WATER POLLUTION CONTROL RESEARCH SERIES
17O2ODHR12/7O
USE OF IMPROVED MEMBRANES
IN
TERTIARY TREATMENT
BY
REVERSE OSMOSIS
ENVIRONMENTAL PROTECTION AGENCY
WATER QUALITY OFFICE
-------
WATER POLLUTION CONTROL RESEARCH SERIES
The Water Pollution Control Research Series describes
the results and progress in the control and abatement of
pollution in our Nation's waters. They provide a central
source of information on the research, development, and
demonstration activities in the Water Quality Office,
Environmental Protection Agency, through inhouse
research and grants and contracts with Federal, State,
and local agencies, research institutions, and industrial
organizations.
Inquiries pertaining to Water Pollution Control Research
Reports should be directed to the Head, Project Reports
System, Office of Research and Development, Water
Quality Office, Environmental Protection Agency, Room
1108, Washington, B.C. 20242.
-------
USE OF IMPROVED MEMBRANES IN
TERTIARY TREATMENT BY REVERSE OSMOSIS
McDonnell Douglas Corporation
Astropower Laboratory
Newport Beach, California 92660
for the
WATER QUALITY OFFICE
ENVIRONMENTAL PROTECTION AGENCY
Program #17020 DHR
Contract #14-12-417
December, 1970
-------
EPA Review Notice
This report has been reviewed by the Water
Quality Office, EPA, and approved for
publication. Approval does not signify that
the contents necessarily reflect the views
and policies of the Environmental Protection
Agency, nor does mention of trade names or
commercial products constitute endorsement
or recommendation for use.
11
-------
ABSTRACT
The purpose of this reverse osmosis study was threefold: (1) to
compare tubular membranes prepared from transesterified (modified)
cellulose acetate with commercially available cellulose acetate
(control), (2) to evaluate the in-situ regenerable membrane reverse
osmosis design on waste water and (3) to evaluate the membranes
on carbon-treated secondary effluents, primary effluents and con-
centrated primary effluents.
The test results were: (1) tubular membranes prepared from trans-
esterified cellulose acetate produced water fluxes slightly gr.eater than
those of membranes prepared from commercially available cellulose
acetate, (2) the in-situ regenerable membranes produced fluxes below
that of tubular units but showed sufficient promise for further develop-
ment, (3) product water flux from operations on carbon-treated second-
ary effluents gradually declined from initial levels of between 15 and
25 gfd. However, product water flux could be maintained near these
initial levels by periodic cleaning with enzyme-active laundry presoak
solution, (4) product water flux,on primary and concentrated primary
effluents declined gradually from initial levels of between 15 and .
25 gfd and stabilized between 4 to 5 gfd on concentrated primary
effluent even with the use of enzyme-active laundry presoak solution,
(5) removal of most waste water constituents was between 90 to 100%
and was generally unaffected by the type of feed water or time of test.
Chloride and nitrate reductions averaged approximately 70 percent.
The test results indicate that it is technically feasible to treat primary
effluents with tubular reverse osmosis process. However, further
development is needed to determine economic feasibility.
This report was submitted in fulfillment of Project Number
17020 DHR, Contract 14-12-417, under the sponsorship of the
Water Quality Office, Environmental Protection Agency.
Key Words: Water pollution, membranes, reverse osmosis,
sewage treatment, waste water treatment, cellulose
acetate, carbon-treated secondary effluent, primary
effluent, solid removal, organic removal, inorganic
removal.
111
-------
CONTENTS
Section Page
I CONCLUSIONS 1
II RECOMMENDATIONS 3
III INTRODUCTION 4
Study Objectives 4
Transesterified1 Modified Cellulose
Acetate 5
Regenerable"Mem.brane 6
IV -MEMBRANE-PREPARATION 7
Modified Cellulose Acetate 7
Tubular Membrane Preparation 7
Regen'era'ble Membrane Preparation 9
V LABORATORY TEST RESULTS 13
VI FIELD TESTING 16
Description of Equipment 16
Operating Conditions 19
Measurements 21
Data Reduction 23
VII OPERATION ON CARBON-TREATED
SECONDARY EFFLUENT 25
VIII OPERATION iON PRIMARY AND
CONCENTRATED PRIMARY EFFLUENT 31
IX DISCUSSION OF RESULTS 42
Membrane Performance 42
Flux Decline 42
Membrane Condition 43
Removal Efficiency 43
Feed Condition 43
Regenerable Membranes 44
X ACKNOWLEDGMENTS 45
XI REFERENCES 46
XII APPENDICES 47
IV
-------
FIGURES
Page
1 TUBULAR MEMBRANE CASTING
APPARATUS 8
2 COMPOSITE TUBULAR ASSEMBLY. 10
3 CONCEPTUAL SHELL AND TUBE DESIGN
(REPEATABLE UNIT) FOR REGENERABLE
REVERSE OSMOSIS PLANT 11
4 REGENERABLE MEMBRANE REVERSE
OSMOSIS UNIT 12
5 FLUX DECLINE USING 3, 000 PPM NaCl 15
6 TEST APPARATUS SCHEMATIC 17
7 FRONT VIEW OF UNIT INSTALLED AT
POMONA 18
8 REAR-VIEW OF UNIT INSTALLED AT
POMONA SHOWING MEMBRANE TUBES 1 8
9 PERFORMANCE OF TUBULAR MEMBRANES
ON CARBON-TREATED SECONDARY
EFFLUENT 27
10 PERFORMANCE OF REGENERABLE UNIT
NO. 5 ON CARBON-TREATED SECONDARY
EFFLUENT 28
11 PERFORMANCE OF MEMBRANE SET la
OPERATED ON PRIMARY AND CONCEN-
TRATED PRIMARY EFFLUENT 33
12 PERFORMANCE OF MEMBRANE SET 2a
OPERATED ON PRIMARY AND CONCEN-
TRATED PRIMARY EFFLUENT 34
13 PERFORMANCE OF MEMBRANE SET 3a
OPERATED ON PRIMARY AND CONCEN-
TRATED PRIMARY EFFLUENT 35
14 PERFORMANCE OF REGENERABLE UNIT
NO. 6 OPERATED ON PRIMARY AND
CONCENTRATED PRIMARY EFFLUENT 37
v
-------
TABLES
No. Pagt
I Average Membrane Performance with
Carbon-Treated Secondary Effluent Feed 26
II Removal of Waste-water Constituents
During Operating with Carbon-Treated
Secondary Effluent 29
III Average Membrane Performance with
Primary and Concentrated Primary
Effluent Feed 32
IV Removal of Wastewater Constituents
During Operation with Primary Effluent 38
V Removal of Wastewater Constituents
During Operation with Concentrated
Primary Effluent—I 39
VI Removal of Wastewater Constituents
During Operation with Concentrated
Primary Effluent—II 40
VI
-------
SECTION I
CONCLUSIONS
1. Tubular membranes prepared from transesterified cellulose
acetate produced product water fluxes*slightly greater than
those of membranes prepared from commercially available
cellulose acetate (Eastman 398-10).
2. The in-situ regenerable membranes produced fluxes below
that of tubular units, and solids built up severely on the
membrane surfaces during operations with concentrated
untreated primary effluent feeds. However, this design
showed sufficient promise for further development.
3. On carbon treated secondary effluent, membranes
prepared from transesterified cellulose acetate provided an
overall average product water flux of 15 gfd. Membranes
prepared from commercially available cellulose acetate
(Eastman 398-10) exhibited an overall average flux of 10 gfd,
both at 600 psig.
4. Product water flux from membranes operated on carbon treated
secondary effluents could be maintained near its initial level of
25 gfd for the (transesterified) modified membranes and 15 gfd
for the E-398-10 (control) membranes by periodic cleaning
(approximately once every 10 days) with an enzyme active
laundry presoak solution (Biz).
.;
5. Product .water flux from membranes operated with periodic
(~10 day interval) Biz cleaning on primary followed by con-
centrated primary effluent feed decreased gradually to 4 to
5 gfd (at 600 psig). Average initial flux was the same as with
the carbon treated secondary e'ffluent.
6. Upon return to straight primary effluent feed, the flux rose to
5 to 7 gfd and continued to increase until the end of the test.
Thus, average flux of an actual reverse osmosis unit would be
greater than the 4 to 5 gfd observed on a concentrated primary
feed.
7. Of the three cleaning procedures used (sponge ball, enzyme
active laundry presoak solution, and 6M Urea solution) for
membrane rejuvenation, the enzyme-active laundry presoak
solution was the most effective under the test conditions
employed.
*A11 wastewater feeds were adjusted to a pH of approximately 5 to 6 with
sulphuric acid.
-------
8. Flux decline rate (slope of log flux-log time plot) on concentrated
primary feed was the same as with pure salt solution.
9. Removal of wastewater constituents was generally unaffected by
the type of feed or time on test and remained essentially constant
.at 90 to 100 percent during operation on carbon treated secondary,
primary, and concentrated (up to 9 times the total solid .concen-
tration of primary effluent) primary effluent.
10. No degradation of any of the membranes was observed during
these tests.
11. It is technically feasible to treat primary effluents with the
tubular reverse osmosis process. However, further develop-
ment is needed to determine the economic feasibility.
-------
SECTION II
RECOMMENDATIONS
These results, coming from a small reverse osmosis test unit, do not
completely indicate full-scale operation. The use of feed recirculation
to simulate high recovery ratios may influence the test results in a way
not experienced in practice.
To take full advantage of these results, operation of an appropriately
sized reverse osmosis pilot plant on primary effluent at a sewage
treatment facility is recommended. The pilot plant should be designed
to specifically establish the necessary operating conditions and
parameters for sustained performance and to provide realistic,
full-scale, cost data on the process. It is estimated that the cost of
this study will be about $250, 000, of which, $120, 000 would be for
engineering, site preparation, laboratory supplies and procurement
of a 20, 000 gfd tubular reverse osmosis unit and $130, 000 for
20 months operation, chemical and physical analyses and four months
for the final report.
-------
SECTION III
INTRODUCTION
Rapidly increasing populations and expanding industrial activities are
placing greater 'demands on fresh water supplies. However, these
supplies are relatively static in availability and in some cases even
decreasing as the result of pollution. The Department of Interior has
recognized the need to augment the nation's natural water resources
through the desalination of brackish and marine waters. Support by
the Department of Interior has produced major advances in the tech-
nology and development of suitable desalination processes. It has
become apparent, however, that perhaps a much better source of
water is reclamation by these methods of municipal wastewater since
it contains far fewer dissolved minerals and is always available rela-
tively near the intended use.
Of the many demineralizing processes, the comparatively low-energy
reverse osmosis process appears-well suited to the renovation of
municipal wastewater. Processes requiring a change of phase, such
as distillation and freezing, are better suited for more saline waters
since their performance and costs are relatively independent of salt
concentration.
Conventional wastewater treatment processes require many steps to
remove wastewater constituents. Treatment processes incorporating
reverse osmosis may be capable of performing a. much superior treat-
ment in fewer operations; dissolved salts, organic substances, and
insoluble suspended matter are all removed in the same operation.
Basic elements of the reverse osmosis process consist of the mem-
brane, a means for providing a high-pressure differential across the
membrane, and a support for the membrane against this pressure
differential. A number of different membrane materials possess the
favorable osmotic properties of relatively high product water flux and
low or no solute transport. However, only cellulose acetate has found
extensive use in desalination.
STUDY OBJECTIVES
The purposes of this study were threefold.
(1) To compare tubular membranes prepared from transesterified
(modified) cellulose acetate with Eastman 398-10 cellulose
acetate (control).
(2) To evaluate the in-situ regenerable membrane reverse
osmosis design on wastewater. In the regenerable unit,
membranes were formed, removed and reformed in place
on porous tubes in a shell and tube configuration.
-------
(3) To evaluate the membranes on carbon treated secondary
effluents, primary effluents and concentrated pri-
mary effluents. The (Concentrated primary effluent
feed was used to simulate the near exit conditions
of an operating reverse osmosis unit. Successful treatment
of primary effluent by reverse osmosis would place this
process in a position favorable to the economics of the over-
all wastewater treatment system as well as eliminate the
relatively difficult to operate biological conventional second-
ary sewage treatment.
Both the tubular membrane and the regenerable membrane units were
assembled and operated in a portable test apparatus at the Ponoma
Water Reclamation Plant.
TRANSESTERIFIED MODIFIED CELLULOSE ACETATE
Since cellulose acetate is the most effective membrane material, its
properties were examined and modifications were made to improve its
stability, and consequently its desalination properties.
Commercially available cellulose acetate commonly used in reverse
osmosis consists mainly of the diacetate (2. 4 acetate); that is, 2.4 out
of the three hydroxyl groups "of each anhydroglucose unit are esterified.
Each unit of glucose has one primary and two secondary hydroxyl
groups.
For manufacturing reasons, the diacetate is made indirectly; that is,
the cellulose is first completely esterified to the triacetate; enough
water is then added to hydrolyze one group. The relative reaction
rates are such that the primary ester linkage is hydrolyzed first,
resulting in an ester having free (unacetylated) primary hydroxyl
groups. This product is satisfactory for the manufacture of lacquer,
which is the principal use of cellulose acetate.
Controlling the esterification process produces a cellulose acetate that
contains more acetylated primary hydroxyl groups and more free
secondary hydroxyl groups.
OH (SECONDARY)
CH2-OH (PRIMARY)
CH2-OH
(PRIMARY)
O-H (SECONDARY
-------
The balance between free and esterified hydroxyl groups can be shifted
towards the thermodynamically more stable equilibrium distribution by
dissolving the cellulose acetate in 98 to'99 percent acetic acid and main-
taining it at 65 to 85 °C for several hours. Malm (Ref. 1-4) has shown
that under proper conditions, no further esterification and little degra-
dation of the cellulose takes place, while the proportion of primary to
secondary ester groups shifts from 1:2 to 1:4, depending on time and
temperature.
Stabilized membranes prepared from this material were hoped to
provide greater product water flux and be more resistant to flux decline,
hydrolysis and other forms of degradation.
REGENERABLE MEMBRANE
Also evaluated in this program was a new, low-operating-cost, design
concept for reverse osmosis equipment. This concept was developed
with the support of the Office of Saline Water (Ref. 5) and consists of a
shell and tube configuration with membranes on the outside of porous
ceramic tubes. The porous ceramic tubes can be removed from the
shell. Membranes are formed on the porous supports by dipping into
a cellulose acetate casting solution followed by gellation with cold water
and a hot water cure. Membranes are removed by dipping the mem-
brane coated porous support tubes into a stripping solution. Membranes
can then be replaced by repeating the membrane forming procedure.
This design promises significant operating cost savings by eliminating
most of the labor cost for membrane replacement.
-------
SECTION IV
MEMBRANE PREPARATION
MODIFIED CELLULOSE ACETATE
Modified (transesterified) cellulose acetate was typically prepared in
400-gram lots, as described in the following procedure.
Into 1, 800 ml of acetic acid was dissolved with stirring 400 gm of dried
Eastman 398-10 cellulose acetate*. The mixture was heated to 85 to
87 °C, then a mixture of 40 ml 10 N HC1 and 400 ml acetic acid was
added. The reaction mix was stirred one hour at 86 ± 1°C. The cellu-
lose acetate was precipitated by pouring the mix into 100 liters of
distilled water. The solids were filtered through a woven cloth filter
and vacuum dried. Upon composite cellulose acetate batch analyses
the acetyl content by weight was 39. 7 percent. The hydroxyl content
by weight was 3. 46 percent of which 37 percent was primary and
63 percent was secondary hydroxyl gr-oups. The remaining hydroxyl
groups were converted (acetylated) to acetyl groups.
TUBULAR MEMBRANE PREPARATION
The casting solution and the tubular casting procedures were based on
the procedures described by Manjikan (Ref. 6) and Loeb (Ref. 7).
Casting solution was prepared by blending 25 percent cellulose acetate
with 45 percent acetone and 30 percent formamide (by weight). The
mix was then rolled overnight to thoroughly blend the material.
Tubular membranes were prepared in precision bore glass tubes
(0. 873-in. ID). The important divergence from the Loeb procedure
was to raise the casting bob within the stationary casting tube and
subsequently air dry the membrane ZO seconds prior to immersing
the tube and membrane into a chilled water (+2°C) gellation bath. The
apparatus is shown in Figure 1.
Membranes formed from the modified cellulose acetate were brittle
and initially could not be fabricated into tubular units, primarily
because the ends could not be flared without tearing. However, air
drying the membrane prior to gellation improved the tensile strength
sufficiently to allow flaring.
In operation, the casting tube is suspended by a yoke attached by cable
to a variable speed motor. A casting bob is lowered.by cable through
the tube. The casting tube is filled at the bottom with 80 to 150 cc of
casting solution, which is held in place by the bob. The tube is lowered
at a constant rate (15 ft/min) past a stationary bob, which is self-
centering and wipes an even film on the tube interior. The film gels
*Eastman Chemical Products Inc. , Kingsport, Tennessee.
-------
TO VARIABLE
SPEED MOTOR
REFRIGERATED WATER
REFRIGERATED
WATER SUPPLY
Figure 1. Tubular Membrane Casting Apparatus
-------
into a membrane as the casting tube enters the cold water bath. The
tubular membranes shrink away from the casting tube and may be
removed when fully gelled.
The tubular membrane is then wrapped with nylon cloth and inserted
into a type 304 stainless steel pressure tube. The membrane ends are
flared to receive the end fittings. This tubular, membrane construction
is illustrated in Figure 2. The tube length is 4-1/2 feet and tube
diameter is 1. 0 inch (OD) and/0. 86-inch ID to yield a membrane area of
approximately 1. 0 square foot per tube. After assembly, the mem-
brane is heat annealed in the tube by circulating hot water through the
tube.
REGENERABLE MEMBRANE PREPARATION
Regenerable membranes were formed directly, on the outside surface
of a porous support, as described in OSW Research Report No. 464
(Ref. 5). The membrane solution of E-398-10 cellulose acetate
(22 percent), formamide (24 percent), and acetone (54 percent) was
cast onto 1/4-inch diameter by 3-fopt long porous ceramic tubes.
These tubes were combined into a'tube bundle and inserted into a
pressure shell. Purified water flowed through the membrane ant1, up
the inside of the tube and out as illustrated in Figure 3. Figure 4 shows
the tube bundle and pressure shell.
The 1/4-inch tubes were sealed into a header section with epoxy. This
header section, shown as the base of'the tube bundle in Figure 4, was
subsequently attached to the pressure shell by a Victaulic* coupling.
The opposite end of the tubes was sealed and the product water flows
through the membrane, into the porous tubular membrane support and
to the header section which also serves as a product collection section
for the tubes.
*Victaulic Company of America, South Plainfield, New Jersey.
-------
NUT
SUPPORT TUBE, TYPE 304 STAINLESS STEEL
2.54 OUTER DIAMETER
0.089 WALL
-7.6
-7.6-
RUBBER GASKET
BEFORE INSTALLATION
DIMENSIONS IN CM
0.16 (TYPICAL)
NYLON\\
WRAPS -^
BODY
SLEEVE
^MEMBRANE
TUBE
,RUBBER
GASKET
AIDING SEAL
DEGREE FLARE FITTING
TYPICALLY
PARKER-HANNIFIN
•TRIPLE-LOK"
Figure 2. Composite Tubular Assembly (Ref. 1}
-------
PRODUCT OUTLET
PRODUCT COLLECTION
VICTAULIC COUPLING
TUBE HOLDING PLATE
POROUS
SUPPORT
STRUCTURE
CELLULOSE
ACETATE
REGENERABLE
MEMBRANE
BRINE FEED
Figure 3. Conceptual Shell and Tube Design (Repeatable Unit) for Regenerable Reverse Osmosis Plant
11
-------
Figure 4. Regenerable Membrane Reverse Osmosis Unit
12
-------
SECTION V
LABORATORY TEST RESULTS
To qualify the modified cellulose acetate, flat-sheet membranes were
prepared and tested at 1,000 psi against a 1-percent NaCl solution.
After 24 hours on test, the product flux was 29. 1 gfd at 96. 3-percent
salt removal. This was considered sufficiently close to the perform-
ance target (30 gfd at 97-percent-removal) to warrant using this
modified cellulose acetate.
Therefore tubular cellulose acetate membranes were made and
mounted into stainless steel tubes for heat treatment and initial
laboratory 24-hour test. This test also included membranes made of
unmodified cellulose acetate (E-398-10). The test results-control-
after 24-hour testing against 3,000 mg/1 sodium chloride at 600 psi
were:
Set A. Modified membranes (open) 54. 6 gfd at 61-percent salt
removal.
Set B. Modified membranes (tight) 20. 1 gfd at 92-percent salt
removal.
Set C. Standard membranes (E-398-10) 18 gfd at 96-percent salt
removal.
Comparison of Set B and Set C indicated only slight difference in
performance between the modified and standard membranes (control).
Additional laboratory tests were made while awaiting completion of the
field test unit. The lab tests established the membrane performance
using NaCl solutions over a select time period for comparison with the
results obtained later with Pomona effluents. Six tubular membranes
were selected at random from the 61-percent removal group (Set A).
These membranes were tested for 2400 hours at 600 psi with a
3,000 mg/1 NaCl solution. All product water flux and salt removal
values were corrected to 77°F. The group average performance was:
24 hours 51.9 gfd 62-percent salt removal
72 hours 47. 6 gfd 61-percent salt removal
1008 hours 38. 7 gfd 59-percent salt removal
1680 hours 26. 4 gfd 78-percent salt removal
2400 hours 21. 1 gfd 84-percent salt removal
A tight tubular membrane, Set B, and a regenerable unit with 5 sq ft
of surface area were also tested on a 3000 mg/1 NaCl solution at
600 psig. The results of these tests are plotted in Figure 5 as log
13
-------
of the product flux versus time. The flux de.cline coefficient, k, is
expressed by
J = J exp (kt)
where
J = product water flux in gallons'/sq ft-day(gfd)
J = initial product water flux in gfd
k = flux decline coefficient in hours
t = time on test in hours
The flux decline rate, m, is the slope of the flux decline curve on a
log flux versus log time plot
log J = log J + m log t
Both of these expressions for flux decline are tabulated below for the
laboratory tests on pure saline solutions (3, 000 mg/1 NaCl at 600 psig).
k m
Modified (loose) tubular membranes -0.059 -0. 115
(Set A)
Modified (tight) tubular membranes -0.066 -0.083
(Set B)
Regenerable membranes -0.077* -0.083
-0.008**
* Initial 600 hours of test
**Final 1000 hours of test
14
-------
en
a
LL
ID
§
10
9
8
7
6
5
4
O TUBULAR MEMBRANE SET A K = 0.059
E) TUBULAR MEMBRANE SET B K = 0.066
A REGENERABLE MEMBRANE K = 0.077
I
I
I
I.
I
I
I
60
180
300
420
540
660
780
• 900
HOURS
1,020 1,080 1,200 1,320; 1.440 1,560 1,680
Figure 5. Flux Decline Using 3000 PPM NaCI
-------
SECTION VI
FIELD TESTING
DESCRIPTION OF EQUIPMENT
A field test unit was built for operation at the Pomona Water Renovation
Facility. A schematic of this field test unit is shown in Figure 6 and
photographs in Figures 7 and 8. The major pieces of equipment are
described in Appendix A.
Municipal effluent from the Pomona Water Renovation Facility flowed
on demand directly into the 500-gallon feed tank. Tank level was regu-
lated by a float valve on the wastewater inlet line.
A centrifugal priming feed pump under the feed tank provided a contin-
uous positive head on the suction side of the high pressure feed pump.
Excess flow from the priming pump not required by the high pressure
feed pump was returned to the tank through a bypass line.
ThepHof the effluent in the feed tank was adjusted to approximately
5 to 6 with sulfuric acid. The pH sensing unit was located in the bypass
line from the centrifugal feed pump. As directed by the pH controller,
sulfuric acid was injected by a metering pump at a point adjacent to
the bypass return, which also provided good mixing of the acid.
Product water from the tubular units was collected in 2-in. diameter
polyethylene tubes mounted around the 1-in. stainless steel membrane
support tubes. These collection tubes were sealed at both ends with a
one-hole (1-in. ID) rubber stopper. The collection tubes were vented
to the atmosphere. The collected product water flowed by gravity to
the flow meters. One-quarter-inch diameter plastic tubes attached to
the bottom of the 2-in. collection tube were used to transport the
product water from the individual tubular membrane units to central
standpipes feeding the flow meter for each membrane set.
Initially, the te'st setup consisted of three sets of tubular membranes
and two regenerable units as follows:
Set 1. Modified cellulose acetate tubular membranes with
60-percent removal (5 ft^).
Set 2. Modified cellulose acetate tubular membranes with
90-percent removal (5 ft^).
Set 3. Standard membranes (E-398-10) with 95-percent.
removal (5 ft^).
Set 4. Regenerable unit No. 3 (5 ft^) with 90-percent removal.
Set 5. Regenerable unit No. 4 (5 it ) with 60-percent removal.
16
-------
SUPPLY
D--
pH
METER
ACID PUMP
* *ir\ T* A Mix
T
JO.
SOLENOID
LVALVE
BACK
PRESSURE
VALVE
TO WASTE
o V
o
AIR
ACCUMULATOR
LJ
FEED PUMP
TO WASTE
IHr-
COMPOSITE \_
M
TUBULAR ASSEMBLY
_T1 SOLENOID
~M VALVE
(MI) (M2) (MS)
TO WASTE
OR FEED TANK
TO FEED TANK
PRODUCT TANK AND
FLOAT OPERATED VALVE
HIGH PRESSURE PUMP
Figure 6. Test Apparatus Schematic
-------
Figure 7. Front View of Unit installed at Pomona
Figure 8. Rear View of Unit Installed at Pomona Showing Membrane Tubes
18
-------
Each tubular membrane set consisted of five tubular units, each
4-1/2 feet long by 0. 86 inches ID, 1. 0 inches OD, providing a mem-
brane surface area of approximately one square foot per tube. The
three types of individual tubular units (from Sets 1, 2 and 3) were
connected randomly in series so that the feed water flowed through
U-bends from one tube to the next.
Two regenerable membranes, each with five square feet of membrane
area were placed in series after the tubular units.
During the first two weeks of operation the following changes were
made.
1. A failure occurred in the regenerable units on the third day
when a tubular membrane loosened and was transported into
these units. Regenerable units No. 3 and No. 4 were replaced
with another membrane bundle, Regenerable Membrane Unit
No. 5, with 13 ft^ of membrane surface area and an average
salt removal capacity of 80 percent.
2. The product water meters proved to be unreliable at the low
flows and pressure heads. Product water production rates
were determined during the remainder of the test by mea-
surement of the product water flow during a given time,
usually one minute, whenever the units were serviced.
OPERATING CONDITIONS
The field test unit was operated continuously for 24 hours a day and
7 days a week throughout the tests. Operating pressure was maintained
at 600 psig. The flow rate through the reverse osmosis unit was main-
tained at 12 gallons per minute. This is equivalent to a flow velocity
past the tubular membranes of 1. 6 feet per second. Reynold's number
in the tubular units was
Three types of feed were tested:
1. Carbon treated secondary effluent
2. Primary effluent
3. Concentrated primary effluent.
Both the carbon treated secondary effluent and primary effluent flowed
continuously into the feed tank, past the membranes, and directly to
the drain. Product water also was allowed to go to the drain. Con-
centration changes of the feed within the test unit were negligible so
that the properties of feed throughout the test unit can be assumed
constant and equal to those of the incoming feed.
19
-------
During the concentrating step all of the primary effluent feed was
returned to the feed tank while the product water was again
allowed to go to drain. This allowed the feed to become more
more concentrated as the product water was removed. No blow-
down of the feed tank was allowed until the total solids concentra-
tion was 9 times that of the incoming untreated primary effluent feed
simulating the near exit conditions of a reverse osmosis system -with
approximately 90-percent water recovery.
The temperature of the carbon treated secondary and primary feed
stream averaged 80°F. Product water flux values were corrected to
77°F (25°C). Temperature of the concentrated primary feed averaged
120°F (49°C) due to the heat added by the high pressure pump during
recirculation of the feed stream. Product water flux values were also
corrected to 77°F (25°C) using a conservative correction factor
(~60 percent of the measured value). Thus, the flux values reported
for 77°F are more likely lower than the actual values. To avoid this
problem a heat exchanger should be used in any future studies where
the wastewater feed is concentrated.
All tests were conducted at a pH of approximately 5 to 6, in order to
reduce the potential for phosphate or carbonate precipitation as well as
reduce the potential for membrane hydrolysis. However, pH values as
high as 8 were reached during a two day malfunction of the pH control-
ler. The sulfuric acid which was injected at the bypass return to the
feed tank by the priming pump maintained the entire contents of the
feed tank at a pH of 5 with the secondary advantage of stabilizing the
wastewater composition.
Three different methods were experimented with to clean and rejuvenate
the tubular membrane surfaces: (1) sponge flush ball, (2) enzyme-active
laundry presoak (Biz) solutions, and, (3) urea solutions. The sponge
flush ball consisted of a cellulose sponge formed into a 1-in. diameter
ball. This ball made a tight fit in the 0. 86-in. diameter membranes
and was forced through the tubular membrane units under approximately
10-psig water pressure from the priming pump. However, the sponge
flush ball would occasionally tear one of the tubular membranes, and
was discarded in favor of a Biz flush which also gave higher flux
recovery. The cleaning procedure followed for both the Biz and urea
flush was to depressurize and drain the system, and then circulate
the desired solution past the membranes for 30 minutes. Flow velo-
city was <^1 ft/sec past the membrane and flow pressure was ~10 psig.
Solution temperature was 77°F (25°C). Concentration of the Biz
cleaning solution was 7. Ig of Biz per liter of water and concentration
of the urea cleaning solution was 360g (6M) of urea per liter of water.
The pH of the Biz solution was 9.2. The urea solution pH was main-
tained at 5.
After cleaning the system was drained and repressurized with the
normal feed. After 30 minutes of operation the product water flux
was determined. Any sampling for chemical analysis was done prior
to cleaning.
20
-------
MEASUREMENTS
Grab samples representing point values were taken for chemical
analysis. The following constituents were monitored and analyzed
in accordance with procedures outlined in the Twelfth Edition of
Standard Methods for the Examination of Water and Wastewaters,
American Public Health Association, Inc. , or FWQA approved altern-
ate methods — pH, electrical conductivity, total dissolved solids, total
solids, total chemical oxygen demand, filtered chemical oxygen demand,
and ammonia.
Electrical conductivity was determined by using a conductivity bridge,
Beckman Model RD 125J with temperature correction, with a 4. 5-ml
capacity cell, Beckman Model G2.
Total dissolved solids were measured by weighing the residue from a
filtered and evaporated sample. A 0. 45p. membrane filter was used.
Total solids were determined in the same manner using an unfiltered
sample.
Total chemical oxygen demand (COD) was determined by potassium
dichromate-sulfuric acid digestion for two hours and ferrous ammon-
ium sulfate titration to the ferroin indicator endpoint; any chloride
present in the sample was complexed with mercuric sulfate. Filtered
chemical oxygen demand was determined in the same manner from a
sample filtered through a 0. 45 p. membrane filter. Initially, a dis-
posable plastic filter unit (Nalgene No. 245) was used. However,
increases in the COD value, attributed to preservatives in the filter,
were observed. Subsequent filtering was done with a pressure filter
(Millipore No. 4004700) and 0. 45p. membrane filter (Millipore HA).
Ammonia contained in samples was distilled from Kjeldahl flasks and
collected in boric acid solution. The amount of ammonia in the distil-
late was determined colorimetrically at 425 fj. wavelength in a spectro-
photometer following Nesslerization.
Samples were taken early in the morning and refrigerated so that
the chemical oxygen demand measurements could be made the same
day since changes in the COD values were observed in samples held
overnight, even under refrigeration. Samples taken for ammonia or
nitrogen analysis were fixed with sulfuric acid (~2 ml/1) immediately
after they were taken.
Product water flux was measured before and 30 minutes after
membrane cleaning to allow the flux to stabilize. The product water
flux only was measured after membrane cleaning because of possible
residues of the cleaner used. All other parameters such as salt,
solids and COD removal were measured prior to membrane cleaning.
Four samples were taken during these tests for a complete chemical
analysis as follows.
21
-------
Sample 1. At the end of the carbon treated secondary effluent
test.
Sample 2. At the end of the primary effluent test.
Sample 3. In the middle of the concentrated primary effluent
test.
Sample 4. At the end (prior to use of urea for membrane
cleaning) of the concentrated primary effluent test.
The following analyses were made on these samples:
MBAS
Total Alkalinity
Total Hardness
Phosphorus
Total Nitrogen
Chloride-
Sulfate
Calcium
Magnesium
Volatile Matter
Methyl blue active substances minimum
detectable level 0.001 mg/1
mg/1 as CaCOg using methyl orange indicator
and comparing against a standard curve.
mg/1 as CaCOg using Eriochrome Black T
indicator and comparing against a standard
curve.
mg/1 of P by the stannous chloride method
in which all phosphorus is converted to ortho-
phosphate and analyzed colorimetrically using
a factor of 0,3263 to convert to phosphorus.
l
mg/1 of free ammonia and organic nitrogen
by the Kjeldahl method.
mg/1 of C 1 by the Mohr titration method using
silver nitrate.
mg/1 of SC>4 measured gravimetrically after
precipitation with barium chloride
mg/1 of Ca by titration with disodium
dihydrogen using muroxide as an indicator.
mg/1 of Mg determined by difference between
calcium and total hardness.
Suspended material - filtrate from Gooch
filter and dried for 1 hour at 500°C
sensitivity 0.01 mg/1.
In order to present comparable product water flux measurements a
plot of temperature versus water flux was prepared and all flux values
corrected to 77°F (25°C).
22
-------
For the carbon treated secondary and primary effluents which were
discharged after passing through the test unit the average feed tem-
perature was 80° and the necessary temperature correction was 0. 97.
The average feed temperature during recirculation (concentrated
primary effluent test) was 120°F (49°C) requiring a temperature
correction factor .of 0. 60.
DATA REDUCTION
Initial examination of the data on product water fluxes indicated that
most test results maybe presented as exponential functions of time.
Therefore flux decline data were calculated using all the data points
presented and are given in the form of coefficients J and k to fit the
general equation:
J = J0 e kt (1)
where J is product water flux in gal/(sq ft) (day), Jo is initial flux
upon attainment of specified concentration, k is the flux decline
coefficient in hours" \ and t is expressed in hours of operation at
desired wastewater concentration condition. The fluxes were converted
to their logarithms and the line of best fit for the data was determined
by least-square regression methods. Average product water flux for
this period was calculated from this line of best fit.
An alternative method of representing the data is by a log-log plot of
product water flux vs time. Using this representation the product
water flux decline may be expressed by the general equation:
log J - log J + m log t (2)
where J, J and t have the same connotation as above. In this
equation m is the slope of the product water flux decline curve.
Both the flux decline coefficient, k, and the slope of the product
water flux decline curve on a log-log plot, m, were calculated using
the least square regression methods and are presented in this report.
A semi-logarithmic plot of log flux vs time, following equation (1),
was chosen to represent the data because it presents the changes
occurring during the test in a simple chronological manner with each
portion of the. test given equal representation.
Membrane performance with respect to constituent removal was
calculated on the basis of the feed to the reverse osmosis unit.
However, at the flow rate used in this test unit (12 gpm) the changes
in feed concentration as it passes through the unit are negligible.
Thus the composition of the feed entering the unit can be considered
representative of the entire unit.
23
-------
A membrane coefficient, pig/(sq cm) (sec) (atm) was calculated and
plotted along with the standard flux units of gfd. The membrane
coefficient corrects or normalizes the flux values for operating
pressure, thus allowing direct comparison of product water flux
measured at different operating pressures,. Several normalizing
coefficients have been used by other workers (Ref. 5, 6 & 7), but
this one has the advantage of yielding numerical values almost
identical to gfd in the very commonly used 600 to 700 psig operating
pressure range.
24
-------
SECTION VII
OPERATION ON CARBON-TREATED SECONDARY EFFLUENT
The tubular membranes described in Section VI were operated on
carbon-treated secondary effluent with and without periodic cleaning
for 1, 185 hours* The overall performance of these membranes on
carbon-treated secondary effluent is summarized in Table I. The
changes in product water flux with time are presented in Figure 9 as
a plot of log J (flux) vs time.
During the first 400 hours of operation (Stage I with no membrane -
cleaning) the product water flux from membranes based on modified
cellulose acetate (Set 1 & 2) dropped from an initial value of 30 gfd
to a more-or-less equilibrium value of 10 to 15 gfd. During the same
period product water flux from control membranes based on E-398-10
(Set 3) declined from an initial value of 20 gfd to an equilibrium value
of 7 gfd.
However, subsequent periodic cleaning with an enzyme-active laundry
presoak solution (Biz) restored and maintained the product water
flux at very near its initial value.
Salt (conductivity), COD, and total solids removed (rejected by the
membranes) remained fairly constant throughout the test. Average
values for Set 3 were 95 to 98 percent. Sets 1 and 2 were lower
(68 to 85 percent), although Set 2 rose to 96 percent during the
latter portion of the test. The lower rejection observed with
Set 1 was attributed to small mechanical leaks and was corrected
in subsequent work with primary effluent.
Changes in product water flux with time for the regenerable membrane
followed the same pattern as the tubular units, Figure 10. Although
flux and salt rejection were somewhat lower, the flux decline coeffi-
cients, k and m, were also lower.
The detailed data for Table I and Figures 9 and 10 are tabulated in
Appendix B.
Near the completion of the operation on carbon-treated secondary
effluent a complete analysis was made of the feed and product streams.
The results are tabulated in Table II.
The carbon treated secondary effluent used in this test was obtained
from the Pomona Reclamation Plant. Just prior to this study,
secondary effluent was treated in four consecutive activated carbon
>'<
'"All carbon-treated secondary effluent feeds were adjusted to pH
of approximately 5 to 6 with sulfuric acid.
25
-------
Table I
AVERAGE MEMBRANE PERFORMANCE WITH CARBON-TREATED
SECONDARY EFFLUENT FEED
Membrane Set No.
Tubular
Regenerable
Flux Decline
-1,
Coefficient*, k (hr )
-0.00258 -0.00084 -0.00334 -0.00178
Slope, m,:**
-0.403 0.295 0.482 0.215
Average Flux, gfd
(Stage I, No Cleaning) 22
Average Flux, gfd
(Stage II, Periodic Cleaning) 29
17
21
11
16
10
Average Percent Removal
(Total Solids)
76
83
97
77
Average Residual Total
Solids. , mg/i
Average Percent Removal
(Total COD)
130
85
100
84
25
97
130
85
Average Residual"COD,mg/l
Average Percent Removal
(Total Salts based on
conductivity)
68
80
95
73
^Calculated for the first 400 hours of test and prior to any cleaning.
**Slope of the flux decline curve on a log-log plot of flux vs time.
Also calculated for the first 400 hours of test.
26
-------
>-
<
0
o
w
(D
x"
it
UI
I
Q
O
cc
OL
D SET 1
A SET 2
O SET 3 (CONTROL)
I I
5.8
oc
m
LU
200
400
600 800
TIME (HR)
1,000
1,200
1,400
Figure 9. Performance of Tubular Membranes on Carbon-Treated Secondary Effluent (600 psig)
27
-------
40
5
29.0 r
23.2
O
01
17.4
5
O
I
y 11.6
LU
z
<
cc
00
5
UJ
5.8
20
h- 10
X
D
cc
UJ
ui
,.
Q
S 4
a.
STAGE I
STAGE II
100
200
300
600
700
800
900 1,000
400 500
TIME (HR)
Figure 10. Performance of Regenerable Unit No. 5 on Carbon-Treated Secondary Effluent (600 psig)
28
-------
Table II
REMOVAL OF WASTEWATER CONSTITUENTS DURING OPERATION
WITH CARBON-TREATED SECONDARY EFFLUENT*
Waste-water
Constituent
Tubular Membrane Set No.
1 2 3 Feed'
MBAS (mg/1)
Percent Removal
<0.00 <0.001 <0.001 0.25
100 100 100
Total Alkalinity (mg/1 as Ca CO_) 7.50 15.00 10.00 30.00
Percent Removal 75 50 60
Total Hardness (mg/ 1 as Ca CO3) 32. 5
Percent Removal 85
32.50 2.50 210.00
85 99
Phosphorus (mg/1)
Percent Removal
3.26 3.58 ' 0.97 11.55
72 69 92
Total Nitrogen (mg/1)
Percent Removal
8.40 6.30 4.20 14.80
43 57 72
Chlorides (mg/1)
Percent Removal
49.6 17.72 3.54 106.35
53 83 97
Sulfates (mg/1)
Percent Removal
36.6 31.68 Trace 281.89
87 89 99
Calcium (mg/1)
Percent Removal
10.0 10.00 0.95 66.00
85 85 99
Magnesium (mg/1)
Percent Removal
1.81 1. 85 Trace 10.89
83 83 99
Volatile matter (mg/1) <0..0
<0.01 <0.01
*Sample taken after 1, 185 hours of operation on carbon-treated
secondary effluent.
**After pH adjustment with sulfuric acid
29
-------
columns, operated in series, for the removal of dissolved organics
and suspended matter. However, during the first portion of this
R. O. study, from 0 to 431 hours (February 3, 1970 until
March 11, 1970), the secondary effluent^was treated in a single
column followed by additional treatment in a small auxiliary column
that acted as a polishing unit. As a result of this mode of activated
carbon treatment, the carbon effluents were of relatively poorer
quality with total COD concentrations ranging from 10 to 30 mg/1 as
compared to 7 mg/1 to 11 mg/1 (after normal,activated carbon opera-
tions were resumed) and total solids averaging 650 mg/1 as compared
to 6l3 mg/1 (after normal activated carbon operations were resumed
Table BV appendix). Thus the R. O. test results during the 0 to 431
hour period were probably adversely affected by the relatively
poorer carbon effluent, especially the product water flux decline.
After March 11, 1970,- the secondary effluent was again treated by
the four-column group, but modified to operated as two 2-stage units
in parallel. However, detention time in each of the two 2-stage units
was the same as in the four-stage consecutive series operation, .thus
carbon effluent quality was approximately equivalent. Carbon effluents
from the two 2-stage units were used from 431 hours until the end of
the carbon effluent test period. •
30
-------
SECTION VIII
OPERATION ON PRIMARY AND CONCENTRATED
PRIMARY EFFLUENT
A new set of tubular membranes of the same dimensions was
prepared from modified cellulose acetate and commercially
available cellulose acetate, Eastman 398-10. Regenerable
membranes using Eastman 398-10 were reformed directly on
the surface of porous ceramic support tubes. These membranes,
identified as follows, were installed on a portable test stand at
the Pomona Water Reclamation Plant and operated on primary and
concentrated primary effluent. *
Set la. Modified cellulose acetate tubular membranes
with 60-percent removal (5 ft2).
Set 2a. Modified cellulose acetate tubular membranes
with 90-percent removal (5 ft^).
Set 3a. Standard tubular membranes with 90-percent
removal (5 ft2).
o
Set 6. Regenerable unit No. 6 (5 ft ) with 90-percent
removal.
The overall performance of these membranes is summarized in
Table III. The changes in product water flux with time are pre-
sented in Figure 11 for tubular membrane set la, Figure 12 for
tubular membrane set 2a and Figure 13 for tubular membrane
set 3a.
The product water flux declined slowly during the first half of the
test (Stage I) period from an initial value of 22 to 27 gfd to 2 to
3 gfd (Figure 11, 12, and 13). Flux then rose and appeared to
stabilize at 5 to 6 gfd, in Stage II.
Stage I and Stage II correspond approximately to the period of
primary and concentrated primary feed respectively. Although
recycling of the feed was instituted during Stage I, higher feed
concentrations were not reached until the end of Stage I. Total
solids content of the feed to the reverse osmosis test unit at the
end of Stage I was five times that of the incoming primary
effluent. During Stage II the total solids content of the feed
continued to rise reaching a maximum of nine times that of the
incoming primary effluent. This concentration simulates the
exit conditions of a reverse osmosis unit operating near 90-per-
cent water recovery.
>'<
''^Primary and concentrated primary effluent feeds were pH
adjusted to approximately 5 to 6 with sulfuric acid.
31
-------
Table III
AVERAGE MEMBRANE PERFORMANCE WITH PRIMARY AND
CONCENTRATED PRIMARY EFFLUENT FEED
Membrane Set No.
Tubular
Regenerable
la
2a
3a
Flux Decline
Coefficient, k, hr~^
Slope, m*
(Stage I)**
Flux Decline
Coefficient, k, hr~^
Slope m*
(Stage II)***
Average Flux, (gfd)
(Stage I)*
Average Flux (gfd)
(Stage II)**
Average Percent
Removal
(Total Solids)
Average Percent
Removal
(Total COD)
Average Percent
Removal
(Total Salts)
-0.00234 -0.00254 -0.00277
-0.438 -0.428 -0.540
-0.00057 -0.00042 -0.000129
-0.080 -0.101 -0.021
14
6
92
95
86
17
6
94
97
97
92
96
89
73
66
* Slope of the flux decline curve on a log -log plot of flux versus
time.
**Primary effluent.
***Concentrated primary effluent, 5'times the average total solids
concentration of untreated primary effluent.
i •
Other performance parameters such as the salt (conductivity)
removal, the COD removal and the total solids (TS) removal
changed very little during the 2, 500 -hour test period including
Stage I and Stage II. Average salt (conductivity) removal was
94 percent for the tight modified cellulose acetate membrane
(Set 2a) and the standard cellulose acetate membrane (Set 3a) ,
and 87 percent for the high-flux modified cellulose acetate
membranes (Set la). The COD removal averaged 98 percent
32
-------
100
CO
CO
1
tr
r
y
r
tT
C
V-
n
LINE OF BEST FIT
200 400
600
800 1,000 1,200 1,400 1,600 1,800 2,000 2,200 2,400 2,600 2,800
Figure 11. Performance of Tubular Membrane Set 1a Operated on Primary and Concentrated Primary Effluent (600 psig)
-------
100
CONCENTRATED
5:1 TO 9:1
•LINE OF BEST FIT
CO
CONCENTRATED
TO 5:1
200 400 600 800 1,000 1,200 1,400 1,600 1,800 2,000 2,200 2,400 2,600 2,800
Figure 12. Performance of Tubular Membrane Set 2a Operated on Primary and Concentrated Primary Effluent (600 psig)
-------
100
OJ
Ul
200
400
600
800
1,000
1,600 1,800 2,000 2,200 2,400 2,600 2,800
1,200 1,400
TIME (HR)
Figure 13. Performance of Tubular Membrane Set 3a Operated on Primary and Concentrated Primary Effluent (600 psig)
-------
for all tubular membranes. The total solids removal averaged
97 percent for the tight modified cellulose acetate membrane
and 91 percent for both the standard and the high-flux modified
cellulose acetate membrane.
Again, membranes were cleaned using (1) a sponge flush ball,
(2) enzyme-active laundry presoak (Biz) solutions and (3) urea
solutions. Initially the sponge flush ball was used, but it would
occasionally tear a membrane. The flux recovery after cleaning
with the laundry presoak solution was higher than with the sponge
ball 'so most subsequent membrane cleanings were made with Biz.
To determine the performance of urea solutions, two successive'
flushed were made with a 6M solution of urea in ambient tempera-
ture'water (^26°C). The urea solution pH was adjusted to 5.
Flux recovery after membrane cleaning with 6M urea
solutions was approximately one half that achieved with Biz
solutions, Figures 11, 12 and 13. However, the use
of the urea solutions at ambient temperatures (26° C) rather
than with heated solutions probably contributed to its reduced
effectiveness.
Performance of the regenerable membrane system on primary
effluent followed the same pattern as the tubular membrane
systems. However, the solids built up severely on the surface
of the membranes during operation with concentrated primary
effluents, Figure 14. Examination of this unit after the test
showed the presence of loose organic material on the membrane
surface. This loose material could be washed away easily with
a water jet, but a brown film remained on the membrane surface.
The average membrane performance is summarized in Table III.
Detailed data for this table and Figures 11 through 14 are tabu-
lated in Appendix C.
A complete analysis of'the product and feed was made at the
conclusion of\ operations on primary effluent and in middle and
at the conclusion of operations on concentrated primary effluent,
Table IV, V, and VI. As with the carbon-treated secondary
effluent, the membranes were effective (85 to 100 percent) in
removing most of the MBAS, hardness, phosphorous, sulfates,
calcium, and magnesium. Nitrogen, nitrate and chlorides
removal was somewhat lower, ranging from 66 to 93 percent.
At the conclusion of the urea cleaning tests, the feed was
switched back to primary effluent for two weeks, followed by
approximately two weeks of concentrated primary effluent feed
(5. times the total solids concentrations of primary effluent).
After the two-week untreated primary effluent test, a sponge
ball was forced through the tubular membrane units. The
solids loosened by the sponge ball, amounting to approximately
100 grams, were collected and analyzed spectrographically,
Appendix D.
36
-------
z _
UJ
LL —
UJ O
CD Z
11.6
CO
5.8
o
J
<
C3
X
OL
UJ
10
9
8
7
<
9 5
Q
2 2
Q.
100
200
300 400
TIME (HR)
500
600
700
Figure 14. Performance of Regenerable Unit No. 6 Operated on Primary and Concentrated Primary Effluent (600 psig)
-------
Table IV
REMOVAL OF WASTEWATER CONSTITUENTS DURING OPERATION
, WITH PRIMARY EFFLUENT*
Waste water
Constituents
MB AS (mg/1)
Percent Removal
Total Alkalinity (mg/1 as CaCO3)
Percent Removal
Total Hardness (mg/1 as CaCO3)
Percent Removal
Phosphorus (mg/1)
Percent Removal
Total Nitrogen (mg/1)
Percent Removal
Chlorides (mg/1)
Percent Removal
Sulfates (mg/1)
Percent Removal
Calcium (mg/1)
Percent Removal
Magnesium (mg/1)
Percent Removal
Volatile Matter (mg/1)
Percent Removal
Tubular
la
<0.001
100
35.00
59
22.50
90
0.94
93
7.00
83
44.21
76
5.34
98
6.00
93
1.81
90
<0.01
100
Membrane
2a
<0.001
100
20.00
76
2.50
99
0.98
92
9.10
78
8.86
95
Trace
99
0.95
99
Trace
99
<0.01
100
Set No.
3a
0.001
100
45.00
47
20.00
91
1.92
85
14.00
66
51.40
73
6.99
97
7.00
92
1.21
93
<0.01
100
Feed**
0.43
85.00
230.00
12.92
40.60
187.43
294.47
85.00
18. 10
0. 10
*After 350 hours of operation - end of run on primary effluent
(once through).
**After pH adjustment with sulfuric acid.
38
-------
Table V
REMOVAL OF WASTEWATER CONSTITUENTS DURING OPERATION
WITH CONCENTRATED PRIMARY EFFLUENT-I*
Wastewater
Constituents
M'BAS (mg/1)
Percent Removal
Total Alkalinity (mg/1 as CaCO )
Percent Removal ^
Total Hardness (mg/1 as CaCOJ
Percent Removal
Phosphorus (mg/1)
Percent Removal
Total Nitrogen (mg/1)
Percent Removal
Chlorides (mg/1)
Percent Removal
Sulfates (mg/1)
Percent Removal
Calcium (mg/1)
Per cent' Removal
Magnesium (mg/1)
Percent Removal
Volatile Matter (mg/1)
Percent Removal
Tubular
la
0. 15
66
10.00
76
100.00
94
6.69
77
26.60
83
129.39
66
225.50
92
27.00
94
7.86
95
<0.01
100
Membrane
2a
<0.001
100
10.00
76
25.00
99
2.93
90
11.20
93
46.08
88
57.84
98
8.00
98
1.21
99
<0.01
100
Set No.
3a
<0.001
100
10.00
76
40.00
98
3. 10
89
16. 80
90
101.03
74
3. 10
99
13.00
97
1.81
99
<0.01
100
Feed To
Membrane**
(Concentrated)
0.45
42. 50**
1,800.00
29.36
16.1.00
382. 86
2,990.37
448.0
164.56
0.70
Feed To
Tank***
(Unconcentrated)
0.21
300.00
200.00
10.31
32.20
170. 16
127.56
52.0
16.94
0.45
GO
vo
*Afte.r 1, 300 hours of operation—middle of Stage II.
**After pH adjustment with sulfuric acid.
!=Not pH adjusted.
-------
Table VI
REMOVAL OF WASTEWATER CONSTITUENTS DURING OPERATION
WITH CONCENTRATED PRIMARY EFFLUENT-II*
Waste-water
Constituents
MBAS (mg/1)
Percent Removal
Total Alkalinity (mg/1 as CaCO3)
Percent Removal
Total Hardness (mg/1 as CaCO,)
Percent Removal
Phosphorus (mg/1)
Percent Removal
Total 'Nitrogen (mg/1)
Percent Removal
Chlorides (mg/1)
Percent Removal
Sulfates (mg/1)
Percent Removal
Calcium (mg/1)
P-ercent Removal
Magnesium (mg/1)
Percent Removal
Volatile Solids (mg/1)
Percent Removal
Nitrates (mg/1)
Percent Removal
Tubular
la
1.03
67
10.00
50
80.00
96
7.01
89
19.60
77
102.80
71
176.94
93
22.00
97
11.49
88
<0.01
100
0.97
82
Membrane
2a
0.12
96
5^00
75
20.00
99
3.10
95
14.00
84
46.08
87
50.61
98
7.00
99
0. 60
99
<0.01
100
0.88
83
Set No.
3a
1.25
60
15.00
25
30.00
99
3.59
94
11.20
87 *
81.53
77
72'. 01
97
9.00
99.,
1.81
98
<0.01
100
1.77
67
Feed To
Membrane**
(Concentrated)
3.12
20.00**
2,000.00
63.62
85.00
354.50
2,658.70
640.00
96.80
92., 00
5.31
Feed To
Tank***
(Unconcentrated)
2. 60
295.00
250.00
14.57
35.00
132.93
88.7
53.00
28.43
48.67
2.65
*After 1, 825 hours of operation, end of Stage II.
#*After pH adjustment with sulfuric acid.
pH adjusted.
-------
At the conclusion of the test program (following operation with
concentrated primary feed) one tubular membrane was removed
for examination prior to any cleaning. Very little residual
material was found on the membrane. This residual material
was removed and amounted to a little over one tenth of a gram.
This material was also analyzed spectrographically, Appendix D.
The large sample collected with a sponge ball cleaning following
the test with primary feed probably came from solids lodging in
fittings and bends of the plumbing since such a small amount of
material was actually found on the membrane following disas-
sembly. The major cation constituents of the residue on the
membrane were iron and sodium. Silicon, lead, barium
and calcium were also present in considerable amount
(>1 percent). Phosphorus was also present in consideraole
quantity.
41
-------
SECTION IX
DISCUSSION OF RESULTS
MEMBRANE PERFORMANCE
Performance (flux and removal efficiency) of tubular membranes
prepared from modified cellulose acetate was slightly better on
primary and concentrated primary feed than that of tubular mem-
branes made from Eastman E-398-10 cellulose acetate control. The
results from earlier tests made with carbon treated secondary efflu-
ent were inconclusive because of problems in sealing the ends of the
modified membranes; some of the tubular membranes prepared from
the modified membranes apparently developed small leaks which did
not change the product water flux, but by allowing a small quantity of
feed water to pass through, substantially reduced constituent removal
efficiency. Five of these tubular membranes constituted a membrane
set and the product water was combined so that a leak in any mem-
brane would change the removal efficiency of the entir.e set.
Membranes formed from the modified cellulose acetate were mechan-
ically weaker and more brittle than membranes prepared from
E-398-10 (control). The ends of the modified membranes could not
be flared for accepting the end fitting seals without cracking. Modi-
fying the casting cycle to provide a 20-second air dry allowed the
tubular membranes made from modified cellulose acetate to be flared
and sealed if done carefully.
The first set of membranes (used for carbon treated secondary
effluent) were made before this procedure was completely developed.
Also initially sponge balls were used to clean the membranes and the
mechanical abrasion occasionally tore the more brittle modified
membranes.
Comparison of results between the high-flux and low-flux transesterified
cellulose acetate membranes indicate that on carbon-treated secondary
effluent the high flux membrane produced a greater product water flux on
the average. However, the flux decline rate was also greater for the
high flux membrane. On primary and concentrated primary effluents,
comparison shows little difference between the two types of membranes.
The type and quantity of membrane foulants were probably different for
the different wastewaters. On carbon treated secondary effluent, the
product.-water flux was maintained near its initial level using the enzyme-
active, laundry-presoak solution. On primary and concentrated primary
effluents, the product water flux declined even with the use of the
enzyme-active, laundry-presoak solution. However, regardless of the
wastewater used, substantial flux decline was observed on the once
through (less than 2 percent product recovery) tests indicating that
membrane fouling is the most important factor in reducing membrane
productivity.
42
-------
FLUX DECLINE
With a carbon treated secondary effluent feed and no membrane clean-
ing the product water flux declined at 4 to 5 times the rate, based on
the slope of a log flux log time plot, of membranes with a pure salt
solution feed. However with periodic membrane cleaning the flux could
be maintained at near its initial level.
With a primary effluent feed and periodic membrane cleaning the flux
declined at a rate (m = 0. 4 to 0. 5) 4 to 5 times that of pure salt solu-
tion. However, during the transition from primary to concentrated
primary effluent feed the flux decline rate, m, abruptly changed to1
equal that observed with pure salt solution (m = 0.08 to 0. 10'j. How-
ever, the product water flux for treating concentrated primary
effluents was relatively low, approximately 5 to 6 gfd. During the
final portion of the test when the feed was changed back to pri-
mary effluent the flux rose by approximately 10 percent and
was still increasing up to the end of the test. This means the
average flux obtained in an actual reverse osmosis system -with
near 90-percent recovery would be Higher than the 5 to 6 gfd
observed with the concentrated primary effluent feed. These
test results are encouraging in that they indicate long term
stability of the membranes when operated at high recovery ratios
on primary effluents.
MEMBRANE CONDITION
Very small quantities of residual solid materials were found on tubu-
lar membrane surfaces following testing with concentrated
primary effluent. This suggests that product water flux decline
is due to deposition of colloidal, dissolved organic and dissolved
inorganic materials in the membrane pores rather than gross, easily
settled, particulate matter on the membrane surface.
REMOVAL EFFICIENCY
The removal efficiency of all the membranes was most encouraging,
generally exceeding 90 percent for most constituents in the waste-
water. Even more encouraging was the ability of the cellulose acetate
membranes to maintain this high removal rate with exposure to con-
centrated primary effluent, high-feed water temperature (49°C),
occasional feed-water pH of 8, and periodic flushing with a strong
enzyme-active laundry presoak solution.
Among the wastewater constituents the nitrates and chlorides were the
most difficult to remove. But even for these materials the average
removal efficiency was over 70 percent.
FEED CONDITIONS
During concentration of the primary effluent almost the entire feed
43
-------
was recirculated and the energy introduced by the high-pressure feed
pump increased the temperature to 49 °C. A .temperature correction
curve was prepared in which the product water flux at the higher tem-
perature was compared to that at 25 °C (77°F). In view of the sub-
stantial temperature correction factor (0. 6) the product water flux
reported is believed to be conservative.
Concentration of the wastewater constituents could lead to a
separation and partial segretation of the suspended solids.
However the high recirculation rate, 12 gpm, and conical feed tank
bottom make this unlikely. Sulfuric acid was used to adjust the pH of
the feed in the feed tank to approximately pH 5 to 6. This produced a
sulfate concentration as high as 3 g/1 during the concentration phase,
and a correspondingly low alkalinity.
REGENERABLE MEMBRANES
The regenerable membrane concept provides a low cost reverse
osmosis system. A simple tube and shell configuration was used in
which the feed was put into the shell side. Thus flow velocities past
the membrane coated tubes were generally much lower and more vari-
able than with the tubular membranes. On carbon treated secondary
and primary effluent feed this resulted in somewhat lower product
water flux. However, during operation on concentrated primary efflu-
ent feed the solids deposition on the membrane surface was severe and
could not be removed by flushing with Biz solution.
As a r'esult of this test, inexpensive flow directors will be incorporated
into the tube and shell design. These flow directors will provide a
constant high-velocity flow of the feed past the membrane. The addi-
tion of these flow directors adds less than a penny per thousand gallons
to the cost of waste water treatment by'this reverse osmopis system.
44
-------
SECTION X
AC KNOWLEDGEMENTS
The program reported here was performed by Astropower Labora-
tory, McDonnell Douglas Corporation at Newport Beach, California,
under the direction of Mr. Gerald Stern, WQO Project Officer.
McDonnell Douglas personnel participating in the program were
Dr. G. A. Guter, Program Manager; Mr. L. M.' Tint, Project
Chemist; Mr. H. K. Bishop, Project Engineer; and Mr. G. F. Schlee
and Mr. R. G. McMillen, Laboratory Technicians.
The complete cooperation and assistance of the County Sanitation
Districts of Los Angeles County and the WQO personnel at Pomona
in providing the municipal wastewaters, space, and service for oper-
ation of the reverse osmosis test unit used in this program are grate-
fully appreciated and acknowledged.
45
-------
SECTION XI
REFERENCES
1. Malm et al. JACS 7£, 2740
2. Malm et al. JACS 12_, 2674
3. Malm et al. JACS 74, 4105
4. Malm et al. JACS 75_, 80
5. Bishop, H. K., Belfort, G. and G. A. Guter, "In-situ
Formation of Regenerative Cellulose Acetate Membranes
on Porous Supports, " Office of Saline Water Research
& Development Report No. 464, Dec 1969
6. Manjikian, S., Loeb, S. , and McCutchan, J. W. ,
"Improvement in Fabrication Techniques for Reverse
Osmosis Desalination Membranes, " Proc. First Intl.
Symp. Water Desalination, U. S. Dept. of Interior,
Office of Saline Water, Wash. D. C. 2: 159, 1965,
7. Loeb, S. , "Sea Water Demineralization by Means of a
Semipermeable Membrane, "Univ. of California, Dept.
of Engr. , Los Angeles, Progr. Rept. No. 6142,
Aug 1961.
46
-------
SECTION XII
APPENDICES
Page No,
A. Description of Equipment ............ 49
B. Performance of Membranes Operated on
Carbon- Treated Secondary Effluent ....... 50
Table BI : Performance of Tubular Membrane
Set ,1 Operated on Carbon-Treated
Secondary. Effluent .......... 51
Table BII: Performance of Tubular Membrane
Set 2 Operated on Carbon- Treated
Secondary Effluent ....... . . . 52
Table Bill: Performance of Tubular Membrane
Set 3 (Control) Operated on Carbon-
Treated Secondary Effluent ...... 53
Table BIV: Performance of Regenerable
Membrane Set 5 Operated on
Carbon- Treated Secondary Effluent . . 54
Table BV: Properties of Carbon- Treated
Secondary Effluent Feed ....... 55
C. Performance of Membranes Operated on
Primary and Concentrated Primary Effluent .... 56
Table CI: Performance of Tubular Membrane
Set la Operated on Primary and
Concentrated Primary Effluent --
Measured Values ..........
Table CII: Performance of Tubular Membrane
Set la Operated on Primary and
Concentrated Primary Effluent --
Calculated Values .......... 58
Table CIII: Performance of Tubular Membrane
Set 2a Operated on Primary and
Concentrated Primary Effluent
Feed -- Measured Values ....... 59
47
-------
Table CIV:
Table CV:
Table CVI:
Table CVII:
Performance of Tubular Membrane
Set 2a Operated on Primary
and Concentrated Primary
Effluent — Calculated Values . .
Performance of Tubular Membrane
Set 3a Operated on Primary
and Concentrated Primary
Effluent — Measured Values . . .
Page No.
60
61
Performance of Tubular Membrane
Set 3 a Operated on Primary
and Concentrated Primary
Effluent — Calculated Values . .
Properties of Primary and
Concentrated Primary Feed . . .
62
63
D.
Table CVIII: Performance of Membrane
Set 6 (Regenerable Unit 6)
Operated on Primary and
Concentrated Primary Effluent
Spectrographic Constituent Analysis of
Deposits on Membrane Surfaces ,
64
65
48
-------
APPENDIX A
DESCRIPTION OF EQUIPMENT
FEED TANK - 500 gallon fiberglass tank 48 by 72 inches manufactured
by Century Fiberglass Company, Santa Ana.
HIGH PRESSURE PUMP - Gardner-Denver (Denver, Colorado), Model
P-Q-2, triplex piston pump with aluminum-bronze
body, stainless steel-valves and colmonoy plungers.
pH SENSOR AND CONTROLLER - Universal Interloc (Santa Ana,
California), Model 1000 M with flow cell.
pH RECORDER - Rustrack millivolt recorder.
ACID PUMP - Precision Chemical Pump Corporation (Waltham,
Massachusetts) Series 1200 pump.
BACK PRESSURE REGULATOR - Marotta Valve Corporation (Santa Ana,
California), Model PRV-533.
MOUNTING RACK - Astropower Laboratory.
PRIMING PUMP - Flotec Model C3P5-1100 with PVC head.
49
-------
APPENDIX B
PERFORMANCE OF MEMBRANES OPERATED ON
CARBON-TREATED SECONDARY EFFLUENT
50
-------
Table BI
PERFORMANCE OF TUBULAR MEMBRANE SET 1 OPERATED
ON CARBON-TREATED SECONDARY EFFLUENT
(Values presented are properties of the product water. )
Time
(Hr)
00 Start
22
104
130
185
240
290
335
381
431
481
688
716
790
791(0
818
883
973
1. 024
1, 075
1, 164
1, 165
1, 185
(a) Adjusted
(b) Cleaned
(c) Flushed
gfd(a)
29.4
21. 3
15.4
13. 1
11.5
11. 1
10. 1
10.0
9.9
14.0
13.4
29.6
32.3
29. 6
33.3
25.3
31.6
28.6
49.6
39.4
18.3
30.8
31.4
32.3
26.2
30.9
29. 5
Conductivity (a)
1 o tal , 1 otal
\cm2 / pH (mg/1) (mg/1)
410
220 127.
280 6.0 2.1 148.
310 2.9 132.
375 5.7 190.
280 5.9 149.
400
260
320
750
350
300
425
410 215.
280
290 136.
195 107.
260 1.5 125.
600
520
350 4.9 168.
300 5.8 148.
340 5.9 4.5 164.
325
310
325
350
to 77°F (25°C).
with sponge ball.
for 30 minutes with a solution of Biz in water (7. 1
Percent Percent
Conductivity COD
Removal Removal
62
79
73 93
72 84
65
54
68
66
68
70
68
65
71
70
77
77 85
46
53
65 57
72
7.0 33
70
72
71
72
g/1) at 25 °C and pH 9. 2 .
Percent
TS
Removal
80
78
81
72
69
74
80
81
72
76
75
51
-------
Table BII
PERFORMANCE OF TUBULAR MEMBRANE SET 2 OPERATED
ON CARBON-TREATED SECONDARY EFFLUENT
(Values presented are properties of the product water. )
Time
(Hr)
00 Start
104
130
185
208
232
290
335
43,(b)
48j(b.c)
554
595
596(c>
642
645
,688
716
790
791(c).
818
883
973
1, 024(c)
1, 075
1, 164
1, I65(c)
1, 185
(a) Adjusted
(b) Cleaned
(c) Flushed
gtt(a:
25. 5
13.0
13.4
15.0
17.0
15.6
29.3
17.7
21.7
30.4
22.2
28.7
24. 1
24.3
22. 2
30.4
23.3
20. 1
19.5
20. 8
ai.4
20.5
20.9
20.0
Conduct! vity(a)
(umho \
cm2 ) PH
150
210 6.0
250
390 6.0
490
220
620
230
250
480
330
425
270
260
330
500
400
180
220 5.8
- 200 5.7
110 6.3
50 7. 2
85
80
to 77°F (25°C).
with sponge ball.
for 30 minutes with a solution
Total Total Percent Percent
COD Solids Conductivity COD
(mg/1) (mg/1) Removal Removal
6.2 90. 87 32
8.7 128. 80 71
5.5 113. 75
231.
44
88
72
79
75
52
188. 67
55
150. 73
159. 69
6.3 170. 71 38
54
64
2.5 78. 82 78
102. 80
1.1 125. 83 84
50. 90
22. 96
92
94
of Biz in water (7. 1 g/1) at 25°C and pH-9. 2.
Percent
TS
Removal
87
81
73
71
70
74
87
83
81
93
96
52
-------
Table Bill
PERFORMANCE OF TUBULAR MEMBRANE SET 3 (CONTROL)
OPERATED ON CARBON-TREATED SECONDARY EFFLUENT
(Values presented are properties of the product -water. )
Time
(Hr)
00 Start
22
104
130
186
208
232
290
335
381
431
481(b.c)
559(0
595
596(c)
642
645
688
716
790
791(0
818
883
973
l,024(c)
1, 075
1, 164
1, 165(c)
1, 185
(a) Adjusted
(b) Cleaned
(c) Flushed
gfd(a<
17.7
14. 2
12.4
9.2
7. 8
7. 2
6.3
3. 8
4. 5
6.8
8.7
10. 6
15.4
15.4
14. 7
16. 1
15.2
16. 2
15.3
15.7
15.0
16. 0
14. 8
15.6
16.0
14. 5
15.6
16. 0
Conductivity (a)
(Eimho x
cm2 ) pH
50
55 6.0
55 5.9
50
70
110
45
50
40
50
20
75
45
60
55
60
45
50 5.7
60 5.7
65 6.0
60 6. 8
70
80
to 77°F (25°C)
with sponge ball.
for 30 minutes with a solution
Total Total Percent Percent
COD Solids Conductivity COD
(mg/1) (mg/1) Removal Removal
32. 95
1.0 38. 95 97
0.3 16. 98
34.
94
93
95
96
95
96
25. 95
93
13. 92
2. 95
3.3 32. 95 67
95
95
1. 96
12. 95
26. 95
28. 94
25. 95
94
94
of Biz in water (7. 1 g/1) at 25°C and pH 9. 2.
Percent
TS
Removal
95
95
98
96
98
99
95
99
98
96
96
96
53
-------
Table BIV
UI
PERFORMANCE OF REGENERABLE MEMBRANE SET 5 OPERATED ON CARBON-TREATED
SECONDARY EFFLUENT
(Values presented are properties of the product water. ) ^^
Time (Hr)
Conductivity^)
mmno ^
V cm /
pH
COD TS Conductivity COD TS
(mg/1) (mg/l) Removal Removal Removal
00 Start
24
48
56
60
106
152
198
248
29?(k)
37o(b)
411(b)
412
459(b)
460
504
533
578(b)
606
671
695
746(b)
797
887(b)
908
(a.) Adjusted
9.3
7.5
9.9
9.1
7.3
6.1
6.1
4.8
7.1
3.7
9.0
23.0
11.4
21.1
6.2
17.3
12. 1
12.6
7.2
8.5
8.0
12.8
4.6
11.4
to 77°F
150
250
115
200
480
370
250
300
210
310
290
240
300
380
425
380
(25°C).
(b) Flushed for 30 minutes with a.
83.
5.6
163.
122.
3.3 154.
0.35
5.6 113.
5.6 150.
174.
7. 1 194.
solution of Biz in water (7. 1
85
87
80
78
66
75
73
75
73 70
74
97
78
74
65
61
70
g/1) at 25 °C and pH 9. 2.
87
76
77
76
82
77
83
66
-------
Table BV
PROPERTIES OF CARBON-TREATED SECONDARY EFFLUENT FEED
Conductivity (a)
Time (Hr)
00 Start
22
104
130
185
208
240
335
381
431
481
554
559
595
642
688
716
790
818
883
973
1,024
1,075
1, 164
1, 185
1 2 I
\ cm /
1,
1,
1,
1,
1;
1,
2,
i,
1,
1,
i,
i,
1,
1,
1,
i,
i,
1,
040
040
040
100
950
070
870
950
020
200
100
100
000
000
980
840
150
100
990
075
150
075
100
250
Total
Solids
(mg/1)
669.
687.
685.
678.
528.
660.
684.
516.
537.
642.
603.
607.
660.
690.
577.
Total
COD
(mg/1) PH
5.5
9.6
29.5 6.0
17.9 5.6
7.4
22.3 5.7
7.8
7.7
6,0
10.1
11.3
5.9
6.6 5.8
6.3
7.7
(a)
Adjusted to 77°F (25°C)
55
-------
APPENDIX C
PERFORMANCE OF MEMBRANES OPERATED ON
PRIMARY AND CONCENTRATED
PRIMARY EFFLUENT
56 ,
-------
Table CI
PERFORMANCE OF TUBULAR MEMBRANE SET la OPERATED
PRIMARY AND CONCENTRATED PRIMARY EFFLUENT--
MEASURED VALUES
(Values presented are properties of the product water)
Time (Hr)
00 Start
11.0
91.0
163. 5
164. 5(b)
282.0
351.0(c)
424.0
471.0
517. 1
614. 5
1,061. 1
l,062.0(b)
1,300.9
(M
1,301. 5* '
1,641.4
l,642.5(b)
1,686.5
1, 825. 1
l,82b.0(d)
1,867.0
1,939.0
1,990. 5
t HI
l,991.5ld)
2, 035.0
2, 056. 0
2, 100.0(e)
2, 101.0(b>
2, 124.0
2, 196.0
2, 220.0
2,244.0
2, 267.0
2, 268.0(b)
2,279.0
2, 300.0
2, 370.0
2, 469.9
/u\
2, 470.5lb)
'2, 537.0
2,616.6
Conductivity!3)
/J^hON COD
gfdv ' \cm / pH (rng/1)
27.8
12.3
14.5
40.7
9-3
7.4
11.6
4.8
3. 15
16.7
9.0
2.6
2.45
7. 2
4.6
7.8
4.6
2. 8
3.4
7. 1
4.9
3.4
4.1.5
3.3
2.55
2.85
5.65
3.35
2.55
2. 8
3.65
5.3
1.65
1.95
1.99
1.81
5.95
6.0
6.05
4.7
3. 5
11.4
7.0
2. 8
140
210
170
550
260
220
400
•480
480
610
890
700
850
680
1000
980
750
690
600
460
165
160
180
170
210
160
170
165
240
400
450
5.6
4.3
6.2
3.5
6.6
6.0
6.3
6.3
6.4
6.3
6.5 34.2
6.8
6.4
6.6
6.8 23.9
6.7 39.8
5.8 35.9
6.3
6.7 8.1
6.6 8.1
6.6
6.6 15.9
6.2 2.8
6.3
6. 1
6.3 5.1
17.4
6.8
7.4
3.6
5.8
Filtered Total .,„ M
COD Solids TDS N 3
(mg/1) (mg/1) (mg/1) mg/1
54
94
102
121
100
96
192
248
230
295
458
454
352
551
553
434. 352
380
376
250 8.8
79
96
141
109
145
66
49
2.5
8.0 95 92
80 1.85
120
206
Notes:
(a) Adjusted to 77°F(25°C).
(b) Flushed for 30 minutes with a solution of Biz in water (7. 1 g/1), pH 9. 2.
(c) Feed recycled.
(d)» Flushed for 30 minutes with a 6M solution of urea adjusted to pH 5.
(e) Returned to one-pass flow-
57
-------
Table CII
PERFORMANCE OF TUBULAR MEMBRANE SET la OPERATED ON
PRIMARY AND CONCENTRATED PRIMARY EFFLUENT --
CALCULATED VALUES
(Based on properties of product water)
Time (Hr)
00 Start
11.0
91.0
163. 5
164.5
Z82.0
351.0(c>
424.0
471.0
517.0
614.5
658.6
779.0
827.2
868.0
1,061. 1
l,062.0(b)
1, 300.9
l,301.5(b)
1,641.4
l,642.5(b)
1,686.5
1, 825. 1
l,826.0(d)
1,867.0
1,939.0
l'.990.5
1.991.5(d)
2,035.0
2,056.0
2, 100.0(e)
2, lQ1.0(b)
2, 124.0
2, 196.0
2,220.0
2, Z44.0
2,267.0
2, 268. 0
-------
Table GUI
PERFORMANCE OF TUBULAR MEMBRANE SET 2a OPERATED ON
PRIMARY AND CONCENTRATED PRIMARY. EFFLUENT--
MEASURED VALUES
(Values presented are properties of the product water)
.Conductivity'^
/ jirnho v
Time (Hr) gfd(a) V cm2 /
00 Start
11.0 21.9 95
91.0 16.1 80
163.5 18.5 50
164; 5*b' 27.2
282.0 11.3 200
if,\
351.0* ' 14.0 80
424.0 13.6 75
471.0 5.1 130
517.0,,, 3.45 170
614.5* ' 11.3 160
658.6 9.45 220
779.0 2.25 520
827.2,., 2.15 360
868.0* ' 6.3 330
1,061.1 6.2 290
( M
1,062.0* ' 8.4
1,300.9 6.15 380
(M
1,301.5* ' 8.35
1,641.4 4.3 390
l,642.5(b) 8.2
1,686.5 6.25
1,825.1 3.7 280
1, 826.0(d) 4.45
1,867.0 3.2 230
1,939.0 2.55 220
1,990.5 2.8 200
l,991-5(cl) 3.2
2,035.0 3.2 65
2,056.0 2.85 70
2, 100.o'e> 3.25 80
/u\
2,101,0*°' 5.85
2,124.0 5.0 80
2,196.0 1.90 120
2,220.0 2.0 90
2,244.0 2.1 100-
2,267.0 2.03
/M
2,268.0*°' 7.4
2,279.0 7.25
2,300.0 7.6
2,370.0 6.3 60
2,469.9'c' 4.3 110
(M
2,470.5* ' 13.0
2,537.0 4.25 150
2,616.6 1.80 100
Notes:
(a) Adjusted to 77°F (25°C).
Total Filtered
COD COD
pH (mg/1) (mg/1)
5.5
4. 1
5.7
3.9
5.9
6.3
6.2
6.6
6.5
6.3
6.7 27.9
6.7
6.5
6.7
6.2 19.9
6.3 15.9
6.2 47.8
6.2
6.7 12.1
6.4 8.1
6.0
6.6 6.0
6.0
6.0
6.3
5.9 2.4
13.5 11.5
6.2
6.4
3.7
6.3
(b) Flushed for 30 minutes with a solution of Biz in water (7. 1 g/1). pH
(c) Feed recycled.
(d) Flushed for 30 minutes with a 6M
(e) .Returned to one-pass flow.
solution of urea adjusted to pH 5.
. Nessler
Total NH - N
Solids TDS WH3 N
(mg/1) (mg/1) -mg/1
22
4
34
35
15
26
55
81
73
101
251
169
147
210
197
150 114
114
108 3.9
34
35
83
55
90
28
14
1.6
45 36
18 0.65
58
73
9. 2.
59
-------
Table CIV
PERFORMANCE OF TUBULAR MEMBRANE SET 2a. OPERATED ON
PRIMARY AND CONCENTRATED PRIMARY EFFLUENT--
CALCULATED VALUES
(Based on properties of the product water)
Time (Hr)
00
11.
91
163
164
282.
351.
424,
471.
517.
614.
Start
.0
.0
. 5
.5
. 0
. 0
;>)
658.6
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
2,
2,
2,
2,
2,
2,
2,
2,
2,
2,
2,
2,
2,
2,
2,
2,
2,
779.
827.
868.
061.
062.
300.
301.
641.
642.
686.
825.
8.26.
867.
939.
990.
991.
0
2(bl
. 0* '
1
o
9
5
4
5
0
6
Percent
, , Conductivity
gfd1 ' Removal
21.
16.
18.
27
11.
14,
13.
5.
5.
11.
9.
2.
2.
6.
6.
8.
6.
8.
4.
8.
6.
3.
4.
3.
2.
2.
3.
3.
2.
.9
. 1
. 5
. 2
. 3
, 0
.6
. 1
.45
3
.45
25
15
3
2
4
15
35
3
2
25
7
45
2
5.5
8
2
2
85
3.25
5.
5.
1.
2.
2.
2.
7.
7.
7.
6.
4.
13.
4.
1.
85
0
90
0
1"
03
4
25
6
3
3
0
25
80
94
95
96
93
94
96
95
94
96
95
90
92
93
95
94
93
95
96
96
95
96
94
94
92
86
90
89
93
92
93
92
Filtered
Percent Percent
COD COD
Removal Removal
96
98
98
95
99
96
97
99
99
92 77
Percent Percent Percent
TS TDS NH3
Removal Removal Removal
97
99
96
97
98
98
97
96
98
97
94
96
96
97
96
96
97 98
98
97 96
97
96
90
93
85
96
98
87
93 94
97 95
93
96
Feed
Concentration
Ratio (TS)
5,
5.
4,
5.
7.
9.
7,
8.
6.
5.
1.
1.
1.
, 1
. 1
.5
.6
4
. 1
. 2
.3
7
, 5
6
.7
,3
(a) Adjusted to 77°F (25"C).
(b) Flushed for -30 minutes with a solution of Biz in water (7. 1 g/1), pH 9. 2.
(c) Feed recycled.
(d) Flushed for 30 minutes with a 6M solution of ur.ea adjusted to pH 5.
(e) Returned to one-pass flow.
60
-------
Table CV
PERFORMANCE OF TUBULAR MEMBRANE SET 3a OPERATED ON
PRIMARY AND CONCENTRATED PRIMARY EFFLUENT--
MEASURED VALUES
(Values presented are properties of the product water)
Conductivity'a) ^..^ ,
, ' Filtered
/ |imno > COD COD
Time (Hr) gfd \ cm ) pH (mg/1) (mg/1)
00 Start
11. .0 27.2 190
91.0 11.7 200
163.5 13.1 180
164. 5(b' 39.2
282.0 8.4 460
351. D'C) 5.6 510
424.0 9.9 230
471.0 5.4 310
517.1 3.25 450
614.5*°' 6.0 400
658.6 8.1 500
779.0 2. 25 590
827.2 2.15 480
868.0*°' 5.0 540
1,061.1 4.7 550
(M
1,062.0*°' 5.9
1,300.9 4.65 630
(M
1,301.5*°' 5.95
1,641.4 4.1 550
1 M
1,642.5*°' 6. -85
1,686.5 5.1
1,825.1 3.9 440
l,826.0(d) 4.45
1,867.0 3.3 440
1.939.0 3.55 360
1,990.5 3.2 280
l,991.5*d) 3.45
2,035.0 3.45 95
2,056.0 3.0 110
2, 100. 0(e' 3.35 110
2, 101.0(b) 5.45
2,124.0 5.2 150
2,196.0 2.50 155
2,220.0 2.35 140
2,244.0 .2.1 150
2,267.0 2.03
(M
2,268.0*°' 6.15
2,279.0 5.85
2,300.0 6.45
2,370.0 5.25 120
i ~\
Z, 469. 9V ' 4.00 220
/u\
2,470.5*°' 9.3
2,537.0 4.45 370
2,616.6 2.65 140
Notes:
(a) Adjusted to 77°F (25°C).
(b) Flushed for 30 minutes with
(c) Feed recycled.
(d) Flushed for 30 minutes with
(e) Returned to one pass flow.
5.8
5.2
6.2
3.7
6.4
6.2
6.3
6.5
6.4
6.4
6.9 32.1
6.9 24.3
6.5
6.6 0.2788
6.9 15.9
6.7 19.9
6.0 35.9
6.5
6.3 16.2
6.4 8.1
6.4
6.5 6.0
6.2
6.2
6.0
6,1 7.3
10.0 8.8
6.9
7. 1
3.8
5.5
a solution of Biz in water (7. 1 g/1), pH
a 6M solution of urea adjusted to pH 5.
Total NH N
Solid TDS NH3 N
(mg/1) (mg/1) ,mg/l
80
96
102
168
118
106
151
236
206
248
302
232
284
312-
277
256 204
204
225
167 5.3
56
74
76
84
118
62
58
2. 2
82 71
60 1.10
123
180
9. 2.
61
-------
Table CVI
PERFORMANCE OF TUBULAR MEMBRANE SET 3a OPERATED ON
PRIMARY AND CONCENTRATED PRIMARY EFFLUENT--
CALCULATED VALUES
(Based on properties of/product -water)
Filtered
Percent Percent Percent Percent Percent Percent
,a- Conductivity COD COD TS TDS NH3
Time (Hr) gfd Removal Removal Removal Removal Removal Removal
00 Start
11.0 27.2
91.0 11.7
163.5... 13.1
87
87
85
164.5(b) 37.2
282.0
351.0(c)
424,0
471.0
517 0
614:5
658. '6
779.0
827.2
868.0(b)
1, 061. 1
l,062.0(b)
1,300.9
l,301.5(b)
1,641.4
1, 642. 5(b)
1,686.5
, 1,825.1
l,826.0(d>
1,867.0
1,939.0
1.-990.5
l,991.5(d)
2, 035.0
2,056.0
(pi
2, 100. O1'
2, 101.0
-------
Table CVII
PROPERTIES OF PRIMARY AND CONCENTRATED PRIMARY FEED
Time
(Hr)
Conductivity
In Out
00 Start
11.0
91.0
163.5
282.0
351.0(c)
424.0
471.0
517. 1
614.5
1,
1,
1,
1,
1.
1,
1,
1,
2,
Z,
2,
2,
2,
2,
2,
2,
2,
2,
2,
2,
2,
658.
779.
827.
868.
061.
300.
461.
686.
825.
867.
939.
990.
035.
056.
100.
124.
196.
220.
244.
267.
279.
370.
469.
537.
616.
6
0
2
0
1
9
4
5
1
0
0
5
0
0
0 Out 745.
107.6 846.
354.6 1,006.
111.6 1,044.
83.7(e> 645.
60.7 1,224.
117.4 222.
107.1 178.
146.
175.
175.
50.
7
6
8
0<<0
6
0
0, >
4(e)
5
6
6
0
7
5
mg/liter NH3 ' N
!n Out In Out
815.
775.
837.
763.
737
576.
771.
610. (£)
609.
678.
733.
652.
575.
628.
572.
792
946
888
1176
841
1199
1825
2116
3258
3682
4184
3922
3760
4314
5479
5267
5531, f)
5076(£)
5086
45Z6
4066 10.7 94.0
1039
954
8ZZ
759
588
684
707
1Z.7
673
600 (f)
13.0
805
1657
pH to
Membrane
6.
3.
5.
3.
6.
6.
5.
5.
5.
6.
5.
6.
6.
5.
6.
6.
5.
6.
6.
6'.
6.
5.
5.
6.
5.
5.
7.
7.
3.
5.
3
1
9
0
4
Z
9
7
9
0
0
3
2
9
4
6
2
9
7
7
4
6
9
5
4
7
6
6
2
3
(a) Primary effluent from plant into tank (make'up for liquid passing through membranes during recycle)
Adjusted to 77°F (25°G).
(b) Feed from tank to reverse osmosis system. Concentrated during recycle .otherwise identical to
feed from plant. Adjusted to 77°F (25°C).
Feed recycled.
(c)
(d)
(e)
Return to one-pass flow.
COD samples filtered through a 0. 45|j. millipore filter.
(f) TDS samples filtered through a 0. 45n miUipore filter.
63
-------
Table CVIII
PERFORMANCE OF MEMBRANE SET 6 (REGENERABLE UNIT 6)
OPERATED ON PRIMARY AND CONCENTRATED PRIMARY EFFLUENT
Time (Hr)
00 Start
11
91
163. 5
164. 5b
282
351C
424
471
517b
545
615
, ^ Percent Percent
/Kmno\ TS Removal Removal
GFDla) \ cm2 / (mg/1) Conductivity TS pH
7.9
3.1
2.5
8.7
3.2
2.4
1.6
1.2
0.9
1.4
0.8
420 188.
370 204.
460 238.
330
620 187.
650 338.
1,100 603.
1,100 671.
1, 000
1,500 840.
72
75
62
89
56
63
56
62
67
66
76 5.
79 5.
73 6.
78 6.
72 6.
67 6.
68 6.
74 6.
7
7
4
0
1
2
5
5
(a) Adjusted to 77°F (25°C).
(b) Flushed for 30 minutes with a solution of Biz in water (7. 1 g/1), pH 9. 2.
(c) Recycle.
64
-------
Appendix D
SPECTROGRAPHIC CONSTITUENT ANALYSES OF DEPOSITS ON MEMBRANE SURFACES
Element
Sponge Ball Solids Removal
After Two Weeks on
Primary Effluent Feed
(% Constituent)
Scrapping of Deposits from
Membrane Surface After Two
Weeks on
Concentrated Primary Effluent Feed
(% Constituent)
en
Chromium
Nickel
Titanium
Molybdenum
Silver
Zinc
Strontium
Sodium
Calcium
Potassium
Silicon
Magnesium
Iron
Phosphorous
Barium
Aluminum
L'ead
Manganese
Tin
Copper
Other Elements
0. 15
0. 052
0. 090
Trace
0. 002
Nil
0. 04
9.0
5.6
.3.6
2. 0
1.6
p. 32
2,4
•0. 11
o: 11
0, 91
0. 36
o: 034
0. 44
Nil
0. 42
0. 056
0. 24
Trace
0. 0066
0. 32
0. 11
6.6
1. 9
Nil
3.2
0. 32
7. 8
7. 4
1. 1
0. 52
3. 8
0. 0048
0. 074
0. 56
Nil
-------
1
Accession Number
w
5
Organization
2
Subject Field & Group
SELECTED WATER RESOURCES ABSTRACTS
INPUT TRANSACTION FORM
i
McDonnell Douglas Corporation
Newport Beach, California 92660
Use of Improved Membranes in Tertiary
Treatment by Reverse Osmosis
10
22
Authors)
Bishop, H.K.
1 A Project Designation
"^ O^
21 Note
Citation
23
Descriptors (Starred First)
Water Pollution, membranes, reverse osmosis, sewage treatment, waste water treatment,
cellulose acetate, carbon treated secondary effluent, primary effluent, solid removal,
organic removal, inorganic removal
25
Identifiers (Starred First)
27
Abstract
The purpose of this reverse osmosis study was threefold: (1) to compare tubular membranes prepared
' from transesterified (modified) cellulose acetate with commercially available cellulose acetate (control),
(2) to evaluate the in-situ regenerable membrane reverse osmosis design on waste water and (3) to evaluate
the membranes on carbon-treated secondary effluents, primary effluents and concentrated primary
effluents.
The test results were: (1) tubular membranes prepared from transesterified cellulose acetate produced
water fluxes slightly greater than those of membranes prepared from commercially available cellulose
acetate, (2) the in-situ regenerable membranes produced fluxes below that of tubular units but showed
suficient promise for further development, (3) product water flux from operations on carbon-treated
secondary effluents gradually declined from initial levels of between IS and 25 gfd. However, product
water flux could be maintained near these initial levels by periodic cleaning with enzyme-active laundry
presoak solution, (4) product water flux on primary and concentrated primary effluents declined
gradually from initial levels of between IS and 25 gfd and stabilized between 4 to 5 gfd on concentrated '
primary effluent even with the use of enzyme-active laundry presoak solution, (5) removal of most waste
water constituents was between 90 to 100% and was generally unaffected by the type of feed water or time
of test. Chloride and nitrate reductions averaged approximately 70 percent.
The test results indicate that it is technically feasible to treat primary effluents with tubular reverse osmosis
process. However, further development is needed to •determine economic feasibility.
Abstractor
HX. Bishop
Institution
McDonnell Douglas Corporation
WR:'02 (REV. JULY 1868)
WRSIC
SEND, WITH COPY OF DOCUMENT. TO! WATER RESOURCES SCIENTIFIC INFORMATION CENTER
U.S. DEPARTMENT OP THE INTERIOR
WASHINGTON. D. C. 20240
* CPOI 1870-389-830
66
-------
BIBLIOGRAPHIC:
McDonnell Douglas Corporation, Use of Improved Membranes in Tertiary Treatment by Reverse
Osmosis, Final Report WQO Program No. 17020 DHR, December 1.970.
ABSTRACT
The purpose of this reverse osmosis study was threefold: (1) to compare tubular membranes
prepared from transesterified (modified) cellulose acetate with commercially available cellulose
acetate (control), (2) to evaluate the in-situ regenerable membrane reverse osmosis design on waste
water and (3) to evaluate the membranes on carbon-treated secondary effluents, primary effluents
and concentrated primary effluents.
The test results were: (1) tubular membranes prepared from transesterified cellulose acetate
produced water fluxes slightly greater than those of membranes prepared from commercially avail-
able cellulose acetate, (2) the in-situ regenerable membranes produced fluxes below that of tubular
units but showed sufficient promise for further development, (3) product water flux from opera-
tions on carbon-treated secondary effluents gradually declined from initial levels of between 15 and
25 gfd. However, product water flux could be maintained near these initial levels by periodic
cleaning with enzyme-active laundry presoak solution, (4) product water flux on primary and
concentrated primary effluents declined gradually from initial levels of between 15 and 25 gfd and
stabilized between 4 to 5 gfd on concentrated primary effluent even with the use of enzyme-active
laundry presoak solution, (5) removal of most waste water constituents was between 90 to 100%
and was generally unaffected by the type of feed water or time of test. Chloride and nitrate
reductions averaged approximately 70 percent.
The test results indicate that it is technically feasible to treat primary effluents with tubular
reverse osmosis process. However, further development is needed to determine economic feasibility,
ACCESSION NO.
KEY WORDS
Water pollution
Membranes
Reverse osmosis
Sewage treatment
Waste water
treatment
Cellulose acetate
Carbon-treated
secondary effluent
Primary effluent
Solid removal
Organic removal
Inorganic removal
BIBLIOGRAPHIC:
McDonnell Douglas Corporation, Use of Improved Membranes in Tertiary Treatment by Reverse
Osmosis, Final Report WQO Program No. 17020 DHR, December 1970.
ABSTRACT
The purpose of this reverse osmosis study was threefold: (1) to compare tubular membranes
prepared from transesterified (modified) cellulose acetate with commercially available cellulose
acetate (control), (2) to evaluate the in-situ regenerable membrane reverse osmosis design on waste
water and (3) to evaluate the membranes on carbon-treated secondary effluents, primary effluents
and concentrated primary effluents.
The test results were: (1) tubular membranes prepared from transesterified cellulose acetate
produced water fluxes slightly greater than those of membranes prepared from commercially avail-
able cellulose acetate, (2) the in-situ regenerable membranes produced fluxes below that of tubular
units but showed sufficient promise for further development, (3) product water flux from opera-
tions on carbon-treated secondary effluents gradually declined from initial levels of between 15 and
25 gfd. However, product water flux could be maintained near these initial levels by periodic
cleaning with enzyme-active laundry presoak solution, (4) product water flux on primary and
concentrated primary effluents declined gradually from initial levels of between 15 and 25 gfd and
stabilized between 4 to 5 gfd on concentrated primary effluent even with the use of enzyme-active
laundry presoak solution, (5) removal of most waste water constituents was between 90 to 100%
and was generally unaffected by the type of feed water or time of test. Chloride and nitrate
reductions averaged approximately 70 percent.
The test results indicate that it is technically feasible to treat primary effluents with tubular
reverse osmosis process. However, further development is needed to determine economic feasibility
ACCESSION NO.
KEY WORDS
Water pollution
Membranes
Reverse osmosis
Sewage treatment
Waste water
treatment
Cellulose acetate
Carbon-treated
secondary effluent
Primary effluent-
Solid removal
Organic removal
Inorganic removal
BIBLIOGRAPHIC:
McDonnell Douglas Corporation, Use of Improved Membranes in Tertiary Treatment by Reverse
Osmosis, Final Report WQO Program No. 17020 DHR, December 1970.
ABSTRACT
The purpose of this reverse osmosis study was threefold: (1) to compare tubular membranes
prepared from transesterified (modified) cellulose acetate with commercially available cellulose
acetate (control), (2) to evaluate the in-situ regenerable membrane reverse osmosis design on waste
water and (3) to evaluate the membranes on carbon-treated secondary effluents, primary effluents
and concentrated primary effluents.
The test results were: (1) tubular membranes prepared from transesterified cellulose acetate
produced water fluxes slightly greater than those of membranes prepared from commercially avail-
able cellulose acetate, (2) the in-situ regenerable membranes produced fluxes below that of tubular
units but showed sufficient promise for further development,. (3) product water flux from opera-
tions on carbon-treated secondary effluents gradually declined from initial levels of between 15 and
25 gfd. However, product water flux could be maintained near these initial levels by periodic
cleaning with enzyme-active laundry presoak solution, (4) product water flux on primary and
concentrated primary effluents declined gradually from initial levels of between 15 and 25 gfd and
stabilized between 4 to 5 gfd on concentrated primary effluent even with the use of enzyme-active
laundry presoak solution, (5) removal of most waste water constituents was between 90 to 100%
and was generally unaffected by the type of feed water or time of test. Chloride and nitrate
reductions averaged approximately 70 percent.
The test results indicate that it is technically feasible to treat primary effluents with tubular
reverse osmosis- process. However, further development is needed to determine economic feasibility
ACCESSION NO.
KEY WORDS
Water pollution
Membranes
Reverse osmosis
Sewage treatment
Waste water
treatment
Cellulose acetate
Carbon-treated
secondary effluent
Primary effluent
Solid removal
Organic removal
Inorganic removal
67
------- |