United Ststw
Environmental Prottctien
Agency
Robert S. K«r Environmental Research
Laboratory
Ada OK 74820
EPA-SOO/2-79-1S4
August 1979
Research and Development
5BPA
Treatment of
Organic Chemical
Manufacturing
Waste water for
Reuse
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development. U.S. Environmental
Protect-on Agency, have been grouped mto nine series. These nine broad cate-
gor-es ,vere established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to 'ester technology transfer and a maximum interface in related fields.
The nine series are:
1 Environmental Healtn Effects Research
2 Environmenfal Protection Technology
3. Ecological Research
4. Environmental Monitoring
5 Socioeconomic Environmental Studies
6 Scienti'ic and Technical Assessment Reports (STAR)
7 ir'teragencv Energy-Environment Researcn and Development
3 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-600/2-79-184
August 1979
TREATMENT OF ORGANIC CHEMICAL MANUFACTURING WASTEWATER FOR REUSE
by
M. Scherm
P. M. Thomasson
L. C. Boone
L. S. Magelssen
Union Carbide Corporation
Chemicals and Plastics Division
Research and Development Department
South Charleston, West Virginia 25303
Grant No. S-801398
Project Officer
T. E. Short, Jr.
Source Management Branch
Robert S. Kerr Environmental Research Laboratory
Ada, Oklahoma 74820
ROBERT S. KERR ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
ADA, OKLAHOMA 74820
-------
DISCLAIMER
This report has been reviewed by the Robert S. Kerr Environ-
mental 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.
11
-------
FOREWORD
The Environmental Protection Agency was established to
coordinate administration of the major Federal programs designed
to protect the quality of our environment.
An important part of the Agency's effort involves the search
for information about environmental problems, management tech-
niques and new technologies through which optimum use of the
nation's land and water resources can be assured and the threat
pollution poses to the welfare of the American people can be
minimized.
EPA's Office of Research and Development conducts this
search through a nationwide network of research facilities.
As one of these facilities, the Robert S. Kerr Environmental
Research Laboratory is responsible for the management of programs
to: (a) investigate the nature, transport, fate and management of
pollutants in ground water; (b) develop and demonstrate methods
for treating wastewaters with soil and other natural systems;
(c) develop and demonstrate pollution control technologies for
irrigation return flows; (d) develop and demonstrate pollution
control technologies for animal production wastes; (e) develop
and demonstrate technologies to prevent, control, or abate pollu-
tion from the petroleum refining and petrochemical industries;
and (f) develop and demonstrate technologies to manage pollution
resulting from combinations of industrial wastewaters or indus-
trial/municipal wastewaters.
This report contributes to the knowledge essential if the
EPA is to meet the requirements of environmental laws that it
establish and enforce pollution control standards which are
reasonable, cost effective and provide adequate protection for
the American public.
W. C. Galegar
Director
Robert S. Kerr Environmental Research Laboratory
111
-------
ABSTRACT
This research program was initiated with the overall objec-
tives of demonstrating the quality of water produced after each
step in the treatment of the raw wastewater of an organic chemical
plant from a best state-of-the-art commercially available process
sequence, and determining the feasibility and economics of renova-
ting this organic chemical wastewater for reuse as boiler feed-
water or cycle cooling water.
A pilot facility was designed and constructed to renovate
5 gallons per minute (gpm) of biological effluent from the Union
Carbide Caribe Incorporated (UCCI) organic chemical manufacturing
plant in Puerto Rico, through best state-of-the-art, commercially
available processes for the removal of suspended solids, residual
organics and dissolved salts. The process sequence consisted of
sedimentation/mixed-media filtration, carbon adsorption, pressure
filtration, reverse osmosis, and ion-exchange and was followed by
a pilot-scale boiler designed to test the reusability of the
renovated wastewater as boiler feedwater at pressures, temperatures
and heat fluxes typical of full-scale manufacturing facilities.
A pilot-scale cooling tower and heat exchangers were designed and
operated to determine chemical treatment requirements for makeup
waters of varying quality taken from different points in the
treatment sequence.
After eight months of continuous operation, the renovation
sequence was evaluated in terms of reduction efficiency across
each process step. A maximum 67 percent water recovery was
achieved. From these pilot data, full-scale renovation facilities
were designed and costs were estimated for a similar sequence
treating 1500 gpm and 3000 gpm. Based upon a +25 percent esti-
mate, not including the cost of sludge or brine disposal, the
total annualized cost of wastewater renovation to boiler feedwater
quality at 67 percent water recovery was estimated to be $2.00/
cubic meter (m*) ($7.50/1000 gallons (gal)) product water in 1978.
The pilot boiler operated successfully at 180,000 BTU/square
foot-hour (ft^-hr), 1500 psig, and 750°F superheat temperature
with renovated wastewater when compared to operation with the
demineralized well-water presently used for feed to the full-scale
plant boilers. The steam condensate derived from renovated waste-
water was alightly more corrosive than that derived from the use
of demineralized well-water as boiler feedwater.
IV
-------
The cooling water test loop, during 30-day test runs,
utilized reverse osmosis permeate, activated carbon effluent,
sedimentation/filtration effluent, and clarified biological efflu-
ent as makeup. Corrosion and heat transfer characteristics with
these makeup waters were found satisfactory only with A-249 stain-
less steel, indicating that special metallurgy would be required
for the use of this renovated wastewater for cooling water.
This report was submitted in fulfillment of Grant No. S-801398
by Union Carbide Corporation under the partial sponsorship of the
U.S. Environmental Protection Agency. This report covers a period
from April 1, 1976, to December 1, 1976; work was completed on
December 1, 1976.
-------
CONTENTS
Foreword iii
Abstract iv
Figures ix
Tables xiii
Abbreviations xv
Acknowledgments xvi
1. Introduction 1
2. Conclusions 3
3. Recommendations . . - 5
4. Wastewater Renovation Facilities . . 6
Facilities Description 6
Biological treatment system 9
Physical-chemical treatment system ... 9
Demineralization treatment system .... n
Waste sludge and backwash handling ... 16
Operation of Experimental Facilities .... 18
Activated sludge 18
Sedimentation/filtration unit 20
Activated carbon adsorbers . 21
Multi-media filter , 21
Reverse osmosis unit 22
Primary deionizer 24
Secondary deionizer 24
Sampling and Analysis 25
Routine samples 25
Special samples 25
Results and Discussion 26
Treatment performance 26
Sedimentation/filtration unit 27
Activated carbon adsorption 27
Activated carbon regeneration 32
Multi-media filtration 37
Waste sludge and backwashes 42
Reverse osmosis 42
Ion-exchange 49
Demineralization waste brines ...... 53
Specific organic analysis 56
Investment Cost and Operating Expense for
Full-scale Facilities 60
Process description 62
Summary of economics 64
vii
-------
CONTENTS (continued)
5. Boiler Test Loop 67
General Facility Description ........ 67
Individual Equipment Description 73
Test Boiler Experimental Approach 77
Sampling and Analysis 78
Results and Discussion 78
6. Cycle Cooling Water Test Loop 88
General Facility Description 88
Cooling Water Test Experimental Approach . . 94
Sampling and Analysis 94
Results and Discussion 98
Corrosion data 98
Average chemical analyses and mass
balance 113
Heat transfer coefficients 116
Biological fouling 124
Summary of cooling water test
conclusions 124
Bibliography 126
Appendices
A. Specific organic analyses - sample concentration
and identification procedures 128
B. Weekly averaged data summaries 132
C. Sizing of full-scale facilities for investment
cost and operating expense estimates 136
D. Analytical methods—deposit analysis 141
E. Calculation of corrosion test coupon
penetration rate
V113
-------
FIGURES
Number Page
1 Block flow diagram for wastewater reuse
pilot plant 7
2 Pilot-plant wastewater renovation facilities
showing activated sludge in foreground and
tertiary facilities in background 8
3 Sedimentation/filtration unit 10
4 Activated carbon columns 12
5 Multi-media filter 13
6 Reverse osmosis unit 15
7 Ion-exchange columns showing weak-base anion
column and mixed bed column 17
8 Waste sludge and backwash handling system
showing waste collection tank with mixer9
and 12-ft. diameter clarifier 19
9 Sedimentation/filtration unit performance data . 29
10 Activated carbon column COD data 30
11 Activated carbon columns percent COD removal
based upon feed to and effluent from each
column 31
12 COD breakthrough curves - fraction of feed COD
remaining (C/Co) 33
13 Activated carbon column cumulative carbon
loadings based upon COD 34
14 Percent organic removal vs. dosage, reactivated
and virgin carbon 38
15 Freundlich isotherm - evaluation of reactivated
carbon from pilot col. #304 treating sedi-
mentation/filtration unit effluent (TOC) ... 39
ix
-------
FIGURES (continued)
Number Page
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
Freundlich isotherm - evaluation of reactivated
carbon from pilot col. #304 treating sedi-
mentation/filtration unit effluent (COD) . . .
Mixed-media pressure filter suspended solids
removal data
Consecutive HO batch runs with spiral-wound
PA membranes
Effect of pH on conductivity rejection
3
Conceptual layout, 5.7 m /min. (1500 gpm)
nominal flow case ...
Wastewater renovation plant schematic diagram . .
Boiler test-loop boiler section
Boiler test-loop condenser section
Boiler test-loop boiler feedwater section ....
Pilot boiler control panel
Backside of pilot boiler showing configuration:
steam drum, boiler drum and mud drum
Boiler steam condensate corrosion coupon test
loop
Test-boiler heater probe deposition
Cycle cooling-water test loop facilities ....
Forced-draft cooling tower and acid/caustic
addition facilities
Test heat exchanger .
Cooling-water test loop heat exchangers and
condensate collection tanks
40
41
47
48
61
63
68
69
70
71
72
76
82
90
91
92
93
-------
FIGURES (continued)
Number
33
34
35
36
37
38
39
40
Corrosion rates for A-214 carbon-steel test
coupons ..................
Run 1 cooling-water test, R.O. permeate with
Zn/POx inhibitor. Heat exchanger tubes
before cleaning .....
Run 1 cooling-water test, R.O. permeate with
Zn/PO4 inhibitor. Heat exchanger tubes
after scale removed
Run 2 cooling-water test, activated carbon
effluent with Cr/Zn inhibitor. Heat
exchanger tubes before cleaning
Run 2 cooling-water test, activated carbon
effluent with Cr/Zn inhibitor. Heat
exchanger tubes after scale removed . . .
Run 3 cooling-water test, activated carbon
effluent with Zn/PO4 inhibitor. Heat
exchanger tubes before cleaning
Run 3 cooling-water test, activated carbon
effluent with Zn/PO4 inhibitor. Heat
exchanger tubes after scale removed .* . ,
99
101
102
103
104
105
106
Run 4 cooling-water test, sedimentation/
filtration unit effluent with Zn/PO4
inhibitor. Heat exchanger tubes before
cleaning
107
41
Run 4 cooling-water test, sedimentation/
filtration unit effluent with Zn/PO4
inhibitor. Heat exchanger tubes after
scale removed
108
42
Run 5 cooling-water test, sedimentation/
filtration unit effluent with Cr/Zn
inhibitor. Heat exchanger tubes before
cleaning
109
XI
-------
FIGURES (continued)
Number Page
43 Run 5 cooling-water test, sedimentation/
filtration unit effluent with Cr/Zn
inhibitor. Heat exchanger tubes after
scale removed 110
44 Run 6 cooling-water test, biological effluent
no inhibitors. Heat exchanger tubes before
cleaning Ill
45 Run 6 cooling-water test, biological effluent
no inhibitors. Heat exchanger tubes after
scale removed 112
46 Water side heat transfer coefficients, run 1,
Zn/P04 inhibitor 118
47 Water side heat transfer coefficients, run 2,
Cr/Zn inhibitor 119
48 Water side heat transfer coefficients, run 3,
Zn/P04 inhibitor 120
49 Water side heat transfer coefficients, run 4,
Zn/PO4 inhibitor 121
50 Water side heat transfer coefficients, run 5,
Cr/Zn inhibitor 122
51 Water side heat transfer coefficients, run 6,
Zn/PO4 inhibitor 123
xii
-------
TABLES
Number Page
1 Average Operating Conditions for Activated
Sludge Pilot Facility 20
2 Cleaning Solutions for RO Membranes 23
3 Routine Wastewater Analysis Conducted on
24-Hour Composite Samples 25
4 Activated Sludge Effluent Summary, Weekly
Averages 28
5 Spent Carbon Reactivation Tests 35
6 Pore Size Distribution for Reactivated and
Virgin Carbon 36
7 Waste Sludge Characterization Summary,
Overall Averages 43
8 Reverse Osmosis Data Summary, Overall Averages. . . 44
9 Inorganic Analysis for Design of Demineralization
System, Multi-media Filter Effluent 50
10 Inorganic Analysis for Design of Demineralization
System, Reverse Osmosis Permeate 51
11 Ion-exchange Data Summary, Overall Averages .... 52
12 Ion-exchange Inorganic Analysis 54
13 Demineralization Waste Brine Characterization ... 55
14 Specific Compounds Analyses - Direct Inject .... 57
15 Additional Specific Compounds Detected in
Ponce Water Reuse Samples 59
16 Investment and Operating Cost Summary 65
Xlll
-------
TABLES (Continued)
Number Page
17 Total Costs of Wastewater Renovation
Facilities 66
18 Characteristics of Boiler Heating Elements 79
20 Steam Generator Probe Deposit Analysis 80
21 Boiler Steam and Return Condensate Analysis .... 84
22 Steam Superheater Probe Deposit 85
23 Boiler Steam Condensate Corrosion Studies 86
24 Chemical Treatment for Cooling Water
Pilot Tests 95
25 Cycle Water Test Loop Corrosion Data 100
26 Average Chemical Analysis, Cycle Cooling
Water Tests 114
27 Chemical Mass Balance, Calculated Relative
Chemical Precipitation in Cycle Cooling-water
Tests 115
28 Heat Transfer Information, Cycle Cooling-water
Tests 117
29 Summary of Acceptability of Makeup Waters
Tested with Various Metallurgies 125
B-l Waste Sludge Characterization 132
B-2 Reverse Osmosis Data Summary,
Weekly Averages 133
B-3 Ion-exchange Data Summary,
Weekly Averages 135
xiv
-------
ABBREVIATIONS
BFW
BOD
BTU
BV/hr.
°C
CA
CE
cm
COD
DWW
°F
ft.
F/M
gm.
gpd
gpm
in.
IWT
Kg
Kwh
1
Ibs.
1pm
m
mg/1
min.
MLSS
MLVSS
PA
psig
PVC
RO
RSS
RWff
SS
SQ. FT.
TOC
UCCI
jimho
UOP
boiler feedwater
biochemical oxygen demand
British thermal unit
bed volumes per hour
degree Centigrade
cellulose acetate
Combustion Engineering
centimeter
chemical oxygen demand
demineralized well-water
degree Fahrenheit
feet
food to microorganism ratio
gram
gallons per day
gallons per minute
inches
Illinois Water Treatment Co.
kilogram
kilowatt hour
liter
pounds
liters per minute
meter
milligrams per liter
minutes
mixed liquor suspended solids
mixed liquor volatile suspended solids
polyamide
pounds per square inch gage
polyvinyl chloride
reverse osmosis
recycle suspended solids
renovated wastewater
suspended solids
square feet
total organic carbon
Union Carbide Caribe, Inc.
microliter
micromho
Universal Oil Products
xv
-------
ACKNOWLEDGMENTS
The following Union Carbide personnel contributed signifi-
cantly to the conception, planning, and implementation of the
experimental program: B. J. Brymer, J. A. Fisher, 0. D. O'Bryan,
C. T. Lawson, R. Ocejo, M. Ramos, R. C. Vaseleski, V. Vega.
J. C. Boesch, W. Goff, J. Myers, and G. M. Whipple assisted in
the full-scale design and cost analysis.
The design and cost estimates for full-scale reverse osmosis
systems were prepared by M. Juar and T. Smith of Universal Oil
Products (UOP).
The contributions of Mr. J. Shook et al., of Betz Environ-
mental Engineers (BEE) in the design, implementation, and data
evaluation of the pilot boiler tests are gratefully acknowledged.
The authors thank Dr. P. E. Des Hosiers, Staff Engineer,
Mr. Leon Myers, Research Chemist, and Dr. T. E. Short, Project
Officer, of the Environmental Protection Agency, who contributed
to the planning of this joint research program.
xvi
-------
SECTION 1
INTRODUCTION
Effective conservation of our natural water resources is
high on the list of national objectives. Toward these objectives,
increasingly stringent wastewater effluent standards, have
recently been imposed. Indeed, the Federal Water Pollution Con-
trol Act Amendments of 1972 (PL 92-500) have set a goal of zero
discharge of pollutants into any navigable water by 1985. The
implementation of a zero discharge goal would make wastewater
reuse mandatory in the very near future. Unfortunately, existing
secondary wastewater treatment facilities in the organic chemi-
cals manufacturing industry are, in general, of the biological
type which produce effluents not meeting the water quality
criteria required of makeup water for most heat exchange systems;
nor is the product suitable feed for typical water treatment
facilities. This project was directed at investigating the
current technology existing in the best state-of-the-art commer-
cially available processes when applied toward the practical
achievement of zero discharge through recycle of product for
boiler feed or recycle cooling water makeup. The concept of
zero discharge of treatment residues was not within the project
scope.
Little information is available in industrial wastewater
treatment literature pertaining to the reuse of renovated waste-
water from large, multi-product organic chemical plants.
Specifically, there are no references to a process sequence of
commercially available technology to produce makeup for high
temperature heat transfer systems or similar high-quality water.
The Environmental Protection Agency's interest in developing
general information which defines the acceptable limits for
pollutants in reclaimed petrochemical wastewater when used as
cooling tower or boiler makeup, coupled with Union Carbide's
additional interest in wastewater recovery as a supplemental
source of raw water led to this jointly funded large-scale pilot
plant.
The pilot plant was designed and constructed to attain a
high rate of water recovery while continuously processing the
secondary effluent from a large organic chemical manufacturing
complex. The sequence of processing steps which had the highest
potential of producing high-quality boiler feedwater was
selected from commercially available technology. Reuse feasi-
bility was demonstrated in two carefully modeled heat transfer
-------
test loops. The cycle cooling water test loop was designed to
provide heat transfer fouling rates, to define effective chemical
treatment programs, and to evaluate the corrosion resistance of
the three most common metals of construction for heat exchangers.
Makeup water for the cooling tower from several sources within
the treatment process was evaluated. The boiler test loop, which
includes all of the appurtenances common to large, fairly
sophisticated boilers, was capable of providing all of the per-
formance data for steam generation over a wide range of pressure,
temperature, and heat flux.
The pilot program's primary objective was to demonstrate the
quality of water each step of the treatment can be expected to
produce from an organic chemical plant's secondary wastewater
treatment system and to determine the operating cost when this
water is renovated for reuse as boiler feedwater or cycle cooling
water makeup.
The wastewater reuse pilot plant was installed in Union
Carbide Caribe, Inc.'s organic chemical manufacturing plant near
Ponce, Puerto Rico, at a cost of $925,000. The experimental
program extended over an eight-month period (April 1976 through
November 1976) and incurred an operating cost of $800,000.
-------
SECTION 2
CONCLUSIONS
The total annualized cost of producing water of boiler
feedwater quality through a renovation sequence consisting of
reactor clarifiers, carbon adsorption, pressure filtration,
reverse osmosis and ion-exchange would be approximately $2.00/
m3 ($7.50/1,000 gal) in 1978, not including primary or secondary
treatment costs or facilities for the handling and disposal of
waste brines and sludges.
Carbon adsorption, including regeneration facilities, make
up approximately 35 percent of the total fixed investment and
greater than 30 percent of the total annual operating expenses.
Reverse osmosis facilities account for approximately 25 per-
cent of the total fixed investment and 30 percent of the annual
operating expense.
Waters of lesser quality than boiler feedwater could be
obtained at significantly reduced costs for use in low
pressure steam systems or as cooling water.
Any impurities that passed through the wastewater renovation
sequence did not noticeably affect boiler deposition.
The amount of waterside deposit within the boiler when
using renovated wastewater was less than that produced using
plant boiler feedwater (demineralized well-water).
The quality of steam produced from the renovated wastewater
was equivalent to that generated from demineralized well-water.
Superheater deposition using renovated wastewater was
equivalent to that produced using demineralized well-water.
Condensate derived from using renovated wastewater as
boiler feedwater appeared slightly more corrosive than condensate
derived from demineralized well-water.
When renovated wastewater was used as cycle cooling water
makeup, chromate treatment was effective in controlling the
corrosion of A-214 carbon steel; however, unacceptable heat
transfer characteristics resulted from the formation of scale.
-------
The use of Zu/PO. corrosion inhibitor resulted in satis-
factory heat transfer on carbon steel, but excessive corrosion
was observed.
Only A-249 stainless steel was effective in maintaining
satisfactory corrosion and heat transfer characteristics with
the makeup waters and treatments tested. Special metallurgy
would be required for the use of this renovated wastewater as
cooling water.
-------
SECTION 3
RECOMMENDATIONS
A high quality secondary effluent is necessary for effective
operation of the tertiary wastewater renovation facilities.
Since most secondary effluents in the organic chemicals industry
are of the biological type with residual organics and solids, it
is strongly recommended that every attempt be made to optimize
the biological system prior to attempting wastewater renovation
for reuse.
Only virgin activated carbon was used in the residual-organic
adsorption step of these wastewater renovation process studies.
Further adsorption tests with regenerated carbon are recommended
before a final judgment is made on the efficacy and economics of
this step of the renovation sequence.
The use of renovated wastewater as cycle cooling water
should be considered in water-short areas, provided existing
materials of construction in the heat exchange system are com-
patible or can be made compatible with the water. Stainless
steel, or similar alloys, were recommended in the present case
when the degree of renovation, beyond conventional biological
treatment, involved only the removal of suspended solids and the
addition of appropriate inhibitors and dispersants to the cycle
water. More extensive renovation may make admiralty brass an
acceptable construction material in some cases.
Because of the cost of renovation, the reuse of wastewater
for high pressure boiler feedwater, while technically feasible,
is not generally recommended. In cases where lesser qualities
of water are acceptable for the production of low pressure steam,
and the cost of the renovation sequence can be reduced, the re-
use of wastewater for boiler feedwater is a viable alternative
and is worthy of consideration.
-------
SECTION 4
WASTEWATER RENOVATION FACILITIES
FACILITIES DESCRIPTION
The pilot treatment facilities, designed to operate con-
tinuously, included an activated sludge unit; a physical-chemical
treatment system consisting of a suspended solids removal unit,
an activated carbon unit, and a multi-media filtration unit; a
demineralization system consisting of a reverse osmosis unit and
primary and secondary deionization units; and a system for the
collection and handling of waste sludge and backwash water for
maximum water recovery. The process sequences selected were
determined by bench-scale studies and literature survey as having
the highest potential for producing boiler feedwater quality
water from an organic chemical plant's process wastewater. The-
processes tested were considered the best state-of-the-art
commercially available. Process design was done by Union Carbide
Corporation's Environmental Engineering Group, with detailed
design assistance from The Austin Company.
Many of the treatment steps incorporated into the pilot
plant renovation processes were accomplished in commercially-
available package units. The package units set the design flow
rates throughout the treatment steps which were, consequently,
greater than the test loop requirements* Cooling water test
loop and boiler test loop makeup-water requirements were
approximately 4.78 liters/minute (1pm) (1 gpm) and 1.89 1pm (0.5 gpm)
respectively. The package units utilized were preassembled and
included the sedimentation/filtration unit; the multi-media
filter unit; the reverse osmosis unit; and the ion-exchange unit.
The pilot facilities were designed to operate continuously
on the raw wastewater from Union Carbide Caribe's manufacturing
units after primary clarification, pH adjustment, nutrient
addition and equalization. For those units requiring backwashing
or regeneration, intermediate storage capacity was provided to
accumulate effluents from these units. These water inventories
were arranged to provide backwashing, rinsing, and regenerant
dilution water for each respective unit, as well as to provide an
uninterrupted feed supply to the subsequent units and test loops.
The pilot plant process sequence is illustrated in the
attached block flow diagram (Figure 1) and a photograph of a
portion of the pilot facilities shown in Figure 2.
6
-------
NEUTRALIZED,
CLARIFIED,
EQUILIZED
BIO- ICLARI-
WASTEWATER 5IEACTOR ' FIER
*J |
FROM EXISTING I I , )
PRIMARY SYSTEM
r "
I
PHYSICAL/CHEMICAL TREATMENT SYSTEM
SUSPENDED SOLIDS
REMOVAL UNIT
I'
I [Primary I (Activated] Mined
jggffPff- (.*-. .FllUr'— "-
Settlert,
t tj_iU U
""I I I
Corbon j Media
Filter
Mumne
To Sewer
Water Recycle
Sludge, Rime
Water, And
Bockwa* Collec-
tion And Handling
Syetem
)EMINERALIZATIONI ggTEM
leveree I Primary Kecondor
ftmoeis peionizer bekmizer
Unit I Unit I Unit
Muloeel
WMk I Mixed
Boee I Bed
Anion , Ion
Cxohangv
:: I I 1
Brine And Spent Reagent
Stream* to Sewer
CYCLE WATER
TEST LOOP
Cooling 1 Heat
N9W9f lEWJfUHlQtf
t
~*
BOILER
Deo»fotor| Bol
+
I 1
1
Boiler | Super- Jconden-
| heater J »er
•+4-
Sludge to
Landfill
Figure 1. Block flow diagram for wastewater reuse pilot plant.
Blowdown
to Sewer
-------
00
Figure 2. Pilot-plant wastewater renovation facilities showing
activated sludge in foreground and tertiary facilities in background
-------
Biological Treatment System
Biological treatment consisted of a very conservatively
designed activated-sludge process comprised of a bioreactor,
aeration .equipment, and a clarifier (Figure 2). The equipment
was sized for 7.5-day detention time, an influent of 28.4 1pm
(7.5 gpm) (11,000 gpd (gallons per day)) of process wastewater,
and a mixed liquor volatile suspended solids (MLVSS) concentra-
tion of 2,500 milligrams per liter (mg/1).
The bio-treatment system consists of a plastic lined earthen
bioreactor and a 3.66-meter (m) (12-ft) diameter clarifier. Air
was diffused into the bottom of the bioreactor through headers
located under static mixers (POLCON) which provided mixing and
distributed oxygen for support of the biological activity. Be-
tween the bioreactor and the clarifier, wastewater flowed through
a level controlled weir box which maintained a constant flow to
an agitated vessel wherein coagulants could be added to enhance
settling of the biosolids. Sludge was recycled to the bioreactor
at a rate necessary to maintain the proper MLVSS concentration
and sludge age. Sludge wastage was measured daily by wasting in-
to a waste s,ludge receiving tank before discharge to the waste
sludge and backwash handling system. The clarified liquor flowed
through an intermediate tank into the physical-chemical treatment
system.
Physical-Chemical Treatment System
The physical-chemical treatment system consisted of a sus-
pended solids removal unit, an activated-carbon absorption unit,
and a multi-media filtration unit. The influent to the physical-
chemical treatment system was 18.9 to 26.5 1pm (5.0 to 7.0 gpm).
Sedimentation/Filtration—
The sedimentation/filtration unit removed suspended solids
by coagulation, sedimentation, and filtration.
This unit was a packaged commercial unit (Neptune Microfloc)
(Figure 3) consisting of an influent flow-splitting box, a rapid
mix tank, a slow-mixed flocculator tank, primary and secondary
tubular settlers, a mixed-media filter, a backwash water storage
tank, and four chemical feed tank assemblies complete with mixers
and metering pumps. Coagulating chemicals could be metered into
the rapid-mix tank along with a controlled flow of clarified bio-
effluent from the flow-splitter box. The effluent from the
flocculator flowed over a weir into the first-stage tube settler,
which contained settling tubes 5.08 centimeters (cm) (2 inches
(in)) in depth and 26.54 cm (24 in) in length. The settling tubes
were inclined at 60°. The settling tubes maintained a sludge
blanket which promoted coagulation and provided a source of
sludge for recycle to the rapid mix tank. Effluent from the
first stage flowed into the second-stage settling chamber for
-------
Figure 3. Sedimentation/filtration unit
10
-------
final clarification. This chamber contained modules composed of
settling tubes 2.54 cm (1 in) in depth by 99.06 cm (39 in) in
length, inclined slightly to facilitate drainage of the sludge
during the backwash cycle. The second-stage tube settler efflu-
ent flowed by gravity into the mixed-media filter which removed
most of the remaining suspended solids.
The filter and second-stage settlers were backwashed, manu-
ally or automatically, based on filter head loss. The backwash
water with the accumulated solids and the wasted sludge from the
first-stage settler was pumped to the waste collection tank in
the waste sludge and backwash handling system.
Activated Carbon Adsorption Unit—
Effluent from the sedimentation/filtration unit flowed into
the carbon adsorption system, which consisted of six 40-cm (16-in)
diameter columns which were 7.6 m (25 ft) high and which had a
4.9-m (16-ft) bed depth (Figure 4). Initial testing of various
combinations of parallel and "series flow revealed that the best
arrangement was a three beds-in-series in an expanded upflow mode.
This arrangement allowed near maximum organic (COD) adsorption.
The three carbon columns in series had an accumulated bed depth
of 14.6 m (48 ft) which when operated at 18.9 1pm (5 gpm) through-
put corresponds to a hydraulic loading of 159 lpm/m^ (3.9 gpm/ft^)
and a total contact time of 0.59 Bed Volumes (BV)/hour (100 minutes)
All column charges were made using virgin granular carbon—
Calgon Filtrasorb 300.
Multi-Media Filtration Unit—
A commercial multi-media filter unit, purchased from Illinois
W^ater Treatment Co. (IWT), was used to remove any carbon fines
from the activated carbon adsorption unit effluent (Figure 5).
This unit, which operated at about 10 psig, consisted of 35.6-cm
(14-in) diameter by 76.2-cm (30-in) deep bed composed of layers
of coal, sand, and garnet. After filtration, the carbon column
effluent entered the adsorber effluent tank which acted as a surge
tank to provide an uninterrupted feed flow to the demineralization
unit as well as backwash water for the activated carbon beds and
the multi-media filter. Spent backwash water was routed to the
waste sludge and backwash collection tank.
Demineralization System
The demineralization treatment system was designed to remove
soluble compounds from the wastewater and consisted of a reverse
osmosis unit, a primary deionizer unit, and a secondary deionizer
unit.
Reverse Osmosis Unit—
The reverse osmosis (RO) unit, purchased from Universal Oil
Products, was designed to operate automatically in a batch mode
11
-------
Figure 4. Activated carbon columns
12
-------
Figure 5. Multi-media filter,
13
-------
and was intended to demineralize up to 90 percent of the influent
water (Figure 6). The unit was also designed to accommodate
modules of hollow fiber, spiral wound or tubular membrane configu-
rations. The RO service pump transferred the filtered water from
the demineralization supply tank into the two-compartment RO feed
tank, from which a suction booster pumped the water through a
10-micron cartridge filter into the suction of the RO cycle pump.
This suction booster-filter step insured an adequate water supply
to the RO feed pump, thus preventing pump cavitation, plus it
protected the membranes from most solids that passed through the
multi-media filter. The RO system was equipped with in-line acid
addition, and pH, flow, temperature, and permeate conductivity
monitors. Automatic alarm and shutdown were provided for low
flow, high temperature, high pH, and high permeate conductivity.
Tubular and spiral wound modules were used during the test
period.
Tubular membrane configuration—The UOP OSMOTIK module uti-
lized a filament-wound fiberglass tube, each tube containing an
integrally formed 520 type cellulose acetate membrane on the inner
surface. The 18-tube module had a membrane surface of 16.7 ft .
These tubes, which were each 88 in in length, were contained with-
in a hexagonal shroud or module, and were arranged for series flow
by means of molded heads. A rod was positioned axially through"
each module to hold the assembly in place. A volume displace-
ment rod was installed in each tube to reduce solids deposition
and fouling of the membrane surface. The membrane unit consisted
of one stage of 52 modules arranged in 13 parallel rows of 4
modules in series per row. The wastewater passed through each
row of modules, being concentrated in each as it flowed back to
the RO feed tank for recycle until the batch was complete. The
permeate was collected in the shrouds and flowed into the RO unit
product tank.
Spiral-wound membrane configuration—Membranes of cellulose
acetate or poly(ether)amide, cast directly on a support sheet,
enclosed a product water channel material which was attached to
a perforated PVC permeate tube. The other three edges of this
laminate were sealed with a waterproof cement. A mesh spacer
which provided a uniform flow channel across the membrane was
placed on this laminate, and the assembly was wound in a spiral
around the perforated rod. Each module contained approximately
80 ft of membrane surface. Three such modules, equipped with
peripheral brine seals, were connected in series and slipped into
a 10.2-cm (4-in) phenolic-lined steel pipe, which acted as a
pressure vessel. The unit consisted of two of these assemblies
connected in series.
During operation, feedwater entered the upstream end of the
first vessel and flowed axially through the first module. Some
purified water flowed through the membrane, down the porous
14
-------
Figure 6. Reverse osmosis unit.
-------
backing and into the central permeate tube. Brine passed through
each module, in series, becoming more concentrated as permeate was
removed. The brine flowed back to the RO feed and recycled until
the batch was completed. The permeate from each module flowed
through the central tube and out of the pressure vessel. Perme-
ate from both pressure vessels combined to form the RO product
water.
Primary Deionizer--
The primary deionizer was a pre-assembled unit purchased
from Illinois Water Treatment Co. (Figure 7). The unit consisted
of a 41-cm (16-in) by 183-cm (72-in) resin container, complete
with all valves, piping, and controls for completely automatic
operation and regeneration. The resin bed contained 0.20 m^
(7.0 ft3) of weak-base anion resin intended to remove the residu-
al organics which might foul the resins in the following secondary
deionizer unit. The weak-base resin also removed the strong an-
ions. Effluent from the primary deionizer was accumulated in an
intermediate (fiberglass) storage tank which supplied feed for
the secondary deionizer as well as regeneration cycle rinse water
for the primary deionizer.
The resin bed was regenerated with a weak caustic solution
which was discharged to the sewer.
Secondary Deionizer--
The secondary deionizer was also a pre-assembled package
unit purchased from IWT and similarly equipped for automatic op-
eration and regeneration (Figure 8). The exchanger was a 20-in
by 72-in vessel containing 0.07 m^ (2.4 ft^) of strong-acid
cation resin and 0.14 m (4.8 ft3) of strong-base anion resin.
This mixed-bed treated 5.7 1pm (1.5 gpm) of water from the inter-
mediate storage tank for removal of the remaining total dissolved
salts. The demineralized water then flowed to a 7,570 1 (2,000
gal) fiberglass storage tank for use as boiler feedwater, dilu-
tion water, and deionizer regeneration cycle rinse water.
When the dissolved-salt concentration in the effluent from
this mixed-bed unit reached the maximum permissible conductivity,
the ion exchanger units were regenerated by an automated backwash,
regenerant, and rinse cycle, and then were air mixed prior to re-
turning the unit to service. The cation resin was regenerated
with sulfuric acid, and the anion resin was regenerated with
caustic simultaneously. The water from the backwash and rinse
cycles was also sewered.
Waste Sludge and Backwash Handling
This system consisted of a series of collection tanks and a
clarifier for handling the waste sludge and backwash streams from
the biological and physical-chemical treatment systems. Waste
16
-------
- 1
Figure 7. Ion-exchange columns showing weak-base anion column (A)
and mixed bed column (B). Fiberglass tank in background
(C) stored the renovated water to be fed to the boiler.
-------
sludge from the biosystem clarifier and the sedimentation/filtra-
tion unit, plus spent backwash water from the sedimentation/
filtration unit, carbon columns and the multi-media filter were
piped to a 9,463-1 (2,500-gal) tank where coagulants could be
added and blended. The wastewater then flowed by gravity to the
center well of a 3.7-m (12-ft) diameter clarifier (Figure 8).
Clarified effluent flowed by gravity from the clarifier to a
2,271-1 (600-gal) tank, from which, depending upon the quality,
the clarified water was recycled to the bioreactor, the micro-
solids removal unit or discharged to the sewer for ultimate dis-
posal. Sludge from the clarifier was transferred to a 379-1
(100-gal) tank for measurement before disposal.
OPERATION OF EXPERIMENTAL FACILITIES
The wastewater renovation pilot facilities were operated on
a continuous flow basis receiving pH adjusted, clarified, and
equalized wastewater from existing primary treatment facilities.
Primary treatment consisted of neutralization, clarification, and
24-hr equalization. Wastewater was pumped continuously from the
equalization basin discharge pump to the pilot plant at the rate
of 7.5 gpm.
The pilot activated sludge facility was constructed based
on a very conservative design to produce the best possible
secondary effluent available from a biological facility. Since
the pilot activated sludge facility was already considered best
available technology for BOD removal, it was not a part of the
studies covered under the grant. Its operation and performance
are, therefore, discussed only briefly. Primary emphasis of these
studies was on tertiary treatment technology.
The pilot facilities were operated, maintained, and super-
vised by UCCI plant personnel with technical assistance from a
research and development technician on site. Routine wastewater
analyses required for daily operational changes were performed
by the shift operators. All analytical analyses on routine
samples were conducted by plant laboratory personnel.
Activated Sludge
Average operating conditions over the entire study for the
activated sludge system are summarized in Table 1.
18
-------
Figure 8. Waste sludge and backwash handling system showing waste
collection tank with mixer (A) and 12-ft-diameter
clarifier (B).
-------
TABLE 1. AVERAGE OPERATING CONDITIONS FOR
ACTIVATED SLUDGE PILOT FACILITY
Reactor pH 8.15
Reactor temperature 31-34°C (88-93°F)
Retention time 7.58 days
Mixed liquor suspended solids (MLVSS) 2,140 mg/1
Mixed liquor volatile suspended 1,883 mg/1
solids (MLVSS)
Recycle suspended solids (RSS) 16,000 mg/1
o
Clarifier overflow rate 100 gpd/ft
Sludge age 55 days
F/M applied 0.18 Ib BOD/day/lb MLVSS
Basin dissolved oxygen 2.0 to 4.0 mg/1
As a routine procedure whenever an inlet total organic car-
bon (TOC) greater than 5,000 mg/1, or phenol greater than 200
mg/1 was indicated, flow was diverted to the panic pond. Normally,
these high peaks lasted only a short time and as the concentration
dropped below these levels, normal flow through equalization was
resumed. Wastewater collected in the panic pond was then slowly
bled back into the equalization basin.
Polyelectrolytes or coagulants were not added to the
secondary clarifier,•though facilities to do so were available.
The very conservative size of the clarifier and already good
settling characteristics of the mixed liquor did not warrant the
addition of these flocculant aids. Sludge was recycled and
wasted continuously from the cl-arifier underflow to maintain a
sludge age of 50 days.
Bio-treated and clarified wastewater was then used as feed
to the subsequent tertiary treatment facilities.
Sedimentation/Filtration Unit
This unit processed 22.7 to 26.5 1pm (6.0 to 7.0 gpm) of
clarified bio-effluent for the removal of suspended solids. Poly-
electrolytes and coagulants were not used throughout the study as
a result of lower solids loading than expected and satisfactory
clarification and filtration without the use of these chemicals.
20
-------
The unit received flow continuously. Solids were removed
intermittently from the bottom of the primary and secondary
settling tubes during filter backwash. The multi-media filter.
operating at a hydraulic loading of 142.5 lpm/m2 (3.5 gpm/ft2)
was backwashed, automatically or manually, when the filter
effluent pump suction exceeded 10 inches mercury vacuum. Back-
wash frequency was approximately once per shift at 45.4 1pm
(13 gpm) for 8 minutes.
Effluent from the sedimentation/filtration unit was
collected in the carbon adsorption feed tank. This tank had a
continuous overflow to insure variable feed to the carbon columns.
Activated Carbon Adsorbers
Three beds in series were operated in an expanded upflow
mode. The adsorbers received flow continuously at 18.9 lpm/m2
(5 gpm/ft2). Carbon contact time through the three-bed sequence
was 100 minutes (0.59 BV/hr overall).
The operating sequence for this three-bed series adsorption
system was as follows: Three beds were placed on-line in series
operation after being charged with virgin carbon. Feed and
effluent COD for each column were monitored closely on 24-hour
composite samples. At the point when the effluent from the
last column in series exceeded some maximum acceptable level
(arbitrarily chosen in this case to be the 60 percent COD
rejection level), the lead column was taken off line for regen-
eration—the second and third columns were advanced to the lead
and second position, and a fresh column, charged with virgin
carbon,was placed in the terminal position. These columns were
kept on-line until the minimum acceptable COD removal (60 percent)
was again reached. At this time the lead column was removed, and
the process was repeated. A total of 6 carbon beds were exhausted
in these studies.
Throughout these studies virgin carbon was used for all
column charges. Exhausted carbon was regenerated in vendor's
laboratories and analyzed for comparison to virgin carbon.
When the pressure drop across any column exceeded 20 psig,
the column was backwashed for about 15 minutes at 45.4 1pm
(12 gpm) to remove the accumulated solids. Backwash water was
supplied from the adsorber effluent tank that stored effluent
from the multi-media pressure filter following the carbon
adsorption unit. Spent backwash water was piped to the waste
sludge and backwash collection tank.
Multi-Media Filter
The multi-media filter bed received 18.9 1pm (5 gpm)
flow directly from the last carbon adsorber in series, entering
21
-------
the top of the bed at 10 to 15 psig. When the pressure drop
across the bed exceeded 8 psig, the operator placed the filter
in backwash for 10 to 15 min at 34.1 1pm (9 gpm) to expand the bed
and remove the accumulated solids. Spent backwash was piped to
the waste sludge and backwash collection tank. Backwash was done
manually, approximately once per shift, and the unit was placed
immediately back in service.
Reverse Osmosis Unit
Cellulose acetate (CA) membranes in the tubular configuration,
CA and polyamide (PA) membranes in spiral-wound configuration were
each operated at one point in these studies. Operating conditions
varied somewhat with each membrane used.
Tubular Cellulose Acetate Membrane—
Initial operation of the RO system utilized the tubular con-
figuration with cellulose acetate membranes. These modules oper-
ated up to 600 psig inlet pressure with a pressure drop of 200 to
250 psig across the membrane surface. An 87.1 1pm (23 gpm) feed
rate resulted in 30.1 1pm (8 gpm) of permeate flow; the remaining
56.8 1pm (15 gpm) concentrate flow was recycled back to the feed
tank. Maximum allowable feed temperature set by the manufacturer
was 43.3°C (110°F). Feed pH was maintained between 4.0 and 5.0
by adjustment of acid feed pumping rate.
When permeate flow began to drop at 600 psig inlet pressure,
the membranes were flushed with demineralized water. When this
failed to restore permeate flow, the system was flushed with one
of two recommended cleaning solutions (Table 2) and rinsed with
demineralized water prior to being placed back in service. When
the unit was inoperative for short periods of time the modules
were flushed with demineralized water adjusted to a pH of 4.0 to
6.0. When inoperative for more than 100 hours, a solution contain-
ing 0.5 percent formaldehyde was maintained in the modules.
RO permeate was collected in a covered tank which had a
continuous overflow.
Spiral-wound Cellulose Acetate Membranes--
These modules were operated under conditions similar to the
tubular membrane configuration, except maximum inlet pressure was
reduced to 500 psig with an 80-psig pressure drop across the mem-
brane. Total flow through the unit was reduced to 22.7 1pm (6 gpm)
permeate and 22.7 1pm (6 gpm) concentrate return. The same pH
adjustment and membrane cleaning procedures were followed as with
the tubular configuration. Shell-in-tube heat exchangers were
added to the concentrate return in an attempt to keep the temper-
ature below 43.3°C (110°F).
22
-------
TABLE 2. CLEANING SOLUTIONS FOR RO MEMBRANES
Solution A - Iron or Heavy Metal Fouling
Deionized Water 45 gal
Citric Acid 7.5 Ib
Triton X-100 (Rohm and Haas) 175 ml
Carboxy Methyl Cellulose 2.5 gm
NH4OH to Adjust to pH 3.0
Formaldehyde (37% formalin) 2,365 ml (Optional)
Solution B - Organic Fouling
Deionized Water 45 gal
Sodium Tripolyphosphate 7.5 Ib
Sodium EDTA (powder) 3 Ib
Carboxy Methyl Cellulose 2.5 gm
Triton X-100 (Rohm and Haas) 175 ml
H2SO4 to adjust pH to 6.0 (4160 HR elements)
10.0 (polyamide elements)
Formaldehyde (37% formalin) 2,365 ml (Optional)
23
-------
Spiral-wound Polyamide Membranes—
The module configuration was the same as the CA spiral-wound
modules with the same operating flows and pressures. Polyamide
membranes, with a maximum allowable temperature of 57.2°C (135°F),
operated safely at 48.9°C (120°F). The pH tolerance of this mem-
brane (pH of 3.0 to 9.0) enabled operation with or without pH ad-
justment .
Primary Deionizer
The primary deionizer was operated intermittently to produce
water free of any residual organics for the secondary deionizer.
Water from the RO product tank was pumped through the bed at 18.9
1pm (5 gpm) and into the fiberglass intermediate tank. The resin
bed was regenerated at the same time as the secondary deionizer.
Regeneration involved: 15-min backwash at 26.5 1pm (7 gpm), addi-
tion of 4 percent caustic solution, slow rinse at 11.4 1pm (3 gpm)
and fast rinse at 22.8 1pm (6 gpm). Final rinse was complete
when the rinse water effluent reached a pH of 9.5 or less. The
resin bed was then ready for service. All flows during regenera-
tion were sewered.
Secondary Deionizer
The secondary deionizer was operated intermittently to pro-
duce deionized hardness-free boiler feedwater. When in service,
a 56.8 1pm (15 gpm) flow was passed through the mixed bed with
51.1 1pm (13.5 gpm) recycled back to the intermediate tank, and
5.7 1pm (1.5 gpm) of effluent was passed to the boiler feedwater
storage tank. The resin bed was considered exhausted when the
effluent exceeded 1.0 iimho conductivity. At this level, an
alarm sounded, and all effluent automatically diverted to the
sewer.
Regeneration of the mixed bed was done manually and involved
several steps:
1. Slowdown to lower water level for backwash.
2. 15-min backwash to separate anion and cation resins.
3. 2 percent H2SO4 fed for 15 min, followed by H2SO4 for
15 min to convert cation resin to hydrogen cycle.
4. 4 percent caustic solution fed to convert anion resin.
5. 18.9 1pm (5 gpm) slow rinse for 15 min.
6. 47.3 1pm (12.5 gpm) fast rinse for 20 min.
7. Air-water mix to blend the resins.
8. Air drain to settle the mixed resins.
9. Final rinse until resin bed is the desired purity.
24
-------
All caustic, acid and water utilized in regeneration were dis-
charged to the sewer.
SAMPLING AND ANALYSIS
Routine Samples
Grab samples of feed and effluent from each of the wastewater
renovation unit operations were taken every 4 hours and combined
to form a 24-hr composite. Samples were kept refrigerated and
sent to the plant laboratory each morning for analysis. The waste-
water analyses listed in Table 3 were conducted according to
Methods for Chemical Analysis of Water and Wastes (1) and Standard
Methods for the Examination of Water and Wastewater (2).
TABLE 3. ROUTINE WASTEWATER ANALYSES CONDUCTED
ON 24-HOUR COMPOSITE SAMPLES
Biochemical Oxygen Demand (BOD-) pH
Chemical Oxygen Demand (COD) Specific Conductance
Total Organic Carbon (TOC) Total Dissolved Solids (TDS)
Volatile Suspended Solids Iron
Phenol Silica
Chloride
Not all analyses were run on every sample. For example, the
emphasis on the filters was suspended solids; on the carbon
columns the emphasis was COD; and the emphasis on demineralization
units was conductivity. Analyses performed on specific samples
are shown in the tables referred to in the section RESULTS AND
DISCUSSION.
Special Samples
Waste secondary sludge and backwash waters from the sedimentation/
filtration unit, carbon columns and mixed-media filter were
sampled periodically and analyzed for solids and heavy metals
content. Grab samples were taken when the units were considered
in normal operation. Samples for COD and suspended solids were
collected in glass bottles; samples for metals analyses were
collected in polyethylene bottles and fixed with 5 ml of concen-
trated nitric acid per liter of sample. Metals analyses were
conducted by atomic adsorption and flame UV, after digestion of
25
-------
the samples and after returning the samples to original volumes.
Analyses of the original samples are expressed as mg/1.
Demineralization System—
Grab samples of the RO feed and permeate and ion exchange
columns feeds and effluents were taken periodically for additional
analyses. Samples were collected in polyethylene bottles and
analyzed for specific anions and cations. These analyses were
not critical to the operation of the demineralization facilities
but were needed for characterization of the wastewater.
Waste Brines—
Reverse osmosis brine and the regeneration brines from the
ion-exchangers were sampled and analyzed periodically in the same
manner as the waste sludges were for solids and heavy metals con-
tent. The RO brine was sampled after a batch had finished being
concentrated and the brine was dumped to the sewer. The ion-ex-
change regeneration brines were sampled several times during each
cycle in the regeneration sequence and composited according to
flow to obtain a sample representative of the total brine to be
discharged.
Specific Organics—
A 24-hr composite sample was taken of the effluent from the
equalization basin, sedimentation/filtration effluent, carbon
adsorption, RO and ion-exchange effluents, for gas chromatographic
analysis for specific organics. The samples were shipped to
research and development laboratories in South Charleston, West
Virginia, for analysis. The analytical procedures employed were
among those generally accepted for the analysis of specific com-
pounds in wastewater; and, in the case of the concentrating
procedure, it was that procedure recommended by EPA (3,4). These
procedures are detailed in Appendix A.
RESULTS AND DISCUSSION
Treatment Performance
As previously stated, the pilot activated sludge basin and
clarifier were constructed only to produce high quality secondary
effluent. As such, they are not a part of these studies covered
by the EPA grant and the operation and performance of these units
will not be discussed in detail.
The pilot activated sludge system was a very conservative
design—long retention time, low clarifier overflow rate, and
ample oxygen supply to meet any conceivable influent demand. The
system was carefully operated to exclude spills, upsets in
nutrient supply, and surges in hydraulic flow-rate. The constant
semitropical ambient temperature and the inherently high amena-
bility of the raw wastewater from this plant to bio-treatment
favored a high level of performance in the pilot activated sludge
26
-------
system. The biological effluent average BOD, COD, and suspended
solids levels for the study period were 31 mg/1, 184 mg/1, and
44 mg/1, respectively (Table 4). The consistent achievement of a
high degree of biological treatment was critical to the successful
performance of the downstream tertiary treatment systems in the
actual water renovation pilot plant. Excessive variability in the
bio-effluent could have led to erratic performance of the suspended
solids removal system, rapid breakthrough in the carbon adsorbers,
excessive filter backwash requirements, and possible fouling in
the reverse osmosis-ion exchange units. In fact, these potential
problems were largely absent due to the consistent performance of
the pilot activated sludge system.
Sedimentation/Filtration Unit—
The sedimentation/filtration unit averaged approximately 70
percent suspended solids removal decreasing the suspended solids
from an average 45 mg/1 to 13 mg/1 (Figure 9). Some COD reduc-
tion 20 to 25 percent, was observed across the unit. COD measure-
ments were made on unfiltered samples and, therefore, the observed
reduction is throught to be primarily insoluble COD—mostly in the
form of bacterial solids.
Suspended solids levels generally experienced in secondary
effluents from organic chemical wastewaters are on the order of
100 to 200 mg/1 and often exhibit poor settling and filtering
characteristics. The low solids levels and the ease with which
these solids were removed make an evaluation of this unit diffi-
cult. Effluent from the sedimentation/filtration unit was fed
directly to the activated carbon columns.
Activated Carbon Adsorption—
The variability in COD load to the carbon columns was
relatively small due to the equalizing effect of primary and
secondary treatment steps. Feed to the carbon columns averaged
140 mg/1 COD. The final effluent from the three beds in series
operation was consistently below 50 mg/1 total COD (Figure 10).
An average 71 percent COD reduction was achieved with virgin
carbon (Figure 11). The lead bed in the series was exposed to
the more readily adsorbed organics and, therefore, removed a
higher percentage of the organics than the two trailing beds,
except when the lead bed approached exhaustion at which point
the second bed in series showed a higher percentage COD removal.
27
-------
TABLE 4. ACTIVATED SLUDGE EFFLUENT SUMMARY
WEEKLY AVERAGES
PILOT PLANT BIOLOGICAL EFFLUENT
Date
4/3-9
10-16
17-23
24-30
5/1-7
8-14
15-21
22-28
29-6/4
5-11
12-18
19-25
26-7/2
3-9
10-16
17-23
24-30
7/31-8/4
5-11
12-18
19-25
26-9/1
2-8
9-15
16-22
23-29
30-10/6
7-13
14-20
21-27
27-11/3
4-10
11-17
18-24
25-12/1
2-8
9-15
16-18
AVG.
BOD
mg/1
22
18
18
21
16
54
16
12
19
32
76
18
62
48
15
6
—
17
65
25
34
50
34
16
23
-
19
24
12
82
78
10
36
23
24
75
10
—
31
COD
mg/1
-
-
—
80
90
121
127
100
100
249
119
128
170
101
104
149
243
134
147
149
158
147
151
143
139
147
109
124
140
149
145
228
228
246
340
205
196
184
TOC
mg/1
134
78
50
86
90
127
64
78
65
39
105
99
-
—
-
-
—
-
-
—
—
—
—
—
-
—
-
-
—
-
-
-
-
-
-
-
-
—
-
SS
mg/1
61
50
29
29
46
47
43
30
21
23
60
48
35
47
23
30
28
72
62
66
68
63
43
43
46
43
40
24
29
25
27
36
72
73
81
91
64
58
44
Phenol *• l^
mg/1
—
_
_
0
2.4
0
0.9
2.3
0
0
0
0
0
0.7
1.2
0
0
0
0.25
0.26
0
0.75
0
0
0
0.3
0
0
0
0
0
0
0
0
0
0
0.27
'Less than detectable limit of test (0.1 mg/1) is recorded as 0,
28
-------
fro
CO
300
_,275
£ 250
a a5
§ 200
I 175
1 15°
g 125
I 10°
^ 75
S 50
S 25
0
SECONDARY CLARIFIER EFFLUENT
SEDIKENTATION/FILTRATIffJ UNIT
EFFLUENT
AVG. - 186
3-9 17-23 1-7 15-21 29-1 12-1828-2 10-16 21-30 5-11 19-25 2-8 16-22 30-6 11-20 28-3 11-17 25-30
10-16 21-30 8-11 22-28 5-11 19-25 3-9 17-23 31-1 12-18 26-1 9-15 23-29 7-13 21-27 1-10 18-21
APRIL
MAY
JUNE JULY AUGUST
DATE, WEEKLY AVERAGES
SEPTEMBER
OCTOBER
NOVEMBER
Figure 9. Sedimentation/Filtration unit performance data.
-------
250
200
_j
V.
56
§ 150
5
50
, FL-CD TO CARBON COLUMNS
• LEAD BED
o SECOND IN SERIES
* THIRD IN SERIES
1 8-111 22-28' 5-11 ' 19-25' 3-9 ' 17-23'31-1 ' 12-18!26-1 ' 9-15'23-29' 7-13121-27 1-10 18-21 1-6
1-7 15-21 29-1 12-18 26-2 10-16 21-30 5-11 19-25 2-8 16-22 30-6 11-20 28-3 11-17 25-30 7-13
MAY JUNE JULY AUGUST SEPTEMBER OCTOBER NOVEMBER DECEMBER
DATE, WEEKLY AVERAGES
Figure 10. Activated carbon column COD data.
-------
AVG, = 71 PERCENT
OVERALL REMOVAL
LEAD BED
SECOND IN SERIES
* THIRD IN SERIES
1-7 ' 15-21' 29-1 ' 12-18 ' 26-2 ' 10-16 ' 21-30' 5-11 ' 19-25' 2-8 ' 16-22 ' 30-6 ' 11-20 ' 28-3 ' 11-17 ' 25-30
8-11 22-28 5-11 19-25 3-9 17-23 31-1 12-18 26-1 9-15 23-29 7-13 21-27 4-10 18-21 1-6
MAY JUNE JULY AUGUST SEPTEMBER OCTOBER NOVEMBER DECEMBER
DATE, WEEKLY AVERAGES
Figure 11. Activated carbon columns percent COD removal
based upon feed to and effluent from each column.
-------
A total of five carbon beds containing virgin granular car-
bon were exhausted at an average loading of 0.352 gram (gin)
(pound (lb)) of COD per gm (Ib) of carbon. A sixth bed, taken off
line at the end of the studies, approximately one week before ex-
haustion, was loaded to 0.275 gm (lb) of COD per gm (lb) of carbon.
Breakthrough curves plotted as the fraction of feed COD remaining
in the effluent from each column, along with the final carbon load-
ings are shown in Figure 12. Cumulative carbon loadings (Figure 13)
as well as the breakthrough curves (Figure 12) indicate that the
carbon adsorption unit operated classically with smooth break-
throughs, regular time periods between column exhaustions, and
consistent carbon loadings.
Activated Carbon Regeneration--
Carbon column #304 was operated continuously as the lead bed
in the series adsorption system from April 14 to June 25 (COD data
were not taken in April). During this period of operation, the
COD removal efficiency of the carbon decreased from 57 percent to
19 percent, and the carbon adsorbed 0.311 gm (lb) COD/gm (lb)
carbon. A sample of the exhausted carbon was removed from column
#304, air-dried, and approximately 11.3 kilograms (kg) (25 lb)
were submitted to the Calgon Corporation Laboratory in Pittsburgh
for reactivation tests.
After reactivating a small aliquot in a laboratory furnace,
the bulk of the sample was reactivated in a pilot-scale rotary
kiln. The reactivation restored 98 percent of the carbon's
iodine number, 92 percent of the molasses number, and increased
the ash content by 62.6 percent (from<^-8 percent to 13 percent),
compared to minimum specifications for virgin Calgon carbon
(Table 5). While the carbon losses on reactivation in the rotary
kiln were quite attractive, 4 percent vs. 5 to 15 percent losses
reported in the literature, this parameter cannot be extrapolated
from one furnace to another.
As an additional check on the behavior of the carbon during
reactivation, the returned reactivated carbon was analyzed for
pore size distribution using the Digisorb 2500 analyzer. Results
shown in Table 6, compare to a similar analysis for virgin
Filtrasorb 400 (similar to, but not necessarily identical to
Filtrasorb 300). Thermal regeneration reduced the carbon B.E.T.
surface area by f^-25 percent, decreasing the surface area in
pores <5Q angstroms in diameter and increasing the surface area
in larger pores. Loss of surface area resulted from burnout of
interstitial carbon and from accumulation of ash that choked off
smaller pores.
32
-------
o
o
o
.0-
.•9.
.8.
§
O
§
H
•CARBON LOADING.
IBS COD/LB CARBON
0.310*
0.311
4-
1-7 15-21'29-1 '12-18'26-2 ' 10-16 21-30 5-11' 19-25 2-8 16-22 30-6 11-20 28-3 '11-17 25-30
a-11 22-28 5-11 19-25 3-9 17-23 31-1 12-18 26-1 9-15 23-29 7-13 21-27 1-10 18-21 1-6
HAY JUNE JULY AUGUST SEPTEMBER OCTOBER NOVEMBER DECEMBER
DATE, WEEKLY AVERAGES
Figure 12. COD breakthrough curves - fraction of feed COD remaining (C/Co).
-------
.5,
< .4
U
Q
8
.1
1 I I I
100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500
CUMULATIVE GALLONS TREATED (M GALS.)
Figure 13. Activated carbon column cumulative carbon loadings based upon COD.
-------
tn
TABLE 5. SPENT CARBON REACTIVATION TESTS
SPENT
Apparent Density, Air Dried 0.670
Apparent Density, Dried 150 °C 0.608
Apparent Density, Reactivated
Iodine Number 581
CC14 Number 23.8
Molasses Number 231
Ash (%) 8.90
Time (Minutes)
Temperature
PILOT- SCALE
LABORATORY ROTARY KILN
REACTIVATED REACTIVATED
mm
-
0.478
845
47.9
324
11.50
.1
—
-
0.456
886
51.3
367
13.01
35
1700° - 1800 °F
TYPICAL
VIRGIN/-,-,
CARBON' '
-
-
0.48
900-975
/s/400
5-8.0
-
(1)
Based on telephone conversation with Mr. Talli at Calgon.
-------
TABLE 6. PORE SIZE DISTRIBUTION FOR REACTIVATED
AND VIRGIN CARBON
(Digisorb 2500 Analyzer)
Pore Diameter
Angstroms
< 20 A
20-30
30-40
40-50
50-60
60-80
80-100
100-150
150-200
200-300
300-400
400-500
500-600
Total B.E.T.
Surface Area
2
Surface Area, m /gin,
Reactivated
533.508
133.226
39.426
19.344
11.746
14.193
8.454
12.650
6.194
5.754
2.785
1.282
0.718
789.278
in Pores of Given Size
Virgin
758.803
198.992
44.703
19.417
9.941
10.144
4.900
5.707
2.644
2.682
1.359
0.593
0.436
1060.390
36
-------
The Freundlich isotherm technique was used to compare the
performance of reactivated carbon to virgin carbon in batch
adsorption tests using pilot-plant bio-effluent after the sedi-
mentation/filtration step. Both carbons were pulverized prior
to the test. Carbon dosages from 0 to 20 gm/1 were contacted
with wastewater aliquots for two hours, filtered, and the
filtrates analyzed for TOC and COD. Percentage removals vs.
carbon dose are shown in Figure 14. Both carbons performed
about equally well. The decrease in percentage COD removal at
the higher carbon dosages probably resulted from the indicated
pH increase induced by the carbon.
Figures 15 and 16 show Freundlich isotherm plots for the
batch adsorption tests. Again, essentially equivalent perfor-
mance is indicated for both virgin and reactivated carbon. The
slope changes in the graphs probably reflect the aforementioned
effect of increasing wastewater pH with increasing carbon dose.
No loss of adsorptive capacity occurred in the carbon during
the exhaustion/reactivation cycle, as indicated by the isotherm
values of (X/M)c :
mg/gm Carbon
Reactivated Virgin
TOC 185 210
COD 1090 795
The apparent differences result from fitting a line to the data
points by eye.
It was concluded that the sample of "exhausted" carbon
from the pilot-plant adsorber was restored to near-virgin con-
dition by thermal oxidative regeneration in a rotary kiln at
Calgon's Pittsburgh Laboratory. Iodine number, molasses number,
and the carbon's performance in batch isotherm tests all indi-
cated this high degree of reactivation.
The furnace operating conditions and observed carbon
losses (4% by volume) cannot be directly extrapolated to larger-
scale furnace systems. However, this study does indicate that
spent carbon from the pilot-plant can be reactivated to near-
original adsorptive properties.
Multi-media Filtration—
Effluent from the last carbon column in series flowed
directly into the multi-media filter bed. Feed suspended solids
numbers are based upon samples taken of the sedimentation/
filtration unit effluent and, therefore, do not include any
biological solids or carbon fines from the carbon adsorber beds.
Feed suspended solids numbers ranged from 5 to 35 mg/1 and
averaged 13 mg/1. Effluent from the filter ranged from NIL to
20 mg/1 and averaged 7 mg/1, approximately 50 percent solids
removal (Figure 17). Actual removals were probably considerably
37
-------
Virgin Carbon
rt
o
e
a>
a
o
d
O
Q
O
O
Virgin Carbon
8 6 pH After
Carbon
Treatment
Reactivated
Virgin Carbon
46 8 10 12 14 16 18 20
Carbon Dose, gro/1
Figure 14. Percent organic removal vs. dosage.
Reactivated vs. virgin caTgon filtrasorb carbon
sedimentation/filtration effluent: TOC = 35 mg/1
COD = 190 mg/1
38
-------
100
§10
a
o
I
u
bD
S
10
- Reactivated Calgon^
• Filtralorb
X Virgin Calgon Flltra-
sorb 400
(X/M)C
Virgin 210 mg TOC/gm
Reactivated 185 mg TOC/gm
\ \
10
Co
-M-L
1 1 I
100
Cg mg TOC/1
Figure 15. Freundlich isotherm - evaluation of reactivated carbon from
pilot col. #304 treating sedimentation/filtration unit
effluent.
39
-------
1000
o
43
h
ct
0
6
Q
8
bo
S
X
100
• Reactivated Carbon
ji Virgin Carbon
(X/M)C,
10
Virgin
Reactivated
795 rag COD/gm
1090 mg COD/gm
i I I t I I I
no r~
l I I i i I I
J tilt
100
1000
Ce «ng COD/1
Figure 16. Freundlich isotherm - evaluation of reactivated carbon from pilot
col. #304 treating sedimentation/filtration unit effluent.
40
-------
34.0
32.0
30.0
28.0
26.0
24.0
22.0
20.0
18.0
16.0
14.0
12.0
10.0
8.0
6.0
4.0
2.0
0
o FILTER INFLUENT
, FILTER EFFLUENT
T
T
-1—•—I—
11-17 25-30
4-10 18-24
HQVENBER
'
17-23' 1-7 ' 15-21 '29-4 '12-18 '26-2 '10-16
1 24-30'
5-11' 19-25 '
10-16 2M-30 8-14 22-28 5-11 19-25 3-9 17-23 31-4 12-18
APRIL NAY JUNE JULY AUGUST
DATE, WEEKLY AVERAGES
2-8 ' 16-22' 30-6 '14-20'28-3
26-1 9-15 23-29 7-13 21-27
SEPTEMBER OCTOBER
Figure 17. Mixed-media pressure filter suspended solids removal data,
-------
higher than this due to the suspended solids that may have been
produced in the carbon beds but not measured in the feed.
Filter run times between backwashes were approximately 8 hours.
A maximum feed concentration of 35 mg/1 SS and an average ef-
fluent of 7 mg/1 with an 8-hour run time corresponds to a filter
loading of 3.4 Kg/m3 (0.21 lbs/ft3) of filter media.
Waste Sludge and Backwashes—
Grab samples of waste secondary sludge and backwashes from
the sedimentation/filtration unit, carbon columns and multi-
media filter were taken periodically and analyzed for COD,
solids and heavy metals. A summary of this data is shown in
Table 7 with more detailed data in Table 1 of Appendix B.
These analyses refer to the concentration of materials in the
original sample. They may be expressed as percent of dry solids
by dividing the concentration of the metal ion by the total sus-
pended solids and multiplying by 100.
The relatively high levels of calcium, magnesium and sodium
result from the seawater based cooling-water system. Copper
and iron levels result in part from material corrosion. The
primary clarification system has pH adjustment for the precipi-
tation of aluminum hydroxide. Incomplete removals in the
primary clarifiers account for the presence of aluminum in the
waste secondary sludge as well as the backwash waters. No
traces of the toxic metals cadmium, mercury or arsenic were
found in any of the samples.
Reverse Osmosis—
Tubular cellulose acetate—Operation began in April and
continued until July 9, 1976 with the tubular cellulose acetate
(CA) membranes. RO performance is summarized in Table 8 which
shows averages taken over the entire test period. From the
very beginning, poor conductivity rejection was experienced and
averaged from a low of 67 percent to a high of 88 percent re-
jection, reducing the conductivity to approximately 1000 umho.
Total dissolved solids rejection ranged from 73 to 97 percent
(Table 2, Appendix B). These rejections were well below that
stated by the manufacturer of these membranes. Conductivity
in the range of 200 to 400 umho was required for satisfactory
life of the downstream ion-exchange resin beds.
Several attempts were made to improve rejections by
changing operating conditions of pH, temperature and pressure
with negligible improvements. The manufacturer upon disas-
sembling a module and analyzing the membranes concluded that
the membranes had been damaged due to excessive heat and/or pH
conditions outside the acceptable range. Heat damage most
probably resulted from exposure to the sun, accounting for the
poor rejections from the start.
42
-------
TABLE 7. WASTE SLUDGE CHARACTERIZATION SUMMARY, OVERALL AVERAGES
Assay
mg/1
(Detectable
Limit)
PH
COD
TSS
TDS
Ca
Li (1.0)
Mg
K (0.1)
Na
Cd (1.0)
Cr04
Cu (0.1)
Fe
Mn (0.1)
Al
Ni (0.1)
Zn (0.1)
Hg (0.05)
As (0.1)
Si02 (Si)
Waste
Secondary
Sludgy
7.9
14241
14468
3491
138
Nil
49
83
1217
Nil
0.4
0.7
24
0.6
130
0.4
2.7
Nil
Nil
(48)
Sedimentation/
Filtration
Unit Backwash
8.4
272
303
3465
58
Nil
5.9
5.0
1088
Nil
< 0.1
0.1
2.3
< 0.1
5.9
< 0.1
0.15
Nil
Nil
2.2
Multi-Media
Filter
8.5
74
11
2988
53
Nil
4.7
0.9
1056
Nil
Nil
< 0.1
0.7
Nil
1.2
Nil
0.1
Nil
Nil
1.2
Carbon
Column
Backwash
8.5
550
161
3485
53
Nil
5.4
0.9
1212
Nil
Nil
Nil
5.2
Nil
2.4
Nil
0.1
Nil
Nil
2.6
-------
TABLE 8. REVERSE OSMOSIS DATA SUMMARY, OVERALL AVERAGES
Assay
PH
Conductivity, umho
Hardness , CaCOg
Fe
TSS
TDS
(TOC) COD
Silica, Si
Chloride
Tubular
Cellulose Acetate
Feed Perm
8.3
5148
61
0.43
14.7
3665
(48)
8.3
-
6.8
1086
9
0.35
12.2
609
(14)
4.7
-
Spiral-wound
Cellulose Acetate
Feed Perm
8.1
4257
66
0.54
8.4
2924
44.5
11.4
97.7
5.8
498
8.5
0.41
0.58
313
10
3.9
29.5
Spiral-wound
Poly amide
Feed
8,1
5858
53
0.29
11.9
4490
47
6.6
241
Perm
7.1
477
0.7
0.14
1.7
334
4
1.9
51
-------
The test period using the tubular CA membranes was plagued
with additional problems including rapid temperature buildup of
the circulating water, faulty pH control, leaky brine seals and
pump failures. One 30-day cooling-water test was conducted
using RO product water produced by the tubular CA system and is
not considered representative of RO performance. No boiler
tests were conducted using renovated water produced during this
time period.
At the manufacturer's recommendations, spiral-wound CA
membranes were installed to replace the tubular CA membranes.
Spiral-wound cellulose acetate—The tubular CA membranes
were replaced on July 10, 1976 with spiral-wound CA membranes.
These membranes were selected primarily because of their much
lower replacement cost than the tubular configuration for the
same membrane surface area, and the fact that suspended solids
were felt at this time to be low enough not to present a
fouling problem in the spiral-wound configuration which has
very small liquid flow channels.
The spiral-wound CA membranes performed very well during
the first two weeks of operation, averaging greater than 95
percent dissolved solids rejection and approximately 70 percent
COD rejection at 80 percent water recovery (Table 8 and Table
2, Appendix B). These removals were greatly improved over the
previous tubular CA operation. However, after this initial
period of operation,rejections decreased to the 85 percent
level. Permeate flow also began to decline. Membrane fouling
became apparent when the recommended cleaning solutions failed
to improve rejections and permeate flow. Organic rejections
(COD) varied between 55 and 100 percent during the period of
operation using the spiral-wound CA membranes.
Several weeks of operation continued with many attempts
to improve rejections and determine the cause of the membrane
fouling. Upon acidifying the feed with sulfuric acid a clear
gel-like substance was noted to frequently form in the feed
tank at a pH < 5.5, but not above this pH. Although originally
thought to be a silica precipitate, this was not confirmed.
The precipitation of any materials in the RO unit must be
avoided in any case to prevent membrane fouling. Disassembly
and examination of one of the modules showed rust particles and
a very fine black slime on the surface of the membrane. Be-
cause the cleaning solutions recommended for iron, heavy metals
and organics were ineffective, particulates appeared to be the
major cause of fouling.
45
-------
A filter after the RO feed tank was originally thought not
to be needed, since suspended solids were very low in the multi-
media filter effluent; also the multimedia filter effluent was
passed through a 10-micron cartridge filter before entering the
RO feed tank. However, very fine particulates either passed
through the filter or precipitated in the feed tank. Iron, re-
sulting from corrosion of the carbon-steel feed tank, contributed
solids to the RO feed-water.
In addition to membrane fouling, high circulating water tem-
perature continued to be a problem. The rapid temperature build-
up of the recirculating water appeared to be unavoidable in this
batch-type process operation.
After 10 weeks of unsuccessful performance, polyamide spiral-
wound membranes were installed. No cooling-water tests or boiler
tests were conducted on water renovated using the spiral-wound CA
membranes.
Spiral-wound polyamide—Spiral-wound polyamide (PA) membranes
were placed in operation on October 7, 1976, based upon manu-
facturers' recommendation and their claims of the ability of these
membranes to tolerate temperatures up to 27.2°C (135°F), pH ranges
from 3 to 9, and greater organic rejections than the CA membranes.
A 10-micron cartridge filter was installed between the RO feed -
tank and the RO pump suction as an additional precaution against
suspended solids fouling the membranes.
The RO unit operated for seven consecutive weeks achieving
consistently ~-91 percent conductivity rejections. Organic
rejections (COD) improved to an average 91.5 percent as well as
near complete rejection of hardness (Table 8 and Table 2,
Appendix B). Feed conductivity during this period ranged from
4000 to 9000 umho.
Rejections were found to decline with each consecutive
batch run as shown in Figure 18. Increased fouling resulted in
higher initial permeate conductivity as well as a more rapid
increase in permeate conductivity as the batch became more con-
centrated. Ideally one would like these curves to be very steep
with a sharp break at some higher percent water recovery. In
all tests conducted, permeate conductivity increased rapidly
at water recoveries beyond approximately 60 to 70 percent.
Test runs were conducted to determine if pH adjustment
significantly affected conductivity rejection (Figure 19). It
was found that by reducing the RO feed to pH ~» 5 that lower
permeate conductivities were achieved with higher water re-
coveries before breakthrough of conductivity. This improvement
46
-------
Run #1
0)
•P
20l
A 40|
60
80
Inlet Pressure
Outlet Pressure
Concentrate and
Permeate Flow
Feed Conductivity
pH 8.3
400 psi
345 psi
6 gpm
4700 umho
100|
I
1
1
I
200 400 600 800 1000 1200 1400 1600
Permeate Conductivity, umhos
Figure 18. Consecutive RO batch runs with spiral-wound PA membranes, 10/7/76.
-------
00
20
40
-------
with reduced pH was observed several times; however, upon re-
duction of the feed pHUmore rapid plugging of the 10-micron
filter between the feed tank and RO pump suction occurred. In
all runs at the reduced pH,. this filter was coated with the gel-
like material previously addressed. The increase in the
precipitation of salts at lower pH conditions and the resulting
increase in filter plugging, or membrane fouling, must be
weighed against the benefit of increased salt rejections. This
fact could play a significant role in the cost effectiveness of
reverse osmosis.
Although the use of the membrane cleaning solutions often
proved ineffective in restoring salt rejection, it was observed
several times that simply rinsing the membranes with demineral-
ized water and letting them soak in demineralized water for a
long period (1-2 weeks) restored the membranes' ability to
reject dissolved salts and increased permeate flux to original
levels. This phenomenon suggests that the membranes are not
irreversibly fouled and that a more rapid cleaning procedure
is needed. This observation was discussed with the RO manu-
facturer who indicated that this has been observed elsewhere,
but as of yet is an unresolved problem.
Saflaples were taken periodically of tne RO feed and per-
meate for more specific inorganic analyses (Tables 9 and 10).
These values were utilized to insure the correct sizing of the
RO unit in the full-scale design. Calcium, magnesium, sodium,
bicarbonate, sulfate and chloride are the primary ions to be
rejected. Sodium and chloride are among the ions most poorly
rejected. The divalent ions were all rejected very well and
would appear to present no problems in a full-scale unit.
Ion-Exchange—
The ion-exchange resin beds were operated only when needed
to produce renovated water for feed to the pilot boiler. The
ion-exchange unit was operated as a complete removal system such
that when the effluent reached unacceptable quality, the beds
were regenerated. The object was not to evaluate ion-exchange
but to produce water satisfactory for feed to the boiler. Data
collected routinely during operation is tabulated as weekly
averages in Table 3, Appendix B and summarized in Table II.
The average analyses over the entire study period indicated
limited effectiveness of the primary deionizer (weak base anion)
in scavenging the low concentrations of organics present. At
these very low COD levels, which are at the very lowest limits
of the analytical test, it is difficult to evaluate the resin's
effectiveness in scavenging organics based upon COD alone.
49
-------
TABLE 9.
INORGANIC ANALYSIS FOR DESIGN OF DEMINERALIZATION SYSTEM
Cn
O
Assay
rag/1
PH
COD
Conduct ivit y (umho )
TDS
Total Hardness(CaC03>
Alkalinity (CaC03).P
M
Si02(Si)
Fe
Mn
Ca
Mg
Na
K
OH
COa
HCO3
SO4
*>
ci-
NO3
Ortho PO4
Phenol
8/24
8.4
41
4125
3318
38
68
1402
(12.0)
0.1
Nil
3.6
40
847
0.7
0
136
1266
350
96
140
9
Nil
8/26
8.4
34
4625
3244
34
84
1598
(17.0)
0.2
Nil
2.3
42
1128
0.6
0
168
1430
452
182
183
35
Nil
Multi-Media Filter Effluent (R.O. Feedwater)
Dates
8/31 9/2 9/7 9/9 9/16 9/21 10/1
8.4
12
2500
1998
126
70
1158
(14.3)
0.1
Nil
35.8
8.2
634
1.2
0
140
1018
252
71
129
15
"•
8.6
45
3250
2150
56
100
1360
(11.5)
Nil
Nil
5.1
4.5
828
1.8
0
200
1160
242
62
129
18
~
8.4
40
4375
2650
36
80
1356
(9.0)
0.1
Nil
3.0
3.3
1210
O.9
0
160
1196
1064
75
23
70
Nil
8.3
55
3625
2525
24
12
1134
(8.0)
0.2
Nil
4.9
4.0
1012
0.6
0
24
1110
756
62
23
35
"™
7.9
_
4000
2336
40
34
576
(2.6)
1.1
Nil
4.4
1.9
752
0.9
0
68
508
196
57
65
40
™
8.0
_
1375
883
150
22
308
(13.0)
0.1
Nil
4.8
10.0
250
1.0
0
44
264
-
36
38
6
Nil
8.4
_
4750
3402
30
84
1180
(8.6)
Nil
Nil
3.0
4.2
1250
0.9
0
168
1012
1600
20
70
25
10/9
8.4
20
4437
3326
36
100
1184
(6.3)
0.2
Nil
3.2
5.0
1150
1.4
0
200
984
829
112
124
-
10/24
8.6
_
5437
4720
38
144
2260
(8.0)
0.5
Nil
2.1
5.0
1800
3.0
0
288
1972
952
136
31
20
11/2
8.7
-
6625
3910
62
104
1224
(5.0)
Nil
Nil
1.7
92.8
1707
7.9
0
208
1016
448
388
116
7
Nil
Avg
8.4
30
4069
2872
56
75
1228
(9.6)
0.2
Nil
9.7
18.3
1047
1.7
0
150
1078
649
108
89
25
Nil
-------
TABLE 10. INORGANIC ANALYSIS FOR DESIGN OF DEMINERALH'.ATIOI SYSTEM
cn
Assay
mg/1
PH
Membrane
COD
Conductivity (umho)
TDS.
Total Hardness (CaCOq)
Alkalinity (CaCO,) P^
3 M
Si09(Si)
Fe 2
Mn
Ca
Mg
Na
K
OH
CO^
'•"-'o
HCO.,
so.
4
Cl3
NO,,
Ortho PO.
Phenol
9/23
5.7
CA
_
179
127
2
0
20
1.0
Nil
Nil
0.3
0.1
20
Nil
0
0
20
7.2
2.0
4.9
1.0
Nil
10/1
6.6
CA
_
488
248
2
0
24
2.9
0.1
Nil
0.1
0.1
86
0.1
0
0
24
3.4
101
55
3.0
Nil
10/6
6.9
CA
16
456
252
Nil
0
52
2.5
0.3
Nil
0.1
0.1
90
0.1
0
0
52
3.4
43
91
Nil
Nil
Reverse Osmosis
Dates
10/9 10/13 10/17
7.5
PA
16
613
396
Nil
0
104
1.6
Nil
Nil
Nil
0.1
127
0.1
0
0
104
3.4
46
79
Nil
Nil
6.6
PA
20
431
288
Nil
0
66
1.1
0.1
Nil
0.1
0.1
92
0.3
0
0
66
3.4
55
77
Nil
Nil
8.3
PA
24
450
306
Nil
8
130
1.3
Nil
Nil
Nil
0.1
95
Nil
0
16
114
5.6
34
22
1.0
Nil
Permeate
10/18
7.7
PA
29
500
312
Nil
0
134
1.3
Nil
Nil
0.6
0.1
119
0.1
0
0
134
11.8
24
62
Nil
Nil
10/22
7.5
PA
-
550
366
Nil
0
154
1.0
Nil
Nil
0.1
0.1
156
0.1
0
0
154
16
30
30
2.0
Nil
10/24
8.0
PA
-
631
384
Nil
0
180
1.0
Nil
Nil
0.1
0.1
187
Nil
0
0
180
27
50
45
Nil
Nil
CA
11/2 Avg^
7.6 6.4
PA
24 16
890 374
430 209
Nil 1.3
0 0
120 32
1.3 2.1
Nil 0.1
Nil Nil
Nil 0.2
1.2 0.1
244 65
0.8 <0.1
0 0
0 0
120 32
10.1 4.6
141 49
93 50
Nil 1.3
Nil Nil
PA
Avg
7.6
-
23
580
355
Nil
1
115
1.2
Nil
Nil
0.1
0.2
146
0.2
0
0
124
10.0
54
58
0.4
Nil
-------
TABLE 11. ION-EXCHANGE DATA SUMMARY, OVERALL AVERAGES
en
K)
Assay
pH.
Conductivity, umho
Hardness , CaCO
Fe
TSS
COD
Silica, SiO2
Chloride
RO Permeate
6.7
516
9.0
0.25
1.0
14
3.8
33
Primary Deionizer^*'
8.2
418
9.4
0.13
1.0
13
3.4
23
Secondary Deionizer^2^
7.4
5
1.3
0.12
0.0
6
0.5
1.8
Weak-base anion exchange
Mixed bed ion-exchange
-------
Samples from the two ion-exchange columns were also taken
periodically for more specific inorganic analysis that would
enable an accurate sizing of full-scale demineralizers based
upon the ions present in the water (Table 12). Primary leakage
through the ion-exchangers was sulfate and chloride ions.
The on-line time for the pilot-scale exchangers ranged
from 28 to 39 hours and averaged 35 hours before the secondary
deionizer (mixed-bed) exhausted, requiring regeneration. Regen-
eration of the primary deionizer with caustic presented no
problems. The secondary deionizer required an excessive amount
of rinse water to clean the bed of regenerate. It was felt that
this was due to poor flow distribution ir? the column which would
cause excessive rinse requirements as a result of inadequate
resin mixing. This fact, along with frequent shutdown of the
reverse osmosis unit prevented a good water balance from being
conducted on the pilot-plant facilities. A theoretical water
balance based upon both the pilot-plant experiences and a con-
tinuous on-line demineralization system is presented and
discussed under ESTIMATES OF INVESTMENT COST AND OPERATING
EXPENSE FOR .FULL-SCALE FACILITIES.
Demineralization Waste Brines—
In addition to the product water from the demineralization
system waste streams are produced that contain a more concen-
trated solution of all the salts that were contained in the
feedstream. These are the very salts that made the original
feedstream unusable and, therefore, their disposal presents a
serious and difficult problem. With a reverse osmosis system
operated at 75 percent water recovery and the regeneration
brines from the ion-exchange beds, the waste brine requiring
disposal will be 30 percent or more of original wastewater flow
entering the renovation facility.
Since the primary objective of these studies was the pro-
duction of water for reuse in boilers, various alternatives
for brine disposal were not evaluated, nor was the cost of
disposing of these brines determined—a very important and
significant factor in the overall cost of wastewater renovation.
Samples of the reverse osmosis brine as well as the ion-
exchange regeneration brines were analyzed primarily for the
various heavy.metals (Table 13). None of the very toxic metals
(Hg, Cd, As) were found in concentrations above the detectable
limit. The waste brines consist primarily of the measured ions
sodium, calcium, magnesium, iron and aluminum and would contain
as well significant levels of chloride, sulfate and bicarbonate.
Although not reported here, the levels of these and other anions
are easily estimated based upon their rejections and water
recovery of the original feedstream.
53
-------
ION EXCHANGE INORGANIC ANALYSIS
01
Assay
mg/1
PH
COD(TOC)
Conductivity (umho)
Total Hardness (CaCOo)
Alkalinity(CaCO3) P
.M
Si02
Fe
Un
Ca
Mg
Na
K
OH
C03
HC03
SO A
Cl
N03
Ortho P04
Phenol
10/1
8.7
-
412
4
8
68
3.8
Nil
Nil
0.2
0.2
81
0.1
0
16
52
14.0
Nil
56
3
Nil
WB Organic Ion Exchanger Effluent Mixed Bed Ion Exchanger Effluent
Dates Dates
10/2 10/13 10/16 10/18 10/22 10/24 Avg 10/1 10/2 10/13 10/16 10/22 10/24 11/2 Av»
8.6
-
419
Nil
12
80
1.6
Nil
Nil
0.1
Nil
85
0.5
0
24
56
23
29
-
Nil
Nil
8-° 8.1 8.8 8.8 9.0 8.6 8.5 6.7 6.6 8.2 7.5 6.7 7.0 7.3
29 57 - 43 - 24 - Nil Nil (3) <8
418 456 331 456 631 446 4 2 4 3 3 1.5 0.9 2.6
Nil 6 Nil Nil Nil 1.4 Nil Nil Nil Nil Nil Nil Nil Nil
4 64 4 18 30 20 0 0 0 0 0 0 00
64 178 94 172 190 121 2 4 4 4 4 25 10 8
1-6 - 0.9 1.4 1.6 1.8 1.9 Nil Nil Nil Nil Nil Nil Nil
Nil Nil Nil Nil Nil Nil Nil Nil Nil Nil Nil Nil Nil Nil
Nil Nil Nil Nil Nil Nil Nil Nil Nil Nil Nil Nil Nil Nil
0.1 0.1 0.3 0.1 0.2 0.2 Nil Nil Nil Nil Nil Nil Nil Nil
1-1 - Nil 0.1 0.1 0.3 Nil Nil Nil Nil Nil Nil Nil Nil
92 95 74 119 200 107 0.5 Nil 3.9 Nil Nil Nil Nil 0.6
0.2 Nil Nil Nil 0.1 0.1 Nil Nil 0.3 Nil Nil Nil Nil <0.1
000 00000000000
8 128 8 36 60 40 000 0 0 0 00
56 50 86 136 130 81 2 4 4 4 4 25 10 8
3.5 6 12 17 37 16 1.2 5.6 3.4 Nil 11.2 Nil 1.1 3.2
41 38 14 29 36 27 24.0 2.9 3.9 3.0 10.0 1.0 10.0 7.8
84 54 52 39 32 53 Nil 0.9 1.8 4.5 0.9 2.2 2.2 1.8
Nil 1 Nil 3 Nil 1 15 Nil Nil Nil Nil Nil Nil Nil
Nil Nil Nil Nil Nil Nil Nil Nil Nil Nil Nil Nil Nil Nil
-------
TABLE 13. DEMINERALIZATION WASTE BRINE CHARACTERIZATION
Assay
auc/1 Reverse Osmosis Brine
(Detectable Dates
Limit) 9/22 10/4 10/9 J.0/16 10/22 10/24
Membrane CA CA PA PA PA PA
% Recovery BO 80 80 80 80 80
pH 7.3 7.2 8.3 8.2 8.8 8.6
COD — -- 181 145 130 155
TS8 28 43 47 33 24 55
TDS 11164 14848 15772 12566 19743 13493
Ca 77 132 150 128 108 130
Li (1.0) Nil Nil Nil Nil Nil Nil
Kg 35 26 38 18.5 15.5 23.5
K 3.0 4.5 6.1 1.9 10.5 3.6
Na 3650 4860 5250 3780 3687 5250
Cd (1.0) Nil Nil Nil Nil Nil Nil
CrO4 0.1 Nil Nil Nil Nil 0.1
Cu 0.2 0.1 Nil Nil 0.1 0.2
Fe 4.0 1.5 0.4 0.2 0.3 0.2
Hn 0.1 0.1 Nil Nil Nil Nil
Al — 2.3 3.1 2.1 6.0 3.5
Ni 0.2 0.1 Nil Nil 0.1
Zn 0.4 0.2 0.5 0.1
Hg (0.05) — Nil — — Nil
As (0.1) — Nil -- — Nil
SiO2 ~ 2.3 4.2 3.5 3.5 3.6
CA PA
Avg Avg
80 80
7.2 8.5
152
36 40
13006 18645
105 124
Nil Nil
31 25
3.8 5.5
4255 4491
Nil Nil
<0.1 <0.1
0.15 <0.1
2.7 0.2
0.1 Nil
2.3 3.7
0.1 Nil
Oi3 0.2
Nil Nil
Nil Nil
2.3 3.4
Regeneration Brine
WB Organic Ion Exchanger
Dates
9/18 10/14 10/23
11.7 12.2 12.2
48 100
22 44 32
13325 18392 . 9528
94 144 114
Nil Nil Nil
0.1 0.8 0.7
10.0 1.9 1.8
8060 6330 4213
Nil Nil Nil
0.1 Nil Nil
0.1 Nil Nil
1.7 4.5 5.7
Nil Nil Nil
1.1 1.5
0.3 Nil Nil
0.1 0.1 Nil
—
. —
1.0 1.0
Avg
12.0
74
33
13748
117
Nil
0.5
4.5
6201
Nil
<0.1
<0.1
4.0
Nil
1.3
<0.1
<0.1
—
—
1.0
Regeneration Brine
Mixed Bed Ion Exchanger
9/18 10/15 10/23
11.3 1.6
37 25
81 38
15498 17003
174 130 108
Nil Nil Nil
3.8 6.1 3.5
2.6 2.4 1.8
3200 4268 3625
Nil Nil Nil
0.1 Nil 0.2
0.1 Nil 0.1
6.9 11.7 15.6
0.1 0.1 Nil
1.8 0.9
0.3 Nil 0.1
0.3 0.4
—
—
4.1 3.0
AVK
6.5
31
60
16250
137
Nil
4.5
2.3
3697
Nil
<0.1
<0.1
11.4
<0.1
—
<0.1
0.3
—
--
3.6
01
Ol
-------
At costal locations, an obvious method of disposal is
through a seawater outfall. The dangers of high salt concen-
trations on the land are avoided and the effects of the brine
discharge are minimal and localized as a result of re-dilution
effects of runoff and river inflows. It is very possible to
design an outfall system with the proper choice of diffuser
that will control the initial dilution of the effluent to meet
a predetermined acceptable level and thus minimize or avoid
environmental problems.
Alternate methods of brine disposal include: solar evapor-
ating ponds, deep-well injection, and concentration/mechanical
drying followed by landfill of solids. Typical costs of
disposal of waste brines using the various methods range from
$1.00 to $10.00 per 1000 gallons of brine (5,6). At a given
location, the geological and meterological conditions will
determine which disposal method is most economical. For coastal
locations, dispersion in an ocean outfall appears to be most
economical. For inland facilities, solar evaporation offers an
environmentally acceptable and economically reasonable disposal
method—provided the evaporation rate exceeds the precipitation
rate (primarily southwestern U.S.). For deep-well disposal
suitable underground formations must exist, which is not the
case in many areas. Mechanical drying requires high energy
consumption and as a result has a very high cost (7,8,9).
Specific Organic Analyses—
Analyses for specific organic compounds were made on
samples of wastewaters from the pilot-plant. The samples in-
cluded: (1) sedimentation/filtration effluent, (2) carbon
column effluent, (3) reverse osmosis effluent and (4) ion-
exchange column effluent. Specific compound identification
and quantification were made using a direct-inject gas chroma-
tographic (GC) procedure which gave a limit of detection of
~1 mg/1. Additional specific compounds present in lower
concentrations (parts per billion range) were identified but not
quantified by a concentrating procedure followed by a gas chro-
matographic/mass spectrometric analyses.
Direct-inject GC analyses—The results of the direct-inject
GC analyses are shown in Table 14. The direct-inject GC study
included analyses for so-called "volatile" organics and gly-
cols. The "volatile" materials are organics detectable by GC
and include such compounds as ketones, esters, alcohols, aro-
matics and organic acids. Most of these compounds were detected
using a 6.1-m (20-ft) column packed with CARBOWAX 20M on Chromo-
sorb W; organic acids, however, required a column packed with
Porapac Q which had been pre-treated with phosphoric acid.
Glycols were analyzed by preparing a trimethylsilyl derivative
followed by GC determination. These procedures are described
in Appendix A.
56
-------
TABLE 14. SPECIFIC COMPOUNDS ANALYSES - DIRECT-INJECT
Compound
Sedimentation/
Filtration
(a)
Effluent, mg/1
Carbon Column
Effluent, mg/1
RO
Effluent, mg/1
Ion Exchange
Effluent, mg/1
en
-3
Acetone
Methyl ethyl ketone
Toluene
Isobutanol
Ethyl benzene
n-But anol
Cumene
Styrene
Acetophonone
Naphthalene
Phenol
Acetic acid
Propionic acid
Isobutyric acid
Butyric acid
Ethylene glycol
Propylene glycol
Diethylene glycol
Ethoxytrlglycol
Triethylene glycol
44
42
2
13
12
3
5
24
8
2
5
(b)
(b)
trace
(a) Materials indicated at <1 rag/1 were present in the biological feed but below
quantifiable levels in the effluent.
(b) RO effluent not analyzed for organic acids or glycols.
-------
In the analyses, twenty specific compounds were identified
and quantified in the feed to the biological system. In gen-
eral, the major components were glycols, organic acids,
alcohols, ketones and aromatic products of the olefins opera-
tion. All of these compounds were reduced significantly or not
detected in the biological effluent (sedimentation/filtration
effluent) indicating excellent removal of these specific com-
pounds,by biological treatment.
Specific compounds were not detected by direct-inject GC
using the CARBOWAX 20M column in samples of the carbon column
effluent, RO effluent and ion-exchange effluent. These results
indicate that if such specific compounds were present, they
were in concentrations less than the detectable limits of the
method « 1 mg/1). Using the acid-treated Porapac Q column,
small amounts of organic acids were found in the carbon column
effluent but were not detected in the ion-exchange effluent,
probably having been adsorbed in the weakly-basic ion-exchange
bed. A small amount of ethylene glycol was indicated in the
carbon column effluent with only a trace (< 1 mg/1) being found
in the ion-exchange effluent. These results show that specific
compounds, if present in the ion-exchange effluent, are in parts
per billion concentrations and not detectable by direct-inject
GC analyses.
Additional specific compound identification—Additional
specific compounds were identified by concentrating the con-
tained organics in the samples followed by GC-mass spectrometer
identification. The analyses were made to identify compounds
present in very low concentrations and not detected in the
direct-inject GC procedures. These specific compounds were not
quantified because of the relatively large amount of work in-
volved. The concentrating procedure employed methylene chloride
in several extractions and is described in Appendix A. The
contained organics in the samples were concentrated at least a
thousand-fold in the operation and thus were contained in the
starting water samples in the parts per billion range or less.
Organic acids would be detected in the procedure; glycols
probably would not.
Seven additional specific compounds were identified
in the effluent from the biological system (sedimenta-
tion/filtration effluent). These compounds as well as those
detected in the carbon column effluent, RO effluent and ion-
exchange effluent are listed in Table 15. Toluene was found in
the carbon column effluent but not in the preceding sedimenta-
tion/filtration unit effluent. Both the RO and ion-exchange
affluents contained only formamide, xylene and ethyl benzene.
The presence of formamide probably resulted from the use of
formaldehyde in the RO membrane cleaning solution for the con-
trol of bacterial growth.
58
-------
TABLE 15. ADDITIONAL SPECIFIC COMPOUNDS DETECTED
IN PONCE WATER REUSE SAMPLES(a)
Sedimentat ion/Filtrat ion
Effluent
Carbon Column
Effluent
RO
Effluent
Ion Exchange
Column Effluent
Acetone
Acetone
Formamide
Formamide
Ethyl benzene
Toluene
Xylene
Xylene
en
Xylene
Styrene
Ethyl benzene Ethyl benzene Ethyl benzene
Xylene
C,QH,4aromat ic
Methyl styrene
C10H20
These compounds contained in ppb range or less,
-------
After treatment with activated carbon, no specific
compounds were detected by direct injection in the GC analysis
using the CARBOWAX 20M column. After complete treatment through
the renovation facilities, the resulting water for feed to the
pilot boiler contained only trace amounts (ppb) of formamide,
xylene, ethyl benzene and ethylene glycol.
INVESTMENT COST AND OPERATING EXPENSE FOR FULL-SCALE FACILITIES
This section presents estimates of fixed investment and
operating expenses for a full-scale tertiary treatment system,
very similar to the pilot-plant, which could produce renovated
wastewater suitable primarily for boiler feedwater. Because
of the very unique characteristics of each chemical plant's
wastewater these economic data should not be construed to re-
flect cost and expenses from an optimum-designed wastewater
renovation facility suited for petrochemical plants in general,
but rather as a basis from which reasonable economic data can
be extrapolated for specific situations.
These data were developed by the Cost Estimating Section
of UCC's Engineering Department and approach the quality of
estimates used for major capital budget requests. Two design
cases were estimated having wastewater flow as their only dif-
ference. This hypothetical facility shown in a conceptual
layout in Figure 20 is proposed for UCCI's organic chemical
manufacturing complex near Ponce, Puerto Rico.
The following general specifications and assumptions were
the basis for developing these costs and expenses:
• The influent to the tertiary treatment facility
is UCCI plant's wastewater after primary and
secondary (activated sludge) treatment. Two
design cases will be considered: Case A—5.7
m3/min (1500 gpm) and Case B—11.4 m3/min
(3,000 gpm). The renovated water is to be used
as boiler feedwater for 1100 psig, 750°F steam.
Average water recovery is 67 percent.
• All backup and support facilities such as
electrical power, steam, instrument air,
roadways, etc., are available at the tertiary
treatment facility battery limits and no
additional general facilities are required.
Also assume that waste sludge and brine dis-
posal facilities are existing and no additional
costs or expenses will be incurred.
60
-------
Battery Limits
Solids
Contact
Clarifiers
Mixed Media
Gravity
Filters
O O
Activated Carbon
O O
Regeneration System
O
o o o o
Activated Carbon
O O O O
Adsorption Columns
O O O O
o o o o
Mixed Media
Pressure
Filter
System
sh Recdrvery System
Reverse Osmosis
System and
Control Building
o o o
o o o
Ion Exchange System
oon
o
o
Scale: 1 inch =
feet
O
O
Figure 20. Conceptual layout, 5-7 m /min. (1500 gpm) nominal flow case - Case A
-------
• Fixed investment costs are based upon fourth
quarter 1978 mechanical completion. Direct
and period operating expenses are 1979
dollars that do not include depreciation.
No credit is taken for any tertiary treatment
investment cost or operating expenses that
might be required as a result of more
stringent effluent guidelines resulting in
some part or all of the tertiary treatment
effluent being disposed of in-some receiving
body of water.
Process Description
With the exception of the sedimentation/filtration unit
the estimated full-scale facility's process and equipment very
closely parallel that of the pilot-plant. In the early phases
of the experimental program it became apparent that the bio-
logical system's residual suspended solids were readily removed
in the sedimentation/filtration step, thus a more conventional
reactor clarifier was used for solids removal in this full-scale
projection.
The clarified secondary effluent would flow into two solids
contact clarifiers where polyelectrolyte would be added. This
clarifier overflow would flow by gravity through three multi-
media filters operated in parallel. The filtered water would
be pumped from a clearwell into the activated carbon adsorption
columns arranged in four parallel lines of three adsorbers in
series with one standby adsorber. Spent activated carbon would
be regenerated on site in a multiple-hearth furnace. Carbon
column effluent would be filtered in two parallel multi-media
pressure filters and pumped to a reverse osmosis system. De-
mineralization of the reverse osmosis permeate would be
completed by ion exchange. Portions of the ion exchange regen-
eration cycle rinse water would be recycled to the reverse
osmosis system. Spent backwash from the gravity filters, pres-
sure filters and carbon columns, flows to a flocculation tank
where a polyelectrolyte would be added. The flocculated back-
wash would then be clarified and recycled to the secondary
treatment system.
Due to equipment limitations, direct scaleup from pilot-
scale to full-scale is not always possible. A water balance
estimated for a wastewater flow of 5.7 m^/min (1500 gpm) based
upon pilot-scale experience and sizing of full-scale facilities
indicates a maximum achievable water recovery of 67 percent
(Figure 21). The remaining water is contained in waste sludge
and brines that would require disposal. Approximately the same
percent of water recovery would be realized through a system
treating 11.4 m3/min (3000 gpm). Higher waste loads or more
62
-------
G)
CO
5.68 (1500) ! Existin«
.k| Secondary
• I Treatment
[ System
L. _
0.55 (11*7)
Coagulation
and
Jlarificatio
0.13 (33) 0.0
Sludge to Disposal
3.83 (1011)
6.23 (16U?)
*
Clarifiers
^.15 (1626)^
0.08 Sludge
s
J
0.60
(12) v
Product Vater to Reuse
(67 % Recovery)
1.72 (1*56)
Ion
Exchange
System
!«•
Multi-
Gravity
Filters
0.31
(82)
^
0.18 (1*8)
29
1*,. 30 (1136)
-------
variable streams would significantly affect the water balance,
yielding decreased water recoveries, due to increased sludge/
brine residues. Water recovery is strongly dependent upon the
blowdown, regeneration, backwash and cleaning requirements of
each unit operation. Because these streams are unavoidable,
complete water reuse becomes increasingly more difficult to
achieve with higher qualities of water required for reuse.
How the projected full-scale wastewater renovation facili-
ties were sized, along with differences from the pilot facili-
ties tested is summarized in Appendix C. References 10 through
20'listed in the Bibliography were utilized for the sizing and
cost estimation of full-scale facilities.
Summary of Economics
Fixed investment and annual operating costs for each treat-
ment step are broken down in Table 16. Total costs and mis-
cellaneous usage rates for the full-scale facilities are
summarized in Table 17.
The total annualized cost of producing water of boiler feed-
water quality through this tertiary renovation sequence would be
approximately $2.00/m3 ($7.50/1000 gallons) in 1978. This does
not include any primary or secondary treatment costs; nor does
it include facilities for the handling and disposal of waste
brines and sludges. The disposal of these waste streams is
expected to be high with little to no additional water recovered.
Carbon adsorption and carbon regeneration facilities make
up approximately 35 percent of the total fixed investment and
greater than 30 percent of the total annual operating expense.
Reverse osmosis accounts for an additional 25 percent of the
total fixed investment and 30 percent of the annual operating
expense. A water reuse facility using these treatment steps
will require reductions in the cost of carbon adsorption and/or
reverse osmosis to significantly improve the sequences' cost
effectiveness. Lesser quality waters for use in low pressure
steam systems or as cycle cooling-water can be obtained at
significantly lower cost.
64
-------
TABLE 16. INVESTMENT AND OPERATING COST SUUMARV
THOUSANDS OF DOLLARS
CASE A: Plant Influent Rate; 5.7 m3/min (1,500 gpm)
CASE B: Plant Influent Rate; 14.4 m3/aln (3,000 gpa)
Fixed Investment
Material and Equipment
Construction Labor
Construction Overhead' '
Engineering and Startup
Contingencies'5*
Total Fixed Investment
Suspended Solids
Removal
Case A Case B
694
380
331
476
470
2351
966
471
412
628
619
3096
Activated Carbon
Case A Case B
1813
459
411
910
898
4491
3195
813
727
1605
1585
7925
Carbon
Regeneration
Case A Case B
941
124
115
400
^95
1975
1012
124
116
424
4J.J1
2095
Pressure
Filtration
Case A Case B
423
150
133
239
^36
1181
673
253
223
390
385
1924
Reverse
Case A
2293
140
143
873
863
4312
Osmosis
Case B
4457
324
336
1667
1646
8330
Ion Exchange
Case A Case B
1439
250
238
649
3207
2249
386
353
1013
1000
5001
Backwash
Case A
140
111
96
118
116
581
Recovery
Case B
193
153
132
162
160
800
Total
CftBfl A CftSB B
7,743
1,614
1,457
3,665
3^619
18.098
12,745
2,424
2,199
5,889
5,814
29,071
Annual operating Costs
Utilities 46 21 44 10
Chemical Flocculant 86 172 - -
Furnace Fuel - 38
Material Replacement - 78
Operating Labor 30 30 30 30 122
Maintenance 143 189 274 483 121
Plant Overhead _47 _BO _20 _29 95
Total Annual Operating 310 477 348 586 464
Cost
(2)
17
71
157
122
128
ill
636
«>
1
80
30
72
43
326
2
161
30
117
J£
385
380
«>
102
<3)
204
49 49
263 502
131 _244
735 1379
<«
49 49
196 305
50 80
375 594
14
35
6
1
25
49
12
85
2.511 4.142
(1)
(2)
(3)
Includes cost of utilities plus chemicals.
Carbon replacement.
Membrane replacement.
Includes supervision, purchasing, accounting, warehousing, material control, temporary buildings and other indirect costs.
Contingencies: 25% of engineering, equipment, and construction cost.
-------
TABLE IT. TOTAL COSTS OF WASTEWATER RENOVATION FACILITIES
Case A
Case B
Plant Influent Flow, m-Vmin (gpm)
Boiler Feedwater Production, m /min (gpm)
Percent Water Recovery
Fixed Investment
Annual Operating Costs
Product Cost^1), $/m3 ($/1000 gal.)
Miscellaneous Production and Usage Rates
Additional Operators Required
Plant Area, m2 (ft2)
Utilities Usage:
Electricity, KWH/yr.
Low-pressure steam, kg/yr (MM Ibs/yr)
Instrument Air ^2', m^/yr (MM ft3/yr)
Waste brine production, m-Vday (gpd)
Waste sludge production, (3) m3/d.ay (gpd)
5.7 (1500)
3.8 (1010)
66.7
$18,100,000
$2,500,000
$2.00 ($7.50)
13
7,300 (78,000)
7,700,000
850,000 (1.9)
(15)
2UH8 (662, UOO)
187 (>+7,500)
ll.lt (3000)
7.6 (2020)
66.7
$29,100,000
$U ,100,000
$1.60 ($6.05)
13
10,800 (116,000)
15,000,000
1,700,000 (3.7)
1*25,000 (15)
U896 (1.32U.800)
(95,000)
(l) Based upon depreciation of fixed investment at 8 percent per year.
(2) Volume at 1 atmosphere pressure and 21°C.
(3) Based upon 0.5 percent solids.
-------
SECTION 5
BOILER TEST LOOP
GENERAL FACILITY DESCRIPTION
The boiler test loop was designed by Betz Environmental
Engineers, Inc. The carbon steel test boiler was similar to the
research boilers built and operated by Betz Laboratories in their
product evaluation studies. The entire boiler test loop was a
skid-mounted package consisting of feed pumps, chemical addition
facilities, deaerating heater, boiler, superheater, steam con-
densers, and sample coolers.
The boiler configuration and supporting facilities are
illustrated in Figures 22, 23, and 24. Photographs of the boiler
control panel and configuration of the boiler drums are shown in
Figures 25 and 26, respectively.
Demineralized water from the boiler feedwater storage tank
was pumped into the steam-heated deaerator where the bulk of the
dissolved oxygen was driven off by heating the water to satura-
tion. Hydrazine to remove the remaining oxygen and internal
boiler water treatment chemicals to prevent scaling and fouling
were added to the boiler feedwater as it left the deaerator.
The test boiler was a two-drum design similar in configura-
tion to a standard "D" type industrial boiler with an external
separator installed between the steam drum and the superheater to
remove entrained water droplets. The boiler was designed to pro-
duce up to 81.7 Kg/hr (180 Ibs/hr) of saturated steam at pres-
sures up to 119.5 Kg/cm2 (1700 psig). Heat input to the boiler
was through electrical resistance heating elements. Boiler out-
put and boiler heat flux were controlled by varying the size and
rating of boiler heaters in service.
Saturated steam leaving the boiler passed through an elec-
trically heated carbon steel superheater. Control of superheat
temperature up to 399°C (750°F) was accomplished by varying the
number of heating elements in service.
Steam from the superheater flowed to surface condensers.
The steam condensate then passed through a corrosion test loop to
monitor the effect of any contaminant on corrosion test coupons.
67
-------
Saturated
steam
sample
Feed Water
to Boiler
Boiler
Continuous
Slowdown
Drai
Boiler
Bottom
Slowdown
High Pressure Shutdown switch
Pressure Gage
Temperature recorder
Temperature recorder-
controller'
Pressure recorder-controller
Slowdown collection and
measuring tank
Level Gage Glass
Level Electrode Chamber
Cooling Water
Boiler Heating Elements
SHHE- Super Heater Heating Elements
T- Slowdown Timer
-i'i Flow Restrictor
LGG-
LEC-
CW-
BHE-
Figure 22. Boiler test-loop toiler section.
68
-------
STEAM FROM
BOILER
BOIL-OFF
CONDENSER
AIR-COOLED
CONDENSER
CONDENSATE
TRIM COOLER
:ONDENSATE
SAMPLE
LEGEM
COOLING
WATER
CORROSION
TEST COUPON
FLOW TOTAL-
IZER-RECORDER
Figure 23. Boiler test-loop condenser section,
69
-------
RENOVATED WASTEWATER
DEMORALIZED WELL-WATER
HEATING STEAM
FROM PLANT
OPERATED O
CONTROL VAUUE
FEEDWATER
STORAGE
TANK
DEAERATING
HEATER
DEAERATOR FEED PUMPS
EED TANK FEED
i PRESSURE
WATER TO
_ER
BOILER FEED PUMPS
TEMPERATURE
CONTROLLER
TEMPERATURE
INDICATOR
PRESSURE GAGE
BACK PRESSURE
REGULATOR
Figure 24. Boiler test-loop boiler feedwater section
70
-------
\
X
Figure 25. Pilot boiler control panel (A). Superheaters can be seen on the left (B)
-------
Figure 26. Backside of pilot boiler showing configuration:
A) steam drum, B) boiler drum and C) mud
drum.
72
-------
All components of the boiler test loop except the pressure
vessels and boil-off condenser were standard commercially avail-
able units.
At the conclusion of each boiler test run, the boiler was shut-
down, and the heating element was removed, photographed, and sent
to the laboratory for scale analysis.
INDIVIDUAL EQUIPMENT DESCRIPTION
Deaerator Feed Pumps
Small sliding vane rotary pumps delivered water from the feed-
water storage tank to the deaerator. Each pump delivered 3.79 1pm
(1 gpm) at 2.8 Kg/cm2 (20 psig) recirculated excess flow back to
the feedwater storage tank.
Chemical Feed Pump and Tank
A 5.68-1/hr (1.5-gph) adjustable rate diaphragm pump injected
boiler water treatment chemicals into the boiler feed pump suction
line. Chemicals were mixed in the chemical feed tank, and the
pumping rate was adjusted to maintain proper chemical concentra-
tions in the boiler. A back pressure regulator held constant
pressure on the pump discharge to maintain accurate pumping rate
control.
Hydrazine Feed Pump and Tank
A dilute solution of hydrazine was pumped by a diaphragm
pump with a capacity of 5.68 1/hr (1.5 gph) into the feedwater as
it left the deaerating heater. The pumping rate was adjusted to
maintain the proper hydrazine residual in the boiler. A back
pressure regulator held constant pressure on the pump discharge
to maintain accurate pumping rate control.
Deaerating Heater
The deaerating heater was a 950-1 (250-gal) atmospheric vessel
designed to heat the incoming water to 100 C (212°F) to remove
the majority of the dissolved oxygen and CO2 from the boiler feed-
water. Water level was maintained by a float operated level con-
trol valve. Outlet water temperature was maintained by the steam
inlet control valve. Steam for the deaerator was supplied from
the plant steam header.
Boiler Feed Pumps
The boiler feed pumps were positive displacement plunger
pumps, each having a maximum capacity of 133 1/hr (35 gph).
Pumping rate was adjustable to match the boiler demands. One
boiler feed pump was designed to operate continuously while the
73
-------
other was turned on and off by the boiler level controller to
maintain proper drum level. The feed pump, operating continuously,
was set for slightly less than boiler requirements; the intermit-
tent pump was set so that the combining pumping rate of both pumps
was slightly greater than boiler requirements. Safety valves on
each pump, set at 133.6 Kg/cm2 (1,900 psig) protected pumps against
overpressure in the event of discharge line blockage.
Boiler
The boiler and superheater were constructed of carbon steel
and had a design pressure rating of 140.6 Kg/cm2 (2000 psig) and
a design superheater steam temperature of 399°C (750°F).
The steam generating section contained six electrical heating
elements and the superheater section contained five heating ele-
ments. Steam output was determined by heating wattage. Charac-
teristics of the different heating elements are shown in Table 18.
TABLE 18.
CHARACTERISTICS OF HEATING
ELEMENTS
Boiler High
Boiler Medium
Boiler Low
Superheater
Wattage
Watts
9650
9600
4915
2200
Heat Release
BTU/hr/sq ft
240,000
180,000
100,000
22,000
Steam Output
Ibs/hr
180
180
90
-
Operating steam pressure was controlled by the steam back
pressure control valve.
Superheated steam temperature was controlled by turning
superheater heating elements on and off in response to the output
signal from the temperature controller. Superheater heating
elements were interlocked with the condensate flow signal to pre-
vent operation unless steam flow past the heating elements
exceeded a preset minimum.
Boiler drum level was maintained by the electrode type level
controller which started and stopped the intermittent operating
boiler feed pump, alarmed and shut off boiler heating elements on
low level, and alarmed and shut off continuous operating feed
pump on high level. A high pressure shutdown and alarm was
activated if boiler pressure exceeded 126.6 Kg/cm2 .(1800 psig).
74
-------
Continuous Slowdown System
Solids concentration in the boiler water was controlled by
the blowdown timer which regulated the amount of time the blowdown
valve was open.
Blowdown was cooled by a small water-cooled heat exchanger
and collected in a graduated measuring pot.
Bottom Blowdown System
Sampling and blowdown of the boiler mud drum were accom-
plished by manually opening the blowdown valve. Blowdown was
cooled by a small water cooled heat exchanger.
Saturated Steam Sampling System
Saturated steam samples were condensed by a small water-
cooled heat exchanger and were collected manually.
Superheated Steam Condensing System
Superheated steam produced by the boiler was condensed in
order to measure corrosivity of the condensate and to accurately
measure boiler output. Where an adequate cooling water supply is
available, this would normally be done with a water-cooled con-
denser. Since the boiler test loop was installed in a water-
short area/ the following equipment was used to minimize water
usage.
Boil-off Condenser—
A stainless steel coil in an open vessel boiled off waste
condensate from the boiler test loop at atmospheric pressure,
thereby removing heat from the steam inside the coil.
Air Cooled Condenser—
Steam and condensate from the boil-off condenser were further
condensed and cooled to approximately 71°C (160°F) by a fan-cooled
air condenser.
Condensate Trim Cooler—
A small water cooled heat exchanger was used to cool the con-
densate to approximately 43°C (110°F).
Corrosion Test Loop
A series of pipe fittings was designed to hold four standard
corrosion test coupons for evaluating corrosion characteristics of
the steam condensate (Figure 27).
75
-------
ro
H-
CTC}
P
hi
0)
to
•si
H-
M
ED
0)
po
3
o
o
S3
0
en
o
§
3
en
H.
o
o
o
O
p
H-
r+
H
-------
TEST BOILER EXPERIMENTAL APPROACH
The primary objective of the test boiler studies was to
determine whether the renovated wastewater would be suitable for
use as boiler feedwater. Normally, the main concern about any
feedwater impurity is its ultimate corrosivity and/or deposition
potential within the primary steam generating area of a boiler.
Acceptable practice conditions are based on the mineral consis-
tency of the water (hardness, iron, copper) and the boiler de-
sign critiera. Because the water in question was derived from
an organic-chemical plant wastewater that might still contain
small quantities of organic chemicals, it was necessary to
establish whether these chemicals would adversely affect not
only the primary steam generating areas of boilers, but also the
superheater section. It was also necessary to evaluate whether
potentially volatile organic chemicals would carry into the
steam supply and condensate systems and cause corrosion or
.deposition problems.
To this end, test boiler studies were completed that com-
pared existing plant boiler feedwater (demineralized well-water)
to renovated wastewater as boiler feedwater. The results of
these experimental boiler studies were then compared to the
actual and documented internal condition of the petrochemical
plant boilers. The various full-scale plant boilers are typical
42.18 Kg/cm2 (600 psig) wasteheat boilers, 77.84 Kg/cm2 (1100
psig) CE tangentially fired power boilers, and 119.52 Kg/cm2
(1700 psig) wasteheat boilers. These have been in service for
six years using the same plant boiler feedwater (demineralized
well-water) used to complete the experimental boiler studies.
Although the test boiler was designed for 119.02 Kg/cm2 (1700
psig) operation, it was necessary to limit actual operating
pressure to 105.46 Kg/cm2 (1500 psig) because of safety concerns
with the drum level gage glass. It became apparent after
repeated failures of the 240,000-BTU/hr/ft2 heating elements
during tests using demineralized well-water that the physical
configuration of the test boiler would not allow evaluation of
the renovated wastewater at these extremely high heat release
rates.
Operating parameters chosen for primary investigation were
105.46 Kg/cm2 (1500 psig) drum pressure, heat transfer rates of
180,000 BTU/ft2>hr and 50 cycles of feedwater concentration with-
in the boiler. The physical testing parameters were selected
based on actual plant operation of the organic chemical manufac-
turing facility. Test boiler design controlled the steam flow
to approximately 72 Kg (160 lb)/hr. Test duration was estab-
lished at 2, 4, or 8 days, depending on individual test
requirements. It was not a purpose of this study to evaluate
treatment chemical performance. All tests were completed using
accepted chemical treatment programs consisting of an oxygen
77
-------
scavenger (applied to the deaerator), a standard coordinated
pH/phosphate internal boiler water treatment plus a proprietary
blend of polymeric dispersants and antifoam agents for scale and
foam control (applied to the boiler steam drum), and a proprie-
tary blend of volatile amines for steam and condensate line
corrosion protection (applied to the boiler steam drum).
SAMPLING AND ANALYSIS
Boiler feedwater, boiler blowdown water, steam, and steam
condensate samples for each experiment were obtained every four
hours during boiler operation. Boiler feedwater was monitored
for hardness (calcium and magnesium ions), pH, conductivity and
total organic carbon (TOC). The boiler blowdown was analyzed
for conductivity, pH, alkalinity, phosphate, silica, iron and
TOC. This was necessary to insure that pH and phosphate were
maintained within the prescribed limits, and to accomplish this
the chemical treatment was adjusted accordingly. Conductivity
and silica analysis were completed for comparison to ABMA (Amer-
ican Boiler Manufacturers Association) prescribed standards.
Steam and condensate samples were analyzed for conductivity,
pH, silica and TOC to monitor general steam purity (sodium,
silica, conductivity) and to determine the degree of organic
carryover (TOC). While condensate analysis was another indi-
cation of steam purity, it was primarily used to monitor
condensate corrosivity. Average values of the pertinent
analyses for each boiler test are shown in Table 19. Analytical
methods were conducted based upon Standard Methods (2), Methods
of Chemical Analysis of Waster and Wastes (1) and Betz Handbook
of Industrial Water Conditioning (4).
Scaling potential within the boiler was a major part of this
study. The determination of deposition rate and volume consisted
of quantitative analysis of the deposit formed on the high heat
transfer area of the electric heat probes. The wet analytical
methods employed for quantification of the deposit are described
in Appendix D. The weight of boiler probe deposits were deter-
mined to enable comparison of deposit weights for the various
experiments and the method is shown in Appendix E. The weights
of probe deposits are listed in Table 20 for each successful
boiler test run.
RESULTS AND DISCUSSION
To reiterate, the primary purpose of the investigation was
to document corrosion and deposition potential of renovated
wastewater when used as boiler feedwater, compared to demineral-
ized well-water. Over twenty boiler tests were attempted.
However, all but fourteen were aborted due to mechanical problems
such as sight glass failures, safety valve failures, feed pump
check valve problems and heating element failure. Eight addi-
tional experiments were considered invalid because of poor
78
-------
TABLE 19. AVERAGE BOILER FEEDWATER AND SLOWDOWN ANALYSIS
PH
FEEDWATER
COND TOC MALK TH PH
BOILER SLOWDOWN
COND TQC PALK MALK PO^ SiOo Fe
Deraineralized
Well-Water
Run 9
Run 17
Run 20
Renovated
Wastewater
Run 10
Run 12
Run 16
7.5
7.7
7.7
2
1
1
7.7 4
7.6 3
8.0 1
8
14
3
10
0
0
0
0
0
2 0
0 0
0 0
10.1
8.3
9.2
9.4
9.5
9.6
332
127
141
313
146
133
17
3
3
15
5
11
42
11
10
14
23
15
65
37
40
48
55
33
27
16
26
7 0.1
5 0
5 0.2
49 8
36 11
16 5
0.2
0.1
0.1
COND = Specific Conductance (JJmho)
TOC = Total Organic Carbon (ppm)
PALK = P Alkalinity (ppm as CaCOs)
MALK = M Alkalinity (ppm as
TH = Total Hardness (ppm as
P04 - Phosphate (ppm)
Si02 = Silica (ppm)
Fe = Iron (ppm)
-------
TABLE 20. STEAM GENERATOR PROBE DEPOSIT ANALYSIS
Demineralized
Weil-Water
Test Duration
(hours)
Probe A
(g/m2)
Probe B
(g/m2)
RUN NO.
9
17
20
96
96
181
2.56
2.08
4.52
2.49
1.33
3.97
Renovated
Wastewater
HUN NO.
10
12
16
62
96
192
0.78
0.65
0.45
1.00
1.18
0.47
80
-------
chemical control. Chemical imbalances were believed to be due
to residuals (magnesium oxide insulation) left after heating
element failure and insufficient chemical cleaning between tests.
Six valid experiments were completed; three in which plant
boiler feedwater (demineralized well-water) was used, three in
which renovated wastewater was used. All of the operational
parameters (previously defined) were constant in these tests,
except that of test duration. Test duration was varied so that
deposition rate comparisons could be made between the feedwaters
employed. The heat transfer rate of 180,000 BTU/ft2-hr was
selected as typical for an average radiant heat section of most
industrial boilers constructed today. The operating pressure was
set at 105.5 Kg/cm2 (1500 psig). Tests were not run at lower
heat flux or pressure levels because past practice has shown that
as test severity is decreased, the potential for deposition with-
in the boiler decreases.
Analyses listed in Table 19 indicate only subtle differences
between the feedwater quality for each experiment. The differ-
ences listed may be in part "real" differences, or may reflect
only analytical variability with the very high water quality. It
must be emphasized that the analytical instrumentation and
methods employed (pH, conductivity, TOC analyzer, alkalinity
measurement) are significantly less sensitive when used for the
analysis of very high purity water. These same variations may
be to a large degree responsible for the differences seen in
deposition weights listed in Table 20. However, all these
differences fall within expected test deviations of boiler probe
deposits.
A comparison of the feedwater analysis between the experi-
ments using demineralized well-water shows little variation. A
similar comparison of the feedwater analyses between the experi-
ments using renovated wastewater also shows little variability
but the levels of TOC are slightly higher. This does not pre-
clude the use of renovated wastewater as boiler feedwater. A
review of the analytical data for the boiler blowdown water shows
no significant difference between any of the experiments report-
ed (demineralized well-water or renovated wastewater). The
differences recorded could easily be due to the subtle differ-
ences in the feedwater constituency or due to slight operational
differences such as cycles of boiler water concentration or
boiler steam flow.
The amount of deposit formed on the electric heater probes
is shown in Table 20 and plotted in Figure 28. The total deposit
listed is the sum of the individual components (calcium, magne-
sium, iron, phosphate and silica) that constitute boiler scale
formed under the conditions employed in these experiments. The
data obtained using demineralized well-water shows acceptable
deviations between individual heater probe deposit weights within
81
-------
oo
K)
w
bC
CO
0
ft
a
«H
o
bo
•H
5.0
4.0
3.0
2.0
1.0
Demineralized Well-Water
Renovated Wastewater
50
100 150 200
Test Duration, hours
Figure 28. Test-boiler heater probe deposition.
-------
single experiments, and between duplicate experiments under
identical operating conditions (Table 20 - Tests 9 and 17). The
differences seen are only subtle and likely to be caused by dif-
ferences in feedwater constituency and/or operating parameters.
The increase in deposit weight with time seen with the demineral-
ized well-water is typical when tests are completed under severe
operating conditions and/or when marginal chemical treatment is
employed. The data obtained using renovated wastewater as feed-
water exhibited similar scatter. However, when the renovated
wastewater was used there was no significant change in deposit
quantity with extended test duration (Figure 28). This is
indicative of less severe operating conditions and optimum chem-
ical treatment. The only difference between the two sets of data
is that of feedwater employed; therefore, it can be concluded
that the use of this renovated wastewater represents a lower
scale potential than the demineralized well-water. Since the use
of renovated wastewater as feedwater posed no greater deposition
problem in the experimental boiler than the demineralized well-
water, and because past experience has shown that the demineral-
ized well-water presents no unusual scale problems in the full-
scale plant boilers, it can be further concluded that the reno-
vated wastewater should present no unusual deposition problems
in the full-scale plant boilers.
In order to assess steam purity and potential superheater
deposition, low heat flux (22,000 BTU/ft2 hr) electric probes
were installed into the boiler system to raise steam temperature
from saturation 314°C (598°F) up to 390°C (750°F) superheat
temperature. Steam samples were taken prior to the superheat
section of the experimental boiler, and samples were taken of
condensed superheated steam (return condensate). These data
shown in Table 21 reveal some degree of contamination, but no
significant difference between tests completed using either
demineralized well-water or renovated wastewater. A review of
the superheater probe deposits in Table 22 show no definitive
differences in deposit quantity between experiments. It cannot
be concluded absolutely from these experiments that either
demineralized well-water or renovated wastewater will not impair
boiler superheater operation. However, because the experiments
completed using either feedwater gave similar results, it is
expected that the renovated wastewater could be used in the
full-scale plant boilers without problems any greater than those
using existing plant boiler feedwater (demineralized well-water).
A corrosion test loop was installed in the return condensate
line of the boiler system. Mild-steel and copper specimens were
placed in the test loop for each boiler experiment conducted.
Specimen weight loss was recorded and converted to average pene-
tration rate. These data recorded in Table 23 using the calcu-
lation shown in Appendix F show a comparison between boiler tests
using renovated wastewater and demineralized well-water. The
data obtained from experiments when demineralized well-water was
83
-------
TABLE 21. BOILER STEAM AND RETURN CONDENSATE ANALYSIS
Boiler Steam
pH COND TOC SiO0
£t
Demineralized
Weil-Water
Run 9 7.8 8 5 0.3
Run 17 7.7 2 0
Run 20 7.9 2 20
Renovated
Wastewater
Run 7 7.6 4 25 0.1
Run 8 7.6 8 80
Run 10 8.0 5 13 0
Run 12 7.9 3 1 0.1
Run 16 8.2 2 20
Return Condensate
pH COND TOC SiO0
/2
7.9 550
7.9 2 - 0.1
8.0 120
7.6 3 5 0.2
7.8 8 6 0.1
8.3 7 12 0.1
8.0 5 1 0.1
8.4 230
COND = Specific Conductance (umho)
TOC = Total Organic Carbon (ppm)
SiO0 = Silica (ppm)
84
-------
TABLE 22. STEAM SUPERHEATER PROBE DEPOSIT
Demineralized Probe Deposit
Well-Water Test Duration (hrs) (g/m2)
Run 9 96 .28
Run 17 96 .28
Run 20 181 .24
Renovated
Wastewater
Run 10 62 .34
Run 12 96 .36
Run 16 192 .21
85
-------
TABLE 23. BOILER STEAM CONDENSATE CORROSION STUDIES
Demineralized Test Duration,
Weil-Water hours
Run 17 96
Run 20 181
Renovated
Wastewater
Run 7 96
Run 8 96
Run 10 62
Run 12 96
Run 16 192
Corrosion Coupon Weight
Loss, mils/year
Mi Id- Steel Copper
6.3
5.3
1.2
1.0
1.3
1.1
3.4
4.5
18.4
16.3
0.8
0.6
2.7
1.9
4.2
4.0
0.6
0.4
-
17.9
17.1
1.0
1.2
2.9
3.5
86
-------
evaluated revealed satisfactory corrosion control of mild steel
and copper specimens (Tests 17 and 20). The data obtained from
the experiments conducted using renovated wastewater show ques-
tionable corrosion control of both mild-steel and copper speci-
mens in experiment 10 and good corrosion control in the other
four experiments (7, 8, 12, and 16). A review of the condensate
analysis does not show cause for the poor results in Test 10.
While it cannot be concluded that condensate derived from reno-
vated wastewater was more corrosive than condensate derived from
demineralized well-water, it would be advisable to monitor cor-
rosion rate when renovated wastewater is initially applied to
the full-scale plant boilers.
From these pilot-scale boiler tests evaluating the feasi-
bility of using renovated wastewater for boiler feedwater it is
concluded that:
• Any impurities that passed through the wastewater treatment
system did not noticeably affect boiler deposition; that is
in comparison to impurities carried through standard plant
demineralizer systems.
• The amount of waterside deposit when using renovated waste-
water was less than that produced using plant boiler feed-
water (demineralized well-water).
• The quality of steam produced from the renovated wastewater
was equivalent to that generated from demineralized well-
water.
• The amount of superheater deposition using renovated waste-
water was equivalent to that produced using demineralized
well-water.
• The condensate derived from using renovated wastewater as
boiler feedwater appeared slightly more corrosive than did
the condensate derived from demineralized well-water.
87
-------
SECTION 6
CYCLE COOLING WATER TEST LOOP
A recycle cooling water pilot plant was operated in con-
nection with the EPA test program to study the use of
treated wastewater as makeup to an open, recirculated cooling
water system.
Four treated wastewaters were studied:
1. Reverse Osmosis Permeate
2. Activated Carbon Effluent
3. Sedimentation/Filtration Unit Effluent
4. Clarified Activated Sludge Effluent
Both chromate and non-chromate corrosion inhibitors were studied
to evaluate corrosion and/or fouling of carbon-steel (A-214),
austenitic stainless steel (A-249) and Admiralty (B-lll) heat
exchanger tubes and test coupons.
An evaluation of any water for makeup to a cycle cooling
water system requires an examination of several cooling system
effects as well as their interactions, including biological
fouling, corrosion and heat transfer characteristics. A par-
ticular water may, for example, produce good heat transfer
but demonstrate excessive corrosivity, preventing its practical
use as cooling tower makeup. Conversely, another water may
produce low corrosion rates due to the formation of a protective
type scale but have poor heat transfer characteristics. The
cycle cooling water test program was designed to enable an eval-
uation of the major factors and effects controlling the use of
a water as makeup as well as their relationships to each other.
In this way an overall acceptability of a particular makeup
water is determined.
GENERAL FACILITIES DESCRIPTION
The cooling water test-loop was designed to permit the eval-
uation of the scaling, fouling and corrosion effects of the
renovated wastewater when used as makeup water. Standard,
commercially available equipment was utilized in the assembly of
88
-------
the test facility. The primary components of the test-loop
illustrated in Figure 29 are a cooling tower, heat exchangers
with appropriate temperature measurement equipment, and steam
condensate collection tanks.
Forced-Draft Cooling Tower
The cooling tower utilized in the test was a standard com-
mercial, packaged, forced-draft cooling tower rates to cool 15
gallons per minute of water from 37.8°C to 29.4°C (100°F to 85°F)
with an ambient wet-bulb temperature of 25.6°C (78°F). The cool-
ing tower was rated for a maximum cooling capacity of 110,000
BTU/hr. The cooling tower, flow meters and acid/caustic storage
tanks and feed pumps are shown in the photograph in Figure 30.
Acid/Caustic Feed Pumps
Positive displacement pumps rated to feed 1.9 liters (0.5
gallons) per hour fed acid or caustic as indicated by the pH
meter/controller to maintain the desired pH of the circulating
water.
Heat Exchangers
A 1.5 hp cooling-water circulating pump took suction from
the cooling tower basin and pumped the cooling water through
three double tube heat exchangers (Figure 31) operated in paral-
lel and returned it to the cooling tower. Heat load was
supplied on the shell side by reducing 200 psig steam to 8 psig.
The three heat exchangers were approximately 76 cm (30 inches)
long having a heat transfer area of 0.30 m2 (0.323 ft2) per
tube. One heat exchanger was tubed with two carbon-steel tubes
(A-214); one heat exchanger with two copper alloy tubes (B-lll)
and the third heat exchanger with two stainless steel tubes
(A-249), thus permitting simultaneous evaluation of the effects
of the water and inhibitors on different materials of construc-
tion. The heat-exchanger tubes were connected to the circul-
ating water lines by a utility hose to prevent galvanic
corrosion at the tube connection point, and to facilitate easy
removal of the tubes for replacement, inspection, or cleaning.
Condensate Collection Tanks
Steam condensed on the shell side of each heat exchanger
was collected in small condensate collection tanks. Each tank
was equipped with level switches and timers to enable measure-
ment of condensate flow for use in subsequent heat transfer
coefficient calculations. Condensate collection tanks can be
seen in the lower right corner of the photograph in Figure 32.
89
-------
CA.E.6OU
l//FILTR»ri»»V
ML 1
.'CAU
AIE IMLET-^
©V
MB. tXHAUST
ADD
Nl
MANUAL
CMEM.'CAL
FEED
1
1
1
L.
i
€
0
POMPi
0
LEGEND
Flow Indicator
Temperature Indicator
Temperature Element
Level Switch
Valve Switch
pH Meter/Controller
Level Control Valve
Hand Switch
COLLECTION
bLDWDOWM
Figure 29. Cycle cooling-water test loop facilities.
-------
Figure 30. Forced-draft cooling tower and acid/caustic addition facilities
-------
Pressure Gauge
8
tso
Hot Water Line
Thermometer
To Drain
Condensate
to Waste
Figure 31. Test heat exchanger.
-------
Figure 32.
Cooling-water test loop heat exchangers (A) and
condensate collection tanks (B).
93
-------
SAMPLING AND ANALYSIS
Data were studied in the areas most important to evalu-
ating the reuse feasibility of renovated wastewater in an
open recirculating cooling water system. The cooling water
test loop was monitored closely by a rigorous sampling and
analysis program consisting of:
• Daily complete laboratory chemical analysis
of makeup and cycle water.
• Field analysis of relevant factors several
times per shift (e.g. chlorides, inhibitor,
residual chlorine).
• Measurement of circulation rate and tempera-
ture differential several times per shift
and calculation of evaporation rate, makeup
and blowdown.
The recirculating cooling water system operated with a cir-
culation rate of 95 1pm (25 gpm) and a temperature differential
on the order of 8.3°C (15°F). Inlet water was about 29.5°C (85°F)
The water velocity through the tubes was about 1.5 m/sec (5 ft/
sec). At four cycles of concentration, the makeup was approxi-
mately 1.5 1pm (0.4 gpm), and the blowdown was about 0.4 1pm
(0.10 gpm). Each test was preceded by an initial pretreatment
of approximately 24-hour duration to effect cleaning and prepara-
tion of tubes for uniform inhibitor treatment.
COOLING WATER TEST EXPERIMENTAL APPROACH
Effluents from the reverse osmosis, carbon column, sedi-
mentation/filtration unit, and secondary clarifier were used
as makeup to the tower. The dissolved solids concentration in
the circulating water was controlled at the desired level by
blowing down a portion of the circulating water. The pH of
the circulating water was controlled automatically by the
injection of caustic or acid with a proportioning pump. Water-
treating chemicals and biocides were added manually.
Each cooling water test was conducted over approximately
30 days. Both chromate (chrome/zinc) and non-chromate (zinc/
phosphate) inhibitors were evaluated. Sodium hypochlorite
was added to maintain 0.2 ppm free chlorine to control bio-
logical growth. The chemical treatments used for each test
are summarized in Table 24.
94
-------
TABLE 2k. CHEMICAL TREATMENT FOR COOLING WATER PILOT TESTS
Test Run
Number Makeup Water
#1 Reverse Osmosis
Permeate
#2 Activated Carbon
Column Effluent
#3 Activated Carbon
Column Effluent
Dates
5/26
to
6/23
(28 days)
6/2U
to
7/28
(3»t days)
8/2
to
9/1
(30 days)
Chemical Treatment ^'
(A) Zinc /Phosphate Inhibitor
(B) Phosphonate Dispersant
(C) Phosphonate Inhibitor
Sodium Hypochlorite
Sulfuric Acid
Sodium Hydroxide
(D) Chrome/Zinc Inhibitor
(B) Phosphonate Dispersant
Sodium Hypochlorite
Sulfuric Acid
Sodium Hydroxide
(A) Zinc /Phosphate Inhibitor
(B) Phosphonate Dispersant
(C) Phosphonate Inhibitor
Sodium Hypochlorite
Sulfuric Acid
Sodium Hydroxide
Chemical Control
Level
3.0 ppm Zn
2 times (A) added
G—10 ppm POlj
pH 6.8-7.2
0.2 ppm Free C12
25-30 ppm CrO,
2 times (D) added
• 0.2 ppm Free Cl2
pH 6.2-6.8
Same as Test #1
#1* Sedimentation/
Filtration Unit
Effluent
9/3 (A) Zinc/Phosphate Inhibitor
to (B) Phosphonate Dispersant
10/3 (C) Phosphonate Inhibitor
Sodium Hypochlorite
(30 days) Sulfuric Acid
Sodium Hydroxide
Same as Test #1
#5
#6
Sedimentation/
Filtration Unit
Effluent
Clarified Activated-
Sludge Effluent
10/12
to
11/9
(28 days)
11/11
to
12/8
(27 days)
(D) Chrome/Zinc Inhibotor
(B) Phosphonate Dispersant Same as Test #2
Sodium Hypochlorite
Sodiun Hydroxide
Sulfuric Acid
Sodium Hydroxide pH 6.8-7.2
Sulfuric Acid
Sodium Hypochlorite 0.5 ppm Free Cl-
71)
All heat exchange equipment was pretreateci with 600 ppm of a zinc/polyphosphate type
inhibitor circulated for k hours at a pH range of 5 to 7. Following pretreatment, the
system was blown down until the phosphate level was at 20 ppm.
95
-------
• Measurement several times per shift of steam,
condensate and cooling water parameters
(pressure, temperature and flow, as appropriate).
The chemical analyses and the frequency collected on the inlet
water and recycle water included:
Analysis
Total Dissolved Solids
Chlorides
Hardness
Alkalinity
Sulfate
Phosphate
Iron
Chromate
Zinc
Calcium
Magnesium
pH
Residual Chlorine
Conductivity
Frequency
Daily
Daily/Hourly
Daily
Daily
Daily
Daily
Daily
Daily
Daily
Daily
Daily
Hourly
I/Shift
Hourly
Sample
24-hr
24-hr
24-hr
24-hr
24-hr
24-hr
24-hr
24-hr
24-hr
24-hr
24-hr
Grab
Grab
Grab
composite
composite/grab
composite
composite
composite
composite
composite
composite
composite
composite
composite
In addition to the chemical analyses, routine measurements
were taken at various points in the process to enable sufficient
data for mass balances and heat transfer calculations.
Corrosion Test Coupons
Test coupons constructed of A-214 carbon steel, A-249
stainless steel and B-lll admiralty brass were exposed to the
recirculating cooling water in the cooling tower sump. Three
racks with four coupons each of the three tested materials were
initially exposed in each test. Coupons were removed in suc-
cessive weekly intervals to observe and quantify the general
corrosion rate as calculated by weight loss per unit time per
unit area.
Heat Exchanger Tubes
Test heat exchanger tubes of A-214 carbon steel, A-249
stainless steel and B-lll admiralty brass were removed following
each test run; the tubes were split, photographed before and
after cleaning, pit depth and density measured, and scale
thickness determined. The chemical composition of the scale
was determined by analysis. A chemical mass balance was per-
formed based on the average chemical analysis to quantify the
chemical precipitation occurring within the system for cor-
relation with other fouling test data. The heat exchangers
were monitored routinely as follows:
96
-------
Measurement Frequency
Steam Flow, Ibs/hour (calculated) Every 2 hours
Steam Pressure, psig Every 2 hours
Steam Inlet Temperature Every 2 hours
Condensate Outlet Temperature Every 2 hours
Water Flow Every 2 hours
Water Inlet Temperature Every 2 hours
Water Outlet Temperature Every 2 hours
Heat Transfer Data
The coefficients of heat transfer were calculated from
water and steam side data. Of the two values, the water side
coefficient data was determined to be the most reliable. This
value was plotted against time to delineate any loss of heat
transfer during the test run. The fouling factor for deposition
(Rg) was calculated as the difference of the reciprocals of the
water side heat transfer coefficients between the termination
and the beginning of the test run. In addition to the above
steam and water measurements, the cooling tower was routinely
monitored:
Measurement Frequency
Recirculating Water Concentration Every 2 hours
Circulation Rate Every 2 hours
Inlet and Outlet Temperature Hourly
Evaporation Rate (calculated) Hourly
Blowdown Rate (calculated) Hourly
Makeup Rate (calculated) Hourly
Chemical Treatment and Controls
The levels of chemical inhibitor, dispersant and residual
chlorine were measured and adjusted once per shift. The levels
of these treatments maintained during each test are summarized
in Table 24.
Biological Fouling
Biological factors were not investigated in depth; however,
observations of the cooling tower fill and wood samples placed
in the cooling tower sump were observed during each test run
and the degree of biological fouling evaluated.
97
-------
RESULTS AND DISCUSSION
Corrosion Data
Corrosion effects were determined by measuring the actual
weight-loss of metal strips (coupons) of known initial weight
removed from the cooling-tower sump each week, and by measur-
ing the degree of pitting on the inside walls of the heat
exchanger tubes. Individual general corrosion rates were
calculated in mils/year and plotted vs. time of exposure in
the tower sump (Figure 33) for A-214 carbon steel.
The effectiveness of the inhibitors tested was established
in a test using activated sludge effluent with only pH ad-
justment and biological growth control (Test 6). The initial
corrosion rate was high and increased during the first week
with no added inhibitors. The decline in corrosion rate
after the first week resulted from the formation of iron
oxides which, after initial corrosion, redeposit to form a
barrier-type corrosion protection. A 6.0 mil/year corrosion
rate is probably the lowest that could be achieved with this
wastewater without inhibitor addition.
Based upon the terminal general corrosion rate of each test,
these cooling-waters, in a circulation system using 16-gage
carbon-steel exchanger tubes and a corrosion allowance of
one-half the wall thickness, would have a projected tube life
as shown in Table 25. Only the activated sludge effluent
without inhibitors showed unacceptable tube life, with RO
permeate being marginal (assume 5 mils/yr as maximum acceptable
corrosion rate). The RO test was not representative and,
therefore, the results should not be concluded as unacceptable.
It follows that RO as makeup would be acceptable with chrome/
zinc treatment in that the activated carbon water which has a
higher total dissolved solids content was acceptable.
In terms of corrosion and pitting of carbon-steel heat
exchanger tubes, pitting was found to be the factor control-
ling exchanger tube life and the selection of chemical treat-
ment. Pit depth and density in the exchanger tube measured
mechanically following each test run are also shown in Table 25.
Photographs of the split heat exchanger tubes, before and
after cleaning are shown in Figures 34 through 45.
It appears that, from the standpoint of pitting, chrome/
zinc was the only acceptable means of treatment for a carbon-
steel system. Since A-249 stainless steel and B-lll admiralty
brass showed negligible corrosion and pitting in all tests,
the data are not presented. These materials would be suitable
for all wastewaters and inhibitors tested from the standpoint
of general corrosion.
98
-------
§
•H
00
o
(4
o
u
16.0
14.0
12.0
10.0
8.0
6.0
4.0
2.0
Test 1
R. O. Permeate
Test 6
Bio-effluent
Test 5
ediraenta-\
. tion/
iltration
Unit
Test 3
Carbon
Eff.
Test 4
Sedimenta-
tion/
. Filtratiorf
Unit
Test 2
Carbon
Elf.
- Zinc/Phosphate Inhibitor
— Chrome/Zinc Inhibitor
1234
Corrosion Coupon Exposure Tine, Weeks
Figure 33. Corrosion rates for A-214 carbon-steel test coupons.
99
-------
TABLE 25. CYCLE WATER TEST LOOP CORROSION DATA
Test
# Cooling Water
1 BO Permeate *3)
2 Activated
Carbon
Effluent
3 Activated
Carbon
H Effluent
o
4 Sedimentation/
Filtration
Unit Effluent
5 Sedimentation/
Filtration
Unit Effluent
6 Clarified
Activated
Sludge
Effluent
Inhibitor
Zn/PO4
Cr/Zn
Zn/P04
Zn/P04
Cr/Zn
None
Corrosion^ '
A-214 Carbon Steel Coupons
Terminal Corrosion Tube /9-.
Rate, Mils/Year Life. Yrsu'
4.5 7.1
1.0 32
1.8 17.1
1.0 32
1.0 32
6.1 5.3
(1 \
Pitting1 ;
Heat Exchanger Tubes, A-214
Carbon Steel
Maximum Pit Pitts Tube ,_,
Depth, Mils/Yr 3d. In. Life^ Yrsu;
144 15-20 .25
< 1.0 < 0.5 > 30
60 4-5 .5
60 2-3 .5
1.2 < 0.5 27
84 10-12 .3
A-249 Stainless Steel and B-lll Admiralty brass showed negligible corrosion and pitting in all test runs.
(2) Based upon 16 gauge carbon steel heat exchangers tubes.
(3) Not representative of good HO treatment, membrane deterioration, chlorides in permeate 100-400 mg/1.
-------
A-249 Stainless
B-lll Admiralty Brass
A-214 Carbon Steel
Figure 34.
Run 1 Cooling-water test, R. O. permeate with Zn/PO4 inhibitor.
Heat exchanger tubes before cleaning.
-------
A-249 Stainless
o
DO
B-lll Admiralty Brass
A-214 Carbon Steel
Figure '35.
Run 1 Cooling-water test, R. 0. permeate with Zn/PC-4
inhibitor. Heat exchanger tubes after scale removed.
-------
o
OJ
A-249 Stainless
B-lll Admiralty Brass
A-214 Carbon Steel
Figure 36.
Run 2 cooling-water test, activated carbon effluent with
Cr/Zn inhibitor. Heat exchanger tubes before cleaning.
-------
A-249 Stainless
B-lll Admiralty Brass
.—.
o
P.
A-214 Carbon Steel
Figure 37.
Run 2 Cooling-water test, activated carbon effluent with
Cr/Zn inhibitor. Heat exchanger tubes after scale removed,
-------
A-249 Stainless
B-lll Admiralty Brass
A-214 Carbon Steel
Figure 38. Run 3" Cooling-water test, activated carbon effluent with
Zn/PC>4 inhibitor. Heat exchanger tubes before cleaning.
-------
A-249 Stainless
B-lll Admiralty Brass
o
05
A-214 Carbon Steel
Figure 39. Run 3 Cooling-water test, activated carbon effluent with
Zn/PC>4 inhibitor. Heat exchanger tubes after scale removed
-------
A-249 Stainless
o
•sj
B-lll Admiralty Brass
A-214 Carbon Steel
Figure 40. Run 4 Cooling-water test, sedimentation/filtration unit
effluent with Zn/P04 inhibitor. Heat exchanger tubes before
cleaning.
-------
A-249 Stainless
o
OC-
B-lll Admiralty Brass
A-214 Carbon Steel
Figure 41. Run 4 Cooling-water test, sedimentation/filtration unit
effluent with Zn/PC>4 inhibitor. Heat exchanger tubes
after scale removed.
-------
A-249 Stainless
o
tO
B-lll Admiralty Brass
A-214 Carbon Steel
.
.,' -•
I
Figure 42.
Run 5 Cooling-water test, sedimentation/filtration unit effluent
with Cr/Zn inhibitor. Heat exchanger tubes before cleaning.
-------
A-249 Stainless
M
H
o
B-lll Admiralty Brass
A-214 Carbon Steel
Figure 43. Run 5 Cooling-water test, sedimentation/filtration unit effluent
with Cr/Zn inhibitor. Heat exchanger tubes after scale removed.
-------
A-249 Stainless
B-lll Admiralty Brass
A-214 Carbon Steel
Figure 44. Run 6 Cooling-water test, biological effluent no inhibitors
Heat exchanger tubes before cleaning.
-------
A-249 Stainless
B-lll Admiralty Brass
to
A-214 Carbon Steel
Figure 45.
Cooling-water test, biological effluent no inhibitors.
exchanger tubes after scale removed.
Heat
-------
Average Chemical Analyses and Mass Balance
The average chemical analyses for each run are listed in
Table 26. These analyses represent the average of daily
chemical tests during each test run. Using this data, the
relative chemical precipitation of total hardness and phos-
phate was calculated based upon Run #4 which demonstrated
the highest hardness/phosphate precipitation. For example,
in Run #1 with 4 cycles of concentration (Cl) and a measured
hardness in the feed of 45 mg/1 the expected recycle concen-
tration would be 4 x 45, or 180 mg/1. Only 145 mg/1 was
measured in the recycle, indicating 35 mg/1 of hardness and
precipitated. In Run #4, 103 mg/1 of hardness precipitated,
the highest degree of hardness precipitation for any of the
test runs. Based upon a scale of 0 to 10 with 103 mg/1 set
at 10 the relative degree of hardness precipitation in Run #1
was 35 x 10 or 3.4. This procedure was used to estimate
1G5
hardness and phosphate precipitation for each test run (Table
27). These numbers indicate the relative degree of hardness
precipitation, but do not necessarily mean that the higher
numbers result in greater scale formation. They do mean
that chemically, conditions were more ideal for scale forma-
tion. Other factors such as temperature, pH and velocity
greatly affect the actual deposition of scale. Table 27
further indicates the relative degree of the formation of
calcium sulfate scale as a function of the solubility pro-
duct of calcium sulfate. In all test runs, except test #1
using RO permeate, the solubility product of calcium sulfate
was exceeded indicating that a portion of the hardness was
precipitated as calcium sulfate. The type of hardness
precipitate will affect the heat transfer as well as corrosion
rates, due to the physical characteristics of the scale formed.
Run #1, RO permeate with Zn/PO4 inhibitor, indicated a
somewhat lesser level of chemical precipitation than the
other runs. The total dissolved solids were considerably
lower in this test run than in other tests.
Comparison of test runs using activated carbon column
effluent (Runs 2 and 3) indicated greater chemical precipitation
in Run #2 using the Cr/Zn inhibitor. However, in Run #2 the
total dissolved solids level in the recycle was nearly 85
percent higher than in Run #3. Had the total dissolved solids
levels been equivalent, Run #3 using the Zn/PO4 inhibitor would
show a higher level of hardness precipitation.
The sedimentation/filtration unit effluent using Zn/PO4
inhibitor (Run #4) showed the highest level of hardness/
phosphate precipitation and exceeded the solubility product
of calcium sulfate the least. Using the Cr/Zn inhibitor with
this makeup water (Run #5) at comparable TDS levels, the
113
-------
TABLE 26. AVERAGE CHEMICAL ANALYSIS
CYCLE COOLING WATER TESTS
Test Run Number
Assay (ppm)
Total Dissolved Solids
Chlorides
Cycles of Concentration
based upon :
Chlorides
TDS
Hardness as CaC03
Total Alkalinity as
CaCO3
Sulfate (SO/i)
Phosphate (PO4>
Iron (Fe)
Chromate (Cr04>
Zinc (Zn)
Calcium (Ca)
Magnesium (Mg)
pH
1
F(D
1668
320
45
170
120
8
2
-
_
12.6
3.2
7.9
R(2)
6685
1300
4.0
4.0
145
60
2200
12
3.0
-
12
40
10
7.2
2
F
3800
145
48
530
1288
17
.18
-
-
13
3
8.3
R
18,600
845
5.8
4.9
207
160
9,612
22
.5
36
5
60
15
7.3
3
F
2733
82
77
1012
387
24
.12
-
—
22
6
8,3
R
10,100
297
3.6
3.7
241
121
5,251
54
.35
-
5.5
67
17
6.8
4 5
F
3001
108
60
1030
585
43
.17
-
-
17
4
8.4
(
R F
20920 i3357
444
178
i
4.1
7.0
143
296
7166
81
.8
-
4.2
40
10
7.5
50
1510
532
14
.2
-
-
14
4
8.6
R
18,104
812
4.5
5.4
264
293
8,040
45
1.1
28
4.4
73
19
7.2
6
F
4348
210
91
1337
812
13
.6
—
-
25
7
8.5
R
19,872
876
4.2
4.5
352
670
6336
23
.9
—
—
98
25
7.4
(1)
(2)
F - Feedwater
R - Recycle
-------
TABLE 27. CHEMICAL MASS BALANCE
CALCULATED RELATIVE CHEMICAL PRECIPITATION
CYCLE COOLING WATER TESTS
Excess of Calcium
Test Run Phosphate '*' Hardness ^ gulf ate Solubilitv Factor
#1.
#2.
#3.
#4.
RO permeate 2.1 3.4
Zn/PO4 inhibitor
Activated carbon effluent 8.1 6.9
Cr/Zn inhibitor
Activated carbon effluent 3.4 3.5
Zn/PO4 inhibitor
Sedimentation/filtration 10 10
Not Exceeded
4.9
3.2
2.6
unit effluent
Zn/PO4 inhibitor
#5. Sedimentation/filtration
unit effluent
Cr/Zn inhibitor
#6. Activated sludge effluent
No inhibitor
pH control only
1.9
3.4
0
2.9
5.3
5.6
^ ' Based upon a scale of zero to 10 with Run #4 showing the highest hardness/phosphate
precipitation.
-------
hardness/phosphate precipitation was relatively low compared to
Run #4; while the calcium sulfate solubility product was ex-
ceeded by more than twice that of Run #4. This data would
indicate that the scale formed while using the Zn/PC>4 inhibitor
would be primarily hardness/phosphate, while that formed using
Cr/Zn inhibitor would be primarily calcium sulfate.
Using the clarified activated sludge effluent with no
inhibitors added and only pH adjustment (Run #6) chemical
analysis indicated hardness/phosphate precipitation as well as
the greatest tendency of all runs to precipitate calcium sulfate.
Heat Transfer Coefficients
The effects of solids deposition on the heat transfer
surfaces were evaluated by observing the decline in heat trans-
fer during each test period. Data representing these obser-
vations is listed in Table 28. Heat transfer coefficients were
calculated daily for both steam and water. Of the two values,
the water-side coefficient was determined to be most reliable.
The water-side coefficients were plotted with time to delineate
the loss of heat transfer during each test run (Figures 46
through 51). The difference of the reciprocal of the terminal
and initial heat transfer coefficients is the heat transfer
resistance of the scale formed during the run, referred to as
the fouling factor (Rs) and listed with the heat transfer co-
efficients in Table 28. The scale thicknesses measured and the
appearance of the scale at the termination of each test run are
also summarized in Table 28.
Based upon a maximum allowable fouling factor of .0010,
both Runs #2 and #5 using Cr/Zn inhibitor demonstrated unsatis-
factory to marginal heat transfer characteristics. Recall that
Run #2 had a higher dissolved solids level in the recycle and
would, therefore, be expected to deposit more scale. B-lll
admiralty brass showed the greatest decline in heat transfer in
all test runs except Run #1 using RO permeate and Run #4 using
sedimentation/filtration unit effluent. A-249 stainless showed
the lowest fouling factors and, therefore, the best heat trans-
fer characteristics in nearly all test runs. Without the
addition of any inhibitors or dispersants (Run #6, clarified
activated sludge effluent) acceptable fouling factors were
measured except with B-ll sdmiralty brass. Recall that cor-
rosion and pitting was high in this test which would support
the fact that without any protective scale formation heat
transfer may be good but corrosion will be unsatisfactory.
One possible explanation for the lower corrosion and
greater decline in heat transfer in the Cr/Zn test runs is
that the formation of hardness/sulfate scale is more dense and
stable than the phosphate/hardness scale formed when the Zn/P04
inhibitors are used. This is an important consideration in the
selection of corrosion inhibitors.
116
-------
TABLE 28. HEAT TRANSFER INFORMATION CYCLE COOLING WATER TESTS
Test
#
#1
#2
#3
04
#3
#6
Heat Transfer Coefficient
(Water Side)
Cooling Water Inhibitor Initial Terminal
A-Z14 A-249 B-ill A-214 A-249 B-lll
Reverse Osmosis Zn/PO. 440 400 385 335 340 373
Permeate
Activated Cr/Zn 460 440 360 240 225 ISO
Carbon
Effluent
Activated Zn/PO. 420 430 410 310 385 210
Carbon *
Effluent
Sedimentation/ Zn/PO, 525 505 570 495 450 470
Filtration *
Dnit Effluent
Sedimentation/ Cr/Zn 540 470 510 345 355 315
Filtration
Unit Effluent
Clarified pH 500 480 550 390 390 310
Activated Sludge Control
Effluent Only
R <11 Scale ...
s Thickness (mils)1*'
A-214 A-249 B-lll A-ZU A-Z4B B-lll
.0007 .0004 .0001 956
.0019 .0022 .0027 7 4 18
.0008 .0003 .0023 4 9 14
.0001 .0002 .0004 3 < 1 4
.0010 .0007 .0012 3 10 3
.0006 .0005 .0014 10 5 16
Scale Physical
Characteristics
Light Brown
Rough Toxture
Soft
Greenish White
Uniform Scale
Soft
Uniform Brown
Deposit
Rough, Brown,
Soft Deposit
Grey White Scale
with Rust Colored
Spots
Thick Rust Colored
Deposi t
(1) The reciprocal of the difference in the terminal and the Initial heat transfer coefficients representing the resistance of the scale formed to heat transfer
(Fouling Factor).
(2)
Measured average.
-------
500
400
A-214 CARBON STEEL
\
MO DATA
END OF RUN
300 ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' * * ' '
ce.
UJ
GO
300
400
300
A-249 STAINLESS
NO DATA END OF RUN
i i i I I I I I I I I I I I I I I I I I I I I I I I I I I T I I I I
B-lll ADMIRALTY BRASS
\
NO DATA
END OF RUN
\
26 30 3 7
MAY
11 15 19 23 27
JUNE
Figure 46. Water side heat transfer coefficients,
Run 1, Zn/PO4 inhibitor.
118
-------
a
e/j
QC
S
400
300
200
400
300
200
300
200
100
A-214 CARBON STEEL
A-249 STAINLESS
N— ' EMDOF
A NO DATA I
i i i i i i i i i i i i i i i *
END OF RUN
B-lll ADMRALTY BRASS
END OF RUN
END OF
i i i i i i i t
24 28
JUNE
2 6 10
14 18
JULY
22 26
Figure 47.
Water side heat transfer coefficients,
Run 2, Cr/Zn inhibitor.
119
-------
500
A-214 CARBON STEEL
400
300
200L
NO DATA
END OF RUN
I I I I I I I I I W I I I
400
300
A-249 STAINLESS
NO DATA
END OF MM
, i I _i I I i i i i t i f i i i
300
B-U1 ADMIRALTY BRASS
200
NO DATA
~\
\l\ ENDOF RUN
I_LJ T I I I I I I I I i i f I i I
6 10 14 18 22 26 30 3
AUGUST SEPTEMBER
Figure 47.
Water side heat transfer coefficients,
Run 3, Zn/PO4 inhibitor.
120
-------
600
500
4001
A-214 CARBON STEEL
NO DATA
END OF RUN
i i i i i i I i i i
600
500
100
500
400
B-lll ADMIRALTY BRASS
30d_i_Lj_
NO DATA
END OE RUN
3 711151923271 5
SEPTEWER OOOBER
Figure 49.
Water side heat transfer coefficients,
Run 4, Zn/P04 inhibitor.
121
-------
A-214 CARBON STEEL
500 -
400
END OF RUN
500 r
B-lll ADMIRALTY BRASS
END OF RUN
12 16 20 24 28 1 59
OCTOBER NOVEMBER
Figure 50. Water slide heat transfer coefficients,
Run 5, Cr/Zn inhibitor.
122
-------
500
too
300
A-2M CARBON STEEL
END OF RUN
I I 1 I LI I I i I I f t _t
I
l_>
1
<-»
OC
£
500
400
300
A-249 STAINLESS
END OF RUN
i 1 I I I I I I I 1 I 1 I I I I t f
B-lll ADMIRALTY BRASS
300
U 15 19 23 27
NOVEMBER
Figure 51.
Water side heat transfer coefficients,
Run 6, Zn/PCL inhibitor.
123
-------
Biological Fouling
Wood samples placed in the cooling tower sump during each
test run were examined for indications of severe biological
fouling. No adverse biological effects were observed in any of
the tests. Observations made of the cooling tower packing, wood
samples and sump suggested normal biological growth. A detailed
study of the biological growth was not undertaken in this pro-
gram. However, biological growth was apparently effectively
controlled with all makeup waters tested.
Summary of Cooling Water Test Conclusions
An examination of biological fouling, corrosion and heat
transfer characteristics and their relations and interactions
with each other is necessary for final evaluation of the use of
a water for makeup to a cooling tower. Based upon the data
collected and observations made in each of these areas, a sum-
mary of the acceptability of each of the various metals tested
for each test run is summarized in Table 29. Conclusions based
upon these findings should be limited to this specific study
and do not necessarily reflect what would have been observed
using other quality makeup waters, metallurgies, test conditions
or chemical treatments.
1. Chromate treatment appeared to be effective in
controlling the corrosion of A-214 carbon steel;
however, the type of scale formed decreased the
heat transfer characteristics to an unacceptable
level, making this treatment unacceptable.
2. The use of Zn/PC>4 inhibitor resulted in satis-
factory heat transfer on carbon steel but
excessive corrosion resulted.
3. A-249 stainless steel and B-lll admiralty brass
appeared to be acceptable in terms of corrosion,
regardless of inhibitor treatment.
4. B-lll admiralty brass was unacceptable on heat
transfer in all but two cases.
5. Biological fouling was effectively controlled
in all test cases.
6. It appears from the data evaluated that only
A-249 stainless steel was effective in main-
taining satisfactory corrosion and heat trans-
fer characteristics with the makeup waters and
treatments tested and that special metallurgy
would be required for the use of renovated waste-
water as cooling water.
124
-------
TABLE 29. SUMMARY OF ACCEPTABILITY OF MAKEUP WATERS TESTED WITH VARIOUS METALLURGIES
Makeup
Water
Metallurgies^ ' Acceptable in Terms of
Treatment
Heat Transfer
Corrosion
Biological Fouling
Overall
Acceptable
Hetalurgy
ND
CJl
RO Permeate Zn/P04
Activated Cr/Zn
Carbon
Effluent
Activated Zn/PO4
Carbon
Effluent
Sedimentation/ Zn/PO.
Filtration
Unit Effluent
Sedimentation/ Cr/Zn
Filtration
Unit Effluent
Clarified
Activated
Sludge
Effluent
A-214, A-249, B-lll
A-214, A-249
A-249
A-214, A-249
A-249, B-lll A-214, A-249, B-lll
A-214, A-249, B-lll A-214, A-249, B-lll
A-249, B-lll
A-249, B-lll
A-214, A-249, B-lll A-249, B-lll
A-249, B-lll
A-214, A-249, B-lll A-249
A-214, A-249, B-lll A-249, B-lll
A-214, A-249, B-lll A-214, A-249, B-lll A-249
A-214, A-249, B-lll A-249
A-214 Carbon Steel
A-249 StainleM Steel
B-lll Admiralty braw
-------
REFERENCES
1. EPA 625/6-74-003, Methods for Chemical Analysis of Water
and Wastes, U. S. Environmental Protection Agency
Technology Transfer, 1974.
2. Standard Methods for the Examination of Water and Waste-
water, 13th Ed., America Public Health Assn., New York,
N. Y., 1971.
3. EPA-600/4-75-007, "Analytical Quality Assurance for Trace
Organic Analysis by Gas Chromatography/Mass Spectrometry."
4. FWPCA 211 (Ref. 3-69, p. 10), Part I, Section E, Project
Schedule.
5. Mcllhenny, W. F., Zeitoun, M. A., LeGros, P. G., "The
Disposal of Waste Brine from Desalting Operations," 25th
Purdue Industrial Waste Conference, XXV, 559, 1970.
6. "Get Zero Discharge With Brine Concentration," Hydrocarbon
Processing, October, 1973.
7. Tofflemire, T. J., VanAlstyne, F. E., "Deep-well Injection,"
Journal WPCF, Vol. 45, 1973.
8. Ricci, L. J., "Injection Wells' Iffy Future," Chemical
Engineering, August, 1974.
9. Stickney, W. W., Fosberg, T. M., "Putting Evaporators to
Work: Treating Chemical Wastes by Evaporation," Chemical
Engineering Progress, April 1976.
10. EPA 625/l-75-003a, Suspended Solids Removal, U. S. Environ-
mental Protection Agency Technology Transfer, January 1975.
11. Process Design Manual for Carbon Adsorption, U. S. Environ-
mental Protection Agency Technology Transfer, October 1971.
12. Metcalf and Eddy, Inc., "Wastewater Engineering," McGraw-
Hill, Inc., 1972.
13. Process Design Manual for Suspended Solids Removal, U. S.
Environmental Protection Agency Technology Transfer,
October 1973.
126
-------
14. Water Treatment Plant Design, American Water Works Associa-
tion, Inc., New York, 1969.
15. Process Design Manual for Carbon Adsorption, U. S. Environ-
mental Protection Agency Technology Transfer, October 1973.
16. Roderick H. Horning, "Control System Treats Dye Plant Waste,"
American Dyestuff Reporter, August 1974.
17. Walter J. Weber, Jr., Physicochemical Processes for Water
Quality Control, Wiley-Interscience, New York, 1972.
18. Gordon Maskew Fair, John Charles Geyer, and Daniel Alexander
Okun, Water and Wastewater Engineering, Vol. 2, John Wiley
and Sons,Inc.,New York,1968.
19. Russell L. Gulp, Gordon-L. Gulp, Advanced Wastewater Treat-
ment, Van Nostrand Reinhold Company, Cincinnati, 1971.
20. R. A. Hutchins, "Thermal Regeneration Costs," Chemical
Engineering Progress, May 1975.
21. Betz Handbook of Industrial Water Conditioning, Betz
Laboratories, Inc., Sixth Edition, 1962.
ADDITIONAL REFERENCES
22. Schwartz, S. M., "Total Wastewater Use and Recycling at an
Aluminum Products Manufacturing Plant," Ind. Water Eng.
12, No. 3, June/July 1975, pp. 18-20.
23. Kaye, J. B., "Reuse-Facility Rehabilitation," Jour. AWWA
63, No. 10, October, 1971, pp. 641-643.
24. Lawson, C. T., and Ledbetter, J. B., et al, "Wastewater
Recovery and Reuse in a Petrochemicals Plant," in Complete
WateReuse - Industry's Opportunity, Proceedings of the Nat.
Conf. on Complete WateReuse, AIChE/EPA, April 1973, pp.
351-359.
25. Brymer, B. J., "Problems of Complete WateReuse in a Chemical
Manufacturing Plant," Ibid, pp. 297-304.
127
-------
APPENDIX A: SPECIFIC ORGANIC ANALYSES - SAMPLE
CONCENTRATION AND IDENTIFICATION PROCEDURES
Direct-Inject GC Analyses—An F and M, Model 810, chromato-
graph with a hydrogen flame ionization detector was used in all
three GC analyses. These analyses included the use of: (a) a
CARBOWAX 20M column for most volatile compounds, (b) a Porapak Q
column for organic acids and (c) an OV-101 column for the glycol-
trimethysilyl derivatives.
(a) Volatile Compounds:
The analysis for the volatile compounds was by direct-inject
of the sample and employed the following conditions.
Column CARBOWAX 20M (10%) on 80/100 AW Chromosorb
W, 20 ft. by 1/8-inch O.D., stainless steel
Sample size 5 jul
Column Temperature 50 to 250°C
Program at 10°C/min
Injection Port 220°C
Temperature
Block Temperature 250°C
Helium flow 37 ml/min
Under these conditions, the components identified, eluted at the
following times.
Elution
Compound Time, min.
Acetone 4.5
Methyl ethyl ketone 5.7
Toluene 10.2
Isobutanol 11.0
Ethyl benzene 11.5
n-Butanol 12.2
Cumene 12.5
Styrene 14.8
Acetophonone 23.1
Naphthalene 24.6
Phenol 34.1
These compounds were identified by a GC-mass spectrometric
procedure after concentration using the method described. Quanti-
fication was established by use of external standards calibrated
in the same range as specific compounds in the samples.
128
-------
(b) Organic Acids:
In the organic acid analysis, the samples were pre treated
prior to GC analysis to remove volatile compounds which could
interfere in the determination. A 100-ml aliquot of the sample
at "-'pH 7-8 was evaporated to dryness using a Rotovac apparatus
at 50 °C and -^5 mm Hg absolute. Volatile organic compounds were
removed from the residues in the Rotovac flask by this procedure.
The residues in the flask were redissolved in 10 ml distilled
water and adjusted to pH 2 with phosphoric acid. This solution
of organic acids was used for GC analysis employing the following
conditions.
Column Porapak Q, 3 ft. by 1/8 in. O.D. , H3P04
treated, stainless steel
Sample size 5
Column Temperature 175 °C
Injection Port 240 °C
Temperature
Block Temperature 270 °C
Helium Flow 37 ml/min
Under these conditions, the contained organic acids eluted at the
following times.
Elution
_ Compound Time , min .
Acetic acid 1.8
Propionic acid 3.8
Isobutyric acid 7.4
Butyric acid 8,6
The acids were quantified using external standards in the
same concentration range as in the samples.
(c) Glycols:
The samples for glycol determinations were concentrated and
reacted with REGISIL to form trimethylsilyl derivatives. These
glycol derivatives were analyzed by GC. This procedure for prepa-
ration of the derivatives (4) comprised charging a 50-ml aliquot
of the sample to a distillation flask along with 150 ml pyridine
containing 15 mg/1 1, 4-butandiol. The 1, 4-butandiol was used as
an internal standard in the GC analysis. The flask was fitted to
a distillation column and condenser, and 195 ml of distillate was
removed at atmospheric pressure. The distillation flask con-
taining 3 to 4 ml of residue was removed from the distillation
column and cooled. REGISIL (1 ml) (Regis Chemical Company) was
129
-------
added to the residual pyridine. The flask was stoppered and
allowed to stand for 15 minutes for reaction of the REGISIL and
contained glycols. The pyridine solution of glycol REGISIL
derivatives was employed in the GC analysis at the following
conditions.
Column OV-101 (3%) 12 ft. by 1/8-inch O.D. , on
100-200 mesh Supelcoport
Column Temperature 50 °C for 10 min. ; 5°C/min. to 180 °C
Sample size 5
Injection Port 240°C
Temperature
Block Temperature 260 °C
Helium Flow 37 ml/min
Under these conditions, the glycols eluted at the following times.
Elution
_ Compound _ Time, min.
Ethylene glycol 12.1
Propylene glycol 13.6
1,4-Butandiol 19.9
Diethylene glycol 22.1
Ethoxytriglycol 26.2
Triethylene glycol 28.0
GC-Mass Spectrometer Analysis — The concentrating procedure
for the GC-mass spectrometer analysis was that recommended by EPA
(3). In the method, a 3-1 sample at -pH 7 was transferred to a
separatory funnel. Fifty milliliters of ethyl ether were added
and the mixture was shaken for 1 minute. The sample then was
extracted three times with 75-ml portions of methylene chloride,
and the extracts were combined. The purpose of the ethyl ether
was to improve the extraction efficiency of the more polar com-
pounds like phenols and acids. The pH of the water layer then
was adjusted to pH 2 using concentrated HC1 and the methylene
chloride extraction was repeated. Ethyl ether was not added a
second time. When the second extraction was completed, the pH of
the water layer was adjusted to pH 12 using saturated NaOH and was
again extracted with methylene chloride. All extracts then were
combined for drying and concentrating.
The ethylene chloride extracts were dried by pouring them
through 2 inches of anhydrous sodium sulfate in a 19-mm ID glass
column. The dried extracts were collected in a distillation
flask and evaporated to -~5 ml using a Snyder column and steam
bath. The concentrated extract then was transferred to an ampul
130
-------
and was further concentrated to~'0.5 ml in a warm water bath under
a stream of clean, dry nitrogen. This concentrated extract was
submitted for the GC-mass spectrometer analysis for identification
of additional specific chemicals in the wastewater samples not
detected by direct-inject GC.
131
-------
APPENDIX B: WEEKLY AVERAGED DATA SUMMARIES
TABLE B-I WASTE SLUDGE CHABACTERIZATIOH
N)
A>|U Waste Secondary Sludge
mg/1 Date.
(Detectable Limit) 9/16 9/22 10/1 10/9
pH 7.5 - 7.9 8.2
COD ... 14241
TSS 13250 20710 11810 12100
TD8 3178 3582 3620 3582
Ca 132 144 146 128
Li(l.O) Nil Nil Nil Nil
Mg 52 47 55 43
K(O.l) 74 86 104 70
Na 1220 1100 1330
Cd. (1.0) Nil Nil Nil Nil
Cr04 0.5 0.3 0.5 0.3
Cu (0.1) 0.9 0.9 1.0 Nil
re 33 22 26 16.5
Hn (0.1) 0.8 0.7 0.9 0.0
Al 170 110 150 90
Hi (0.1) 0.5 0.5 0.5 Nil
Zn (0.1) 2.4 1.7 3.6 3.1
Hg (0.05) Nil Nil Nil Nil
As (0.1) Nil Nil Nil
SiOa (81) (62) (38) (62) (29)
Avg
7.9
14241
1446«
3491
138
Nil
49
83
1217
Nil
0.4
0.7
24
0.6
130
0.4
2.7
Nil
Nil
(48)
Bed imenta t ion/m trat ion
Unit - Backwash
Dates
19/16 9/23 10/1 10/9
8.4 8.5
272
410 335 164
3360 3658 3378
52 64 62 54
Nil Nil Nil Nil
5.0 4.9 7.8 5.8
14.0 2.0 2.1 2.1
1O10 1015 1202 1125
Nil Nil Nil Nil
0.2 Nil Nil Nil
0.1 0.1 0.4 Nil
2.3 2.8 1.9 2.1
0.1 0.1 0.1 Nil
6.8 8.6 4.4 3.8
0.1 Nil 0.1 Nil
0.1 0.2 0.2 0.1
Mil Nil Nil
Nil Nil Nil Nil
2.0 2.0 1.6 3.0
Avg
8.4
272
303
3465
58
Nil
5.9
5.0
1088
Nil
<0.1
0.1
2.3
<0.1
5.9
<0>1
0.15
Nil
Nil
2.2
Multi-Media Filter
Backwash
Dates
9/16 9/24 10/1 10/9
8.6 - 8.4 8.5
74
9 6 8 20
2631 2732 3360 3230
38 • 64 56 54
Nil Nil Nil Nil
4.5 3.3 5.4 5.5
5.0 0.5 0.9 1.5
965 785 1214 1260
Nil Nil Nil Nil
Nil Nil Nil Nil
Nil Nil 0.3 Nil
0.7 0.4 0.5 1.0
Nil Nil Nil Nil
1.0 0.8 1.8 1.2
Nil Nil 0.1 Nil
0.2 0.1 Nil 0.2
Nil Nil Nil -
Nil Nil Nil -
1.0 1.5 1.0
AVB
8.5
74
11
2988
53
Nil
4.7
0.9
1056
Nil
Nil
<0.1
0.7
Nil
1.2
Nil
0.1
Nil
Nil
1.2
Carbon Column
Backwash
Dates
9/29 10/3 10/15
8.5 8.4 8.6
103 1000
25 22 437
3542 3410 3504
50 60 50
Nil Nil Nil
4.1 6.1 6.1
0.7 1.8 0.2
1240 1260 1138
Nil Nil Nil
Nil Nil Nil
Nil Nil 0.1
2.3 0.4 13.0
Nil Nil Nil
2.7 2.1 -
Nil Nil Nil
0.1 0.2 0.1
Nil Nil Nil
Nil Nil Nil
1.3 1.0 5.5
Avg
8.5
550
161
3485
53
Nil
5.4
0.9
1212
Nil
Nil
Nil
5.2
Nil
2.4
Nil
0.1
Nil
Nil
2.6
-------
WEEKLY AVERAGES
Date
4/4-11
26-30
5/1-7
8-14
15-21
22-28
5/6/29-4
6/5-11
12-18
19-25
6/7/26-2
7/3-9
AVG.
10-16
17-23
24-30
7/8/31-4
8/5-11
12-18
19-25
8/9/26-1
9/2-8
9-15
16-22
23-29
9/10/30-6
AVG.
10/7-13
14-20
21-27
10/11/28-3
11/4-10
11-17
18-27
AVG.
MeBbrane Feed
Tubular Cell- 8.1
ulose Acetate 8.5
8.5
8.4
8.4
8.4
8.6
8.5
8.3
7.8
8.1
7.8
8.3
Spiral Cell- 5.5
ulose Acetate 8.0
8.3
7.8
8.5
8.3
8.3
8.5
8.4
8.3
8.4
8.3
8.4
8.1
Spiral 8.3
Polyaaide 8.6
8.7
8.5
8.6
8.4
5.5
8.1
PH
Pen
6.8
6.5
6.9
7.0
6.8
7.3
7.1
7.4
6.9
6.5
6.3
5.6
6.8
4.4
5.6
6.2
5.9
5.6
5.4
4.8
6.1
6.2
6.5
5.8
6.2
6.2
5.8
7.1
6.6
7.6
7.5
7.7
8.0
5.1
7.1
Conductivity
Feed
5703
4377
5133
4857
4876
5156
5444
5289
5223
4956
5289
5477
5148
8273
5567
5200
4767
3627
3213
3656
3453
3133
3387
3133
3067
4867
4257
9623
5144
4876
4133
5133
6300
5800
5858
Perm
1257
809
1167
1192
1121
957
1657
1733
1029
553
824
731
1086
324
328
607
461
297
467
431
611
535
792
515
367
733
498
680
480
445
367
493
462
413
477
% KEH
78.0
81.5
77.3
75.5
77.0
81.4
69.6
67.2
80.3
88.8
84.4
86.7
78.9
96.1
94.1
88.3
90.3
91.8
85.5
88.2
82.3
82.9
76.6
83.6
88.0
84.9
88.3
92.9
90.7
90.9
91.1
90.4
92.7
92.9
91.9
Hardness
Feed
88
100
40
44
44
75
59
64
54
69
52
41
61
63
29
17
-
68
93
75
93
92
93
74
62
33
66
85
32
20
33
69
75
60
53
Pen
36.0
5.0
4.0
8.0
3.0
7.0
11.0
10.0
4.0
5.0
5.0
6.0
9
0.0
0.5
3.0
_
2.0
41.0
1.2
24.0
12.0
13.0
2.0
3.0
0.5
8.5
0.0
0.0
0.0
0.0
4.0
1.0
0.0
0.7
Fe
Feed
0.28
-
_
0.00
0.08
0.08
0.08
0.29
1.82
1.22
0.24
0.23
0.43
1.56
0.09
0.26
1.50
0.30
0.12
0.14
0.11
0.51
0.27
0.90
1.00
0.29
0.54
0.25
0.51
0.37
0.25
0.08
-
-
0.29
Pen
0.30
_
0.13
0.10
0.02
0.03
0.02
0.58
0.36
1.49
0.52
0,29
0.35
0.34
0.48
0.38
1.10
0.45
0.80
0.05
0.18
0.61
0.10
0.59
0.05
0,15
0.41
0*.04
0.20
0.03
0.25
0.08
-
0.21
0.14
Feed
20.0
6.0
7.0
29.0
3.3
6.7
65.7
6.0
5.7
4.7
9.2
13.5
14.7
15.0
4.7
5.7
22.0
10.0
8.3
6.0
8.3
5.5
7.0
7.0
4.0
6.0
8.4
5.5
8.0
0.0
17.5
10.3
20.0
22.0
11.9
TSS
Pen
13.0
2.0
18.8
40.3
0.7
5.0
57.3
4.7
1.0
0.0
3.3
0.3
12.2
0.0
0.3
0.0
2.0
1.0
0.8
0.0
1.0
1.0
0.0
0.0
0.0
1.4
0.58
0.0
0.0
0.0
4.0
4.7
-
1.5
1.7
(Continue^)
133
-------
(TABLE B2 continued)
Date
4/4-11
26-30
5/1-7
8-14
15-21
22-28
5/6/29-4
6/5-11
12-18
19-25
6/7/26-2
7/3-9
AVG.
10-16
17-23
24-30
7/8/31-4
8/5-11
12-18
19-25
8/9/26-1
9/2-8
9-15
16-22
23-29
9/10/30-6
AVG.
10/7-13
14-20
21—27
10/11/28-3
11/4-10
11-17
18-27
AVG.
Membrane Feed
Tubular Cell- 4582
ulose Acetate 3212
3472
2986
3090
3719
3934
3717
4069
3660
3657
3881
3665
Spiral Cell- «««
ulose Acetate J787
3690
3312
2468
2254
2040
2337
1933
1822
2038
2141
3350
2924
Iodide 3326
3013
4127
4575
4950
4490
TDS
Pen
699
459
608
673
526
721
992
997
572
111
520
431
609
156
188
382
372
249
244
254
400
317
426
330
282
469
313
485
298
310
320
311
288
329
334
% REM
84.7
85.7
82.5
77.5
83.0
80.6
74.8
73.2
85.9
97.0
85.8
88.9
83.3
97.7
95.0
89.6
88.8
89.9
89.2
S7.5
82.9
83.6
76.6
83.8
86.8
86.0
89.3
93.8
91.0
91.2
89.4
92.5
93.7
93.4
92.6
(TOC)COD
Feed Pen % REM
(109)
(120)
( 39)
( 43)
( ID
( 25)
( 28)
(Nil)
( 28)
( 70)
37
59
( 48)
34
27
68
176
39
29
32
23
34
35
17
34
30
44.5
134
37
31
38
27
37
8
47
(23) 78.9
(13) 89.2
( 5) 87.7
(22) 49.5
( 7) 37.7
(16) 36.0
(11) 60.4
(Nil)
(25) 9.1
(16) 77.9
0 100.0
6 89.2
(14)
12
7
8
25
0
13
11
0
0
9
27
12
5
10
14
5
5
0
5
0
2
4
70.4
64.7
72.5
88.2
85.8
100.0
55.2
65.6
100.0
100.0
74.3
-
64.7
83.3
77.5
89.6
86.5
90.2
100.0
81.5
100.0
75.0
91.5
Silica
Feed Pern
4.0
5.0
5.0
7.2
11.0
11.1
4.9
15.9
8.6
9.8
8.3
17.0
8.5
9.0
8.8
12.1
18.0
13.3
12.3
10.8
9.5
10.4
11.9
7.0
11.4
9.4
5.8
5.2
6.2
4.8
5.9
8.7
6.6
1.0
7.0
3.8
5.1
4.6
4.6
5.7
5.5
6.8
3.8
4.1
4.7
1.0
0.0
4.7
2.4
4.2
6.7
8.0
5.6
5.6
3.4
2.5
4.0
3.2
3.9
1.5
2.2
1.3
2.6
1.0
1.7
2.7
1.9
Chloride
Feed Pen % REM
-
229
164
147
119
72
62
92
76
56
46
49
47
111
97.7
264
114
106
182
467
381
171
241
-
38
29
57
59
27
18
23
30
20
17
14
9
43
29.5
43
28
28
80
86
52
41
51
-
83.4
82.3
61.2
50.4
62.5
71.0
75.0
60.5
64.3
63.0
71.4
80.9
61.3
83.7
75.4
73.6
56.7
81.6
86.4
76.0
78.8
134
-------
TABLE B-3 ION-EXCHANGE DATA SUMMARY, WEEKLY AVERAGES
LO
Ul
4/26-30
pH - RO Permeate ,.,
Primary Deionizer .»,
Secondary Delonizer
Conductivity, umho - RO. Permeate
Primary Deionlzer >2)
Secondary Delonizer
Hardne» - RO Permeate,.,
Primary Deionlzer ,n\
Secondary Deionizer
Iron - RO Permeate ,,,
Primary Delonizer ,-.
TSa - RO Permeate ,n
Primary Delonizer J2)
Secondary Deionizer
COD(TOC) - RO Permeate,.,
Primary Deionlzer ,«v
Secondary Deionlzer
Silica - RO Permeate ...
Primary Delonizer ,«\
Secondary Delonizer
Chloride - RO Permeate,,,
Primary Delonizer ,,.
Secondary Deionlzer
6.5
6.8
4.3
607
738
-
4.7
7.0
4.0
_
-
i.'e
0.0
0.0
(26)
(20)
(10)
_
-
-
5/1-7
6.8
6.9
-
897
752
-
22
3.4
0.13
3.2
2.6
(15)
(8)
7
-
-
5/8-11
6.9
7.1
-
950
1030
-
12.2
8.0
-
_
-
2
4.7
-
(16)
(13)
4.5
3.5
-
7/22-28
6.4
8.0
6.6
411
275
S
3.0
5.5
1.0
0.33
0.10
0 04
0.0
0.0
0.0
24
15
2
3.0
2.8
0.0
52
8/9-17
5.6
8.1
6.9
467
314
9
41.0
26.0
0.5
0.8
0.07
0. 06
0.8
0.7
0.0
13
18
4
6.7
5.4
1.1
18
20
2
8/22-31
6.2
9.0
7.6
360
283
4
3.0
17.0
3.0
0.05
0.13
0. 00
0.0
1.0
0.0
12
21
9
4.0
8.0
9
18
1
9/1-6
6.3
8.8
8.9
285
171
9
12
28.0
1.0
0.41
0.12
0, 06
0.0
0.0
0.0
26
8
2
5.4
2.S
0.0
12
12
1
9/14-27
5.8
8.9
8.0
445
274
-
2.4
7.0
1.0
0.50
0.50
0. 67
0.0
0.0
0.0
26
28
1
3.6
3.2
2.1
14
11
0
10/1-18
6.5
8.6
7.1
402
329
3
0.3
0.5
0.0
0.20
0.21
0. 21
0,0
0.3
0.0
12
2.8
1.9
0.8
27
22
< 1
10/23-29
7.7
9.0
8.9
387
359
-
0.0
0.0
0.0
0.04
0.00
0. 00
0.0
0.0
0.0
10
7
16
2.0
5.3
0.5
33
43
3
11/1-9
8.0
8.7
8.0
425
324
2
3.2
2.5
0.0
0.08
0.00
0. 00
4.5
3.6
0.9
4
3
6
1.0
0.6
0.1
79
46
6
11 14-20
7.8
87
7.7
560
161
4
4.0
7.5
2.0
0.00
0.00
0. 00
0.00
0.0
0.0
0
11
6
2.1
0.4
0.0
53
15
2
AVG.
6.7
82
7.4
516
418
5
9.0
9.4
1.3
0.25
0. 13
0. 12
1.0
1.0
0.0
14
13
6
3.R
3.4
0.5
33
23
1.8
Weak Baae Anton-Exchange
<2> Mixed Bed Ion-Exchange
-------
APPENDIX C: SIZING OF FULL-SCALE FACILITIES FOR
INVESTMENT COST AND OPERATING EXPENSE ESTIMATES
Suspended Solids Removal—
Suspended solids removal was accomplished in the pilot-plant
using a package sedimentation/filtration unit that included floc-
culation, tube settlers and multi-media gravity filtration.
Because of the low solids levels experienced and the ease with
which they were removed in the pilot-plant these functions were
estimated to be performed in the full-scale system by two solids
contact clarifiers operated in parallel followed by three multi-
media gravity filters operated in parallel.
Solids contact clarifier— Each clarifier was sized for 75
percent of design flow. Diameter and depth are a vendor's recom-
mendation based on flow rate and wastewater characteristics.
Equipment Data
Case A
Case B
Plant influent, m /min (gpm)
3
Design flow, m /min (gpm)
Diameter, m (ft)
Depth, m (ft)
5.7 (1500)
6.3 (1658)
13.7 (45)
4.7 (15.5)
11.4 (3000)
12.6 (3315)
19.8 (65)
4.7 (15.5)
Multi-media gravity filters—Three filters would be operated
in parallel. Each filter was sized for 50 percent9of design flow
at a hydraulic loading of 0.16 m3/min*m (4 gpm/f t ) . Filter
dimensions were chosen to accommodate 4.1 m (13.5 ft) diameter
rotary spray surface washes.
Equipment Data
Case A
Case B
Plant influent flow,
m /min (gpm)
3
Design flow, m /min (gpm)
Filter area (each),
m2 (ft2)
Filter width, m (ft)
Filter length, m (ft)
Polyelectrolyte @ 20 mg/1,
Kg/yr (Ibs/yr)
Polyelectrolyte cost,
$/Kg
5.7 (1500)
6.1 (1605)
18.9 (203)
4.7 (14)
4.42 (14.5)
11.4 (3000)
12.2 (3210)
37.4 (403)
4.7 (14)
8.76 (28.7)
66,000 (145,500) 122,000 (270,000)
1.32 (0.60)
1.32 (0.60)
136
-------
Activated Carbon Adsorption—
Effluent from the multi-media gravity filters would be
pumped from the filter clear-well through a series of activated
carbon columns. The columns would be arrayed in parallel trains
of four columns in each. Within each train, three columns would
be in service in series with the fourth on standby. When the
lead column is exhausted, it would be taken out of service and
the spare column placed in service as the final column. The
carbon in the exhausted column would be regenerated on site. As
in the pilot plant, the carbon bed depth in the full-scale columns
was designed to be 4.88 m (16 ft) and the height of the cylindri-
cal section of each column was 7.62 m (25 ft). A hydraulic
loading of 0.163 m3/min-m2 (4 gpm/ft2) is applied to the full-
scale design.
Equipment Data
Case A
Case B
Plant influent flow, m /min (gpm)
3
Design flow, m /min (gpm)
Number of parallel trains
Column diameter, m (ft)
Bed area,
(ft2)
5.7 (1500)
5.8 (1526)
4
3.35 (11)
35.3 (280)
11.4 (3000)
11.8 (3052)
7
3.66 (12)
73.6 (793)
Activated Carbon Regeneration—
The spent carbon would be transferred from the columns to a
dewatering and furnace feed tank. Regeneration would take place
in a multiple-hearth furnace equipped with an afterburner and a
wet scrubber. Regenerated carbon would be accumulated in a
regenerated carbon storage tank. Virgin carbon makeup received
in bulk would be inventoried with the reactivated carbon. The
carbon would be transferred intra-unit by a water slurry. The
furnace capacities were sized for a process stream COD removal
of 93 mg/1, a carbon capacity of 0.334 Kg COD/Kg carbon and 60
percent on-line time.
Equipment Data
Case A
Case B
3
Plant influent flow, m /min (gpm)
Carbon exhaustion rate, Kg/day
(Ibs/day)
Furnace capacity, Kg/day (Ibs/day)
Carbon attrition <§ 7% loss per
regeneration, Kg/day (Ibs/day)
Carbon cost, $/Kg ($/lb)
5.7 (1500) 11.4 (3000)
2300 (5070) 4600 (10,140)
3800 (8377) 7600 (16,755)
161 (354) 322 (708)
1.32 (0.60) 1.32 (0.60)
137
-------
Multi-media Pressure Filtration—
Effluent from the carbon columns would flow under pressure
through two multi-media filters in parallel. The filtered water
would flow into a surge tank with sufficient capacity to provide
wash water for the carbon columns and pressure filters while main-
taining a uniform flow into the reverse osmosis unit. Polyelec-
trolyte would be added just before the pressure filters via
in-line static mixers. The pressure filters are sized for 50
percent of the plant influent flow and a hydraulic loading of
0.204 m3/min-m2 (5 gpm/ft2).
Equipment Data
Case A
Case B
Plant influent flow, m /min
(gpm)
3
Design flow, m /min (gpm)
2 2
Filter area, each, m (ft )
Bed width, m (ft)
Bed length, m (ft)
Bed depth, m (ft)
Backwash storage tank,
m3 (gal)
Polyelectrolyte @ 20 mg/1,
Kg/yr (Ib/yr)
Polyelectrolyte cost, $/Kg
5.7 (1500)
5.7 (1500)
14.8 (160)
3.05 (10)
4.85 (16)
1 (3.3)
490 (130,000)
11.4 (3000)
11.4 (3000)
29.7 (320)
3.05 (10)
9.76 (32)
1 (3.3)
590 (155,000)
61,000 (134,000) 122,000 (269,000)
1.32 (0.60)
1.32 (0.60)
Reverse Osmosis—
Effluent from the pressure filters would be pumped from the
pressure filter backwash tank to the reverse osmosis (RO) system.
According to the manufacturer of the pilot-scale RO unit, the
full-scale system would be arrayed in parallel trains each rated
at 0.852 m /min (225 gpm). Unlike the batch-mode pilot operation,
the full-scale reverse osmosis system would operate on a continu-
ous flow, once-through basis using the spiral-wound polyamide
membranes. The design salt rejection is 95 percent for 75 percent
water recovery. Conservative design flow rates were used—6.4 m /
min (1700 gpm) for the 5.7 m3/min nominal flow case and 12.9 m3/
min (3400 gpm) for the 11.4 m3/min nominal flow case. The entire
system is located indoors.
Membrane replacement cost, included in the operating ex-
penses, are based on a life expectancy of three years at an
average cost of $0.046/m3 of product water ($0.175/1000 gal).
138
-------
Brine disposal facilities are not included in investment cost
or operating expenses. However, viable solutions may be: ocean
water disposal, solar evaporation, deep-well injection and mechan-
ical crystallization and land fill. Likely, viable solutions for
the Ponce, P. R. area are ocean (Caribbean Sea) disposal or solar
evaporation.
Ion-Exchange—
Demineralization would be completed in an ion-exchange sys-
tem. The reverse osmosis unit permeate would flow through cation
exchange beds into the degasifier column and be collected in a
clear well. The degasified water would then be pumped through
weak base beds and strong base beds and into the demineralizer
product storage tank.
The cation resins would be regenerated with sulfuric acid and
rinsed with water from the clear well. The anion resins would be
regenerated with sodium hydro.xide and rinsed with demineralized
water. The spent acid, and caustic would combine in a sump and
be disposed of with the brine from the reverse osmosis unit. The
final rinse waters would be comingled with the influent feed to
the reverse osmosis unit. Storage facilities are provided at the
site for 93 percent sulfuric acid and 50 percent caustic soda.
Equipment Data
Plant influent flow, m /min (gpm)
3
Design flow, m" /min (gpm)
Resin beds:
diameter, m (ft)
height, m (ft)
Number of beds:
in-service
(standby)
cation
weak anion
strong anion
Regeneration frequency (beds/day)
cation
weak anion
strong anion
Volume of beds:
3 3
cation, m (ft ) „
weak anion, m Lft )
strong anion, m (ft )
Case A
5.7 (1500)
4.21 (1110)
2.74 (9.0)
3.05 (10.0)
2 (1)
2 (1)
1 (1)
2
2
1/7 days
Case B
11.4 (3000)
8.42 (2220)
2.74 (9.0)
3.05 (10.0)
4
4
2
4
4
2/7 days
U)
(1)
(1)
8.5 (300) 8.5 (300)
9.9 (350) 9.9 (350)
7.1 (250) 7.1 (250)
139
-------
Equipment Data
Case A
Case B
Degasifier: number of columns
diameter, m (ft)
height, m (ft)
packing depth, m (ft)
2.13 (7.0)
3.35 (11.0)
2.13 (7.0)
2.13 (7.0)
3.35 (11.0)
2.13 (7.0)
Backwash Recovery—
Spent backwash from the gravity filters, pressure filters
and carbon columns would be treated in a backwash recovery
system prior to being recycled to the secondary treatment
system. Spent backwash streams would be combined in a col-
lection sump. Water would be pumped out of the sump to a two
compartment, agitator tank, where a polymer flocculant would
be added in the first compartment and with flocculation
taking place in the second compartment. The flocculated
stream would flow to a clarifier with the clarifier overflow
recycled to the head of the secondary treatment system.
Equipment Data
Design (influent flow m /rain (gpm)
o
Maximum flow, m°/min (gpm)
Rapid mix retention time, seconds
Flocculation retention time,
minutes
Rapid mix chamber volume,
m3 (gal)
Flocculation chamber volume,
m3 (gal)
Clarifier overflow rate,
m3/day-m2 (gpd/ft2)
Clarifier diameter, m (ft)
No. of clarifiers
Case A
5.7 (1500)
0.95 (250)
30
10
0.48 (125)
0.5 (2500)
204 (500)
9.14 (30)
1
Case B
11.4 (3000)
1.8 (475)
30
10
0.96 (250)
19.0 (5000)
204 (500)
12.8 (42)
1
140
-------
APPENDIX D: ANALYTICAL METHODS--DEPOSIT ANALYSIS
THE DETERMINATION OF CALCIUM AND MAGNESIUM
(METHODS FOR BOILER SCALE ANALYSES)
PRINCIPLE
A solution of the deposit (from the boiler probe) is
aspirated into a flame where metal ions are converted into an
atomic vapor which is capable of absorbing radiation. The energy
removed by those atoms in the ground state is a measure of con-
centration of the metal of interest.
SCOPE
The procedures are suited to hydrochloric acid solutions of
the deposits. The methods have a sensitivity of 0.2 ppm for
either metal and exhibit a precision in the order of ^ 0.1 ppm
over the 0-5 ppm range. No direct interferences are known;
however, best accuracy can be obtained by preparing standards
similar in composition to the samples.
REAGENTS AND SUPPLIES
1. Stock solution of calcium (1000 ppm) Fisher Chemical Index
SO-C-191.
2. Stock solution of magnesium (1000 ppm) Fisher Chemical Index
SO-M-51.
3. Lanthanum Oxide (Matheson Coleman and Bell) Catalogue number
LX45-8229.
Dissolve 58.6 gms of La203 in 400 ml of 50% HC1 and dilute
to one liter with double distilled water.
4. Hydrochloric Acid, concentrated.
5. Double distilled water.
6. Acetylene, commercial grade, cylinder.
EQUIPMENT
1. Atomic Absorption Spectrophotomer. (Perkin-Elmer 403 is
suitable) equipped with Boling or suitable burner.
141
-------
2. Recorder or other readout accessory.
3. Hollow cathode tube; combination Ca and Mg available from
Perkin-Elmer.
PROCEDURE
1. General Procedure
a. Instrumentation
The analyst should familiarize himself with the manufac-
turer's operating instructions for the particular instrument in-
volved. In general, after choosing the correct hollow cathode
lamp, it should be allowed a 15-minute warm-up period. During this
time, selection of the proper wavelength is made; slit adjustments
are carried out, and the hollow cathode tube current is adjusted.
Follow the manufacturer's recommendations for lighting and regu-
lating the flame so that stable conditions result. Standards may
now be run and calibration curve can be constructed, or for those
instruments which read directly in connection (P-E-403), set the
curve corrector to read out the proper concentrations on the
digital readout.
b. Preparation of Standard Solutions and Calibration
Working from the stock solutions of the appropriate metal,
standards are prepared to cover the working areas of interest.
For best results calibration standards should be prepared fresh
each time a run is made. Beginning with the blank and after
stable instrumental conditions have been obtained, aspirate each
of the standards from low to high and record the data. This can
be done by means of a recorder, or if the equipment is so equipped,
by means of the readout device.
2. Determination of Calcium
a. Instrumental Parameters
Aspirate the samples using direct readout, or compare the
generated signals to a previously prepared calibration curve and
report results as ppm Ca using proper factors if the sample was
diluted.
3. Determination of Magnesium
a. Instrumental Parameters
(1) Hollow cathode tube Calcium-magnesium
(2) Wavelength 2852A (UV)
(3) Burner Boiling, rotate to 55° setting
(4) Oxidant Air: Flow Meter Setting = 55
(5) Fuel Acetylene: Flow Meter Setting = 35
142
-------
(6) Flame conditions Reducing
(7) Slit Setting 4
(8) Readout Time 10 seconds
b. Optimum concentration range
(1) 0-15 ppm (use scale setting 0.5A)
(2) 0-5 ppm (use scale setting 0.25A)
c. Preparation of Standards
Prepare dilutions of the stock magnesium solution for the
concentration range desired. Pipet 20 ml of each standard and
5 ml of lanthanum solution into a plastic vial and mix well.
Establish the calibration curve. It has been found convenient to
preset the highest working standard at approximately 75% of full-
scale deflection for either range.
d. Sample Analysis
Since magnesium is run on the sample that has been used to
determine calcium, directions are the same as previously
described.
143
-------
DETERMINATION OF TOTAL AND SOLUBLE IRON
OPTIMUM CONCENTRATION RANGE:
0.1 - 5.0 mg/1 using the 2483A line. For iron concentration
below 0.1 mg/1 use the extraction procedure. For iron concentra-
tion above 5 mg/1 dilute samples with deionized water.
APPARATUS REQUIRED:
Water bath
Perkin-KDmer 303 or 403 Atomic Absorption Spectrophotometer
CHEMICALS REQUIRED:
1. Concentruled Hydrochloric Acid, Reagent Grade
2. 1000 ppm le standard. Dissolve 1.000 g reagent grade iron
wire in lit ml (1+1) HNOg. Dilute to 1 liter with deionized
water. Ore ml equals 1 mg Fe.
3. 10 ppm Fe standard. Pipet 10.0 ml of the 1000 ppm Fe
standard into a 1 liter volumetric flask. Dilute to
volume with deionized water.
4. 0.1, 0.5, 1.0, 2.0 and 5.0 ppm Fe standards. Pipet 1.0,
5.0, 10.0. 20.0 and 50 ml of the 10 ppm Fe standard into
100 ml volumetric flasks. Dilute to 100 ml with the
deionized water. These solutions are 0.1, 0.5, 1.0, 2.0
and 5.0 ppm Fe standard, respectively.
PROCEDURE FOR TEST:
a. Total iron - shake sample and proceed as in (c).
b. Soluble Iron - filter sample through a 0.45 micron
membrane filter and proceed as in (c).
c. To each 100 ml of sample in the polyethylene bottle
add 1 ml of concentrated HC1. Heat the fixed sample
in water bath at 90-95°C for four hours. Allow to
cool to room temperature.
Set up Atomic Absorption Spectrophotometer according
to the following parameters:
144
-------
1. Iron hollow cathode lamp
2. Wavelength - 2483A (248-UV)
3. Slit - 4 (7A)
4. Type burner - Boling (3-slot)
5. Fuel - acetylene
6. Oxidant - air
CALCULATION OF RESULTS:
Set up 0.1, 0.5, 1.0, 2.0 and 5.0 ppm Fe standards to readout
0.1, 5.0, 1.0, 2.0 and 5.0 on the digital readout. Aspirate the
fixed samples directly from the polyethylene bottles and report
as ppm iron (Fe).
145
-------
ORTHOPHOSPHATE
TOTAL PHOSPHATE
APPARATUS REQUIRED:
Filter photometer
Erlenmeyer flask, 125 ml
Pipettes, 1 ml
Pipettes, 2.5 ml
Pipettes, 5 ml
Pipettes, 10 ml
Beakers, 50 ml
CHEMICALS REQUIRED:
Hydrochloric acid, concentrated, C.P.
Molybdate reagent
Phenolphthalein indicator
Stannous chloride, crystals
Standard phosphate solution, 45 ppm PO4
Sodium hydroxide, 7N
Sulfuric acid, 50%
PROCEDURE FOR TESTS:
A fresh concentrated stock solution of stannous chloride
should be prepared once each month. For this purpose add 12 gms
stannous chloride crystals to 88 gms of C.P. concentrated hydro-
chloric acid. Store in an amber bottle away from light. Keep
container tightly closed.
The dilute stannous chloride reagent used in this test must
be prepared fresh daily. The dilute reagent consists of 1.0 ml of
concentrated stannous chloride diluted to a total volume of 40 ml
with distilled water.
This procedure employs a wavelength of 610 mu and a light
path of 5 mm. Prepare calibration curves for the photometer using
successive dilutions of the phosphate standard to adequately cover
the range of phosphate in the samples to be tested. Two curves
are required—one for orthophosphate and one for total phosphate.
The dilutions of the standard should be treated in exactly the
same manner as that shown below for analysis of the water samples.
Each time a determination is made the calibration curves
should be checked to establish a correction factor. This
146
-------
procedure is necessary to insure that the results are accurate
since reagent age and stability as well as temperature can affect
the results. Each curve should be checked with phosphate-free
water and also at a dilution of the phosphate standard that
approximates the middle of the phosphate range covered by the
curves. It is very important that the "check" samples are
analyzed at the same time, under the same conditions and treated
in the same manner as the actual water samples. Do not omit any
of the steps as the conversion procedure, etc.
Phosphate must be determined on a filtered sample, using a
filter paper such as Whatman No. 5. Discard the first 10-20 ml
of filtrate since there is a slight adsorption of phosphate by
fresh filter paper.
TO DETERMINE ORTHOPHOSPHATE:
Prepare a "zero" reference blank by adding to a beaker; 5 ml
of clear sample, 10 ml molybdate reagent and 2.5 ml distilled
water. Use this solution to set the photometer at "zero" immedi-
ately prior to test.
To a second beaker add 5 ml clear sample, 10 ml molybdate
reagent and 2.5 ml dilute stannous reagent. Allow to stand one
(1) minute and then immediately obtain photometer dial reading.
TO DETERMINE TOTAL PHOSPHATE:
In order to determine total phosphate, it is necessary to
convert all polyphosphate in the sample to orthophosphate. Place
two 25-ml clear samples in separate 125—ml Erlenmeyer flasks.
Ojie sample is to be used as a blank and the other for analysis.
To each flask add 2.5 ml 50% sulfuric acid. Boil both the blank
and the sample vigorously for at least 30 minutes. Add distilled
water periodically so that the volume does not fall below 5 ml.
If the volume does fall below 5 ml, the sample must be discarded.
If it is not convenient to observe the sample continuously during
boiling, reflux condensers may be employed.
Cool the blank and the sample to room temperature. Add 3
drops phenolphthalein indicator to each flask and neutralize with
7N sodium hydroxide (approximately 5 ml will be required) until a
faint permanent pink appears. Add 50% sulfuric acid, drop by
drop, until the solutions turn colorless.
Since some heat may be generated during neutralization, re-
cool and then adjust the volumes to exactly 25 ml with distilled
water. A precipitate may form at this point but do not filter.
The total phosphate now can be determined. Measure 5 ml of
the blank and 5 ml of the sample after shaking to be sure a
representative amount of any precipitate is included. Proceed in
exactly the same manner as shown for orthophosphate.
147
-------
CALCULATION OF RESULTS:
The orthophosphate and total phosphate values in parts per
million as PO4 are obtained directly from their respective cali-
bration curves. The polyphosphosphate concentration is obtained
by subtracting the value for orthophosphate from the value for
total phosphate.
148
-------
DETERMINATION OF SILICA
0.0-3.0 ppm
APPARATUS REQUIRED:
Filter photometer
Pipettes, 5 ml
Beakers, 100 ml
CHEMICALS REQUIRED:
Ammonium molybdate reagent (low range)
Oxalic acid, 3%
Amino-Naphthol-Sulfonic acid
Silica standard, 50 ppm SiO2
PROCEDURE FOR TEST:
This procedure employs a wavelength of 690 mu and a cell with
a light path of 40 mm. Prepare a calibration curve for the photo-
meter using successive dilutions of the silica standard to
adequately cover the anticipated range of silica in the samples
to be tested. The dilutions of the standard should be treated in
exactly the same manner as that shown below for analysis of the
water samples.
Each time a determination is made the calibration curve
should be checked to establish a correction factor. This pro-
cedure is necessary to insure that the results are accurate since
the reagent age and stability as well as temperature can affect
the results. The curve should be checked with silica-free water
and also at a dilution of the silica standard that approximates
the middle of the silica range covered by the curve. All reagents
as well as the "check" samples and the actual sample to be
analyzed must be at the same temperature.
The amino-naphthol-sulfonic acid reagent used in this test
is not stable and should be prepared once each week. Dissolve
1.0 g of 1-amino, 2-naphthol, 4-sulfonic acid in 4.5 ml of IN
sodium hydroxide. Add with 60 g sodium bisulfite and 2 g sodium
sulfite to 900 ml distilled water. Dilute to 1.0 liter with dis-
tilled water.
Prepare a "zero" reference blank. To a beaker, add 50 ml of
the clear sample, 5 ml oxalic acid, 5 ml sulfonic acid and 5 ml
149
-------
distilled water. Use this blank to set the photometer at "zero"
immediately prior to the test of a sample.
To a second beaker, add 50 ml samples, and 5 ml ammonium
molybdate reagent. Allow to stand approximately 5 minutes. Add
5 ml oxalic acid reagent. Wait approximately one minute and then
add 5 ml sulfonic acid reagent. Allow to stand exactly two min-
utes and immediately obtain dial reading.
CALCULATION OF RESULTS:
The silica concentration in parts per million as SiO2 is
obtained by reference to the prepared silica calibration curve.
150
-------
DETERMINATION OF SILICA
3-50 ppm
APPARATUS REQUIRED:
Perkin-Elmer 403
REAGENTS REQUIRED:
1. 1000 ppm SiO2 standard - Obtain from Betz Lab Stock Division.
2. 150 ppm Si02 standard - Dilute 150 ml of the 1000 ppm SiO2
standard to 1 liter with deionized water.
3. 50-25-5 ppm SiO2 standards - Pipet 50.0, 25.0 and 5.0 ml of
the 1000 ppm standard into 1-liter volumetric flasks and dilute
to volume with deionized water.
4. 3 ppm standard - Pipet 20 ml of the 150 ppm Si02 standard
into a 1-liter volumetric glask and dilute to volume with
deionized water.
PROCEDURE:
Set up the Atomic Absorption Spectrophotometer according to
the following parameters:
1. Silicon hollow cathode tube.
2. Wavelength - 2516A (252-UV)
3. Slit - 4 (7A)
4. Type burner - nitrous oxide
5. Fuel - acetylene
6. Oxidant - nitrous oxide
Set the mode on absorbance and with the 150 ppm Si02 stan-
dard, adjust burner position for maximum absorbance.
CALCULATION OF RESULTS:
Then switch to concentration mode and set up the 50, 25, 10
and 3 ppm calibration standards to read 50, 25, 10 and 3 on the
digital readout. Aspirate settled samples directly from the poly-
ethylene sampling bottles and read ppm SiO2 from the digital
readout.
NOTE: Reject Si02 values below 3.0 ppm. They must be analyzed
colorimetrically.
151
-------
A) Analytical results on deposit analysis are reported as fol-
lows :
Calcium as Ca = ppm
Phosphate as PO4 = ppm
Magnesium as Mg = ppm
Silica as
^
Total Iron as Fe = ppm
Insolubles = mg
sample size = 0.25 1
2
heat transfer area = 0.017 m
2
B) Conversion of deposit- weight (from ppm to g/m )
1. Calcium (Ca)
ppm Ca x (1 g/l)/1000 ppm x 0.25 1 x 1/0.017 m2 =
g/m2 Ca
2. Phosphate (PCK — > P2°5^*
ppm P04 x 142 (P205)/190 (2PO4> x (1 g/l)/1000 ppm x
0.25 1 x 0.017 m2 = g/m2 P2°5'
3. Magnesium (Mg — > MgO)
ppm Mg x 40 (MgO)/24 (Mg) x (1 g/l)/1000 ppm x
2 2
0.25 1 x 0.017 m = g/m MgO
4. Silica (SiOg)
ppm SiO2 x (1 g/l)/1000 ppm x 0.25 1 x 0.017 m2 =
g/m2 Si02
152
-------
5. Total Iron (Fe
ppm Fe x 160 (Fe2O3)/112 (2 Fe) x (1 g/l)/1000 ppm x
0.25 1 x 0.017 m2 = g/m2 Fe2O3
6 . Insolubles
2 2
mg x 1 g/1000 mg x 1/0.017 m «= g/m insolubles.
153
-------
APPENDIX E: CALCULATION OF CORROSION TEST
COUPON PENETRATION RATE
CORROSION TEST COUPON PENETRATION RATE CALCULATION
In calculating the penetration per year from the test cor-
rosion coupons the following formula is applied:
Avg. P (mils per year) = Weight Loss x 0.061
Sp. Gravity x A x time x 1/365
Avg. P = Average penetration (mils per year).
Weight Loss = Loss of weight of specimen in milligrams.
0.061 = Cubic inches per cubic centimeter.
Sp. Gravity = Specific Gravity of material used.
A = Area in square inches (normally for our specimen
3 sq. in.).
Time = Time of exposure in days.
1/365 = Reciprocal of days in year.
Employing the above formula, reduce all constant values to
one single factor so that average P (mils per year) will
equal weight loss in milligrams divided by the number of
days exposed times a factor.
, _. , ., . weight loss „
Avg. P (mils per year) = dayg exposed x f
Listed below are the various metals and their specific
gravities that Betz Laboratory presently employs with the proper
factor.
Metal Sp. Gravity f (Factor)
Admiralty 8.52 0.871
Low Carbon 7.84 0.946
Steel
Copper 8.95 0.829
Aluminum 2.76 2.69
Cast Iron 7.0 1.06
Brass 8.49 0.874
154
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/2-79-184
3. RECIPIENT'S ACCESSION*NO.
4. TITLE AND SUBTITLE
Treatment of Organic Chemical Manufacturing
Wastewater for Reuse
5. REPORT DATE
August 1979 issuing date
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Michael Scherm, Patrick M. Thomasson, Lester C. Boone,
Lawrence S. Magelssen
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Union Carbide Corporation
Chemicals and Plastics Division, R&D
Box 8361, Technical Center
South Charleston, West Virginia 25303
10. PROGRAM ELEMENT NO.
1BB610
11. CONTRACT/GRANT NO.
S801398
12. SPONSORING AGENCY NAME AND ADDRESS
Robert S. Kerr Environmental Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Ada, Oklahoma 74820
13. TYPE OF REPORT AND PERIOD COVERED
FINAL 4/76 to 12/76
14. SPONSORING AGENCY CODE
EPA/600/15
15. SUPPLEMENTARY NOTES
16.
ABSTRACT
This research demonstrated the quality of water produced by each step of a state-of-
the-art, commercially available process sequence and determined the feasibility and
economics of renovating organic chemical wastewater for reuse as boiler feedwater
or cycle cooling water. The 5-gpm pilot facility, Iq.cated in Puerto Rico in the
organic chemical manufacturing plant of Union Carbide Caribe Inc., consisted of
sedimentation/filtration, carbon adsorption, pressure filtration, reverse osmosis,
and ion-exchange. A pilot-scale boiler tested the product water as boiler feedwater
at pressures, temperatures, and heat fluxes typical of full-scale manufacturing
facilities. A pilot-scale cooling tower and heat exchangers determined feasibility
as cycle cooling water makeup and chemical treatment requirements for makeup waters
of varying quality from different points in the treatment sequence. The pilot
boiler operated successfully at 180,000 BTU/ft2-hr, 1500 psig, and 750°F superheat
temperature with renovated wastewater. The cooling water test-loop studies
indicated that special metallurgy would be required for the use of this renovated
wastewater for cooling water. The total annualized cost of wastewater renovation
to boiler feedwater quality at 67 percent water recovery, not including the cost
of sludge or brine disposal, was $7.50/1000 gallons of product water in 1978.
7.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Activated Sludge Process
Activated Carbon Treatment
Osmosis
Ion Exchangers
Filtration
Petrochemistry
Cooling Water
Recycle
Reuse
Boiler Water
68D
3. DISTRIBUTION STATEMENT
Release to Public
19. SECURITY CLASS (ThisReport)
TTnr-1 goo-i f i
21. NO. OF PAGES
r^c&li?^
171
20. SECURITY CLASS JThispage)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
155
trtJ.S. GOVERNMENT PRINTING OFFICE:1979-657-060-5372
------- |