Muni

W W
W Ci L v7l

   \\-7C
Demineral
by Tubular Reverse
Osmosis Process

-------
                RESEARCH REPORTING SERIES

Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:

      1.  Environmental Health Effects Research
      2.  Environmental Protection Technology
      3.  Ecological Research
      4.  Environmental Monitoring
      5.  Socioeconomic  Environmental Studies
      6.  Scientific and Technical  Assessment Reports (STAR)
      7.  Interagency Energy-Environment Research and Development
      8.  "Special" Reports
      9.  Miscellaneous Reports

 This report has been assigned to the ENVIRONMENTAL PROTECTION TECH-
 NOLOGY series. This series describes research performed to develop and dem-
 onstrate instrumentation, equipment,  and methodology to repair or prevent en-
 vironmental degradation from point and non-point sources of pollution. This work
 provides the new or improved technology required for the control and treatment
 of pollution sources to meet environmental quality standards.
 This document is available to the public through the National Technical Informa-
 tion Service, Springfield, Virginia 22161.

-------
                                                EPA-600/2-78-167
                                                September 1978
WASTEWATER DEMINERALIZATION BY TUBULAR REVERSE OSMOSIS PROCESS
                              by

                        Ching-lin Chen
                        Robert P. Miele
       County Sanitation Districts of Los Angeles County
                  Whittier, California  90607
                    Contract No.  14-12-150
                         Project Officer

                        Irwin  J.  Kugelman
                  Wastewater Research  Division
           Municipal  Environmental  Research  Laboratory
                     Cincinnati,  Ohio   45268
           MUNICIPAL ENVIRONMENTAL RESEARCH LABORATORY
               OFFICE OF RESEARCH AND DEVELOPMENT
              U.S.  ENVIRONMENTAL PROTECTION AGENCY
                     CINCINNATI, OHIO  45268

-------
                                 DISCLAIMER
     This report has been reviewed by the Municipal Environmental  Research
Laboratory, U.S. Environmental Protection Agency, and approved for publication.
Approval does not signify that the contents necessarily reflect the views and
policies of the U.S. Environmental Protection Agency, nor does mention of
trade names or commercial products constitute endorsement or recommendation
for use.

-------
                                  FOREWORD


     The Environmental Protection Agency was created because  of  increasing
public and government concern about the dangers of pollution  to  the  health
and welfare of the American people.  Noxious air, foul  water, and spoiled
land are tragic testimony to the deterioration of our natural environment.
The complexity of that environment and the interplay between  its components
require a concentrated and integrated attack on the problem.

     Research and development is that necessary first step in problem solution
and it involves defining the problem, measuring its impact, and  searching  for
solutions.  The Municipal Environmental Research Laboratory develops new and
improved technology and systems for the hazardous water pollutant discharges
from municipal and community sources, for the preservation and treatment of
public drinking water supplies, and to minimize the adverse economic, social,
health, and aesthetic effects of pollution.  This publication is one of the
products of that research; a most vital communications link between the
researcher and the user community.

     One of the goals of wastewater treatment is renovation of wastewater  so
that it can be reused.  It is expected that partial demineralization of
conventionally treated wastewater will be required  if the wastewater is reused
for any purpose which requires high quality water.  Among the techniques for
demineralization that which is newest but shows the most potential is reverse
osmosis.   In this process water is forced through a membrane which can reject
salts.  The permeability of these membranes is low  so high pressure is required
to achieve an economical production rate.  Special  configuration of the mem-
brane and  its support system are required to withstand the high pressure and
maintain a high ratio of membrane surface to system volume.  In the studies
reported in here a reverse osmosis system using tubular support configuration
was tested for its efficacy in demineralization of  secondary effluent.  This
type of configuration is best when water containing suspended solids is being
treated as the large  passages resist clogging.
                                       Francis T. Mayo, Director
                                       Municipal Environmental Research
                                        Laboratory
                                      m

-------
                                 ABSTRACT


    A 16 1pm (4.2 gpm) tubular reverse osmosis pilot plant, manufactured by
the Universal Water Corporation, Del Mar, California, was operated at Pomona
Advanced Wastewater Treatment Research Facility for a total of 10,940 hours.
The purpose of the study was to investigate the applicability of the tubular
reverse osmosis process to wastewater demineralization.

    The study was basically divided into three different phases.  The first
phase, which covered a period of 1,700 hours of operation, was considered a
start-up period.  The optimum pilot plant operating conditions and the
necessary plant modifications were developed and accomplished during the
initial operation period.  In the second phase, the pilot plant operational
goal was the study of the membrane life.  The feed was carbon-treated sec-
ondary effluent.  This study lasted 7,165 hours until the operating membrane
modules reduced from the original 32 modules to 12 modules as a result of
membrane module failure.  It was subsequently decided to make use of the re-
maining 12 modules in the system for exploring the behavior of the membrane
process in treating primary effluent.  The system treated primary effluent
during the third phase for a period of 2,074 hours.

    The experimental results of the first study period have already been
reported in the EPA WATER POLLUTION CONTROL RESEARCH SERIES(l).  Therefore,
this report only summarizes the experimental data obtained in the last two
phases of studies.  The results clearly indicate that the Universal Water
Corporation's tubular reverse osmosis system is not practical for wastewater
demineralization.

    This report was submitted by County Sanitation Districts of Los Angeles
County in fulfillment of Contract No. 14-12-150 under the partial sponsor-
ship of the Municipal Environmental Research Laboratory, Office of Research
and Development, U.S. Environmental Protection Agency.  Work was completed
in January, 1971.
                                     iv

-------
                                   CONTENTS

Foreword	    in'
Abstract	     1v
Figures	     vi
Tables	-    vii
Acknowledgement 	   viii

   1. Introduction  	      1
   2. Conclusions 	      3
   3. Recommendations 	      4
   4. Pilot Plant Description 	      5
   5. Pilot Plant Operations  	      8
   6. Results and Discussions	     12

References	     26

-------
                                    FIGURES

Number                                                                   Page

   1   Schematic flow diagram of the reverse osmosis
       pilot plant 	 .,.,,,,,     6
       Layout of Universal Mater Corporation reverse
       osmosis pilot plant 	
       Flux rate and salt rejection variations with
       operation time	«	    13

       Product water flux rate variation with
       operation time	', , .  , ,	, .  ,  ,    14

       Performance of tubular reverse osmosis
       process treating primary effluent .,,,,,,,,,.,,.,    17

       Module replacement in Universal Water Corporation
       reverse osmosis pilot plant   .,.,.,..«,,,,..«.    21
                                      vi

-------
                                   TABLES

Number                                                                   Page

  1    Summary of Chemical Analyses for Pilot Plant Operations
       with Carbon-Treated Secondary Effluent	       16

  2    Effects of the Brine Velocity upon the Operations
       with Primary Effluent 	 ,,,,,.,,,       18

  3    Summary of Chemical Analyses for Pilot Plant
       Operations with Primary Effluent  . .  .  .  ,	       19

  4    Visual Examination of Failed Modules  .  ,  	       22

  5    Laboratory Analysis of Module Failure 	 ,  .       23

  6    Complete Analysis of the Membrane Module
       after a Total of 10,940 Hours of On-Stream Operation   ....       25
                                      vn

-------
                              ACKNOWLEDGEMENTS
    This study was jointly sponsored by the U. S. Environmental Protection
Agency and the County Sanitation Districts of Los Angeles County.

    Mr. James'Gratteau and Mr. Harold H. Takenaka, former project engineers
at Pomona Advanced Wastewater Treatment Research Facility, were instrumental
in initiating the pilot plant study.

    The efforts of the laboratory and the pilot plant operating personnel of
the Pomona Advanced Wastewater Treatment Research Facility are also gratefully
acknowledged.
                                    viii

-------
                                 SECTION 1

                                INTRODUCTION
     The operation of the 16 1pm (4.2 gpm) tubular reverse osmosis pilot
plant, manufactured by the Universal Water Corporation, Del Mar, California,
was initially started on December 23, 1968.  During the initial  1,000 hours
of experimental run, the pilot plant system was operated at a low water re-
covery, approximately 40 percent, to test the operating conditions of the
physical components of the system - pumps, pH controller, modules, flow
metering system, and sampling system - along with monitoring the perfor-
mance of the membranes while operating on carbon-treated secondary effluent.
The water recovery was set at a low value so that the problems of feed water
concentration and insufficient brine velocities could be minimized.

     After the first 1,000 hours of operation, the piping of the system was
modified to accommodate operations at a higher water recovery of 75 percent.
The arrangement of the eight rows of modules (4 modules per row, connected
in series) was also revised from the original 5-3 array to a 3-2-2-1 array
to maintain sufficient brine velocities throughout the system.  However,
during the initial period of this intended long-term process study, some
serious mechanical and operational problems were encountered, and some
membrane modules were found seriously damaged.  Consequently, the study had
to be temporarily suspended after the first 700 hours of operations.  The
mechanical design of the entire system was thoroughly reviewed by the
Universal Water Corporation and the Los Angeles County Sanitation Districts'
engineers, and some necessary modifications were made on the system to pre-
vent the recurrence of the mechanical problems.  All the damaged membrane
modules were replaced with new membrane modules, and the module arrangement
in the system was further revised to the final 3-2-1-1-1 array, as shown in
Fi gure 1.

     On November 5, 1969, the pilot plant operation of the 16 1pm (4.2 gpm)
Universal Water Corporation tubular reverse osmosis system was put on
stream again.  The objective of this study was to establish membrane life.
Information obtained from previous experimental runs was used to establish
operating condition (1).  A daily tap water flush and a weekly enzyme-
detergent cleaning cycle were used to maintain product water flux rate.
During the first part of this experimental run treating carbon-treated
secondary effluent, modules which became damaged were replaced with the end
modules in the system as time progressed.  Operation in this manner was
continued to determine whether the frequency of module failure could be
intensified or not.

                                      1

-------
    Because of an increasing frequency of module failure, the decision was
made to suspend permanently the pilot plant operation after treating
carbon-treated secondary effluent for a total of 7,165 hours.  On October
15, 1970, an experimental run with the remaining 12 operating modules in
the tubular reverse osmosis system was commenced to treat the primary efflu-
ent.  This was done with a sole purpose of determining the behavior of the
tubular membranes in treating the primary effluent, which lasted a period
of 2,074 hours of on-stream operation.

    This report summarizes the 7,165 hours of operations of the tubular
reverse osmosis pilot plant on the carbon-treated secondary effluent and
the last phase of operation on the primary effluent.

-------
                                  SECTION 2

                                 CONCLUSIONS
    Based on the experimental results of this pilot plant study, the
following conclusions are made:

    A. The tubular membrane modules manufactured by the Universal Water
Corporation were very susceptible to membrane collapse, and they had to be
handled carefully.  The system should be designed and operated in a manner
which would avoid a siphon situation, which could result in membrane
collapse.

    B. For treating the carbon-treated secondary effluent, the weekly
enzyme-detergent cleaning cycle and the daily tap water flushing seemed to
be successful in controlling the product water flux rate decline.  However,
it was necessary to increase the enzyme-detergent cleaning frequency from
once a week to three times a week when treating the primary effluent.

    C. The pilot plant system was able to achieve a 95 percent salt rejec-
tion initially for treating either a carbon-treated secondary effluent or a
primary effluent.

    D. There was less flux decline between membrane cleanings when operat-
ing at a brine velocity of 1.4 mps (4.5 fps) rather than at 0.64 mps
(2.1 fps).

    E. The product water flux rate was reduced about 30 percent after al-
most one year of on-stream operation with carbon-treated secondary effluent.

    F. Because of the serious problems associated with the frequent mem-
brane module failure, it is impractical to apply the Universal Water
Corporation's tubular reverse osmosis system to wastewater demineralization.
The physical configuration of the module and the entire system design have
to be completely revised to avoid any mechanical failure in the membrane
modules.

-------
                                 SECTION 3

                              RECOMMENDATIONS
    Since the frequency of the membrane module failure in the Universal
Water Corporation's tubular reverse osmosis system is so high, the applica-
tion of this system to the demineralization of the wastewater, particularly
the carbon-treated secondary effluent, is certainly not practical.  However,
some encouraging results regarding the demineralization of the primary efflu-
ent by this system have been revealed by this rather short-term study.

    It is recommended that additional efforts be pursued to treat the pri-
mary effluent directly with the tubular reverse osmosis system.  However, the
new pilot plant system should incorporate the following features to assure
satisfactory performance.

    A. Adjust the pH of the cleaning solution to the neutral range be-
tween 7.0 and 7.5, and the pH of the feedwater to 6.0.  This pH adjustment
would not only prevent hydrolysis of membranes but prevent extensive deterio-
ration of the nylon cloth membrane backing material.

    B. Employ a daily cleaning cycle to maintain a constant higher product
water flux level.

    C. Recycle brine to obtain higher water recovery.

    D. Design the entire system properly to prevent the mechanical failure
which may in turn cause membrane collapse.

    The main objectives of the new study should be two-fold:

    A. To determine the feasibility of the demineralization of the primary
effluent with a tubular reverse osmosis process; and

    B. To obtain a reliable and realistic process cost estimate on a long-
term operation basis.

-------
                                 SECTION 4

                          PILOT PLANT DESCRIPTION


     The tubular reverse osmosis pilot plant consisted of thirty-two tubular
modules, 122 cm (48 in) long by 10.2 cm (4 in) in diameter.   Eighteen 1.3 cm
(0.5 in) diameter plastic perforated tubes were connected in series and
placed within a 10.2 cm (4 in) diameter PVC pipe to form a single module that
contained approximately 0.65 sq m (7 sq ft) of preformed membrane for a total
of 20,8 sq m (224 sq ft) of membrane area available in the entire system.

     Figure 1 shows a schematic flow diagram of the tubular reverse osmosis
pilot plant.  The carbon-treated secondary effluent was chlorinated to pro-
vide 1 to 2 mg/1 of total residual chlorine and acidified to a pH close to 5
using ^SOi, before it was fed into the system.  The system was arranged in a
3-2-1-1-1 array to maintain sufficient brine velocities in the downstream
modules.  The necessary provisions for a daily tap water flush, a weekly en-
zyme-detergent cleaning cycle, and a chlorinated tap water flush during the
downtimes were made.  The modifications and changes made to the piping and
valving of the original pilot plant system are shown in Figure 2.  Two check
valves were installed to prevent a siphon from developing, which could result
in reduced pressure within the module and cause possible membrane collapse.
One check valve was installed at the inlet of the unit to prevent backflow
during any unscheduled shutdown.  The other check valve was installed in a
tee connection off the final brine line.  When normal flow conditions existed,
this valve remained closed.  Should pressure in the brine line decrease as a
result of backflow through the system (pipe failure, open valve, etc.), this
check valve would open and allow air to enter the system to avoid a positive
external pressure.  This check valve would be weight loaded to open to the
atmosphere, should the pressure within the system drop below 0.69 Kg/sq cm
(10 psi).  The other valves shown in Figure 2 were used during tap water flush-
ings and enzyme-deteraent cleaning cycles.  A spring-loaded back pressure
regulator was installed in the bypass line to correct the problem of pressure
variations during the operation.

-------
                                                             ACID INJECTION
                                                             CARBON FEED
                                                             CHLORINATION
                                                                 FLOW
                                                             CONTROLLER
                                                                           BRINE
                                                                       M  •*
                                                                       PRODUCT
                                                         •TAP WATER
  NOTE:
    EACH ROW CONTAINS 4
    TUBULAR MODULES         CHLORINATION


Figure I. Schematic flow diagram of the reverse osmosis pilot plant.

-------
CONTROLLER-7


 FINAL ^f
 BRINE
        AIR
  CHECK VALVE-
 MAIN
 FEED"
1—CHECK VALVE-WEIGHT LOADED
        K30QO:
        -
        -oooo-
        -oooo-
rQOOO-
                     r*-OPEN VALVE
              o

              £
              a:
              ui

              i
              o
              3
              Q
              o
              or
              a.
      I
        TAP WATER

        CHLORINATED TAP WATER

        CLEANING SOLUTION
Figure 2. Layout of Universal Water Corporation reverse osmosis

   pilot plant.

-------
                                 SECTION 5
                          PILOT PLANT OPERATIONS
DEMORALIZATION OF CARBON-TREATED SECONDARY EFFLUENT
     The pilot plant Initially was operated with a carbon-treated secondary
effluent as the feed water.   The operating conditions and membrane cleaning
procedures employed for this study are described in the following sections.
Operating Conditions
     A summary of the initial operation conditions is shown below:
     A.  Feed Pressure          :41.4 Kg/sq cm (600 psi)
     B.  Feed pH                Maintained at 5 using sulfuric acid
     C.  Total Residual Chlorine
         in Feed Water          :Maintained at 1 to 2 mg/1
     D.  Feed Flow              :16 1pm (4.2 gpm)
     E.  Product Flow           :12 1pm (3.2 gpm)
     F.  Brine Flow             :4 1pm (1.0 gpm)
     G.  Membrane Product Flux
         Rate (at 25° C)        :834 1/sq m/day (20.5 g/sq ft/day)
     H.  Water Recovery         :76 Percent
     Among the above operating conditions, the feed pressure and the brine
flow rate were the major controlling parameters for the pilot plant operation.
Membrane Cleaning Procedures
     A daily tap water flushing and a weekly enzyme-detergent cleaning cycle
were conducted to maintain the product water flux rate.  The procedures for
the various cleaning steps are described as follows:
                                     8

-------
     A.  Daily Tap Water Flushing

     The pilot plant membrane module system was flushed as a whole with tap
water for 15 minutes once a day.  This tap water flushing was the sole clean-
ing process for the membrane module system on the non-chemical cleaning days,
while on chemical cleaning days it served as a rinsing process after the chem-
ical cleaning cycle.

     B.  Weekly Enzyme-Detergent Cleaning

     The enzyme-detergent cleaning solution was made up by adding 1.42 kilo-
grams (50 ounces) of a commercial enzyme-detergent, Biz, into 139 liters (50
gallons) of tap water.

     The main feed pump was used to fill the module system with the cleaning
solution.  The modules which were not being flushed remained soaking in the
cleaning solution while others were being flushed in the following sequ3nce:

     1.  Row 1 to 3, as indicated in Figure 1, were flushed for 15 mi/iutes
         at a feed pressure of 3.5 kg/sq cm (50 psi).

     2.  Rows 4 and 5 were flushed for 15 minutes at a feed pressure of 3.5
         kg/sq cm (50 psi).
     3.  Rows 6 to 8 were flushed for 15 minutes at a feed pressure of 3.5
         kg/sq cm (50 psi).

     During the enzyme-detergent cleaning cycle, the cleaning solution was re-
cycled for the specified time in the first three rows and then the same clean-
ing solution was applied to the next rows until the sequence was completed.
After each enzyme-detergent flushing, a 15-minute tap water flushing was also
conducted.

     C.  Occasional Acid Flushing

     Whenever a malfunction in the pH control occurred during an unattended
period, the module system was flushed with an acidified feed to remove the
chemical precipitates to recover the product water flux rate.  The acid flush-
ing process consisted of depressurizing the pilot plant system and flushing
with an acidified feed, pH between 2 and 3, for 30 minutes.   If necessary, the
enzyme-detergent and tap water flushing could be employed to provide addition-
al cleaning of the membrane surface.

DEMINERALIZATION OF PRIMARY EFFLUENT

     The pilot plant operation for the demineralization of carbon-treated
secondary effluent was switched  to the demineralization of primary effluent
after a total period of 7,165 hours of on-stream operation.  The total number
of membrane modules had been reduced from the initial 32 modules to 12 modules
at the outset of the primary effluent demineralization study.  The arrangement
of the membrane modules was changed from the  3-2-2-1 array to 3 in parallel
array to accommodate the remaining 12 modules  (4 modules per  row) for the
study of membrane behavior on the primary effluent.  The operating conditions
and the membrane cleaning procedures were also properly modified as follows:

-------
Operating Conditions
     A summary of the initial operating conditions is shown below:
     A.  Feed Pressure              :  41.4 Kg/sq cm (600 psi)
     B.  Feed pH                    :  Maintained at 5 using sulfuric acid
     C.  Total Residual Chlorine
         in Feed Water              :  Maintained at 1 to 2 mg/1.
     D.  Feed Flow                  :  11.0 1pm (2.9 gpm).
     E.  Product Flow               :  3.1 1pm (0.8 gpm)
     F.  Brine Flow                 :  7.9 1pm (2.1 gpm)
     G.  Membrane Product Flux
         Rate (at 25° C)            :  570 1/sq m/day (14 g/sq ft/day)
     H.  Water Recovery             :  28 percent
     No pretreatment other than acidification and chlorination was performed
on the primary effluent before being fed into the reverse osmosis pilot plant
system.
Membrane Cleaning Procedures
     A daily tap water flush and a three-times-a-week enzyme-detergent clean-
ing cycle were conducted to maintain the product water flux rate.  The
enzyme-detergent cleaning cycle was performed on Monday, Wednesday, and
Friday.  The procedures for the various cleaning steps are described as
follows:
     A.  Daily Tap Water Flushing
     The pilot plant membrane module system was flushed as a whole with tap
water for 15 minutes once a day.  This tap water flushing was the sole clean-
ing process for the membrane module system on the non-chemical cleaning days,
while on chemical cleaning days it served as a rinsing process after the
chemical cleaning cycle.
     B.  Three-Times-A-Week Enzyme-Detergent Cleaning
     The enzyme-detergent cleaning solution was prepared in the same manner
as the one for the demineralization of carbon-treated secondary effluent.
                                     10

-------
     The main feed pump was used to fill the module system with the cleaning
solution while others were being flushed in the following sequence:

     1.  Row 1 was flushed for 15 minutes at a feed pressure of
         3.5 Kg/sq cm (50 psi).

     2.  Row 2 was flushed for 15 minutes at a feed pressure of 3.5
         Kg/sq cm (50 psi).

     3.  Row 3 was flushed for 15 minutes at a feed pressure of 3.5
         Kg/sq cm (50 psi).

     During the enzyme-detergent cleaning cycle, the cleaning solution was
recycled for the specified time in the first row and then the same cleaning
solution was applied to the next rows until the sequence was completed.
After each enzyme-detergent flushing, a 15-minute tap water flushing was
also conducted.

     C.  Occasional Acid Flushing

     Whenever a malfunction in the pH control occurred during an unattended
period, the module system was flushed with an acidified feed to remove the
chemical precipitates to recover the product water flux rate.  The acid flush-
ing process consisted of depressurizing the pilot plant system and flushing
with an acidified feed, pH between 2 and 3, for 30 minutes.  If necessary,
the enzyme-detergent and tap water flushing could be employed to provide
additional cleaning of the membrane surface.
                                      11

-------
                                  SECTION 6

                           RESULTS AND DISCUSSIONS


SYSTEM PERFORMANCE

Operation with Carbon-Treated Secondary Effluent

     The system performance for the first 800 hours of operation with the
carbon-treated secondary effluent is summarized in Figure 3.  As indicated
in Figure 3, the product water flux rate during this period between 180 and
640 hours of on-stream operation first recovered and then dropped off after
each enzyme-detergent flush.  This time period also coincided with the upset
of the secondary treatment plant.  A large dose of hexavalent chromium en-
tered the plant and caused considerable damage to the microbial population
of the plant.  As a result, the quality of the secondary effluent deterio-
rated and this caused a corresponding deterioration of the carbon effluent,
which was in turn fed to the reverse osmosis pilot plant.  Apparently, the
daily tap water flush and the weekly enzyme-detergent cleaning cycles were
not able to cope with the fouling conditions on the schedule employed.

     At about 635 hours of operation, the system was temporarily shut down
and flushed with chlorinated tap water for 24 hours while the pump packing
was replaced.  When the system resumed operation, the product water flux
rate was found to increase from 537 1/sq m/day (13.2 g/sq ft/day) to 692
1/sq m/day (17.0 g/sq ft/day).  This increase in flux was clearly attri-
buted to the 24-hour chlorinated tap water flush.  Subsequently, the flux
stayed relatively constant.  During this latter period, the activated sludge
plant functioned satisfactorily.

     It was concluded that, except for the period of treatment plant upset,
the daily tap water flush and the weekly enzyme-detergent cleaning cycle
were successful in maintaining the product water flux rate.  In order to
maintain product water flux rate during periods of treatment plant upset,
the frequency of the enzyme-detergent flush should be increased.

     Figure 4 summarizes the entire series of operation with the carbon-
treated secondary effluent.  Initially the product water flux rate decreased
from 851 1/sq m/day (20.9 g/sq ft/day) to 651 1/sq m/day (16 g/sq ft/day)
at 300 hours of operation.  Between 300 hours and 5,800 hours, the product
water flux rate remained essentially constant at 651 1/sq m/day (16 g/sq ft/day)
At 5,800 hours, the product water flux rate increased to 733 1/sq m/day
(18 g/sq ft/day).  This increase was attributed to some small leaks which
                                    12

-------
  22
-20
•o
o.
01
   18
tr

X
   16
o  14
:D
o
o
a:
a.
   12
                                                                       100
         lgpd/ft*=40.7lpd/m*
ENZYME-DETERGENT

      FLUSH
                                                                       95
                                                                                     0)
             100
           200
300      400      500


   HOURS ON  STREAM
600
700
                                                                       90


                                                                          3}
                                                                          m

                                                                          m
                                                                          o


                                                                       85 O
                                                                       80
                                                                                  75
800
Figure 3. Flux rate and salt rejection variations with operation time.

-------
  10'
111
Sc
aa
IT
UJ
o
o
cc
Q.
                             I  I
     SLOPE =-0.04
                                        I   I  I  II
                                                              i  i  i  i
                                                      t
                                                24 HR. CHLORINATED
                                                TAP WATER FLUSH
                                                8 ENZYME  DETERGENT
                                                CLEANING CYCLE
FEED WATER '
FEED PRESSURE -
MEMBRANE AREA:
INITIAL PRODUCT WATER FLUX .'
                      CARBON EFFLUENT
                      600  psi (4l.4Kg/sqcm)
                      224  ft2 (20.8 m*)
                      20.9  gpd/ft* (85llpd/ma)
             i
I    i
                 i  i  i i  i
i    i  I   i i I  i
    10
                       IOZ
                                          \ol
                                       HOURS ON STREAM
Figure 4. Product water flux rate variation with operation time.

-------
developed in the membrane and eventually resulted in several  module failures.
The flux decline slope was calculated several times during the experimental
run.  It changed from -0.06 at 1,250 hours of operation to -0.04 at the end
of 7,165 hours of operation.  The corresponding values for a  spiral-wound
membrane system under similar operating conditions were -0.09 and -0.07 (2).
These decline slope values seem to indicate that the flux decline in a spiral-
wound membrane system was more rapid than a tubular membrane  system.

     During the entire series of operation with the carbon-treated second-
ary effluent, the water recovery remained at about 70 to 75 percent.  The
salt rejection averaged 97 percent until some leaks started to develop at
about 5,760 hours of on-stream operation in some of the modules.  This caused
the salt rejection to decrease to about 90 percent at the end of 7,165 hours
of operation.

     A summary of all routine chemical analyses run on the feed water, pro-
duct water, and brine water throughout the study is shown in  Table 1.  The
overall rejection of the inorganic ions, as measured by the reduction in IDS,
was about 95 percent.  The least rejected ion was nitrate, which was only
53.8 percent.

Operation with Primary Effluent

     Figure 5 shows the variations of the salt rejection and  the product water
flux rate with the time of on-stream operation.  Both product water flux rates
before and after membrane cleaning are shown in the figure.  As indicated in
the figure, the brine velocity in the last module was maintained at 0.64 mps
(2.1 fps) for the first 800 hours of operation, while for the latter part of
the operation, the brine velocity was increased to 1.4 mps (4.5 fps).  The
brine velocity was increased in an effort to reduce the amount of flux decline
due to membrane fouling between cleanings.  The experimental  data as shown  in
Table 2 confirm this.  There was less product water flux decline between membrane
cleanings at a brine velocity of 1.4 mps (4.5 fps) than at 0.64 mps (2.1 fps).
As indicated in Figure 5, the membrane cleaning was able to achieve recovery of
virtually all of the decline between cleanings.  When the flux rates just prior
to membrane cleaning and just after cleaning were plotted and the points in
each class were connected, a value for flux decline due to other factors than
short-term fouling was obtained.  As shown in Figure 5, the slope of the flux
rate decline was about -0.0125.  This decline rate apparently was not affected  by
the brine velocity.

     During the entire period of operation, the salt rejection seemed to re-
main at approximately 95 percent level.  Table 3 shows the averages of the
water quality analyses conducted on the feed, product, and brine samples.
The percent  rejections of the  various  constituents are also shown in
the table.
                                      1,5

-------
           TABLE 1.   SUMMARY OF CHEMICAL ANALYSES FOR PILOT PLANT
             OPERATIONS WITH CARBON-TREATED SECONDARY EFFLUENT
Parameter
Na
K
Ca
Mg
Cl
SOH
POrP
NH3-N
N03-N
Turbidity, JTU
Total COD
TDS
Feed
(mg/1)
123
12.8
46.3
25.8
105
306
10.5
14.9
2.6
5.9
17.1
687
Product
(mg/1)
7.8
1.1
1.2
0.4
11.5
3.2
0.15
1.1
1.2
0.09
1.6
34.7
Brine
(mg/1)
373
38.4
138
101
344
1016
36.6
46.0
6.5
11.0
53.7
2240
Rejection
93.7
91.4
97.4
98.4
89.0
99.0
98.7
92.6
53.8
98.5
90.6
94.9
NOTES:  1.  All analyses were run on grab samples taken at 8:00 a.m.

        2.  Feed samples taken after acidification with H2SOit;
            hence, values of S0i+ shown in table include sulfate
            contributed by acid addition.
                                     16

-------
2S 100
Z
O
1 °°
b
<
CO
80

X
=> 20
_l
U.
o:«
UJ£
b^
8l 10
o>
1 0
— i — \ — r Mill i i i i i i i i i I i
Q
_ o 9 ft»o • ® o® * ® o*o***e®o* •••••*•• ' ••• oo e ® 9
^ 099\ * 99 * 00* **« «** * °* *
lgpd/ftz = 40.7lpd/m*
® 1 fps s 0.305 mps



O BEFORE CLEANING
• AFTER CLEANING


_ _ _ -..-it
• > * • V * • ••••• •• \ .••
• •* • 0"Wo .•*•
• o *° o o °° °*o° ^ o ° oo oft«
°°oo\ «d« o o^,«0 o ~w- - « o k oo -o" o
0 0 o"w°0°o w ° " v ° V— SLOPE=-aOI25
0
-------
        TABLE 2.  EFFECTS OF THE BRINE VELOCITY
       UPON THE OPERATIONS WITH PRIMARY EFFLUENT
    Brine velocity, fps              2.14           4.5


Product water flux, gpd/ft2
   Before Biz cleaning                3.7           6.1
   After Biz cleaning                13.1          13.5
Water recovery, %
   Before Biz cleaning                9.1           6.1
   After Biz cleaning                26.7          13.
Salt rejection, %                    94            95


Biz frequency per week                3             3
No. of operating modules
 (32 original modules)
NOTES:  1.  1 fps     =0.305 mps

        2.  1 gpd/ft2 =40.7 lpd/m2

        3.  Biz       = enzyme detergent
                          18

-------
           TABLE  3.   SUMMARY  OF CHEMICAL ANALYSES  FOR  PILOT PLANT
                      OPERATIONS WITH  PRIMARY  EFFLUENT
Parameter
Na
K
Ca
Mg
Cl
SO*
PO^-P
NH3-N
N03-N
Total COD
TOC
TDS
Turbidity, JTU
Feed
(mg/1)
126
9.8
55.9
13.3
184
245
6.2
22.6
1.7
110
34.3
760
42.9
Product
(mg/1)
11.6
0.7
0.4
0.5
13.6
1.3
0.9
1.6
0.6
12.6
5.0
36.0
0.1
Brine
(mg/1)
146
12.2
67.7
15.7
205
266
7.2
24.9
1.7
124
37.6
883
47.9
Rejection
90.8
92.8
99.3
96.3
92.6
99.5
85.5
93.0
64.7
88.5
85.5
95.3
99.7
NOTES:  1.  Analyses were run on once-a-week grab samples taken at 8:00 a.m.

        2.  Feed samples taken after H2SQn addition; therefore SO^
            contribution by acid is also included in the result.
                                      19

-------
MEMBRANE MODULE STABILITY

     A  total of 17 membrane modules failed during the entire operation with
the carbon-treated secondary effluent between November, 1969 and August, 1970,
which covered a total of 7,165 hours of on-stream operation.  A summary of
the position and the date of each module replacement is shown in Figure 6.
The results of an on-site visual examination to determine the cause of the
module  failure for six of these modules are included in Table 4. 'Three of
the six modules showed no visible signs of membrane damage.  These modules
were sent back to the Universal Water Corporation for additional studies
to determine the cause of failure.  Table 5 shows the analytical results of
two failed modules, as reported by the Universal Water Corporation.  These
results indicate that the failure of the modules was due to the failure of the
tube, not the membrane, and that this failure was not related to operational
procedure.

     During August, 1970, some failed modules were replaced with modules
from row 8 (the downstream end row of the pilot plant system), instead of
obtaining replacement modules from the Universal Water Corporation, and the
operation of the system continued with a reduced loading of modules.  This
was done because the frequency of module failure seemed to be increasing.
The operation in this manner was continued to determine if the frequency of
module failure would be intensified or not.

     There are several possible causes for the increased number of module
failures during the month of August, 1970:

     A.  Hydrolysis of the membrane caused by the enzyme-detergent flush-
ing at a pH of about 10 could have possibly weakened the membrane.

     B.  Physical  attrition of the membrane at a point where the membrane
had been heat-treated and flared to form a water-tight seal.  This  area
was considered weak, and it was susceptible to this  type of failure.

     C.  Exposure to high levels of chlorine at approximately 10 mg/1
occurred on several occasions due to a malfunctioning chlorinator.   This
problem was corrected as soon as it was discovered,  within 24 hours, but
it could have already caused some membrane  degradation.

     D.  Physical  jarring of the modules while replacing failed modules with
new replacement modules or with modules from row 8,  could have caused  mem-
brane collapse.

     E.  Fatigue failure of the membrane caused by the daily depressurization
and repressurization for the tap water flush.

     One or several of the above could be responsible for the module failure.

     The experimental  run treating primary  effluent  was  concluded after
2,074 hours of operation on January 20,  1971  with  four modules  remaining.
One of these four modules was sent to  the Universal  Water Corporation  for
                                      20

-------
              AIR
 CONTROLLER

  FINAL
  BRINE
.ER-7


*4h	
               KXXXD-:'
    CHECK VALVE-
  MAIN
  FEED'
•CHECK VALVE-WEIGHT LOADED
                               OPEN VALVE
               -00'
               OOOO-
        rQQQO-
                                      a
                                      o
                                      I
                                      5
                                      a
MODULE REPLACED
    (I)
    (2)
    (3)
    (4)
    (5)
    (6)
    (7)
    (8)
    (9)
                  DATE OF REPLACEMENT (1970)
                   3/10
                   1/13, 3/10, 3/16, 7/22, 8/11
                   8/17, 8/25, 8/27, 8/31
                   2/5
                   3/16
                   7/31
                   7/22
                   8/25, 8/31
                   6/22
Figure 6. Module replacement in Universal Water Corporation
      reverse osmosis pilot plant.
                     21

-------
TABLE 4.  VISUAL EXAMINATION OF FAILED MODULES
Date Replaced
1/13/70
2/5/70
3/10/70
3/10/70
3/16/70
3/16/70
Position
Exit,
Inlet
Inlet
Exit,
Exit,
Exit,
in System
Row 8
, .Row 5
, Row 8
Row 8
Row 4
Row 8
Reason for Failure
Tear at 0-ring seal.
Eccentric U bend. Physical
attrition of membrane not
apparent.
Physical attrition of
membrane not apparent.
Tear at 0-ring seal.
Not apparent.
Not apparent.

-------
             TABLE 5.   LABORATORY ANALYSIS OF MODULE FAILURE  (3)
       Module
         Analytical  Results
Exit Module of Row 8
It showed signs of membrane failure
in a pressure tube, which would not
be related to pilot plant operating
procedures.
Exit Module of Row 4
It showed signs of erosion adjacent
to the element "0" ring seal.   It
is possible that some membrane
collapse also occurred.  The mem-
brane showed negligible signs of
chemical hydrolysis, and tensile
tests indicated that the yield
and break point were within 10%
of the standard, and were con-
sidered nominal.
                                      23

-------
module failure analysis.   The results reported by the Universal  Water
Corporation are shown in Table 6.   It is believed that the membrane and nylon
backing had been exposed to extreme pH conditions and chemical  attack.   The
nylon cloth which served both as a lateral transfer and a support bridge
across the perforation of the plastic tube, failed due to chemical  decomposi-
tion.  This, in turn, caused rupture of the weakened membrane at perforation
site.  The perforation was made to collect the product water through the mem-
brane supporting plastic tube.  In addition, possibly due to cyclic exposure
to very low and very high pH conditions, the membranes were hydrolyzed, re-
sulting in poor performance and loss of mechanical strength.

     Another cause of module failure that had been observed was  the collapse
of the membrane tube due to negative pressure or siphon.  This type of failure
was not always immediately apparent.  In collapsing and then returning to its
original form, the tubular membranes developed a longitudinal ridge that
served as a stress point.  These areas could puncture any time after pressuri-
zation.  It could be immediately or it could take days or weeks.  Generally,
a repeat of siphon conditions in the same tube caused immediate  failure.  To
avoid or minimize this type of failure, some necessary provisions for siphon
prevention should be incorporated in the plant design.
                                      24

-------
         TABLE 6.  COMPLETE ANALYSIS OF THE MEMBRANE MODULE AFTER A
              TOTAL OF 10,940 HOURS OF ON-STREAM OPERATION (4)
       Module  Part
    Analytical Results
Module  support structure
including tubing, headers,
and sleeves
 No  damage,  reuseable.
 '0" ring seals
 Partial  deterioration - needs re-
 placement.

 Damage due  to:  (a) Cold flow;
	(b) Chemical attack,
Membrane
 Partial  hydrolysis and compaction-
 no mechanical failure.

 Original  performance in 9/23/69:

 (a)   Product water flux rate =22.5
      gpd/ft2 (916 lpd/m2) @ 25°C.

 (b)   Salt rejection = 97.5%

 Performance on 1/23/71:

 (a)   Product water flux rate =13.0
      gpd/ft2 (529 Ipd/m2) 


-------
                                REFERENCES
1.  Smith, J.M., Masse, A.N.,  and Miele, R.P.,  "Renovation of Municipal
    Wastewater by Reverse Osmosis," EPA Water Pollution Control  Research
    Series 17040 (May 1970).

2.  Chen, Ching-lin and Miele, Robert P., "Demineralization of Carbon-
    Treated Secondary Effluent by Spiral-Wound  Reverse Osmosis Process,"
    Final Report to the Office of Research and  Development, U.S.  Environ-
    mental Protection Agency,  Contract No. 14-12-150.

3.  Shippey, F.R., Universal Water Corporation, Del  Mar, California,  Private
    Communication.

4.  Manjikian, S., Universal Water Corporation, Del  Mar, California,  Private
    Con-muni cation.
                                     26

-------
                                  TECHNICAL REPORT DATA
                           (Please read Instructions on the reverse before completing}
 REPORT NO.
 EPA-600/2-78-167
                                                           3. RECIPIENT'S ACCESSION NO.
 TITLE AND SUBTITLE
                                                           5. REPORT DATE
 Wastewater  Demineralization by Tubular Reverse
 Osmosis  Process
               September  1978 (Issuing Datpl
              6. PERFORMING ORGANIZATION CODE
 AUTHOR(S)
 Ching-lin  Chen and Robert P. Miele
                                                           8. PERFORMING ORGANIZATION REPORT NO.
 PERFORMING ORGANIZATION NAME AND ADDRESS
 County Sanitation Districts  of  Los  Angeles County
 Whittier,  California  90607
                                                           10. PROGRAM ELEMENT NO.
                  1BC611
              11. CONTRACT/GRANT NO.
                                                                14-12-150
2. SPONSORING AGENCY NAME AND ADDRESS
 Municipal  Environmental Research  Laboratory--Cin.,OH
 Office of  Research and Development
 U.S. Environmental Protection Agency
 Cincinnati,  Ohio  45268
              13. TYPE OF REPORT AND PERIOD COVERED
               Final Report 1/70 - 6/71
              14. SPONSORING AGENCY CODE
                  EPA/600/14
S. SUPPLEMENTARY NOTES
 Project Officer:   Irwin J. Kugelman    513-684-7631
6. ABSTRACT
 A 16 1pm  (4.2  gpm)  tubular reverse  osmosis pilot plant was  operated at Pomona
 Advanced  Wastewater Treatment Research Facility for over one  year.   The purpose of
 the study was  to investigate the  applicability of the tubular reverse osmosis process
 to wastewater  demineralization.   The  pilot plant operational  goal  was the study of
 the membrane life.   The feed was  carbon-treated secondary effluent.  This study
 lasted 7,165 hours  until the operating membrane modules  reduced from the original
 32 modules  to  12 modules as a result  of membrane module  failure.   It was subsequently
 decided to  make use of the remaining  12 modules in the system for exploring the
 behavior  of the membrane process  in treating primary effluent.   The system treated
 primary effluent during the third phase for a period of  2,074 hours.  Repeated
 membrane  failure resulting from poor  mechanical design of the system obviated
 drawing significant conclusions.
17.
                                KEY WORDS AND DOCUMENT ANALYSIS
                  DESCRIPTORS
                                              b.IDENTIFIERS/OPEN ENDED TERMS
                                                                            COSATl Field/Group
 Demineralizing
 Desalting
 Water Reclamation
   Membrane Fouling
  13B
18. DISTRIBUTION STATEMENT

 Release to  Public
 19. SECURITY CLASS (ThisReport)
   Unclassified
»1. NO. OF PAGES

  35
                                               2O. SECURITY CLASS (Thispage)

                                                Unclassified
                                                                          22. PRICE
EPA Form 2220-1 (Rev. 4-77)
27
                                                        . GOVERNMENT PRINTING OfflCE: 1978- 7B7-140/1429

-------