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  EffluentI                              39

VI       Removal  of Wastewater Constituents
         During Operation with Concentrated
         Primary  EffluentII                             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  1C.  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 (+2C)  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 77F.  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 80F. Product water flux values were corrected to
 77F (25C).  Temperature of the concentrated primary feed averaged
 120F (49C)  due to the heat added by the high  pressure pump during
 recirculation of the feed stream.   Product water flux values were also
 corrected to 77F (25C)  using a conservative  correction factor
 (~60 percent of the measured value).  Thus, the flux values reported
 for 77F 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  77F (25C).   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 500C
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 77F (25C).
                                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 120F (49C)  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 (^26C).  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 operationmiddle 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 (49C),
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 (77F).  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 77F (25C).
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 77F (25C).
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 25C 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 77F (25C)
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 25C 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 77F


150
250
115
200
480
370
250

300


210
310
290

240
300
380
425
380
(25C).
(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 77F (25C)
                               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 77F(25C).
  (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 77F (25C).
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 77F (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 77F (25C).
(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 77F (25G).
    (b)  Feed from tank to reverse osmosis system.  Concentrated during recycle .otherwise identical to
        feed from plant.  Adjusted to 77F (25C).
        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 77F (25C).
(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

-------