EPA-600/2-77-169
September 1977
Environmental Protection Technology Series
DEMORALIZATION OF
SAND-FILTERED SECONDARY EFFLUENT BY
SPIRAL-WOUND REVERSE OSMOSIS PROCESS
Municipal Environmental Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
-------
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-600/2-77-169
September 1977
DEMORALIZATION OF SAND-FILTERED SECONDARY EFFLUENT
BY
SPIRAL-WOUND REVERSE OSMOSIS PROCESS
by
Chi ng-1in Chen
Robert P. Miele
County Sanitation Districts of Los Angeles County
Whittier, California 90607~
Contract No. 14-12-150
Project Officer
Irwi n 0. 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 publica-
tion. Approval does not signify that the contents necessarily reflect the
views and policies of the U.S. Environmental Protection Agency, nor does
mention of trade names or commercial products constitute endorsement or
recommendation for use.
11
-------
FOREWORD
The Environmental Protection Agency was 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 pre-
servation and treatment of public drinking water supplies, and to mini-
mize 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 waste-
water so that it can be reused. It is expected that partial deminerali-
zation of conventionally treated wastewater will be required if the waste-
water 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 permiability of these
membranes is low so high pressure is required to achieve an economical
production rate. Special configuration of the membrane 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 a spiral membrane-support configura-
tion was tested for its efficacy in demineralization of secondary
effluent.
Francis T. Mayo, Director
Municipal Environmental Research
Laboratory
-------
ABSTRACT
A 22.7 cu m/day (6,000 gallons/day) spiral-wound reverse osmosis pilot
ilant, was operated at the Pomona Advanced Wastewater Treatment Research
:acility on the sand-filtered secondary effluent. The pilot plant study
(as conducted under optimum operating conditions based on previous studies.
luring the first year of operation, all the system performance parameters,
;uch as salt rejection, water recovery, and product water flux rate, were
only slightly decreased from their initial values. However, the salt
rejection and product water flux rate were substantially reduced to almost
half of their initial values after a two year operation period. During this
same two year period, the water recovery was found to decline about 15 per cent
of its initial value.
A cost estimate for a 37,850 cu m/day (10 MGD) plant for August, 1973
cost figures indicated that for membranes with only one-year life the process
cost was about 16.5(^/1,000 liters (63.6^/1,000 gallons). However, the cost
could be substantially reduced to 12.4<£/1,000 liters (47.5^/1,000 gallons)
for membranes with two-year life. Both cost estimates did not include the
costs for sand filtration pretreatment and brine disposal.
This report was submitted in fulfillment of Contract No. 14-12-150 by
the County Sanitation Districts of Los Angeles County under the sponsorship
of the U.S. Environmental Protection Agency. This report covers a period
from July 1971 to June 1973.
IV
-------
CONTENTS
Foreword ..................... i i 1
Abstract ..................... "> y
Figures ...................... vi
Tables ...................... Y! ]
Acknowledgments .................. vm
1. Introduction ................ 1
2. Conclusions ................. 3
3. Recommendations ............... 5
4. Pi lot Plant Description ........... 6
5. Pilot Plant Operation ............ 10
6. Results and Discussions ........... 13
7. Process Cost Estimate ............ 33
References 37
-------
FIGURES
Number
1 Configuration of spiral-wound membrane
module
2 Schematic flow diagram of the reverse
osmosis pilot plant
3 Salt rejection and feed pressure variation
vs. operational time under constant
flux rateoperation 14
4 Salt rejection and product water flux variation
vs. operation time under constant
operating pressure 20
5 Monthly averages of the R.O. pilot plant
performance parameters (corresponding to
the period of 5900 to 17528 hours of
operation) 29
VI
-------
TABLES
Number pag<
1 Evaluation of the Lead Module Membrane
in Pressure Vessel No. 1 at 4,450
Hours of Operation 19
2 Average Water Quality Characteristics
(October 15, 1971 to March 2, 1972) 31
3 Typical Ion Rejection Values (%) at
Different Period of Operation 32
4 Process Cost Estimate for 37,850
Cu M/Day (10 MGD) Spiral-Wound
Reverse Osmosis Plant 36
vn
-------
ACKNOWLEDGMENTS
This study was jointly sponsored by the U.S. Environmental
Protection Agency and the County Sanitation Districts of
Los Angeles County.
Mr. Harold H. Takenaka, former U.S. EPA Project Engineer at
Pomona Advanced Wastewater Treatment Research Facility, was in-
strumental in initiating the pilot plant study.
The advice and suggestions given by Dr. James E. Cruver of
Gulf Environmental Systems Company during the course of the
study were important contribitions to the success of the study.
The untiring efforts of both the operating and laboratory
staff of the Pomona Advanced Wastewater Treatment Research
Facility are gratefully acknowledged.
vn
-------
SECTION 1
INTRODUCTION
As part of the continuing development program for the
application of the reverse osmosis process to wastewater de-
mineralization, the Gulf Environmental Systems Company con-
ducted a special research study, on a short-term basis, on the
effects of various pretreatrnent systems on the membrane perform-
ance under the U.S. EPA Contract No. 14-12-831. The study was
mostly performed at the Pomona Advanced Wastewater Treatment
Research Facility concurrently with other reverse osmosis pilot
plant studies.
The Gulf Environmental Systems Company concluded their pre-
treatment study on the contract expiration date of July 31,
1971. They indicated in their contract final report that an
activated carbon adsorption pretreatment was clearly not neces-
sary for a successful reverse osmosis system operation.'1) In
the same report, they further demonstrated that a sand filtra-
tion process could provide an equally satisfactory pretreatment
for a reverse osmosis system operation.
This study was initiated to confirm the findings of the
Gulf Environmental Systems Company on an extended long-term
basis, and also to achieve the following specific objectives:
A. To establish the effective life of the membrane of
a spiral-wound reverse osmosis system in demineralizing a
sand-filtered secondary effluent:
B. To determine the reliability of the process perform-
ance ; and
C. To obtain the operating and design data for making
a realistic process cost estimate.
The study was conducted with the same reverse osmosis pilot
plant previously used by the Gulf Environmental Systems Company
in their pretreatment study. The pilot plant was a spiral-
wound membrane system and had a nominal production capacity of
22.7 cu m/day (6,000 gallons/day). The pilot plant had accumu-
lated a total of 2,503 hours of on-stream operation before the
Gulf Environmental Systems Company terminated their study on
July 31, 1971. This extension study had added another 15,025
-------
hours of operation to the pilot plant to accomplish a total of
17,528 hours (equivalent to a two-year period) of on-stream
operation. The study was formally completed as of June 11,
1973.
-------
SECTION-2
CONCLUSIONS
The following conclusions can be drawn from the pilot plant
study:
A. The spiral-wound reverse osmosis system operated suc-
cessfully on the secondary effluent of Pomona Water Reclamation
Plant with only sand filtration pretreatment. The effective
membrane life for the operation was approximately one year.
B. The pilot plant maintained 90 percent or more of salt
rejection, 326 1/sq m/day (8 gal/sq ft/day) or more of product
water flux, and 75 percent or more of water recovery under an
average operating pressure of 34.5 Kg/sq cm (500 psi) during
the first year of operation.
C. The product water flux rate decline was controlled by a
daily air-tap water flushing and a three-times-a-week chemical
cleaning. Three types of cleaning solutions, namely, Biz
enzyme detergent, sodium perborate and sodium ethylenediaminete-
traacetate (EDTA), were found to perform equally well. However,
as recommended by the membrane manufacturer at the middle of the
study, only EDTA cleaning solution was used for the membrane
cleaning during the second year of operation.
D. The water flux rate decline could be minimized by main-
taining a minimum brine flow of 15 1/min (4 gpm).
E. The water recovery could be enhanced by a partial re-
cycling of the brine to the feed stream.
F. The overall reductions in the salt rejection, water re-
covery, and product water flux rate during the two years of on-
stream operation were approximately 51 percent, 15 percent, and
50 percent, respectively.
G. The product water quality prior to the start of serious
membrane deterioration was excellent.
H. The total process cost estimate for a 37,850 cu m/day
(10 M6D) plant is about 16.54/1,000 liters (63.64/1 ,000 gallons)
However, if the membrane effective life can be improved from one
year to two years, then the process cost can be substantially
-------
reduced to 12.44/1,000 liters (47. 5cVl ,000 gallons). Both cost
estimates do not include the costs for sand filtration pretreat-
ment and brine disposal.
I. A comparison of total process costs, including pre treat-
ment costs, between two different pretreatment schemes for the
spiral-wound reverse osmosis process indicates that the sand
filtration pretreatment scheme is somewhat less expensive than
the carbon adsorption pretreatment scheme.
-------
SECTION 3
RECOMMENDATIONS
The relatively short membrane life as concluded from this
pilot plant study on the wastewater demineralization is rather
discouraging. An optimum membrane life would be three years,
if the process is to be practical and economical for the appli-
cation to the wastewater demineralization.(2) Therefore, it is
highly recommended that further studies be pursued primarily in
the areas of membrane improvement. Other parameters such as
pretreatment methods, membrane cleaning solution and frequency,
feed pressure, brine recirculation , membrane module configura-
tion, and brine flow rate should also be thoroughly evaluated
and investigated.
-------
SECTION 4
PILOT PLANT DESCRIPTIOI
The 22.7 cu rn/day (6,000 gpd) spiral-wound reverse osmosis
pilot plant consisted of four steel pressure vessels 3.0 m (10
ft) long by 10 cm (4 in) in diameter, each of which contained
three ROGA spira1-wound membrane modules. Each module had a
5.6 sq m (60 sq ft) of high flux membrane cast on D-601 sail-
cloth backinn, 1.1 mm (0.045 in) polypropylene Vexar brine
spacers, and melamine-treated tricot product water channels.
The average water permeability coefficient value of the membrane
modules was 2.5xlO"s g/sq cm/sec/atm. Figure 1 shows the con-
figuration of the spiral-wound membrane module. The total mem-
brane area in the pilot plant system was 67 sq m (720 sq ft).
The schematic flow diagram of the 22.7 cu m/day (6,000 gpd)
reverse osmosis pilot plant is shown in Figure 2. The sand-
filtered secondary effluent was chlorinated to provide 1 to 2
mg/1 of residual chlorine and acidified to pH close to 5 using
sulfuric acid before being fed to the reverse osmosis system.
The pilot plant system was in a 2-1-1 array, as shown in Figure
2, to maintain sufficient brine flow rates in the downstream
modules.
The brine was partially recycled to maintain an apparent
water recovery at the level of 75 to 80 percent. This brine re-
cycling slightly increased the inorganic and organic matters in
the feed and thus might cause adverse effect on membrane foul-
ing. A minimum brine flow of 15 1/min (4 gpm) was maintained
during the system operation to avoid concentration polarization
of the membranes.
An Apco back pressure unit was used to regulate the system
operating pressure. Sufficient sample valves were installed on
the pilot plant system, so that samples from the raw feed (sand-
filtered secondary effluent), blended feed (mixture of sand-
filtered secondary effluent, chlorine solution, sulfuric acid,
and recycled brine), brine and product streams could be taken
regularly. Instrumentation was included to measure the tem-
perature and the pressure of the blended feed, brine, and prod-
uct streams. A proportional chemical feed pump was used to add
-------
BACKING
MATERIAL
SEE DETAIL-A
PERMEATE
TUBE
SPACER
GLUE
MEMRANE
DETAIL-A
Figure I. Configuration of spiral-wound membrane module.
7
-------
CO
SECONDARY
EFFLUENT
PUMP
RECYCLED BRINE
PRESSURE
VESSEL I
PRESSURE
VESSEL 2
PRESSURE
VESSEL 3
PRESSURE
VESSEL 4
"I
•
1
WASTE
'BRINE
PRODUCT
WATER
Figure 2. Schematic flow diagram of the reverse osmosis pilot plant.
-------
sulfuric acid to the blended feed stream for pH control. The
pump rate was regulated by a pH controller. An Advance gas
chlorinator was employed for chlorine addition.
The pilot plant system was designed and constructed to be
cleaned regularly without disassembly of manifolds. A chemical
cleaning solution was made up in a cleaning tank and was then
circulated through the pilot plant by a centrifugal booster pump.
Enough valves were provided in the pilot plant system so that
each pressure vessel could be cleaned individually or the system
could be cleaned as a series-para 11 el array. Tap water or air-
tap water mixture could be introduced for flushing just ahead of
the pressure vessel array during cleaning cycle or downtimes.
The pressure sand filter used in the pretreatment system was
a standard package designed by L.A. Water Conditioning Company,
City of Industry, California. The filter was 76.2 cm (30 in) in
diameter and had about 45.7 cm (18 in) depth of sand. The sand
bed with an effective size and a uniformity coefficient of
approximately 0.5 mm and 1.6, respectively, was supported by a
layer of graded gravel. The hydraulic loading rate of the sand
filter was maintained at 2 Ips/sq m (3 gpm/sq ft). The filter
was normally backwashed once a day.
-------
SECTION 5
PILOT PLANT OPERATION
OPERATING CONDITIONS
At the beginning of this study, the pilot plant was operated
at a constant product flux rate of 407 1/sq m/day (10 gal/sq
ft/day) to simulate the actual plant operation. The operating
pressure was frequently adjusted to maintain such constant flux
rate operation. The operating pressure was found to vary from
24.2 Kg/sq cm (350 psi) to 48.3 Kg/sq cm (700 psi) to produce a
407 1/sq m/day (10 gal/sq ft/day) flux rate during the first
5,900 hours of operation. Due to the lack of an adequate control
mechanism to automatically make the necessary operating pressure
adjustment, the constant flux mode of operation was converted to
a constant 34.5 Kg/sq cm (500 psi) operating pressure mode of
operation starting at 5,900 hours of on-stream operation.
The initial performance parameters under the constant
operating pressure operation were as follows:
A. Product water flux rate (adjusted to 25°C) : 488 1/sq
m/day (12 gal/sq ft/day).
B. Water recovery (defined as "100 x flow rate of product
stream/flow rate of raw feed stream:) : 80 percent.
C. Salt rejection (defined as "100 x conductivity of
product stream/conductivity of blended feed stream") :
97 percent.
The sand-filtered secondary effluent was chlorinated to
provide 1 to 2 mg/1 chlorine residual and acidified to a pH
close to 5 using sulfuric acid for biological growth and chemi-
cal precipitation controls, respectively. The minimum brine
flow was regulated at 15 1/min (4 gpm) and the recycled brine
flow was maintained at 10 1/min (2.7 gpm).
MEMBRANE CLEANING PROCEDURES
The reverse osmosis pilot plant operation was started out
with a daily tap-water flushing and a twice-a-week chemical
10
-------
solution cleaning cycle. This membrane cleaning schedule was
practiced routinely throughout the first 2,500 hours of on-
stream operation. However, it was found necessary to increase
the frequency of the chemical solution cleaning to three times
a week to maintain the desired performance level. This new
cleaning frequency was equivalent to a cleaning interval of 815
liters of product water per square meter of membrane surface
area (which was about 20 gallons per square foot of membrane
surface area).
Three types of chemical cleaning solutions--Biz enzyme-
detergent, sodium perborate and sodium ethylenediaminetetraace-
tate (EDTA)--were tested during the first 7,900 hours of on-
stream ope r at ion.
However, the Biz enzyme-detergent (contained some small
amount of sodium perborate) and the sodium perborate solution
were subsequently found corrosive to the cellulose acetate mem-
brane according to Cruver'3) of the Gulf Environmental Systems
Company.
Consequently, only the EDTA cleaning solution was used as
the cleaning agent for the membranes throughout the rest of the
pilot plant operation. The concentrations and constituents of
the various cleaning solutions used were as follows:
A. Biz enzyme-detergent solution:
a. 2"'. of Biz enzyme-detergent
B. Sodium perborate solution:
a. 2% of sodium perborate
b. 0.15% of Triton X-100 (non-ionic detergent)
c. 0.0015% of carboxy methyl cellulose
C. EDTA solution:
a. l/o of EDTA (tetra sodium salt)
b. 0.15% of Triton X-l00
c. 0.0015% of carboxyl methyl cellulose
The pH of all the cleaning solutions was adjusted to 7.5 to
8-0 to minimize the membrane hydrolysis reaction. The cleaning
solutions were prepared with warm tap water (40°C to 60°C) to im-
prove the cleaning efficiency.
The chemical cleaning solution was first flushed through all
the reverse osmosis pressure vessels for five minutes, then the
solution was flushed through each individual pressure vessel for
10 minutes, and finally the solution was flushed through the
11
-------
entire system again for another five minutes. Therefore, the
total flushing time for the pilot plant was about 50 minutes.
The flushing rate for the cleaning solution was regulated at the
level of 37.9 1/min (10 g p m) . After the chemical solution clean-
ing cycle, the system was thoroughly rinsed with tap water in the
same procedure as the chemical cleaning solution. Therefore, the
downtime required for daily membrane cleaning ranged from 50 to
100 minutes depending on whether the chemical cleaning was prac-
ticed along with the daily tap-water flushing or not.
SAMPLING AND MONITORING PROCEDURES
During the week days (Monday through Friday), the following
operating parameters were monitored routinely:
A. The conductivities of the feed, product, and brine
streams of each pressure vessel .
B. The rate of product flow from each pressure vessel.
C. The pressure difference through each pressure vessel.
Daily (including Saturday and Sunday) measurements were per-
formed on the following parameters:
A. The temperature of the blended feed water.
B. The pH values of the blended feed and brine streams.
C. The chlorine residual of the blended feed streams.
D. The total pressure difference through the pilot plant
system.
E. The total product flow before and after the membrane
cleaning.
F. The conductivities of the blended feed and the final
product streams.
G. The total and recycled brine flows.
H. The feed operating pressure.
I. The total on-stream operation time.
Besides the above routine daily monitoring procedures, some
grab samples from the raw feed, blended feed, product and brine
streams were also taken every Thursday at 8:00 A.M. for water
quality analyses.
12
-------
SECTION 6
RESULTS AND DISCUSSIONS
CONSTANT PRODUCT FLUX OPERATION
Figure 3 shows the variations of the salt rejection and the
feed pressure during the first 5,900 hours of on-stream operation
The product flux was maintained at a constant rate of 407 1/sq
m/day (10 gal/sq ft/day) during the first 5,900 hours of opera-
tion. As indicated in Figure 3, the feed pressure required to
maintain the above constant product flux rate was approximately
31.7 Kg/sq cm (460 psi) during the first 1,000 hours of opera-
tion. However, the required feed pressure rapidly increased to
as high as 47.6 Kg/sq cm (690 psi) during the next 400 hours of
operation. Because of this high feed pressure development, the
system was depressurized for a 72 hour period after 1,430 hours
of operation. During the depressurization period, a constant
tap water flushing through the system was maintained. After
this depressurization treatment, the feed pressure was reduced
to the previous level of 31.7 Kg/sq cm (460 psi) to maintain the
407 1/sq m/day (10 gal/sq ft/day) product flux. The pressure
remained at about 27.6 to 34.5 Kg/sq cm (400 to 500 psi) until
2,200 hours of operation.
Between 2,200 to 3,600 hours of operation, the system re-
quired 31.1 to 38.0 Kg/sq cm (450 to 550 psi) to maintain the
constant flux of 407 1/sq m/day (10 gal/sq ft/day). Then it
became necessary to gradually increase the feed pressure to as
high as 48.3 Kg/sq cm (700 psi). Two 24 hour and two 72 hour
depressurization periods on tap water between 3,600 and 4,450
hours of operation were not successful in reducing the feed
pressure to the initial operating level.
At the 3,844 hours of operation, the membrane modules were
inspected. Several brine seals were "blown," so they were re-
taped in place. The "blown" brine seals probably resulted from
the high operating feed pressure. At 4,000 hours of operation,
the membrane modules were again inspected. The inspection
showed that the modules were very clean with the barber-pole
type outer wrap tape and the brine seals were in excellent con-
dition. However, the condition of the brine channels could not
13
-------
100
o
LU
-s
LU
(T
_J 90
CO
600
CL
- 500
CO
CO
LU
cr
CL
Q
LU 400
LU
.•
PRODUCT FLOW
5 GPM-»-
-»~4.8 GPM
LOST pH CONTROL
B-BIZ FLUSH
E-EDTA FLUSH
S-SODIUM PERBORATE FLUSH
^-72 HR. TAPWATER
>4 HR. TAPWATER
I PSI = 0.069 kg/sqcm
I GPM = 3.8 I /min *
.
.*
• •
• • • • •
.•.
••
..
SS SS SS SS SS SSSS
1 1 1 \ I
SE EE BBB
200
400
600 800 1000
HOURS ON STREAM
1200
1400
1600
Figure 3. Salt rejection and feed pressure variation vs. operational time under constant flux
rate operation.
-------
100
o
UJ
-»
UJ
(T
(O
90
600
CO
o.
- 500
UJ
(O
(O
UJ
ac
CL
o
UJ 400
UJ
u.
B B B
B B B S
J_
_L
B B B
I
BBB
_ I
SSS SSS SSS
1
L
1600
1800
2000
2200 2400 2600
HOURS ON STREAM
2800
3000
3200
Figure 3. Continued.
-------
2 100
o
o
LU
UJ
4.8 GPM PRODUCT FLOW 4-400'PSI FEED PRESSURE
<
CO
90
• •
600
CO
Q.
- 500
a:
to
CO
LU
cr
o.
o
LU 400
LU
A
t
r
SS SSS
SSS S B S SBBBB B E E EE E E
1 1 1 1
t
SB IBB BBB B
3200
3400
3600
3800 4000 4200
HOURS ON STREAM
4400
4600
4800
Figure 3. Continued.
-------
100
o
o
UJ
3
cr
<
CO
90
600
10
^500
DC
CO
CO
UJ
OC
CL
O
UJ 400
UJ
u.
400 PSI FEED
PRESSURE
450 PSI FEED PRESSURE
B B
BBB SSS SSS SSS SSS SSS
_L
I
4800
5000
5200 5400
HOURS ON STREAM
5600
5600
6000
Figure 3. Continued.
-------
be determined without opening the modules. The reason for the
high feed pressure required to maintain 407 1/sq m/day (10
gal/sq ft/day) product flux was not readily apparent.
After the 72 hour depressurization at 4,450 hours of opera-
tion, it was decided to operate the system at 27.6 Kg/sq cm
(400 psi) feed pressure for a month period in an attempt to im-
prove the "A" (water permeability coefficient) value. At this
time, the lead module in the pressure vessel No. 1 was replaced
with a new module. The old module was sent to the Gulf
Environmental System Company for membrane evaluations. The re-
sults are shown in Table 1.
During the month long period of constant 27.6 Kg/sq cm (400
psi) feed pressure operation, the product water flux rate (at
25°C) increased from about 305 to 366 1/sq m/day (7.5 to 9
gal/sq ft/day) which was an encouraging trend. The constant
feed pressure was subsequently increased to 31.1 Kg/sq cm (450
psi) for another month of trial operation. The product water
flux rate was improved from 366 to 448 1/sq m/day (9 to 11
gal/sq ft/day) as a result of this feed pressure increase.
Therefore, the system operation was converted from a constant
product flux rate operation to a constant feed pressure opera-
tion after 5,900 hours of on-stream operation.
CONSTANT FEED PRESSURE OPERATION
At all times after 5,900 hours of operation, the reverse
osmosis pilot plant system was operated at a constant 34.5
Kg/sq cm (500 psi) feed pressure. The variation of the salt re-
jection and the product water flux rate throughout this series
of constant feed pressure operation are shown in Figure 4. As
indicated in this figure, the initial product water flux rate
was about 448 1/sq m/day (11 gal/sq ft/day). However, the flux
rate started to decline at about 7,000 hours of operation.
At 7,266 hours of operation, the No. 2 pressure vessel was
inspected because its salt rejection was about 3 percent lower
than the other three vessels. The inspection showed that two of
the three modules had experienced "blown" brine seals. The
three modules looked relatively clean and the outer wrap tapes
were in excellent condition. The blown brine seals were retaped
in place and the system was put back in operation. Experience
at Pomona had shown that the barber-pole type of outer wrap tape
could successfully maintain module integrity when used in treat-
ing municipal wastewater.
As shown in Figure 4, there was a sudden decrease in prod-
uct water flux rate at 7,445 hours of operation. This might be
the result of excessive membrane fouling due to the poor
18
-------
TABLE 1
EVALUATION OF THE LEAD MODULE MEMBRANE IN PRESSURE
VESSEL NO. 1 AT 4,450 HOURS OF OPERATION
Feed Solution
and
Membrane Condition
2,000 mg/1 NaCl Solution
Before Membrane Cleaning
Feed
Press ure
(psi)
415
"A" Value Salt Rejection
(g/cm2/sec/atm) %
1.25X.O-5 93.4
2,000 mg/1 NaCl Solution 420 1.81X10'5 94.3
After Membrane Cleaning
Notes: 1. The membrane was cleaned with an enzyme-detergent (Biz) at 60°C
with pH adjusted to 7.
2. The module was opened after the tests, and there was a brownish
film on the membrane with very little, if any,seal ing evidence.
3. 1 psi = 0.069 Kg/sq cm.
-------
100
_ I
Q
• ••
• •i
~~ ^ — '~"~*~~****~~W^ • _ • w
•
o
LU
UJ
cr
5
<
90h B-BIZ FLUSH
~~»-72HR. TAPWATER
12
O
o
lO
CO
. 10
X
B!
cc
UJ
1
cr
a.
•• • *•••«•
•• • . * * • - • _ _ .. •.
E-EDTA FLUSH
S- SODIUM PERBORATE FLUSH
I-GFD = 40.7 l/sqm/day
• •
ss sss ssssss sss sss sss sss sssss
I I I I I I I
5800 6000 6200 6400 6600 6800 7000 7200 7400
HOURS ON STREAM
Figure 4. Salt rejection and product water flux variation vs. operation time under constant
operating pressure.
-------
100
UJ
-------
100
o
UJ
—>
UJ
(T
<
CO
90
80 •
O
o
IO
CM
PO Q
- 10
X.
a:
UJ
I
8
8
IT
O.
E E E E E
I
EEE EEE EEEEE
I I i
' t
EEE EEEE E E.EE E
900O
9200
9400
9600 9800
HOURS ON STREAM
lopoo
10,200
10,400
10,600
Figure 4. Continued.
-------
ro
oo
100
o
UJ
UJ
o:
|3 90
<
80
O
o
IO
CVJ
a
u.
°. .0
X
-------
100
o
t-
o
UJ
tr
90 •
80
O
e
IO
CM
O
* ,o
s
cr
UJ
E E
• •• •
EEE
• E E
E
EEE EEE EEE EEE EEE EEE EEE E
12,200
12,400
12,600
12,800 13,000 13,200
HOURS ON STREAM
13,400
13,600 13,800
Figure 4. Continued.
-------
ro
en
90
g
o
UJ
UJ
80
70
o
o
\f>
CM
O
* .0
X
cc
UJ
O
EE EEE EEE EEE
6
13,800
EEE EEE EEE EEE EEE
_J _ I - 1 - 1
EEE EE
14,000
14,200
I4.4OO 14,600
HOURS ON STREAM
14,800
15,000
I5.20O 15,400
Figure 4. Continued.
-------
80
O
UJ
->
UJ
cr
CO
60 -
O
O
m
W
o
u.
® 10
X
oc.
IU
I
O
O
O
• •
• • •
1 - -. •
EEE EEE EEE EEE EEE EEE
EEE E £• • E.EE
_L
I
I
15,400
15,600
15,800
16,000 16,200 16,400
HOURS ON STREAM
16,600
16,800
17,000
Figure 4. Continued.
-------
IV)
70
O
UJ
UJ
cc
<
U)
60
50
O
o
10
CM
8
X
IT
UJ
I
O
O
o
cr
a.
EEE EEE EEE
17,000
17.2OO
17,400
17,600 17,800
HOURS ON STREAM
18,000
18,200
18,400 18,600
Figure 4. Continued.
-------
secondary effluent quality. The turbidity and the dissolved
chemical oxygen demand (DCOD) of the sand-filtered secondary
effluent were as high as 22 JTU and 76 mg/1, respectively, dur-
ing the two week activated sludge plant upset period between
7,421 and 7,737 hours of pilot plant operation. In spite of the
poor quality secondary effluent, the module cleaning, using so-
dium perborate solution, was able to maintain the product water
flux rate between 265 to 366 1/sq m/day (6.5 to 9 gal/sq ft/day).
At 8,096 hours of operation, the membrane modules in the
system were inspected again. The results of the inspection
showed that feed ends of the lead modules in pressure vessels
No. 1 and No. 2 were covered with suspended solids and other de-
bris, which were easily removed by applying a high pressure
stream of tap water. No serious scaling to the extent of caus-
ing a severe restriction of the brine flow was found. The brine
seals and the outer wrap tape were in good condition. A 72 hour
depressurization on chlorinated tap water allowing the module
inspection was able to improve the product water flux rate
si i gh11y.
As indicated in Figure 4, the salt rejection maintained at
95 percent or more for the first 7,400 hours of on-stream opera-
tion. Since then, the salt rejection gradually decreased to
about 90 percent at 9,000 hours of operation. Similarly, the
product water flux rate was maintained at 407 1/sq m/day (10
gal/sq ft/day) or higher practically throughout the first 7,400
hours of operation, and then it decreased to about 326 1/sq
m/day (8 gal/sq ft/day) at 9,000 hours of operation. This
amounted to a 6 percent reduction in salt rejection and a 20
percent reduction in product water flux rate during the first
year pilot plant operation.
The salt rejection was maintained at the level of 90 per-
cent between 9,000 and 10,600 hours of operation. However, the
product water flux rate was found to decrease from the level of
326 1/sq m/day (8 gal/sq ft/day) to the level of 285 1/sq m/day
(7 gal/sq ft/day) during the same operation period.
During the period between 10,600 and 12,600 hours of opera-
tion, the product water flux rate was found to hold steady at
the level of 285 1/sq m/day (7 gal/sq ft/day) while the salt re-
jection was found to decrease rapidly from 90 percent to 80 per-
cent .
The product water flux rate started to reverse its down-trend
at 12,600 hours of operation (in the month of November, 1972),
as shown in Figure 5. This up-trend continued for a period of
3,200 hours, and it was finally decreased again at 15,800 hours
of operation.
28
-------
100
WATER RECOVERY
PRODUCT WATER FLUX
I GFD = 40.7 l/sq m/doy
/ / ,
/72 /72 /72
MONTH/YEAR
Figure 5. Monthly averages of the R.O. pilot plant performance parameters (corresponding to the
period of 5900 to 17528 hours of operation).
-------
The phenomenon of the product water flux rate reversal
could very well indicate that some physical destructions of the
membrane structure had taken place inside the membrane modules.
The damaged membranes could lose the ability to function. Con-
sequently, some brine water would be able to flow through the
membrane and contaminate the product water, thus causing a poor
salt rejection. After 15,800 hours of operation (corresponding
to the month of March, 1973), the salt rejection, product water
flux rate, and the water recovery were all found to decrease
rapidly, as shown in Figure 5. The pilot plant was finally
terminated on June 11, 1973, after a total of 17,528 hours of
on-stream operation. The total reduction in salt rejection,
water recovery, and product water flux rate during the two
years of on-stream operation period were approximately 51 per-
cent, 15 percent, and 50 percent, respectively.
WATER QUALITY CHARACTERISTICS
The typical water quality characteristics of the sand-
filtered secondary effluent, which was used as the raw feed to
the reverse osmosis pilot plant study, are shown in Table 2.
The concentration of the nitrate ion in the raw feed was so low
that the 30 percent rejection, as indicated in Table 2, might
not demonstrate the true rejection capability of the reverse
osmosis system. It is also shown in Table 2 that both organic
matter and turbidity were quite effectively removed. However,
those high rejection values during the initial period of opera-
tion were found to decrease rapidly after approximately 9,000
hours of on-stream operation.
Table 3 shows the typical ion rejection values at t
different periods of operation. As indicated in Table 3, :.
chloride and nitrate rejections dropped to zero after about
two years of operation.
30
-------
TABLE 2
AVERAGE WATER QUALITY CHARACTERISTICS
(October 15, 1971 to March 2, 1972)
Analysi s
Na
K
Ca
Mg
Cl
so,
P04-P
NH3-n
N03-N
TCOD
DCOD
TDS
Turbidity
(JTU)
Raw F
(mg/
94.
12.
57.
11.
80.
65.
10.
18.
0.
31 .
25.
496
2.
eed
1)
5
0
2
4
6
0
3
6
09
7
1
1
Bl ended
Feed
(mg/1)
175
21
103
21
155
534
19
33
0
57
50
1 ,127
3
.0
.9
.9
.8
.4
.0
.8
.1
.10
.1
.0
.3
Product
(mg/1)
13.
1 .
0.
0.
18.
4.
0.
1 .
0.
0.
0.
52
0.
8
4
94
34
9
9
26
8
07
95
56
03
Bri ne
(mg/1)
323
41
211
41
301
1 ,038
37
63
0
99
85
2,099
8
.0
.2
.3
.8
.0
.5
.6
.16
.8
.8
.0
Rejecti on
U)
92
94
99
98
88
99
99
95
30
98
99
95
99
Notes: 1. Raw Feed is sand-filtered secondary effluent.
2. Analysis on once a week grab samples taken at
8:00 A.M.
3. Difference between raw feed and blended feed is
due to HjSO^ addition, chlorination and brine
reci rculat ion .
31
-------
TABLE 3
TYPICAL ION REJECTION VALUES (:;) AT
DIFFERENT PERIOD OF OPERATION
Ion —
Sodi urn
Potassi urn
Ammonia Nitrogen
Cal ci um
Magnes i um
Chi ori de
Nitrate Nitrogen
Sulfate
Phosphate
Total COD
Dissolved COD
TDS
Appro
6,500
90
94
95
99
98
84
30
99
98
99
99
95
ximate Hours
12,000
72
80
75
90
85
33
13
88
85
85
86
79
on Stream
17,500
43
38
39
40
48
0
0
47
33
48
37
41
32
-------
SECTION 7
PROCESS COST ESTIMATE
The pilot plant studies at Pomona Research Facility have
successfully demonstrated the technical feasibility of apply-
ing the spiral-wound reverse osnosis system to the deminerali-
zation of the sand-filtered secondary effluent. A process cost
estimate for this application is prepared on the basis of the
pilot plant operation results. The major assumptions made for
this cost estimate are listed as follows:
A. The blended influent TDS for the reverse osmosis
system is about 1200 mg/1;
B. The water recovery for the process is about 80
percent;
C. The product water flux rate is approximately 407
1/sq m/day (10 gal/sq ft/day) at 25°C;
D. The process is capable of rejecting 90 or more
percent of the influent TDS;
E One percent EDTA solution is used as the membrane
cleaning solution, with pH of the solution ad-
justed to 7.5 to 8.0 wijh sulfuric acid at a
temperature of 40 to 60 C;
F. The membrane cleaning is performed three times a
week, or at an interval of 815 liters of product
water per square meter of membrane area (20 gallons
per square foot of membrane area);
G. The membrane life is one year;
H. The capital cost is amortized for 20 years at 5 per-
cent interest rate; and
I. The estimate is based on August, 1973 material and
construction costs.
33
-------
The initial capital cost including the feed pumps, mem-
branes, pH controllers, chlorinators, chemical feed pumps,
booster pumps, brine recirculation pumps, and a post treatment
system for final pH adjustment is about 3.66 million dollars for
a 37,850 cu m/day (10 MGD) spiral-wound reverse osmosis plant.
The membrane cost alone is about 1.15 million dollars.
Since the membrane has only one year of useful life, the
membrane replacement cost is estimated to be about 8.24/1,000
liters (31.54/1,000 gallons). However, if the membrane life can
be improved to two years, then the membrane replacement cost can
be substantially reduced to 4.04/1,000 liters (15.44/1,000
gal 1ons ).
The annual maintenance material cost is based on 5 per-
cent of the capital cost, excluding the cost of membranes. The
labor requirements include:
A. One man-hour per cleaning schedule for a 378.5 cu
m/day (0.1 MGD) section of the plant with a labor
rate of $10,000 per year; and
B. Three man-years for operating the 37,850 cu m/day
(10 MGD) plant at the same labor rate of $10,000
per year.
The total power cost (14/kwh) for the 37,850 cu m/day (10
MGD) plant operation is estimated to be about 2.04/1,000 liters
(7.84/1,000 gallons). The unit costs for the various chemicals
used in the reverse osmosis process are estimated as follows:
A. EDTA = $1.21/Kg ($0.55/lb);
B. Triton X-100 non-ionic detergent = $0.84/Kg ($0.38/lb);
C. Carboxy methyl cellulose - $0.97/Kg ($0.44/lb);
D. Sulfuric acid » $0.04/Kg ($0.02/lb); and
E. Chlorine = $0.09/Kg ($0.040/lb).
According to the above chemical costs, the total expenses
for the process operation will amount to 2.444/1,000 liters
(9.14/1,000 gallons). This total chemical cost can be broken
down into the following three different categories:,
A. Membrane cleaning - 1.54/1,000 liters (5.64/1,000
ga11ons ) ;
34
-------
B. Acidification for chemical precipitation control -
0.8^/1,000 liters (3.0c/l,000 gallons); and
C. Chlori nation for biological growth control =
O.U/1,000 liters (0.5c/l,000 gallons).
Table 4 summarizes the various items of the process cost
estimates. As indicated in Table 4, the total process cost is
approximately 1 6.6C/1,000 1iters (63.6/1,000 gallons) for one
year membrane life. The cost can be reduced to about 12.4c/
1,000 liters (47.5^/1,000 gallons) by improving the membrane
life to two years. Both cost estimates do not include the costs
for the sand filtration pretreatment and the brine disposal.
This pilot plant study has shown that the secondary efflu-
ent from an activated sludge plant can be successfully de-
mineralized by a spiral-wound reverse osmosis process with only
sand fi 1 tration pretreatment, instead of the carbon adsorption
pretreatment. ('*) The difference in the total process costs, in-
cluding the pretreatment costs, between the sand filtration
(IcVl.OOO liters or 4
-------
TABLE 4
PROCESS COST ESTIMATE FOR 37,850 CU M/DAY (10 MGD) SPIRAL-WOUND
REVERSE OSMOSIS PLANT
Amortization of Capital c/1 .000 gallons c/1 ,000 1iters
$3.66 X 10G;
20 years @ 5% 8.8 2.3
Operation and Maintenance
Chemicals (H2S04, C12
and cleaning agent) 9.1 2.4
Membrane Replacement
One-year membrane life 31.5 8.2
Two-year membrane life 15.4 4.0
MaintenanceMaterials 3.4 0.9
Power 7.8 2.0
Labor 3.0 0.8
Total Process Cost:
One-year membrane life 63.6 16.6
Two-year membrane life 47.5 12.4
36
-------
REFERENCES
1. Cruver, James E., Beckman, James E., and Bevege, Eleanor,
"Water Renovation of Municipal Effluents by Reverse
Osmosis." Final Report to the Office of Research and
Monitoring, Environmental Protection Agency. Project
#EPA 17040 EOR Contract #14-12-831 (February, 1972).
2. Dryden, Franklin D., "Mineral Removal by Ion Exchange,
Reverse Osmosis, and Electrodialysis. " Presented at
workshop on Wastewater Reclamation and Reuse,
South Lake Tahoe, California (June, 1970).
3. Cruver, James E., Gulf Environmental Systems Company,
San Diego, California. Private communication.
4. 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. Environmental Protection
Agency. Contract No. 14-12-150.
37
-------
TECHNICAL REPORT DATA
read Infirm lions on the rci inr hclnri- t ii>/if!ftint;
13. RECIPI ENT'S ACCE SSI Of* NO.
EPA-600/2-77-169
•1 II i Li. AIMU SDH 1 I I Lt
Demineralization of Sand-Filtered Secondary Effluent
by Spiral-Wound Reverse Osmosis Process
All I HOXi;,)
Ching-lin Chen and Robert P. Miele
LJ PLHI UHMING ORGANIZATION NAMt AND ADDRESS
County Sanitation Districts of Los Angeles County
Whittier, California 90607
5 REPORT DATE
September 1977( Issuinq Date]
6. PERFORMING ORGANIZATION CODE
!. PERFORMING ORGANIZATION REPORT NO
1O. °ROGRAM ELEMENT NO.
IBC611
'11. CONTRACT GRANT NO.
14-12-150
SPONSORING AGS NCY NAME ANO ADDRtSS
Municipal Environmental Research Laboratory--Cin.
Office of Research & Development
U.S. Environmental Protection Agency
Cinti, OH 45268
13. TYPE OF REPORT AND PERIOD COVERED
0H I £1 n aj Report 7/71-6/73
14. SPONSORING AGENCY CODE
EPA/600/14
15. SUPPl t Ml N I AHY NOTES
Project Officer: Irwin J. Kugelman (513-684-7631)
16. ABSTRACT
A 22.7 cu m/day (6,000 gallons/day) spiral-wound reverse osmosis pilot
plant, was operated at the Pomona Advanced Wastewater Treatment Research Facility
on the sand-filtered secondary effluent. The pilot plant study was conducted under
optimum operating conditions based on previous studies. During the first year of
operation, all the system performance parameters, such as salt rejection, water
recovery, and product water flux rate, were only slightly decreased from their
initial values. However, the salt rejection and product water flux rate were
substantially reduced to almost half of their initial values after a two year
operation period. During this same two year period, the water recovery was found
to decline about 15 per cent of its initial value.
A cost estimate for a 37,850 cu m/day (10 MGD) plant for August,1973 cost
figures indicated that for membranes with onjy one-year life the process cost was
about 16.5471,000 liters (63.5<£/1000 gallons"). However, the cost could be sub-
stantially reduced to 12.4<£/1,000 liters (47.5c/l,000 gallons) for membranes with
two-year life. Both cost estimates did not include the costs for sand filtration
pretreatment and brine disposal.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
Demineralizing
Desalting
Filtration
Water Reclemation
b.IDENTIFIERS-OPEN ENDED TERMS
Membrane Fouling
Reverse Osmosis
COSATI Held/Group
13B
8. LJIL; rmuunoN STATEMENT
Release to Public
19 SECURITY CLASS r This Report i
Unclassified
21 NO. OF PAGES
46
20 SECURITY CLASS / This
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
38
US GOVtOrtMIPHIBIIIICOfflCE 1977- 7S7-OS6/6S14
------- |