the initial feed, final mixed composite permeate
and final concentrate are detailed in Table 26 for both RO tests. Significant
differences in the RO feed contaminant levels are observed for the two tests.
(It is also of note that the contaminant concentrations in the San Leandro
ultrafiltrate (see Table 19) were much higher than those observed in either
of the Chicago ultrafiltrate samples.)
73
-------
25
CO
£
•a
o
s-
CL.
20
15
10
IX
O - First Sample
D - Second Sample
Feed Temperature - 17 to 27°C
Inlet Pressure - 24,3 atm
First Sample
Second Sample
\r
2X 4X 10X
Feed Volumetric Concentration
20X
FIGURE 11. PRODUCTIVITY VS. FEED VOLUMETRIC CONCENTRATION FOR DUPONT B-9 PROCESSING OF
DEWEY AND ALMY CHICAGO PLANT ULTRAFILTRATE
-------
TABLE 26. ANALYTICAL DATA AND RO MODULE REMOVAL EFFICIENCIES DURING
PROCESSING OF THE GRACE CHICAGO PLANT ULTRAFILTRATE
en
Test
Number
1
2
Assay
Total Solids
COD
BOD
Total Solids
COD
BOD
Initial
Feed
(mg/i, )
1,270
1,890
450
6,840
12,200
2,100
Final
Concentrate
(mg/A )
22,500
27,000
31, 800
82,800
110,500
18,800
Final
Composite
Permeate
(mg/l )
48
282
58
261
1,190
800
Removal
Efficiency
(%)
96.2
90.3
87.1
96.2
90.2
61.9
-------
The DuPont B-9 permeator removal efficiencies for total solids and COD
were essentially identical in both tests and were 96% and 90%, respectively.
For BOD, which is a much less precise analysis, the removal efficiency de-
clined from 87% in the first test to 62% in the second. Assuming the same
RO removal efficiencies for total solids and COD and a 75% removal
efficiency for BOD, RO treatment of the San Leandro UF permeate (to a 10X
concentration factor) would result in a product water with approximately
150 mg/fc total solids, 3,700 mg/£ COD and 2,200 mg/Jl BOD. Since all of the
total solids would be dissolved, the RO product water would satisfy the con-
tractor's recommended values for BPTCA, BADCT and BATEA for suspended solids.
It would not, however, achieve the values recommended for either BOD or COD.
The RO product water from the pilot plant tests was not assayed for
total cyanide or phenolic compounds. Reverse osmosis is capable of
concentrating either cyanide or phenols in the UF permeate when processing
takes place under basic (pH > 8) conditions. It is assumed that a properly
designed RO system could consistently produce a product water meeting the
San Leandro Municipal Discharge Limitations of 1.0 mg/& total cyanide and
1.0 mg/£ phenolic compounds.
RO Module Standard Salt Rejection Tests--
The B-9 salt rejection data for both batch pumpdowns are presented in
Table 27. A decline in module salt rejection to only 50% was noted after
the second processing period. The B-9 permeator was cleaned per DuPont's
recommendation with a solution of citric acid and then treated with PTA,
a proprietory DuPont product. A slight increase in NaCl rejection, to
61.9%, resulted from the cleaning/treatment operation. The decline in
module performance is believed to be related to the suspended matter in the
feed stream, however the exact mechanism of the module failure is not known.
Carbon Adsorption
The equilibrium adsorption isotherms for Filtrasorb 400 granular
activated carbon (Calgon Corp.) at 20°C for the second sample of Grace
Chicago plant ultrafiltrate are shown in Figures 12 and 13 for BOD and
COD removal, respectively. In each figure the logarithm of the contaminant
loading (mg adsorbed per gram of carbon) is plotted against the logarithm
of the equilibrium contaminant concentration. The points fall reasonably
close to a straight line in both cases, indicating agreement with the
Freundlich isotherm expression.
The adsorption isotherm for BOD removal shown in Figure 12 indicates
an equilibrium loading of 2g BOD/g carbon at the initial ul trafiltrate
BOD level of 1,325 mg/£. The adsorption capacity of the carbon decreases
rapidly from this point. At a BOD concentration of 1,000 mg/Jl the carbon
loading reduces to 0.1 g BOD/g carbon.
An even greater decline in adsorptive capacity is observed in the
isotherm for COD removal shown in Figure 13. The initial COD concentration
of the untreated ultrafiltrate, C0, was 12,460 mg/& and the equilibrium
76
-------
TABLE 27. STANDARD SALT REJECTION TEST DATA FOR DUPONT B-9 HOLLOW-FIBER MODULE DURING
PROCESSING OF DEWEY AND ALMY CHICAGO PLANT ULTRAFILTRATE
Cumulative
Operating
Time (hrs)
0
3.6
7.6
...
Feed
Concentration
NaCI (ppm)
1800
5750
5900
5250
Temperature
(°0
25
27
28
22
Permeate
Flowrate
(m3/day)
13.9
11.1
13.9
11.4.
Feed
Flowrate
(m3/day)
28.0
29.0
28.3
30.5
Conversion
(%)
49.7
38.1
49.0
37.5
Rejection
(%)
96.1
95.5
50.0
61.9
Intrinsic
Rejection Remarks
.972
.964. After test
n
After test
#2
After citric
acid cleaning
and PTA treat-
ment
-------
5000
1000 —
500-
o
US
o
Q
O
CO
O)
= ioo_
a
o
CO
50-
10
1 1 1 1 1 I 1 1
x/m at C - 2 g BOD/g Carbon
—
—
1 1 1
;
I
0
ii _
/ '
9 i 1
: ii :
-
i
' 0 •
1
1
: I: :
/ 1
' 1
—
! 1 1 I 1 | 1 1
1
1
1
1 C = 1325 mg/1
1
1 I 11
100
500 1000
BOD Concentration (mg/1)
5000
FIGURE 12.
EQUILIBRIUM ADSORPTION ISOTHERM AT 20°C FOR BOD REMOVAL FROM
DEWEY AND ALMY CHICAGO PLANT ULTRAFILTRATE (SECOND SAMPLE).
78
-------
50,000
10,000
5000
s
e>
o
o
en
c.
•5
§
o
o
o
1000
500
100
x/m at CQ = 11.5 g COD/g Carbon Q
1000
o
ol
o
CQ = 12,400 r,ig/l
I I L
5000 10,000
COD Concentration (mg/1)
50,000
FIGURE 13.
EQUILIBRIUM ADSORPTION ISOTHERM AT 20°C FOR COD REMOVAL FROM
DEWEY AND ALMY CHICAGO PLANT ULTRAFILTRATE (SECOND SAMPLE).
79
-------
loading at this concentration as determined from the isotherm, was 11.5g
COD/g carbon. A 20% reduction in the ultrafiltrate COD concentration to
10,000 mg/& would result in nearly a 100-fold loss in adsorptive capacity
to 0.12 mg COD/g carbon.
The steep slope of the isotherms of Figures 12 and 13 indicates that the
Chicago plant ultrafiltrate is composed of a small amount of strongly
adsorbed material and a larger amount of weakly adsorbed material. From
isotherms of this nature, we can predict rapid breakthrough of BOD and COD
to occur during processing of this ultrafiltrate through carbon columns.
Similar results were obtained for a TOC isotherm performed with the
first Chicago ultrafiltrate sample.
Ozonation
A 20 liter (5 gal) sample of the San Leandro ultrafiltrate was ozonated
by U.S. Ozonair Corporation (South San Francisco, California). The
analytical data from ozonation tests with a single catalyst and with 3
catalysts plus electro-coagulation are given in Table 28. For total solids,
suspended solids and total cyanide the loading in the wastewater increases
following ozonation. The increases observed in the solids assays are
believed to be a result of conversion of volatile solids into non-volatile
solids resulting from catalysis and electro-coagulation (12). The increase
in total cyanide is likely the effect of positive interferences in the assay.
BOD and COD reduction for the two tests averaged approximately 9% and
80%, respectively. While the COD reduction was significant the ozonated
product water COD content (12,000 to 16,000 mg/H) was still an order-of-
magnitude above the value recommended in the draft development document.
Clearly, the BOD level was above the recommended value.
The UF permeate total freon extractives and, zinc content were below the
San Leandro Municipal discharge standards before ozonation and therefore the
96% removal efficiencies observed for these contaminants are not of primary
importance- On the other hand, the reduction of phenolic compounds from
47 mg/£, to <0.15 mg/£, representing a >99% reduction, is highly significant.
Thus, with ozonation post-treatment an effluent meeting all San Leandro
Municipal Discharge Limitations has been demonstrated in a laboratory test to
be readily achievable.
Alternative Oxidation Processes
Three alternative oxidation processes; chlorination, hydrogen peroxide
oxidation and potassium permanganate oxidation were considered for post-
treatment of the UF permeate to effect further cyanide and phenolic
compound reductions. Chlorination was ruled out as viable alternative
since chlorinated phenols are toxic and may import a disagreeable odor to
the water. Hydrogen peroxide treatment was judged unsatisfactory because
it is most attractive economically when the range of cyanide in the
80
-------
TABLE 28. ANALYTICAL DATA FOR OZONATION OF SAN LEANDRO PLANT ULTRAFILTRATE
CO
Ozonation with 1 Catalyst
Assay
Total Solids (mg/i)
Total Suspended Solids (mg/£)
Total Freon Extractives (mg/i)
BOD (mg/H)
COD (mg/i)
Total Cyanide (mg/S.)
Phenotic Compounds (mg/Z.)
Zinc (mg/J.)
pH (units)
UF Permeate
7,200
64
130
5.780
76,700
0.56
47
2.2
8.4
Level
14,500
140
4.0
5,190
12.100
1.5
0.13
0.09
9.4
Removal Efficiency,.- X
_^_
—
96.9
10.2
84.2
—
99.7
95.9
Ozonation with 3 Catalysts plus
El ectrocoagul ati on
Level
33,100
290
4.0
5,340
16,700
1.1
0.10
0.08
9.8
Removal Efficiency, %
96.9
7.6
78.2
99.8
96.4
-------
wastewater TOO to 1000 mg/£ (as sodium cyanide). For San Leandro the
range is 1 to 10 mg/£. The dosage of potassium permanganate required to
oxidize the phenolic compounds in the San Leandro ultrafiltrate was
excessive, therefore KMnO, oxidation was also deemed economically prohibitive.
Post-Treatment Summary
Six post-treatment processes were studied in varying degrees. None of
the processes were deemed capable of lowering ultrafiltrate BOD and COD
loadings to Federal discharge standards. All six processes are able,
however, to produce an effluent meeting local municipal discharge regulations.
Carbon adsorption, hydrogen peroxide oxidation and potassium permanganate
oxidation were judged to be uneconomical for polishing of the San Leandro
UF permeate and are therefore eliminated from further consideration.
Chlorination was eliminated from consideration on technical grounds since it
produces chlorinated phenols.
Only reverse osmosis and ozonation are thought to be both technically
and economically viable. The economics of each will be discussed in a
subsequent section.
DEWATERING TESTS
Introduction
Promulgation of future "zero discharge" regulations may prevent
disposal of the concentrate from adhesives and sealants wastewater treatment
facilities in sanitary landfills. Unless the sludge contains a minimum of
35% solids there may be leaching of soluble contaminants into ground waters.
Also the cost of sludge disposal is a significant treatment expense.
Therefore, several methods of dewatering the settling tank bottoms were
investigated at San Leandro.
Gravity Sedimentation
In an effort to improve the performance of the gravity sedimentation
currently employed at San Leandro, the Lamella Gravity Settler manufactured
by Parkson Corporation was investigated. The basic principle of utilizing a
series of inclined settling plates in close proximity to each other increases
the settling area ten times that of a conventional unit. Another advantage
of this system is a simplified sludge removal technique and use of a low
amplitude vibrator pack to further thicken the settled solids.
A 20 liter (5 gal) sample taken from the underground sump at
San Leandro was tested at the Parkson Laboratory. The sample had a sus-
pended solids loading of 3,000 mg/£ and was therefore, representative of the
raw effluent from the plant. The results of this test were not encouraging
since only a 41 underflow concentration was achieved. Even with the addition
of flocculating aids the settling properties were not significantly improved.
82
-------
Gravity Filters
As "pre-concentration" steps a Bauer Hydrasieve and a SWECO Centrifugal
Wastewater Concentrator/Vibro-Energy Separator Combination were tested at
San Leandro. Blinding of the screens on both the Hydrasieve and the Vibro-
Energy Separator preclude their use with adhesives and sealants wastewaters.
Pressure Filters
A sample taken from the sump in San Leandro could not be filtered using
an open mud discharge filter press without any filter aids. The filterability
was improved slightly through the addition of a coagulant at low pH. A good
break was achieved but the gelatinous nature of the created floe greatly
inhibited filtration.
These tests were conducted by personnel of Industrial Filter and Pump
Mfg., Chicago, Illinois who estimated that the amounts of filter aid, fly-
ash, and/or other coarse material necessary to dewater this sludge may be
economically prohibitive.
Centrifugal Filters
A liquid cyclone was evaluated in San Leandro for sludge thickening but
was ineffective due to the small differences in the specific gravities of the
suspended solids.
ELECTROCQAGULATION TEST
An electrocoagulation process (Swift Environmental Systems Co.) was
tested both as an alternative to, and a pretreatment for ultrafiltration.
Essentially all of the wastewater generated by the San Leandro Plant in one
week was processed by the electrocoagulation system. A portion of the
electrocoagulation system effluent was then processed by the UF system.
Details of the electrocoagulation test are presented in Reference 13, UF
operating performance is discussed below.
The UF membrane flux vs. time curves while processing the pretreated
feed are given in Figure 14. Average flux levels were 2.3 m3/m2-day
(55 gfd) for the HFD membrane and 2.1 m3/m2-day (50 gfd) for the HFM
membranes. For both membrane types, the flux varied considerably during
the 18 hour test period. These variations in flux levels are attributed
to the intermittent nature of the UF system operation (six hours per day
for three days), the variability of the plant effluent and changes in the
electrocoagulation process chemical dosages. However, neither these flux
variations nor the relatively high flux levels can be associated solely
with the electrocoagulation process. For example, during run #6 (see
Figure 7) flux for both the HFD and HFM membranes ranged from 1.4,to
> 2.7 m3/m2-day (35 to >65 gfd) with averages of >2.1 m3/m2-day (50 gfd).
The duration of run #6 was 172 hours.
83
-------
CO
3.0
•o
i
CM
e
x
-------
The analytical data obtained during UF processing of the electro-
coagulation system effluent are presented in Appendix A. Average permeate
loadings and removal efficiencies for the UF process (type HFM membranes
only) are listed below (Table 29).
TABLE 29. AVERAGE UF SYSTEM REMOVAL EFFICIENCIES
FOLLOWING ELECTROCOAGULATION PRETREATMENT
Average UF
Permeate Average Removal
Concentration Efficiency
Assay (mg/Ji) (%)
Total Freon Extractives
Total Suspended Solids
BOD5
COD
Total Cyanide
Phenolic Compounds
Zinc
20.5
34,0
6,480
15,700
4.6
61.7
0.59
5.1
62.4,
32.2
2.6
—
27.9
85.0
Essentially, no reduction in total freon extractives or COD is evident. The
suspended solids levels in the UF permeate ranged from 32 to 54,mg/£ which is
above the average value observed during ultrafiltration of the total plant
effluent without pretreatment. The destabilizing mechanism of the electro-
coagulation process is most likely responsible for this increased secondary
precipitation in the permeate stream.
The increase in the removal efficiency for phenolic compounds was -30%.
However, the San Leandro Municipal Discharge Limit for phenols was still
exceeded. BOD reduction was also -30%. As had been observed in all other
instances, the BOD level in the treated effluent was still significantly
above the proposed federal guidelines.
Zinc levels in the UF permeate were below the discharge level and ranged
from 0.32 to 1.0 mg/£. Interferences in the total cyanide assays are
indicated from the analytical data.
Based on the above discussion and the data in Reference 13, the
following conclusions are drawn relative to electrocoagulation treatment of
the San Leandro Plant effluent.
-- Electrocoagulation is not recommended as a pretreatment
to UF since improvements in final effluent quality are
modest. Also, no significant increase in UF flux is
observed.
85
-------
The plant wastewater flow is only 19 m /day (5000 gpd). The
electrocoagulation process is better suited for total waste-
water flows of >38 m3/day (10,000 gpd).
The high variability of the San Leandro effluent makes electro-
coagulation treatment difficult. Continuous operator attendance
would be predicted.
While electrocoagulation treatment (or pretreatment) is not
recommended at San Leandro, it may be suitable, in terms of
both effluent quality and cost-effectiveness, at larger
plants.
86
-------
SECTION 8
SUMMARY OF ULTRAFILTRATION SYSTEM
OPERATION AT THE DEWEY AND ALMY
CHICAGO, ILLINOIS, PLANT
INTRODUCTION
The Chicago, Illinois plant of the Dewey and Almy Chemical Division of
W.R. Grace & Co. manufactures the division's line of container sealing
compounds and produces Darex and Rock Processing Chemicals for the
Construction Products Division. Although the Chicago plant produces a
similar line of products as the San Leandro, California plant, there are
wide variations in product mix and hence differences exists in the type and
quantity of pollutants generated between the two plants. This dissimilarity
was sufficient to warrant separate piloting of waste treatment schemes at
both locations.
In 1971, a Dissolved Air Flotation (DAF) system was installed to reduce
the oils and grease loading in the Chicago plant effluent. The system im-
proved the quality of the plant's effluent in terms of oil and grease content,
however, due to the effluent's variability from hour to hour, there were
frequent periods during which the pollutant levels were excessive. A
chemical treatment pilot system was installed in conjunction with the DAF
unit in 1973 after an initial feasibility study by Dearborne Environmental
Engineers (a division of W.R. Grace) and successful bench-scale tests.
Experience with the pilot system, discussions with chemical treatment system
users, and computer simulations indicated that this treatment method still
had some disadvantages. In particular, the system was unable to handle
certain chemical compositions, thus the effluent would be out of specification
(i.e., greater than 100 ppm oils and grease) 5% to 10% of the time. To
obtain a higher degree of compliance over that attainable with chemical
treatment an alternative method, ultrafiltration, was proposed, piloted, and
finally accepted.
OVERVIEW OF UF SYSTEM OPERATION
A full-scale ultrafiltration system was installed in the Chicago plant
under Metropolitan Sanitary District (MSD) of greater Chicago Permit Number
74-602 in 1974,at a capital cost of $180,000. This cost includes the UF
system and all associated controls; platforming around the system; tanks
with capacities of 0.95 m3 (250 gal), 9.5 nr1 (2,500 gal), 18.9 mj (5,000 gal),
26.5 m3 (7,000 gal) and 75.8 m3 (20,000 gal); transfer pumps; piping with
87
-------
electrical tracer lines; a building to house the secondary sump; and, all
installation expenses. The UF system is an Abcor, Inc. Model UF 384,FEG
incorporating 78.5 m2 (845 sq. ft.) of Type HFD* membrane in a tubular
configuration. The time-weighted average permeate flow (during 1976) has
been 71.1 m3/day (18,750 gpd) while concentrating the retained pollutants to
an average of 10% total solids.
A summary of UF system separation performance for a number of
contaminants is presented in Table 30. The removal efficiency data in
Table 30 are calculated on a UF concentrate, rather than plant effluent
basis. Therefore, the more meaningful data are the average contaminant
loadings in the permeate. The oil and grease (hexane extractives) loading
has averaged 35 mg/£ compared to an MSD specification of 100 mg/£; the
suspended solids concentration has averaged 24,mg/£. Iron and,zinc concen-
trations in the permeate have averaged 0.8 mg/£ and 1.25 mg/£, respectively.
Similar to the San Leandro results, modest reductions in BOD are achieved by
the UF membranes and average BOD and COD loadings are in excess of the
contractor's recommendations for BPCTCA, BADCT, and BATEA guidelines.
It has been observed in Chicago that permeate quality does not degrade
as the process waste material builds in concentration. As expected,
however, permeate flow decreases with increasing solids concentration. A
regression analysis was performed on data collected during the first two
months of system operation to model the relationship between permeate
throughput and percent total solids. The following relationship was
derived:
Js = 1.324.log (38.52/Cb) (9)
where,
3 2
Js = permeate flux, m /m -day
C^ = bulk solids concentration, % total solids.
This model, together with plant operating conditions, aids the UF system
operator in predicting when the system should be shut down, the concentrate
scavenged, and the membranes cleaned.
The concentrated wastes are disposed of by contract hauling and the
waste sludge is chemically treated by the scavenger prior to its ultimate
discharge. EPA regulations maintain, however, that liabilities reside with
the plant producing the sludge for all damages incurred by any party as a
result of adverse environmental effects, directly or indirectly related to
the disposed wastes.
OPERATING DATA AND COSTS FOR 1976
For the year 1976 the UF system processed 15,463 m3 (4,080,000 gal) of
wastewater averaging 71.1 m3/day (18,750 gpd). The UF system throughput
and the permeate oils and grease loading are summarized in Table 31 by month
_
Note: Type HFM membranes were not commercially available in 1974.
88
-------
TABLE 30. AVERAGE UF PERMEATE QUALITY DURING PROCESSING OF CHICAGO PLANT EFFLUENT.
00
Assay
Total Hexane Extractives
Biodegradable. Hexane Extractives
Non-Biodegradable Hexane Extractibles
Total Suspended Solids
BOO
COO
Iron
Zinc
Average
Concentration
in UF Concentrate
(mg/A)
3,580
3,290
242
1,640
1,290
21,200
7.5
104.0
Average
Concentration
in UF Permeate
(mg/n)
35
31
4
24
1,200
6,080
0.8
1.25
Average
Removal
Efficiency*
99.1
99.1
98.3
98.5
7.0
71.3
40.0
98.8
MSD
Specification
(mg/A)
100
50
15
Removal efficiency calculated on concentrate, rather than feed, basis.
-------
TABLE 31. AVERAGE CHICAGO PLANT UF SYSTEM OPERATING DATA DURING 1976.
Month
January
February
March
April
May
June
July
August
September
October
November
December
Average
Total
Permeate
Flow (m3)
1,189
1,920
1,619
1,603
2,269
1,137
1,694
594
1,114
1,230
459
633
1,288
% of
Design
Capacity
75.6
121.9
99.3
116.6
116.0
75.5
105.9
55.5
•?6.1
59.1
53.9
67.0
86.8
Hexane
Extractives
(mg/£)
55
33
43
39
45
106 *
77
N.A.**
58
32
31
34
51
* Major product spill incurring 2 days out of compliance.
month averaged 30 mg/£.
Remainder of
**
Not available.
90
-------
and broken down to weekly averages in Appendix B. The system throughput
averaged 86.8% of the design capacity in 1976. In 1975 this figure was
110.9%. The problems which led to lower flux levels stemmed from accidental
product and/or raw material spillage resulting in either a fouling of the
membrane or a lowering of feed pH. On one occasion approximately 90.7 kg
(200 Ibs) of a latex were inadvertently spilled. This resulted in a non-
removable coating of the membrane necessitating total membrane replacement.*
Due to the design and layout of the Chicago plant and pollution control
system, rain water and material spillage from one railroad track gets
carried into the system's collection network. This results, on a rainy day,
in low pH; the level of acidity being proportional to the severity of the
rain storm. Low pH (<7) has been found to decrease the permeation rate by
as much as 28%. If a further decrease in pH occurs (to <5.5) fouling results
and the system must be shutdown and cleaned.
The operating costs incurred during 1976 are presented in Table 32. For
comparison, both the 1975 and 1976 operating costs are given. The total
costs were $3.34/m3 ($12.66/1000 gal) in 1975 and $3.51/m3 ($13.23/1000 gal)
in 1976. The increase of $0.15/m3 ($0.57/1000 gal) in 1976 over 1975 is
attributable to a 5% increase in total operating time and increased power
costs. The 18% lower throughput in 1976, resulted for the cost saving
measure of increasing the UF concentrate solids loading before scavenging.
By operating at a higher system conversion a $0.14/m3 ($0.53/1000 gal)
reduction in operating costs was realized. The most significant operating
expense at the Chicago plant is surfactant addition. This process step
accounts for 26.5% of the total waste treatment operating costs but, as
observed in the San Leandro field demonstration test, it will not be required
at all manufacturing sites. The labor costs of $0.45 to $0.50/m3 ($1.7 to
1.9/1000 gal) are associated almost entirely with sludge handling activities
and do not represent operator requirements for the UF system per se^.
In conclusion, the Chicago Plant's water pollution control system,
utilizing ultrafiltration as the means of treatment has:
1. Reduced the concentration of oils and grease and other
major pollutants to well within MSD specifications.
2. Maintained a 99% compliance level with MSD specifications.
3. Required minimal operator attention.
Furthermore, unlike chemical treatment, ultrafiltration is capable of
accepting tertiary treatment equipment in the form of ozonation or reverse
osmosis. In light of the national goal to eliminate all industrial dis-
charges by 1985, this adaptability is a decided advantage.
*Again, it must be noted that Type HFM membranes which can be solvent
cleaned were not commercially available at the time of the Chicago UF
system installation.
91
-------
TABLE 32. ANNUAL OPERATING COSTS FOR CHICAGO PLANT WASTE
TREATMENT SYSTEM
Item
Payrol 1
Sewer Tax
Surfactant
Sludge Removal
UF Concentrate Removal
Analytical Laboratory
Power
Abcor, Inc. Services
Membrane Replacement
Miscellaneous
Total
Total n)3 Permeated
Design Capacity, m3 , @
Total Operating Hours
% Time on Stream
% of Design Capacity
$
7,163
1,346
14,306
5,195
2,849
Charges 1 ,684
9,760
2,366
8,693
999
$54,361
81.75 m3/day
1976 FY Costs
$/m3
0.46
0.09
0.93
0.34
0.18
0.11
0.63
0.15
t 0.56
0.06
$3.51/m3
15,463
17,813
5,223
59.6
86.8
$
9,579
1,484
16,677
6,100
6,087
2,409
6,822
1.498
8,965
3,208
$62,829
1975 FY Costs
$/m3
0.51
0.08
0.89
0.32
0.32
0.13
0.36
0.08
0.48
0.17
$3.34/m3
18,810
16,964
4,973
56.8
110.9
92
-------
SECTION 9
FULL-SCALE SYSTEM DESIGN AND ECONOMICS
FULL-SCALE SYSTEM DESIGN
Based on the experimental results of the field demonstration program,
the preliminary posttreatment process experimentation and the operation of
the ultrafiltration system at the Grace Chicago Plant a full-scale treat-
ment system has been conceptualized. A P&I drawing of the proposed treat-
ment system for adhesives and sealants manufacturing wastes is presented in
Figure 15. The various plant waste streams flow into the plant sump where
partial equalization takes place. A transfer pump, signaled by a level
controller in the sump, passes the plant effluent to the first of a series
of settling, flotation, settling tanks. These tanks have a combined
capacity of 1-1/2 days of waste equalization.
The overflow from the final settling tank is pumped to the ultra-
filtration system feed tank. A pH monitor/controller is signaled by a
probe in the UF feed tank and caustic is added as needed to maintain the
waste pH above 8.0. The stability of the latices and oil emulsions is
believed to improve when basic conditions are maintained. The pH adjust-
ment step is, therefore, a precautionary step against latex destabilization
and subsequent UF membrane fouling. This precaution may not be necessary
at all manufacturing sites.
The pH adjusted feed is then circulated through the ultrafiltration
membranes. Because of the nature of the waste stream and the possibility of
severe latex fouling, tubular modules are preferred. The tubular modules
can treat very dirty wastewaters and can be mechanically cleaned if
necessary. Of the two Abcor, Inc. membrane types tested, the HFM membrane
is preferred over the HFD membrane because of its greater resistance to
environmental attack (i.e., its ability to be solvent cleaned, if
necessary).
The UF system is operated in a semi-batch concentration mode. That is,
fresh feed is continuously added to the UF feed tank until the desired
system conversion (95%-98%) is nearly attained. At this point the contents
of the UF feed tank are batch processed to the maximum conversion achievable.
The concentrate is then drained from the UF system and hauled away by
scavenger to land fill. The UF permeate is continuously withdrawn throughout
the entire processing period.
93
-------
Plant Wastes
Transfer Pump
_. ) I—, k
I Transfer
JLeVeT
Control
Plant Sump
Equalization: Settling/Rotation Tanks
Caustic
Tank and]
Pump
Vent ^
Ozonated Ozone
Product Water Contactor
Discharged "**• •
to Sewer
Option 3
Ozone
Generator
I
Air .—L-,
Dryer | I
»
g
Air in
Controller i_P
Level
Control
Feed
Tank Circulation
Pump
Option 2
RO Concentrate
RO
Feed High
Tank Pressure
Pump
Disposal
Reverse
Osmosis
Modules
to.Disposal
Ultrafil-
tration
Modules
UF
Permeate.
Option 1
Discharge to
Sewer
RO Permeate
for Reuse and/or
Discharge to
Sewer
FIGURE 15. PROPOSED TREATMENT SYSTEMS FOR ADHESIVES AND SEALANTS MANUFACTURING WASTES.
-------
The ultrafiltration system is constructed of carbon steel. The
membrane shells (see Figure 1) are made of PVC plastic. The UF membranes can
easily be removed from their shells and soaked in a solvent bath, if
necessary.* A semi-automatic detergent cleaning package including clean tank
valving arrangements and timer-controlled alarms for signalling the system
operator is included with the UF system.
The preferred ultrafiltration system operating conditions are a feed
circulation rate of 163.5 m3/day (30 gpm), an inlet pressure of 3.4.atm
(50 psig) and a temperature of 32 to 43°C. Operation above 43°C may increase
latex destabilization and promote membrane fouling. Operation below 32°C
will lower membrane flux and cause an increase in the membrane area
requirement.
Three options are considered for posttreatment of the UF permeate. The
first option is to discharge the UF permeate directly to the sewer and
allow Municipal Authorities to treat it biologically. It has been determined
by the preparers of the draft "Development Document for Effluent Limitations
Guidelines and Standards of Performance, Miscellaneous Chemicals Industry,
Adhesives and Sealants Industries" (2) that BOD and COD are pollutants
compatible with publicly-owned treatment plants. If the levels of non-
compatible pollutants (e.g. phenolic compounds and cyanide) are below local
standards it is anticipated that ultrafiltration will produce an effluent
acceptable to Municipal Treatment.
The second posttreatment option involves the use of ozonation for
oxidation of phenolic compounds and cyanide to acceptable discharge levels
and for partial reduction in the BOD and COD loading of the ultrafiltrate.
Dried, compressed air is passed through a corona discharge to generate the
ozone on-site. The UF permeate is contacted with the ozone in a reaction
vessel sized for the proper residence time. This residence time and the
dosages of ozone required were not investigated during this program. It is
expected that the high dissolved solids content of the ozonated product
water will preclude its reuse within the plant and therefore, the ozone
contactor effluent is discharged to the sewer.
The third posttreatment option is processing of the UF permeate by
reverse osmosis. A polyamide membrane (PA-300, UOP, Inc.) in a spiral-
wound configuration is potentially preferred. This membrane is selected
A highly sophisticated system capable of solvent cleaning in situ, semi-
automatic spongeball cleaning and designed with an explosion proof
electrical system is available. The need for such a sophisticated system
is questionable and, in most cases, will not justify the 4Q%-60% capital
cost increase over a standard system.
*
95
-------
because of the high pH of the UF permeate. The spiral-wound geometry is
chosen over the hollow-fiber module design because the secondary
precipitation of suspended solids, which was observed periodically in the
UF permeate, may readily foul a hollow-fiber module.
The reverse osmosis membrane separation system is operated much the
same as the UF system and is therefore not described. A system conversion of
90% to 95% is anticipated and it is projected that the RO product water will
be reuseable within the plant for cooling water makeup. However, reuse has
not been demonstrated during this program. Alternatively, the RO product
water can be discharged directly to the sewer.
FULL-SCALE SYSTEM ECONOMIC PROJECTIONS
Economic projections are presented for the three options discussed
above, namely:
Option 1: fqualization->Ultrafiltration-> Discharge
Option 2: Equalization-^Ultrafiltration-^Ozonation -^-Discharge
Option 3: Equal ization->Ultrafiltration-^Reverse Osmosis -»•
Reuse and/or Discharge
Cases are presented for adhesives and sealants manufacturing facilities dis-
charging 3.8 m3/day (1,000 gpd), 18.95 m3/day (5,000 gpd) and 75.8 m3/day
(20,000 gpd) of wastewater. It is believed that a range of treatment
systems of this magnitude will encompass the majority of manufacturing sites.
For the smallest plant (discharging 3.8 m3/day) it was assumed that all
manufacturing activities would occur during a single 8-hour shift. While
the treatment system could operate 24,hours/day, it was designed on a single
shift basis and therefore has an actual capacity of 11.4,m3/day (3,000 gpd).
The system to be used with plants discharging 18.95 m3/day is designed for
complete waste processing in two shifts (16 hours). This system therefore
has the capability of treating 28.4lm3/day (7,500 gpd). Thus, if either
plants discharging 3.8 rn3/day (1000 gpd) or plants discharging 18.95 m3/day
(5000 gpd) expand their manufacturing operations, the originally purchased
UF system will be able to handle the increased wastewater flow. The
largest system (treating 75.8 m3/day) is designed for 3 shift (24,hours)
operation.
The design bases used for calculating the purchased equipment and annual
operating costs are given in Table 33.
Purchased EquipmentCost_Proje^ti_ojTS_
Estimated purchased equipment costs for each unit process (equalization,
ultrafiltration, ozonation and reverse osmosis) are presented in Table 34.
Excluded from these cost projections are installation costs which will be
highly site specific. If all utilities (power, water, and sewer connections,
96
-------
TABLE 33. DESIGN BASES FOR PROJECTIONS OF UNIT PROCESS
PURCHASED EQUIPMENT AND OPERATING COSTS.
Total Plant Discharge (m3/day)
Item 3.8 18.95 75.8
EQUALIZATION
1. Holding time, days
ULTRAFILTRATION
1. Design flux, m3/m2-day
2. No. of shifts operated
3. Membrane area required, m2
4. Expected membrane life, years
5. Membrane replacement costs, $/m2
6. Pump efficiency, %
7. Membrane type
8. Membrane configuration
9. Membrane area per module, m2
10. Module operating pressure, atm
OZONATION
1. Maximum power requirement
projected by manufacturer, kw/day
1.5
1.23
1
9.28
3
367
70
Abcor, Inc.
Tubular, 0
0.2
3.4
5
1.5
1.23
2
23.2
3
367
70
, Type HFM
.025m dia.
5
1.5
1.23
3
61.9
2
367
70
(non-eel lulosic)
x 3.05m long
5
REVERSE OSMOSIS
1. "Assumed flux, m3/m2-day 0.95 0.95 0.95
2. No.of shifts operated 1 2 3
3. Membrane area required, m2 12 29.9 79.8
4, Expected membrane life, years 332
5. Membrane replacement costs, $/m2 80.8 80.8 80.8
6. Pump efficiency, % 70 70 70
7. Membrane type UOP, Inc., PA-300
8. Module configuration 2 Spiral-wound,o.im dia. x 0.97m long
9. Membrane area per module, m 5.57
GENERAL
1. Operating labor, hours/day 1 2 4.
2. Supervisory labor, hours/day 0.5 1 2
3. Power, $/kwh 0.04. 0.04, 0.04,
4, Sludge and UF concentrate disposal
costs, $/m3 inlet to system (basedon
actual Chicago Plant operating costs) 0.19 to 0.34
5. RO concentrate disposal costs,
$/m3 inlet to system 0.65
97
-------
TABLE 34, ESTIMATED PURCHASED EQUIPMENT COSTS FOR SELECTED UNIT
PROCESSES OF VARIOUS CAPACITIES (THOUSANDS OF DOLLARS)
Actual Capacity (m^/day)
Item 11.4, 28.4, 75.8
EQUALIZATION
1. Settling/flotation tanks with 1-1/2 days
retention 0.7 4.6 6.3
2. Transfer pumps, before and after item 1 1.0 1.5 2.0
1.7 6.1 8.3
ULTRAFILTRATION
1. Ultrafiltration system 20.0 45.0 80.0
2. pH control loop 1.8 1.8 1.8
21.8 46.8 81.8
OZONATION
1. Smallest capacity system of U.S. Ozonair
Corp, including ozone generator, air
dryer, air compressor, pump, contactor
and automatic controls. 14,0 14,0 14.0
REVERSE OSMOSIS
1. Reverse osmosis system (estimated from
Reference (11)) 15.0 25.0 35.0
98
-------
etc.) are in place, and no new facility need be constructed, installation
costs might be as low as 25% of the purchased equipment cost. On the other
hand if significant site preparation is required, installation costs might
be as high as 75% to 100% of the purchased equipment costs.
For ease of discussion the costs for the three system options are
summarized in Table 35. For the smallest size system, designed to treat
3.8 m-Vday (1000 gpd) during 1 shift operation, purchased equipment costs
range from $23,500 to $38,500 depending on the degree of treatment selected.
Since the wastewater flow from these plants is so small, equalization and
UF processing is the most practical treatment scheme. This would provide
excellent "pretreatment" prior to discharge to a municipal sewer. The
purchased equipment cost per m3 of waste treated, based on Option 1 costs,
is $6,184,($23.50/gpd).
The purchased equipment costs for the 18.95 m3/day (5000 gpd) systems
range from $52,900 to $77,900. The costs per m3 of waste treated are
$2,792 ($10.58/gpd), $3,530 ($13.38/gpd) and $4,110 ($15.58/gpd) for
Options 1, 2 and 3, respectively.
A range of purchased equipment costs from $90,100 to $125,100 is
estimated for treatment systems handling 75.8 m3/day (20,000 gpd) of waste-
water. These costs translate to costs of $1,189/m3/day ($4,5/gpd) for an
Option 1 system, $1,373/m3/day ($5.2/gpd) for an Option 2 system and
$1S650/m3/day ($6.26/gpd) for an Option 3 system.
Operating Cost Projections
Table 36 contains the annual operating cost projections for the three
system capacities of each unit process. These costs were calculated using
the design bases presented in Table 33 and are summarized in Table 37 by
treatment system option. In both Tables 36 and 37 the costs given for
disposal of the settling/flotation tank sludges and the UF concentrate are
based on the actual Chicago Plant operating data for 1976. The costs for
the RO concentrate disposal assume 95% conversion and a hauling cost of
$0.013/1 Her ($0.05/gal). Thus, the RO concentrate disposal costs are
$0.65/m3 of inlet wastewater to the system.
Reviewing Table 36, it can be observed that the two most significant
contributors to the operating costs are sludge disposal and labor. The
sludge disposal costs may be significantly reduced if an effective means
of sludge dewatering is identified. The labor costs, as observed from the
Grace Chicago UF system experience, are mainly associated with sludge
handling and not with the UF system. This labor can, therefore, be
unskilled and for non-union plants the annual operating costs will lower
significantly.
99
-------
TABLE 35. ESTIMATED PURCHASED EQUIPMENT COSTS FOR THE THREE
TREATMENT SYSTEM OPTIONS (THOUSANDS OF DOLLARS)*
Actual Sys. Cap, (m^/day)
Option 11.4, 28.4, 75.8
1. Equal ization-^U1trafiltration-> Discharge 23.5 52.9 90.1
2. Equalization -^Ultrafil tration->Ozonation->
Discharge 37.5 66.9 104.1
3. Equalization -»-Ultrafiltration->Reverse
Osmosis-/Reuse and/or Discharge 38.5 77.9 125.1
Does not include installation costs which are highly site specific.
100
-------
TABLE 36. ESTIMATED ANNUAL OPERATING COSTS FOR UNIT PROCESSES OF VARIOUS
CAPACITIES (THOUSANDS OF DOLLARS)*
o
Actual Capacity (nr/day)
Item
EQUALIZATION
1 . Sludge Disposal
2. Capital amortization @ 10% for 10 yrs
ULTRAFILTRATION
1. Membrane Replacement
2. Power
3. Chemicals, cleaning and pH adjustment
4, Concentrate Disposal
5. Capital amortization @ 10% for 10 yrs
OZONATION
1 . Power
2. Capital amortization @ 10% for 10 yrs
REVERSE OSMOSIS
1. Membrane Replacement
2 . Power
3. Concentrate Disposal
4. Capital amortization @ 10% for 10 yrs
GENERAL
1. Operating Labor @ $7.5/hr plus 75%
fringe and OH
2. Supervisory Labor @ $15/hr plus 75%
fringe and OH
11.4
1.01
0.17
1.14
0.28
0.4
0.6
2.18
0.42
1.4
0.45
0.10
1.94
1.5
3.4
3.4
28.4
2.51
0.61
2.83
1.28
1.4
1.4
4.68
0.83
1.4
0.90
0.20
4,82
2.5
6.8
6.8
75.8
6.70
0.83
11.3
5.2
4.6
3.7
8.18
1.25
1.4
3.4,
0.30
12.90
3.5
13.6
13.6
* Annual operating costs assume 260 days/year of operation
-------
TABLE 37. ESTIMATED ANNUAL OPERATING COSTS FOR THE THREE TREATMENT
SYSTEM OPTIONS (THOUSANDS OF DOLLARS)*
Actual System Capacity (m^/day)
Option 11.4 28.4 75.8
1. Equal ization->Ultrafiltration->Discharge 8.84 20.8 52.8
2. Equal ization ->U1 trafil tration -^Ozonation ->.
o Discharge 11.0 23.7 56.8
3. Equalization ->-Ul trafil tration -^Reverse
Osmosis-^Reverse and/or Discharge 13.2 29.9 74.3
* 90% of the treatment system operating and supervising labor is credited to the
equalization/UF system.
-------
The costs of operating a treatment system for a plant discharging
3.8 m3/day (1000 gpd) are estimated to be $8.91/m3 ($33.8/1000 gal) for
Option 1; $ll.l/m3 ($42.0/1000 gal) for Option 2; and $13.4/m3 ($50.8/1000
gal) for Option 3. For systems treating 5 times this amount of wastewater
(18.95 m^/day) the operating costs are $4.22/m3 ($16.0/1000 gal) for
Option 1; $4.81/m3 ($18.2/1000 gal) for Option 2; and$6.07/m3 ($23.0/1000
gal) for Option 3. The highest capacity system, treating 75.8 m3/day
(20,000 gpd) of wastewater has associated with it operating costs of
$2.68/m3 ($10.1/1000 gal) for Option 1; $2.88/m3 ($10.9/1000 gal) for
Option 2; and $3.78/rrP ($14.3/1000 gal) for Option 3.
In none of the above cost projections for Option 3 (equalization,
ultrafnitration, reverse osmosis, reuse and/or discharge) are any credits
given for water reuse. The RO product water is believed to be reusable
within the plant for cooling water makeup and area washdowns and may have
other uses as well. However, until reuse is demonstrated application of
credits may be premature. Also, the amount of credit to be applied would
be highly site specific and would depend upon water use charges, sewer
surcharges and the degree of water reuse achievable.
Comparison of Options 1, 2, and 3 Costs with the Preliminary BPCTCA,
BADCT and BATEA Technology Costs
The Draft Development Document which presents the recommended Effluent
Guidelines for the Adhesives and Sealants Industries (2) gives capital and
operating costs for double-effect liquid evaporation of Subcategory B
(water-based adhesive solutions containing synthetic and natural materials)
and Subcategory C (solvent solution adhesives and cements generating
contaminated wastewaters). These costs are given only for plants of a
single size in each Subcategory and are given in August, 1972 dollars. The
same level of treatment is required to meet all three sets of standards:
BPCTCA, BADCT and BATEA. Table 38 summarized the Development Document's
cost projections. The Development Document also indicates that for a
Subcategory B plant discharging 37.9 m3 of waste per day (10,000 gpd),
56,700 kg (125,000 Ibs) of sludge (dry-weight basis) will be generated from
double-effect liquid evaporation and that for a Subcategory C plant
discharging 22.7 m3/day (6,000 gpd) of waste, 567 kg (1,250 Ibs) of sludge
will be generated. It is not clear, however, whether sludge disposal costs
are included in the annual operating cost estimates.
Table 39 presents a comparison of Option 1, 2 and 3 system costs with
the costs for double-effect liquid evaporation. The letter's costs were
left in 1972 dollars. Using the costs derived during this program for a
plant discharging 18.95 m3/day (5,000 gpd) of waste and the Development
Document's costs for a plant discharging 22.7 m3/day (6,000 gpd) of waste
as the closest point of comparison and leaving the Development Documents
Costs in 1972 dollars the following observations are made:
103
-------
TABLE 38. SUMMARY OF DEVELOPMENT DOCUMENT FOR EFFLUENT LIMITATIONS
GUIDELINES PROJECTED COSTS FOR TREATMENT OF SUBCATEGORY
B AND C ADHESIVES AND SEALANTS WASTEWATERS (2)*
o
Subcategory
B.
Water-based adhesives
solutions containing
sythetic and natural
materials
Plant
Wastewater
Flow (m'/day)
37. D
Proposed
Treatment
double-effect
liquid evaporation
Capital
Costs, $
412,000
Annual Operating Costs
Operating & Maintenance
Energy & Power
Capital Recovery plus
return 9 10S, for 10 yrs
Total
Breakdown, $
65 ,000
57,000
67 ,000
189,000
C. Solvent solution
adhesives and cements
generating contaminated
wastewaters
22.7
double-effect
liquid evaporation 324,000
Operating & Maintenance
Energy & Power
Capital recovery plus
return @ 10%, for 10 yrs
Total
56,000
30,000
53,000
139,000
* Costs are presented in August, 1972 dollars.
-------
TABLE 39. COMPARISON OF OPTION 1, 2 AND 3 SYSTEM COSTS WITH DOUBLE-
EFFECT LIQUID EVAPORATION COSTS
o
en
Option 1
Option 2
Treatment System Description
. Equalization+Ultrafiltration-"
Discharge
. Equal ization->-Ul trafiltratlon*
Ozonation ^Discharge
Option 3. Equalization -'•Ultraf Miration*
Reverse Osmosis -^Reuse and/or
Discharge
Development Document Proposed Treatment
Equalization -^Double-Effect
Liquid
Evaporation -*Reuse and/or
Discharge
Daily
Wastewater
Discharge, m3
3.8
18.95
75.8
3.8
18.95
75.8
3.8
18.95
75.8
22.1
37.9
Estimated
Capital
Costs, $
23,500
52,900
90,100
37,500
66,900
104,100
38,500
77,900
125,100
324,000*
412,999*
Estimated
Annual
Operating
Costs, $
8,800
20,800
52,800
11,000
23,700
56,800
13,200
29,900
74,300
139,000*
189,000*
Estimated
Operating
Costs,
$/m3 processed
8.91
4.22
2.68
11.1
4.81
2.88
13.4,
6.07
3.78
23.6*
19.1*
Costs presented are in August, 1972 dollars
-------
1. The capital cost for double-effect liquid evaporation
is 6 times the Option 1 system cost, 4,8 times the
Option 2 system cost and 4 times the Option 3 system
cost.
2. The operating costs for double-effect liquid evaporation
were estimated at $23.6/m3 ($89.4/1000 gal) in 1972.
Energy and power costs, which have clearly risen since
1972, make up 30% of this operating cost. The estimated
operating costs for the Option 1, 2 and 3 systems are $4,22/m3
($16.0/1000 gal), $4.81/m3 ($18.2/1000 gal) and $6.07/m3
($23.0/1000 gal), respectively.
106
-------
REFERENCES
1. Skeist, I. Handbook of Adhesives, Second Edition, Van Nostrand,
New York, 1977.
2. Roy F. Weston, Inc., "Development Document for Effluent Limitations
Guidelines and Standards of Performance Miscellaneous Chemicals
Industry", Draft Report for EPA9 February, 1975.
3. Douglas, G,, "Modular Wastewater Treatment System Demonstration for
the Textile Maintenance Industry", Final Report for EPA Contract
No. FYV12120, EPA-660/2-72-037, January, 1974.
4, Higares, J.R., Letter from City of San Leandro to Dewey and Almy
Chemical Company, June 3, 1977.
5. Thomas, J.M., and Thomas, W.S., Introduction tothe Principles of
Heterogenous Catalysis, Academic Press, New York, p. 32 (1967).
6. Goldstein, M., "Economics of Treating Cyanide Wastes" Pollution
Engineering, March, 1976.
7. Mauk, C.E. and Prengle, H.W., "Ozone with Ultraviolet Light Pro-
vides Improved Chemical Oxidation of Refractory Organics," Pollution
Engineering, January, 1976.
8. Telephone conversation, Technical Representative, Capital Controls,
Colmar, Pennsylvania.
9. Kibbel, W.H., Jr., et.al., "Hydrogen Peroxide for Industrial Pollution
Control".Industrial Wastes, November, December, 1972.
10. Telephone conversation, Technical Representative, DuPont Chemicals,
Wilmington, Delaware.
11. McNulty, K.J., et.al., "Reverse Osmosis Field Test: Treatment of
Copper Cyanide Rinse Waters", EPA-600/2—77-170, August, 1977.
12. Bambenek, R.A., Letter from Hubeneco, Inc. to D. Steiner, W.R. Grace,
Chicago, Illinois, August 24, 1977.
13 Trelease, D.S., "Evaluation of LECTROCLEAR™ Treatment of Dewey and
Almy's San Leandro Plant Wastewater", Swift Environmental Systems,
January 27, 1977.
107
-------
TABLE Al. ANALYTICAL DATA FROM UF CONCENTRATION OF SAN LEANDRO PLANT EFFLUENT, TEST #1
««l
o
CO
Sampl Ing
SUtlon Date
swp e/16
a/20
8/16-8/20
8/23
8/26
8/23-8/26
After 8/16-8/20
5"11"" 8/23-8/27
UF Concen-
trate 8/16
B/20
8/16-8/20
a/21
8/26
8/23-8/27
"enneat, ""«
a/20
8/16-8/20
8/23
8/26
8/23-8/27
San Ulndro Municipal
Discharge Limits
Total Non-Polar
freon Freon
EKtractibles Extractives
1,050
468
575
364
952
568
—
755
11.000
4,580
6,650
18,800
9,320
167
174
161
147
205
164
408
96
110
116
196
95
—
110
800
268
408
2540
1030
<>
<5
<5
6
9
6
300/1002
PH
t.i
8.4
?.?
7.1
8.3
7.8
7.7
8.4
6.3
6.5
7.5
a. 5
a.o
8.9
6.6
7.2
7.7
8.9
8.1
>6.0
ASSAY (»g/i)
Total
Suspended .on '»'•' '«* ToUl .
Solids "W5 Arienic Od»iu» Ctiro«li» Cyanide Cyanide1 Lead
3580
3380
9850 13,500 «O.Z <0.2 <0.5 0.36 '•' *'
924
9350
4650 14,000 <0.2 <0.2 «0.5 0.44 °-*
-------
TABLE A2. ANALYTICAL DATA FROM UF CONCENTRATION OF SAN LEANDRO PLANT EFFLUENT, TEST #2
Saapt 1ng
SUUon
Sump
After
Settling
UF Concen-
trate
UF
Permeate
Total non-Polar
freon Froon
Date Cxtricttblos Cxtractfbles pH
8/30
9/3
8/30-9/3
9/6-9/10
9/14
9/15
9/14-9/1?
8/30-9/3
9/6-9/10
9/14-9/1?
8/30
9/3
8/30-9/3
9/6-9/10
9/14
9/15
9/14-9/17
8/30
9/3
8/30-9/3
9/6-9/10
9/14
9/15
9/14-9/17
San Leandro Municipal
Discharge Limits
720 204
572 100
608 192
3070
136 36
672 204
1150 248
—
584
...
•J20 28
2140 964
1680 664
4310
4180 3990
3600 1840
6140 4750
34 <5
71 <5
55 <6
61
60 14
28 <5
47 <5
300/100Z
8.8
8.7
8.8
9.4
7.8
7.2
7.3
8.6
8.7
7.6
8.9
9.0
9.0
9.1
9.1
7.6
8.3
7.2
9.8
7.8
6.7
8.2
7.1
7.9
>6.0
Total
Suspended Knn Tfotal Free Total .
Solids ou"5 Arsenic Cadmius Chrosiua Cyanide Cyanide*
J060
2580
2900 14.000 <0.2 ,200
11,600 30.000 0.35 <0.2 <0.5 3.9 1.5
<5
$
11 10,000 '-0.2 <0.2 <0.5 0.55 9
14 10.500
-------
TABLE A3. ANALYTICAL DATA FROM UF CONCENTRATION OF SAN LEANDRO PLANT EFFLUENT, TEST #3
Station
Siaop
Settling
UF Concen-
UF
Permeate
San lesndro
DHchar9e LI
Date
J/20
9/23
9/20-9/24
9/20
9/20
9/23
9/20-9/24
9/20
9/23
9/20-9/24
Municipal
CDltl
Total Kon-Polar Total KSMS (.9rt)
Freon Freon Suspended jgi, total Free Total
tutractlbles Extractlbles pH Solids 5 Arsenic Cad.lu, Chromium Cyanide Cyanide Lead Mercury Phenols Hue
24305
30605
28005
—
24 705
2140S
2300S
2345
2605
2375
300/1 M?
124 8.3 16BO — — — ™ — — — —
312 9,0 2950
136 9.1 5750 10,000 6.o — — o.i o.e 0.5 — ,.„ 10 oov , 03 3o
Interference tn analysU suspected.
J300/ »g/t Oil and Crease of an leal or Kgetsble origin, 100 sig/l 011 and Orease of alneral or petrolewi ortjln.
J1.0 ing/t phenolic cc«WM»vds ulilch cannot be removed ay the Ssency's w«ste«ater trtatsent process.
*lii»jle dlluttd 1:10 to slni»1l« Interference^ detection limit Increases.
5Frtc* contanlnated. Value reported !l raxluuin possible oil and jreast.
-------
TABLE A4. ANALYTICAL DATA FROM UF CONCENTRATION OF SAN LEANDRO PLANT EFFLUENT, TEST #4
Sampling
Station Date
Simp 9/29
10/1
9/29-10/1
After
SettHnj 9/30
trate 9/29
10/1
9/29-10/1
UF
Peraeat* 9/29
10/1
9/29-10/1
San Leandro Municipal
Discharge Limits
~1 " -""
2KM/«/l Oil and Grease of
31.0 ng/t phenolic compounds
total
Mon-Polar
Freon Freon
Extractives Extractives
1,1205
4?8S
494?
...
4Uj5
5225
866s
95
655
22s
456
100
132
—
184
100
S20
15
11
10
300/100Z
animal or vegetable origin, 100 stg/a Oil
which cannot
be removed by the Agency's
PH
11.0
6.1
8.3
8.5
8.0
9.1
8.9
8.3
9,7
B.9
>6.0
Total ASSAtS(»g/t )
Suspended .., ' Total Free Total
Solids 5 Arsenic Cadmium Chromium Cyanide Cyanide Lead Mercury Phenols Zinc
34,000
14,000
4.760 S.3M <0,Z <0.2 <0.5 <2.6* 0.9 <\ .001 13 35
2'OT "' "'
17,500
4,280
4.360 1S.600 0.2 <0.2 <0.5 <2.6* 4.5 2.4 ,002 63 170
12 — --- — — — ... — .;.
91
18 10.000 <0.2 <0.2 <0.5 0.65 3.8 <1 .001 49 0.61
0.1 0.2 0.5 — 1.0 1.0 0.01 l.o3 3.0
and Grease of mineral or petroleun origin.
wastewater treatment process.
Fr»«n contaminated. Valye reported Is mttsaa possible oil and greise.
-------
TABLE A5. ANALYTICAL DATA FROM UF CONCENTRATION OF SAN LEANDRO PLANT EFFLUENT, TEST #5
Sampling
Station Date
Su"» 10/14
10/15
10/18
10/22
10/18-10/22
After
Settling 10/14
10/15
10/18-10/22
Uf Concen-
trate 10/14
10/16
10/18
10/22
.10/18-10/22
UF
Cerate ,„,„
10/15
HFO 10/1B
10/22
10/13-10/22
10/14
10/15
HFH 10/18
10/22
10/18-10/22
San Leandro Municipal
Discharge Limits
'l ' f t It
Total
freon
ExtractlMei
725S
3020S
41405
1 370s
U20
—
—
—
9005
2040s
2000s
72305
(870
2545
SI5
13
37
22
H65
110S
17
38
20
300/1 00?
Xon-Polar
freon
Eatractlblej
2035
984s
1060s
2205
760s
...
—
...
23#
566
48(5
1760s
30)0$
<5S
<55
14s
165
135
<5*
(5
165
13s
17«
°H
9.2
9.5
8.6
5.0
9.2
7.8
9.3
9.4
7.2
e.e
9.)
9.4
9.0
7.2
9.3
9.1
9.4
B.7
7.0
9.0
9.1
9.3
8.8
>6.0
ASSA1S(ng/M
Total
Suspended .„„ Total Free total 1
Solids °gD5 Artenlc Cadnlun Chromlun Cyanide Cyanide lead Hercunr flienoU 7tnc
1.950 980 <0.2 <0.2 <0.5 -5 — — '5'8 M
800
1,800 " — — — — — — — — ' —
3,180
7,000 3.600 •» «»
11.700
11,400
62,600
27.900 13.000
-------
TABLE A6. ANALYTICAL DATA FROM UF CONCENTRATION OF SAN LEANDRO PLANT EFFLUENT, TEST #6
Total
Non-Polar
Sampling Freon Frcan
Station Date Extractives Extractives
Sunp
After
Settling
UF Concen* *
trat*
UF
PermaU
urn
HFH
11/2 321
11/1-11/5 1820
11/8 4650
11/1! 21.400
11/8-11/12 22.000
11/1-11/5
11/8-11/12
11/2 846
11/1-11/5 1790
11/8 4600
11/12 13.800
11/8-11/12 13.900
11/2 SO
11/1-11/5 67
11/8 99
11/12 62
11/8-11/12 65
11/2 58
11/1-11/5 38S
11/6 107
11/12 116
11/8-11/12 65
San leandro Municipal
Discharge Limits
35
735
2370
500
820
...
—
334
10(0
1170
3750
4170
a
7
20
12
<5
<5
9
<5
<5
<5
300/100Z
ASSAYS (mg/t)
Total
pM
9.3
9.3
9.5
(.3
9.2
9.3
a.a
8.9
9.1
9.0
9.6
9.3
a.a
9.0
9.0
9.6
8.5
a.a
9.0
a. 9
8.7
8.6
>6.0
Suspended .fin Total
Solids """S Arsenic Cadmium Chromium
940
9,900 9,900 <0.2
70,800
17,700
18,400 10,500 <0.2
4,820
1,130
7,540
11,900 6.600 <0.2
18,400
50,100
33,100 17,200 <0.2
a ... ... ... ...
19 6,400 <0.2
9
<5
<6 10,600 <0.2
12 — ... —
12 6,400 <0.2
<5
-------
TABLE A7. ANALYTICAL DATA FROM THE UF CONCENTRATION OF LECTROCLEAR EFFLUENT, TEST #7
ASSAYS (mg/1)
Sampling
Station
Swift
Lectroclear
Ef f 1 uent
UF Concentrate
UF Permeate
HFM
Date
12/14
12/15
12/16
12/14
12/15
12/16
12/14
12/15
12/16
Total
Freon
Extract! bles
20.7
17.2
26.0
51.6
50,0
40.8
19.2
15.8
26.4
pH
—
8.8
8.6
8.6
7.4
Total
Suspended
Solids
188
59
85
490
274
350
32
26
44
BOD5
5,610
12,500
6,020
9,800
3,210
9,750
COD
12,200
20,200
15,500
19,900
12,300
19,100
Total
Cyanide
0.8
4.4
13.4
5.8
0.6
8.6
Phenolic
Compounds
246
29.5
9.9
145
8.0
11
158
18.0
9.0
Zinc
3.5
9.5
2.8
7.9
8.9
5.2
1.0
0.42
0.34
HFD
12/16
San Leandro Municipal
Discharge Limits
25.4
300
7.3
>6.0
54
8,790 19,100
5.0
1.0
11
1.0
0.32
3.0
-------
TABLE A8. ANALTYICAL DATA FROM UF CONCENTRATION OF SAN LEANDRO PLANT EFFLUENT
WITH SURFACTANT ADDITION, TEST #8
Sampling
Station
Sump
Feed
After
Sett! tng
UF Concentrate
UF Permeate
Date
2/14
2/18
2/14-2/18
2/22
2/25
2/22-2/25
2/14
2/18
2/14-2/18
2/22
2/25
2/22-2/25
2/14
2/18
2/14-2/18
2/22
2/25
2/22-2/25
2/14
2/18
2/14-2/18
2/22
2/25
2/22-2/25
Total Non-Polar
Freon Freon
Extractibles Extractives
338
668
144
58
288
162
410
50
438
111
322
181
580
668
1,810
344
503
557
156
214
299
196
147
156
no
68
45
16
20
54
160
11
226
11
74
40
64
230
394
31
78
96
9.0
14
11
7.0
10
7.0
pH
8.7
8.9
8.9
8.4
7.9
8.7
9.2
8.7
9.2
8.3
8.7
8.4
9.2
8.6
9.0
8.8
9.3
8.8
9.2
8.6
8.9
8.7
8.9
8.7
Total
Solids
9,350
13,200
9,030
1,980
6,520
3,310
6,180
7,330
4,520
9,490
4,890
6,130
4,440
91 ,800
36,500
137,000
54,000
93,900
2,260
8,100
3,410
6,700
3,340
5,080
Total
Suspended
Solids
1,580
7,040
4,720
320
2,100
3,300
3,620
546
932
1,440
1,650
1,370
1,080
72 ,000
17,000
91 ,000
34,000
61 ,300
27
80
27
132
48
80
BOD
7,270
5,640
7,560
2,460
12,300
6,860
6,110
5,280
4,120
13,000
12,400
12,000
4,900
17,200
11,640
24,800
16,400
22,200
3,990
5,160
5,840
8,860
9,010
10,200
Soluble
BOD COD
6,030 33,900
3,760 28,600
7,480 23,800
3,140 6,350
17,200 30,200
7,400 14,900
3,140 21,300
4,680 13,100
2,220 16,800
12,700 31,000
10,900 20,900
13,200 23,900
3,680 17,100
13,200 181,000
10,160 140,000
17,200 200,000
14,000 143,000
13,300 196,000
3,990 13,300
4,620 23,800
5,130 14,000
10,200 21,800
10,400 17,000
10,300 19,500
Soluble
COD
30,900
20,600
19,300
5,520
26 ,500
12,000
15,000
1 1 ,000
13,100
26,800
18,500
20,000
14,800
149,000
84,100
183,000
91 ,200
138,000
13,000
22,800
13,900
20,600
16,100
18,600
Phenolic
Compounds
638
6.0
235
68
2.8
110
18
205
80
210
46
118
35
375
77
600
170
136
60
49
32
388
96
160
Zinc
18
14
19
24
2.4
12
40
23
22
18
23
22
27
1,100
250
2,000
620
24
10
16
8.5
1,100
3.4
8.5
San Leandro Municipal ^ *>
Discharge Limits
300/100
>6.0
— — —
___
...
— —
—
1.0
3.0
300 mg/1 Oil and Grease of animal or vegetable origin, 100 mg/1 Oil and Grease of mineral or petroleum origin.
1.0 mg/1 phenolic compounds which cannot be removed by the Agency's wastewater treatment process.
-------
cn
TABLE A9. ANALYTICAL DATA FROM UF CONCENTRATION OF SAN LEANDRO PLANT EFFLUENT LESS
CPD STREAM, TEST #9
,
Sampling
Station
Sump
Feed After
Settling
UF
Concentrate
UF
Permeate
Date
3/3
3/4
3/7
3/11
3/7-3/11
3/3
3/4
3/7
3/11
3/7-3/11
3/3
3/4
3/7
3/11
3/7-3/11
3/3
3/4
3/7
3/11
3/7-3/11
Total
Fro on
Extractibles
2,320
228
119
7,160
2,610
816
577
97
667
736
6,670
14,900
7,130
10,200
8,490
149
132
234
195
953
Non-Polar
Freon
Extractibles
520
56.4
-
-
-
201
248
-
-
2,160
7,600
-
-
2.0
43.6
-
-
-
PH
8.4
8.7
6.9
8.3
7.7
8.5
8.6
8.2
8.3
8.3
9.0
9.0
9.1
8.1
8.3
8.6
8.6
9.2
8'. 3
8.5
Total
Solids
17,200
1 ,810
-
-
-
5,350
4,420
-
-
40,400
66,100
-
-
1,460
1,820
-
-
-
A
Total
5usix-niV,i
Si,i ids
1! ,)00
6(,b
594
57, ',00
11 ,200
3,170
2,190
242
2,510
2,370
24 ,900
54,700
95,400
69,500
49,000
12
12
28
24
16
,'-,',AY'-, (.n.|/e.)
nor;
,;J,-,j
4, MO
-
-
5,0/0
4,4'!0
5 , 1 :(0
_
6,740
6,150
7,700
12,500
4,700
5,290
-
-
3, MO
Soluble
\t!':?,
5 ,2 Ml
3 , 7HO
-
-
-
4,760
5,040
_
-
5,540
6,310
-
-
5.040
5.600
-
-
-
Soluble Phenolic
COO
',") ,!100
•~> ,H?0
-
-
26 ,«)()
18,900
17,300
_
16,400
95,900
143,000
-
114,000
10,600
12,900
-
-
11 ,400
COO Corvounds
18,000 5
5,;;:o 12
0.
1 .
10
13,!<00 22
12.500 7,
1
4,
6,
58,500 45
66,000 8.
r.
9.
6
10,500 5
12,400 0
6
0
0
,50
.0
.5
.3
.5
. 3
.0
.0
.0
.0
.6
.8
.5
Zinc
48
32
250
75
180
18
28
2
1.1
38
280
140
450
750
3SO
0.6
1.4
0.9
11.0
1.6
Total
Cyanide
.
-
2.5
3
3
.
„
5
5. 5
4.0
7.0
4.0
_
4
4.5
4.0
300/100
300 mg/8. oil and grease of animal or vegetable origin, 100 mq/t, oil and grease of ir.ir.eral or- petroleum origin.
1.0 mg/i phenolic compounds which cannot be removed by the Agency's wastewater treatment process.
3.0
-------
TABLE A10. ANALYTICAL DATA FROM UF CONCENTRATION OF SAN LEANDRO PLANT EFFLUENT LESS "Z"
STREAM, TEST #10
Sampl i ng
Station
Sump
Sump After
Settling
UF
Concentrate
UF
Permeate
(New Membranes)
UF
Permeate
(Original
Membranes)
San Leandro
Municipal Discharge
Date
4/29
4/26-4/29
5/6
5/2 -5/6
4/29
4/26-4/29
5/6
5/2 -5/6
4/29
4/26-4/29
5/6
5/2 -5/6
4/29
4/26-4/29
5/6
5/2 -5/6
4/29
4/26-4/29
5/6
5/2 -5/6
Limits
Total
Freon
Extractibles
(mg/A)
1,280
626
880
1,290
793
483
1,230
712
5,450
1,490
22,700
16,300
124
294
74
139
106
205
160
298
300/100*
pH
(units)
8.5
9.4
9.2
9.1
9.3
9.7
9.2
9.1
9.3
9.6
8.9
8.9
9.4
9.8
9.0
9.0
9.5
9.7
9.0
9.0
>6.0
Total
Solids
(mg/A)
3,700
8,000
7,890
8,180
14,300
14,100
9,240
10,100
86,600
44,200
160,000
124,000
7,810
7,000
2,960
5,100
9,700
8,370
3,240
5,550
Total
Suspended
Solids
(mg/A)
2,370
2,390
5,580
5,230
2,910
2,970
4,820
4,400
62,600
20,700
140,000
94,800
154
150
138
136
192
132
108
186
BOD
(mg/A)
.-
1,390
__
1,850
._
3,750
--
1,880
5,470
7,760
..
2,230
--
1 ,300
2,500
--
1,500
COD
(mg/A)
13,900
--
25,700
,.
24,400
--
21,100
86,100
340,000
._
10,900
—
6,570
- -
13,800
—
6,990
Phenolic
Compounds
(mg/A)
6.6
106
175
44
140
103
22
45
80
72
32
98
112
56
28
89
143
52
30
89
1.0+
Zinc
(mg/A)
47
65
39
55
no
100
55
50
2000
650
2800
2100
16
14
9.5
16
23
18
9.6
14
3.0
*300 mg/A oil and grease of animal or vegetable origin, 100 mg/A oil and grease of mineral or petroleum origin.
tl.O mg/A phenolic compounds which cannot be removed by the agency's wastewater treatment process.
-------
TABLE All. ANALYTICAL DATA ROM UF CONCENTRATION OF SAN LEANDRO PLANT EFFLUENT,
MAXIMUM CONCENTRATION TEST, TEST #12
oo
Sampling
Station
Sump
Feed
After
Settling
UF
Concentrate
UF
Permeate
(New Membranes)
UF
Permeate
(Original
Membranes)
San Leandro
Date
5/27
6/3
5/27-5/30
5/31-6/3
5/27
6/3
5/27-5/30
5/31-6/3
5/27
6/3
5/27-5/30
5/31-6/3
5/27
6/3
5/27-5/30
5/31-6/3
5/27
6/3
5/27-5/30
5/31-6/3
Municipal Discharge Limits
Total
Freon
Extracti bles
(mg/l)
1,320
1,530
1,400
2,600
301
1,150
328
526
587
3,810
856
1,190
63
116
73
571
65
124
69
208
300/100*
pH
(units)
8.2
8.9
9.1
8.7
8.8
8.5
8.9
8.6
8.5
8.8
8.9
8.9
8.1
8.4
8.7
8.4
8.1
8.3
8.7
8.4
>6.0
Total
Solids
(mg/S.)
19,500
38,100
29,200
49,100
4,320
10,300
4,750
9,110
16,700
28,900
10,800
22,600
2,500
3,820
3,320
4,560
3,270
3,820
3,360
4,450
-
Total
Suspended
Solids
(m<|M)
9,950
30, -100
27,700
47,200
.1,320
4,510
1,030
3,240
13,200
28,300
6,880
17,800
40
102
44
84
160 "
106
46
118
-
BOO
(m
-------
T/5
Week
Ending
1-07-76
1-14776
1-21-76
1-28-77
2-04776
2-11-76
2-18-76
2-25-76
3-03-76
3-10-76
3-17-76
3-24776
3-31-76
4707.76
4714776
4T21-76
4728-76
5-05-76
5-12-76
5-19-76
5-26-77
6-02-76
6-09-76
6-15-76
6-22-76
6-29-76
>BLE Bl. WE
m3
Per Week
143
157
298
287
304
498
548
479
471
433
206
487
493
436
93
360
713
419
497
454
353
432
357
193
305
321
IEKLY SUMMARY
Operating
Hours/Week
95
91
68 -
107
100
119
135
115
121
122
82
151
123
125
40
82
156
99
117
125
101
103
122
80
126
128
OF CHICAGO PLAN"
Throughput vs.
% of Design
44,2
50.7
128.6
78.6
89.0
122.8
119.0
122.0
114,0
104,0
73.7
94.6
117.4
102.2
68.3
128.9
134,0
124,1
124,5
106.6
102.4,
123.1
85.7
70.7
70.9
73.4.
F UF OPERATING
Cumulative
143
301
599
886
1,190
1,690
2,240
2,720
3,190
3,620
3,820
4,310
4,800
5,240
5,330
5,690
6,400
6,830
7,320
7,780
8,130
8,560
8,920
9,110
9,420
9,740
DATA DURING
% of Design
44,2
47.3
69.1
71.9
75.6
85.3
91.7
95.9
98.2
98.9
97.1
96.8
98.6
98.8
98.1
99.6
102.5
103.6
104.8
104,9
104.8
105.6
104.6
103.6
102.1
100.8
1976.
Oil & Grease
ppm
33
67
50
88
40
59
40
30
21
38
42
29
63
37
41
45
32
42
59
65
54
6
11
4T
63
292*
APPENDIX E
\s*
o
o
~o
1—
z
— ^
-n
o
— i
3=1
-------
TABLE Bl. (CONTINUED) WEEKLY SUMMARY OF CHICAGO PLANT UF OPERATING DATA DURING 1976
ro
o
Week
Ending
7-06-76
7-13-76
7-20-76
7-27-76
8-03-76
8-10-76
8-17-76
8-24-76
8-31-76
9-07-76
9-14T76
9-21-76
9-28-76
10-05-76
10-12-76
10-19-76
10-26-76
11-02-76
11-09-76
11-16-76
11-23-76
11-30-76
12-07-76
12-14776
12-21-76
12-28-76
12-21-76
m^ Operating Throughput vs.
Cumulative
Per Week Hours/Week % of Design m6
402
451
384
431
173
135
127
237
96
193
387
313
220
298
222
190
305
214
140
123
44.
System Down
254
244
System Down
System Down
135
107 110.2
120 110.3
120 93.8
94 134.5
69 73.4
68 58.0
62 60.0
119 58.5
65 43.1
59 96.0
137 92.9
132 69.6
101 63,9
137 63.8
117 55.7
118 47.3
127 70.4
111 56.7
103 39.9
105 34,3
41 31.4,
- Replaced one half of ul
113 65.9
127 56.4,
- Membranes Fouled
- Membranes Fouled
37 106.6
10,100
10,600
11,000
11,400
11,600
11,700
11,800
12,100
12,200
12,400
12,800
13,100
13,300
13,600
13,800
14,000
14,300
14,500
14,700
14,800
14,900
trafiltration
15,100
15,300
15,500
% of Design
101.1
101.5
101.2
102.1
101.5
100.7
99.9
98.6
97.6
97.6
97.0
96.1
95.3
94,3
93.3
92.1
91.5
90.6
89.5
88.4
87.9
Membranes
87.4
86.6
86.8
Oil & Grease
ppm
8
52
221*
48
58
N.A.**
N.A.
N.A.
N.A.
N.A.
54
64
130*
43
28
32
27
30
27
35
N.A.
23
48
30
* Raw material spillage
** Not A\
Bailable
- low molecular rosins
-------
TECHNICAL REPORT DATA
(Please read luiunictions on the reverse before completing}
REPORY NO.
EPA-600/2-78-176
3. RECIPIENT'S ACCESSIOf»NO.
TITLE AND SUBTITLE
Treatment of Wastewaters from Adhesives and Sealants
Manufacture by UHrafiltration
5. RFPORT DATE
August 1978 Issuing date
6. PERFORMING ORGANIZATION CODE
AUTHOR(S)
M.H.Kleper, R.L.Goldsmith,!,V.Iran (Walden Div. of Abcor
D.H.Steiner,J.Pecev-ich,M.A.Sakillaris (W.R. Grace)
8. PERFORMING ORGANIZATION REPORT NO.
. PERFORMING ORGANIZATION NAME AND ADDRESS
Walden Div. of Abcor,
850 Main Street
Wilmington, MA 01887
10. PROGRAM ELEMENT NO.
Inc. Dewey and Almy Chem, Div,
W. R, Grace and Co.
55 Hayden Avenue
Lexington, MA 02173
1BB610
11. CONTRACT/GRANT NO.
S804350010
2. SPONSORING AGENCY NAME AND ADDRESS
Industrial Environmental Research Lab. - Cinn, OH
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
13. TYPE OF REPORT-ANO-PERIOD COVERED
Task Final 3/75-7/77
14. SPONSORING AGENCY CODE
EPA/600/12
15. SUPPLEMENTARY NOTES
16. ABSTF
UTtrafnitration was proven to be a viable unit process for separating adhesives
and sealants manufacturing wastewaters into a low volume concentrate stream and a high
volume permeate stream. The UF permeate was characterized by the following average
contaminant loadings: 100 mg/£ total freon extractives, <7.4 mg/£ non-polar freon
extractives, < 27 mg/£ (typically <5 mg/£) suspended solids, 0.43 mg/£'free cyanide,
3.6 mg/£ total cyanide, 8,900 mg/£ BOD, 36,600 mg/£ COD, 44.6 mg/£ phenolic compounds
and 1.5 mg/£ zinc. A treated effluent of this quality is acceptable for discharge
under the San Leandro Municipal Discharge Limitations with the exception of the pheno-
lic compound and total cyanide loadings. Surcharges would be imposed, however, based
on the suspended solids and BOD loadings.
If significant levels of phenolic compounds and cyanide are not present in a
particular plant's wastewater discharge, ultrafiltration is judged capabls of meeting
local Municipal Discharge Standards. When phenolic compounds and cyanide are present
at significant levels either ozonation or reverse osmosis are considered the preferred
post -treatment processes. None of the treatment system options investigated is con-
sidered capable of reducing adhesives and sealants manufacturing plant wastewater BOD
and COD loadings to the recommended Effluent Limitations Guidelines.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
Adhesives
Latex
Sealers
b.lDENTIFIERS/OPEN ENDED TERMS
c. COSATI field/Group
Effluent Guidelines
Wastewater Treatment
Carbon Absorption
Reverse Osmosis
Ultrafiltration
68D
18. DISTRIBUTION STATEMENT
Release To Public
19. SECURITY CLASS (This
Unclassified
135
20. SECURITY CLASS (This page)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
ft U.S. GOVERNMENTPRINTING OFRCE--1978—757-140 / 1 458
121
------- |