Reverse Osmosis Treatment to Remove
Inorganic Contaminants from
Drinking Water
Charlotte Harbor Water Association, Inc.
Harbour Heights, FL
Prepared for
Environmental Protection Agency, Cincinnati, OH
Dec 87
-------
removals were expected by all KO membranes. The test data confirmed the
expectations: the results from four membrane tests showed excellent removals
of greater than 98 percent. The conclusion is, therefore, that uranium Is
extreaely well removed by RO treatment.
REFERENCES
1. National lucerim Primary Drinking Water Regulations, USEPA, Ofc Wtr
Supply, EPA, Wtr. Prog. Fed. Reg., 40:248 (Dec 24, 1975).
2. Fox, K.R. Removal of Inorganic Contaminants from Drinking Water by
Reverse Osmosis. Unpublished Report, USEPA, ORD, Clnti, Oil (1981).
3. Hindln, E.; Dans ton, G.H.; & Bennett, R.J. Water Reclamation by Reverse
Osmosis Bull. 310 Tech. Ext. Service, Washington State University,
Pullman, WA (Aug. 1968).
A. Johnston, H.K. & Llm, H.S. Removal of Persistent Contaminants from
Municipal Effluents by Reverse Osmosis. Res. Rprt. No. 85. Ontario Min.
Envir. Toronto, Ont. (1978).
5. Mixon, F.O. The Removal of Toxic Metals from Water by Reverse Osmosis.
R&D Progr. Rept. 889. Ofc. Saline Water, DOl, Washington, D.C. (1977).
6. Huxstep, M. R. Inorganic Contaminant Removal from Drinking Water by
Reverse Osmosis. EPA-bOO/2-81-115. WERL., ORD, Cincinnati, OH
(Oct 1981).
7. Standard Test Method of Operating Characteristics ot Reverse Osmosis
Devices, D 4194-82. ASTM, Phil. PA (1982).
8. Methods for Chemical Analysis of Water and Wastes. EPA-600/4-79020.
Environmental and Support Laboratory, ORD, Clnti, OH, March (1979).
9. Cook, M.B. and Schnare, D.W. Amended SDWA Marks New Era in the Water
Industry. Jour. AWWA, 78:8:66 (Aug 1986).
10. Sorg, T.J., Forbes, R.W., and Chambers, D.S. Removal of Radium-226 from
Sarasota County, FL, Drinking Water By Reverse Osmosis. Jour. AWWA.
72:4:239 (Apr 1980).
-------
» 11 *», -•, <¦: '"¦ «-.-». h 'v 14 < '' ' • ¦ G 1
O
o
o
c
o
L
u:
innHHihft«-iimi
-------
PBdd-147780
EPA/600/2-87/109
December 1987
REVERSE OSMOSIS TREATMENT TO REMOVE INORGANIC OONTAMINANTS
FROM DRINKING WATER
by
Martin R. Huxstep
Charlotte Harbor Water Association, Inc.
Harbour Heights, Florida 33950
Thomas J. Sorg
Drinking Water Research Division
Water Engineering Research Laboratory
Cincinnati, Ohio 45268
Cooperative Agreement
No. CR-807358
Project Officer
Thomas J. Sorg
Drinking Water Research Division
Water Engineering Research Laboratory
Cincinnati, Ohio 45268
WATER ENGINEERING RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
-------
TECHNICAL REPORT DATA
(Phsu rttd Inttnicitoiu on the ret tne before eomfisunt)
1. REPORT NO. 2.
EPA/600/2-87/109
4. TITLE AND SUBTITLE
REVERSE OSMOSIS TREAT MEM TO REMOVE IK ORGANIC
CONTAMINANTS FROM DRINKINC WATER
5 REPORT DATE
December 1987
S. PERFORMING ORGANIZATION CODE
7. AUTKOH(S)
Marc in R. Huxstep Thomas J. Sorp,
Charlotte Harbor Water Assn U.S. EPA
3 PERFORMING ORGANIZATION REPORT NO
». PERFORMING ORGANIZATION NAME AND ADDRESS
Charlotte Harbor Water Associations
Harbour Heights, FL 33950
10 PROGRAM ELEMENT NO.
II.contract/grant no.
CR 807358
12. SPONSORING AGENCY NAME AND ADDRESS
Water Engineering Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, OH 45268
13. TYPE OF REPORT AND PERIOD COVERED
14. SPONSORING AGENCY CODE
EPA/6Q0/U
16. SUPPLEMENTARY NOTES
Project Officer: Thomas J. Sorg 513-569-7370
111 ABSTRACT
The purpose of this research project was to determine the removal of inorganic
contaminants from drinking water using several 'state-of-the-art' reverse osmosis
membrane elements. A small 5 KCPD reverse osmosis system was utilized and five
different membrane elements were studied individually with the specific Inorganic
contaminants added to several natural Florida ground waters. Testing of each con-
taminant was conducted for a period of 1 - 13 days during which both operational
and chemical data were collected.
This report presents the results of the tests for the removal capabilities
of various reverse osmosis membfane elements for the following inorganic contami-
nants: fluoride, cadmium, mercury, chromium (III and VI), arsenic (III and IV),
selenium (IV and VI), nitrate, nitrite, lead, uranium, radium, molybdenum and
copper. Removal data were a^so collected on naturally occurring susbtances, i.e.
total hardness, chlorides, total dissolved solids and in. some cases sodium and
calcium. / \
The fine reverse osmosis membrane elements selected for the study were:
(1) Toray SC 3100, (2) Filmtec BW30-4021, (3) Dow low pressure 5K, (4) Dupont
B-9 Model 0440-042. and (5) Hvdranautics P/H 4040 LSY-1FC1.
17. KEY WORDS ANO DOCUMENT ANALYSIS
1. DESCRIPTORS
b.IDENTIFIERS/OPEN ENDEO TERMS
c. COSATI Field/Group
18. DISTRIBUTION STATEMENT
Release to the Public
19 SECURITY CLASS iThJ Reportf
unclassified
21. NO of PAGES
Li
20. SECURITY CLASS (Tha pcgej
unclassified
22 PRICE I /%£¦
CPA Fot» 2220-1 ("•»• *-Ti) pnivioui (dition i» obsolctc
1
-------
DISCLAIMER
"Although che Information described In this document has been funded wholly
or in part by the United States Environmental Protection Agency through assis-
tance agreement number CR-807358 '.o Charlotte Harbor Water Association, Inc.,
It has not be'.'n subjected to the Agency's required peer and administrative
review and therefore does not necessarily reflect the views of the Agency and
no official endorsement should be inferred. Mention of trade names or commer-
cial products does not constitute endorsement or recomnendation for use."
11
-------
FOREWORD
The U.S. Environmental Protection Agency Is charged by Congress with
protecting the Nation's land, air and water systems. Under a mandate of
national environmental laws, the agency strives to formulate and Implement
actions leading to a compatible balance between human activities and the
ability of natural systems to support and nurture life. The Clean Water Act,
the Safe Drinking Water Act, and the Toxic Substances Control Act are three
of the major congressional laws tnat provide the framework for restoring and
maintaining the integrity if our Nation's water, for preserving and enhancing
the wate-r we drink and for protecting the environment from toxic substances.
These Jaws direct the EPA to perform research to define our environmental
problems, measure the impacts and search for solutions.
The Water Engineering Research Laboratory is that component of EPA's
Research and Development program concerned with preventing, treating and
managing municipal and industrial wastewater discharges; establishing prac-
tices to control and remove contaminants from drinking water and prevent
It's deterioration during storage and distribution; and assessing the nature
and controllability of releases of toxic substances to the air, water and
land from manufacturing processes and subsequent uses. This publication Is
one of the products of that research and provides a vital communication link
between the researcher and the user community.
This report presents information on the application of several reverse
osmosis membrane elements to renove inorganic contaminants from drinking
water. The data presented are helpful In solving small community problems in
meeting the inorganic drinking water regulations.
Francis T. Kayo, Director
Water Engineering Research Laboratory
ill
-------
ABSTRACT
The purpose of this research project was to deteniine the removal of
inorganic contaminants from drinking water using several 'state-of-the-art'
reverse osmosis membrane elements. A small 3785 L/d (1000 gpd) reverse os-
mosis system was utilized and five different raenbrano elements were studied
Individually with the specific Inorganic contaminants added to several nat-
ural Florida ground waters. Testing of each contaminant was conducted ror a
period of 1 - 13 days during which both operational and chemical data were
collected.
This report presents the results of the tests for the removal capabil-
ities of various reverse osmosis membrane elements for the following inorgan-
ic contaminants: fluoride, cadmium, mercury, chromium (£11 and VI), arsenic
(III and V), selenium (IV and VI), nitrate, iiitrite, lead, uranlua, radium,
molybdenum and copper. Removal data were also collected on naturally occurr-
ing substances, i.e. total hardness, chlorides, total dissolved solids and in
some cases sodium and calcium.
Reverse osmosis membrane elements selected for the study were as follows:
1. Toray SC 3100
2. Filmtec BW30-4021
3. Dow low pressure 5K
4. Dupont B-9 Model 0440-042
5. Hydranautics P/N 4040 LSY-1KCJ
This report was submitted in fulfillment of Cooperative Agreement
CR-807J58 by the Charlotte Harbor Water Association, Inc. under the sponsor-
ship of the U.S. Environmental Protection Agency. This report covers the
period March 1980 to March 1985 and the work, was completed as of April 1985.
lv
-------
CONTENTS
Foreword Ill
Abstract lv
Figures vl
Tables vll
Abbreviations and Symbols vlll
Acknowledgments lx
1. Introduction
2. Conclusions
3. Reverse Osmosis Pilot Plant System
System components......................
Reverse osmosis test unit
Feedwater chemistry.
Feedwater pre treatment
Contaminant addition
Monitoring Instrumentation
Sampling ports
4. Pilot Plant Experiments and Data Collection
System operation
System performance data
Water sample collection
Chemical analyses. 6
Quality control 8
Test schedule 8
Membrane element operating specifications 8
5. Results 13
Introduction 13
System operation performance ....13
Natural constituents . .......13
Tor ay membrane . 14
Fllmtec membrane ...20
Dow membrane 20
DuPont membrane *....31
Hydranautlcs membrane 31
6. Summary and Discussion 43
General 43
Natural substances 43
Specific concamlnats 45
References......... * 50
v
-------
FIGURES
Number Page
1 Flow Diagram of CHWA 19m^/day reverse osmosis research unit . . 6
2 Removal of TDS with Toray membrane ............... 17
3 Rejection (percent) of TDS with Toray membrane 18
4 Removal of TDS with Fllmtec membrane 22
5 Rejection (percent) of TDS with FilraU'C membrane 23
6 Removal of TDS with Dow membrane 28
7 Rejection (percent) of TDS with Dow membrane 29
8 Removal of TDS with Dupont membrane 35
9 Rejection (percent) of TDS with Duponc membrane 36
10 Removal of TDS with Hydranautlcs membrane 40
11 Rejection (percent) of TDS with Hydranautlcs membrane 41
12 Effect of pH on fluoride removal 47
vi
-------
TABLES
Number Page
1 Typical chemical analysis of feedwater (CHWA potable water) 4
2 List of chemical analyses and analytical methods . 9
1 Contaminant groups 10
4 Toray SC 3100 technical specifications 11
5 Fllmtec BW 30-4021 technical specifications 11
6 Dow RO 5K technical specifications 11
7 Dupont B-9 Model 0440 technical specifications 12
8 Hydranautlcs P/N 4040-LSY-IFC1 technical specifications 12
9 Summary of Toray membrane operational data 15
10 Summary of Toray membrane test data 16
11 Summary of contaminant removal with Toray membrane 19
12 Summary of Fllmtec membrane operational data . 21
13 Summary of Fllmtec membrane test data........................... 24
14 Summary of contaminant removal with Fllmtec membrane 23
15 Summary of Dow membrane operational data 26
16 Summary of Dow membrane test data ......... 27
17 Summary of contaminant removal with Dow membrane.......... 30
18 Summary of Dupont membrane operational data 33
19 Summary of Dupont membrane test data 34
20 Summary of contaminant removal with Dupont membrane 37
21 Summary of Hydranautlcs membrane operational data 38
22 Summary of Hydranautlcs membrane test data 39
23 Summary of contaminant removal with Hydranautlcs membrane 42
24 Summary of reverse osmosis pilot plant tests 44
vll
-------
LIST OF ABBREVIATIONS AND SYMBOLS
Abbreviations
WEKL - Water Engineering Research Laboratory
CHUA - Charlotte Harbor Water Association, Inc.
DWRD - Drinking Water Research Division, USEPA
EMSL - Environmental Monitoring & Support Laboratory, USEPA
EPA - Environmental Protection Agency
EQL ' Environmental Quality Laboratory
gpd - Gal ions per day
Kgpd - Thousand gallons per day
*pm - Gallons per minute
L/s - Liters per second
m(3) - Cubic meters
mg/L - Milligrams per liter
NIPDWR - National Interim Primary Drinking Water Regulations
pCl/L - Picocurles per liter
psig - Pounds per square inch gage
kPa - Kilopascals
RO - Reverse Osmosis
TDS - Total dissolved solids
TU - Total hardness
Symbols
F
-
Fluoride
Cd
-
Cadmium
Hg
-
Mercury
Cr
-
Chr otalum
As
-
Arsenic
Se
-
Selenium
NO3
-
Nitrate
N03(N)
-
Nitrate nitrogen
Pb
-
Lead
U
-
Uranium
Ra
-
Radium
Mo
-
Molybdenum
Cu
-
Copper
NO 2
-
Nitrite
NO2W
-
Nitrite nitrogen
vlll
-------
ACKNOWLEDGEMENTS
Mr. Wil-Ham D. Darby and Mr. Paul L. Bray ton, Charlotte Harbor Water
Association, Inc., were responsible for project management. Alexander Padva,
Ph.D., Ralph Montgomery, Ph.D. and Arnold M. Hartley, Ph.D., Environmental
Quality Laboratory, Port Charlotte, Florida, were responsible for the EQL
chemical analyses and provided extensive consultation on all laboratory
analyses and quality control eftorts.
Special acknowledgement is given to Mr. Ben Mohlenhoff, Mr. Stu McLellan,
Mr. Chip Harris and Mr. William K. Hendershaw of Basic Technologies, Inc. for
providing engineering support as well as contributing both time and effort with
regard to presentation of specifications of available state-of-the-art mem-
brane elements. Mr. Ben Mohlenhoff and Mr. Chip Harris also contributed
their time and efforts to technical problems associated with the servicing of
the RO test system.
The help of Herb Braxton, Mark Guttadauro, and Jia Tenhundfeld, USEPA,
In the processing and tabulation of all the data is also acknowledged.
Ik
-------
SECTION 1
INTRODUCTION
Reverse osmosis (RO) is a relatively new uater treatment process; It has
been applied successfully for desalting brackish water for domestic use for
less than two decades. RO systems typlci'ly operate at 5520 kPa (800 pslg)
for sea water applications (35,000 mg/l Total Dissolved builds (TDS)J and at
2760 kPa (400 pslg) for brackish Jater applications with TDS ranglng from
1,000 to 15,000 mjl.
During the late 1970'S, progress was made in membrane technology wherein
advancements not only occurred with the traditional 2760 kPa (400 pslg), 90
percent TDS rejecting membranes, but even more significantly with the reduced
pressure membranes thjt require approximately 1830 kPa (200 pslg) to achieve
high TDS rejection in excess of 90 percent. These oembranes also operate in
a wider range of feed water pH and thus are capable of increasing applications.
The major advantage, however, is the greatly reduced energy requirements and
therefore significantly lo^er operating costs.
R0 is effective for the removal of most dissolved solids; specific
removal in most cases is dependent upon the weight, size, and valence of the
Ionic specie. Extensive studies have been conducted to ascertain the efficacy
of RO to reject the common water constituents such as sodium, chloride, sul-
fate, TDS, calcium, etc., however, very limited experimentation has been per-
formed to evaluate the effeetive'ess of R0 to remove from drinking water many
of the heavy metals and other inorganic contaminants listed in the National
Interim Primary Drinking Water Regulations (NIPDWR) (1). The investigations
have generally consisted of laboratory studies and most results have not been
verified on either a pilot plant or full scale level (2-5).
The objective of this research project was to determine the rejection of
the inorganic contaminants listed in the NIPDWR using several state-of-the-art
R0 membrane elements. Limited tests were also conducted with several contam-
inants also being considered for regulation In the future. Because of various
problems associated with the specific chemistry of the raw water, some of the
contaminants were not Investigated. This project was a continuation of a
similar project that was reported by Huxstep (6).
This final report describes the R0 test system and components, experi-
mental procedures, and results of tests with fluoride, nitrate, arsenic (III
& V), selenium (IV & VI), chromium (III & VI), cadmium, mercury, lead, uran-
ium, radium, molybdenum, copper and nitrite.
1
-------
SECTION 2
CONCLUSIONS
The primary objective of this study was the development of reverse
osmosis treatment data on drinking water contaminants regulated by che USEPA
using several state-of-the-art RO membranes. Using spiked Florida ground
waters, five RO membranes were used In the study with individual tests last-
log from one to 13 days. Operating conditions for each membrane varied
according to the manufacturer's operating specifications. Although the five
RO nembranes were operated under different conditions (pressure, recovery
rate), rejections of the natural substances measured in the test waters and
th» spiked contaminants werj generally in agreasent for all membranes.
Considering the test data from all four membranes as a whole, the contam-
inants (natural and spiked) can be grouped according to removal capability as
fol lows:
Highly removed (above 9 5 percent) - As+5, Ca, Cd, Cr+3, Cr+6, Cu,
Pb, Mo, Na, Ra, Se+4, Se+6, U,
hardness, TDS
Moderately removed (85 - 94 percent) - F, Cl~, NO^, NO2
Poorly removed (below 85 percent) - As+3, Hg(I)
Wide variation In removals occur with four contaminants: As+3, Hg, P,
and NO3. Because nitrite tests were limited to a two day test with one
membrane, no general conclusion on variability for nitrite removal can be made.
The variation in removals of these contaminants occurred among membranes and
within each membrane test. The reason for the variation is concluded to be
the chemistry of the contaminants and water matrix, membrane material, and
test conditions. In the case of meicury, analytical procedures may also have
contributed to the variation in results.
2
-------
SECTION 3
REVERSE OSfOSIS PILOT PLANT SYSTEM
SYSTEM COMPONENTS
The reverse 06taosls pilot plane system was housed In the Charlotte Harbor
Water Association, Inc. (CHUA) water treatment plant facilities located in
Harbour Heights, Florida. The system consisted of a 19 iP/d (5 kgpd) reverse
osmosis module with a high pressure pjmp, a 378.5 L (100 gal) stainless steel
tank with a low pressure pump which acted as a feedwater source, pretreatment
in the form of 5 micrometer filtering, a cooling unit for temperature stabil-
ization, and a disposal line through which spent water was directed to a dis-
posal pond. The RO module and feed water tank occupied an area of approxi-
mately 5.4 (58 sq ft).
After extensive consideration of the primary intent of this project, the
system was altered from a standard flow configuration with no recirculation
to a continuous reelrculation mode of operation by returning both permeate ar.d
concentrate to the feedwater holding tank. Because considerable heat was
generated by this system design, a heat exchange unit was installed. This
system, shown in Figure 1, Is similar to that defined by the Ai>TM "Standard
Test Method for Operating Characteristics of Reverse Osmosis Devices." (7)
The RO test system was obtained from a previous U.S. EPA research project
and refurbished by Basic Technologies, Inc., Riviera Beach, Florida.
Reverse Osmosis Test Unit
The RO test system was a 19 tn? (5kgpd) high pressure 2760 kFa (400 psig)
unit with a single fiberglass reinforced plastic pressure vessel into which a
single 4 inch membrane element could be loaded. Three of the membrane ele-
ments tested were provided by the manufacturer already contained within a
pressure vessel ready for operation. TTte actual permeate rapacity of this
system was dependent upon several factors, the most obvious being Che speci-
fic membrane element being tested atvl the recovery (permeate to feedwater
ratio) at which the element was being operated.
Feedwater Chemistry
Initially, the test water used was the same raw water used by CHUA
having a TDS of 1900-2000 mg/L. However, after having experienced several
problems caused by the relatively high sulfate content (550 mg/L) of this
well water, the decision was made to switch to CHWA finished potable water
wherein the sulfate concentration was considerably lower (£0 mg/L). A typi-
cal chemical analysis of the CHUA finished water is presented in Table 1.
3
-------
Table 1. TYPICAL CHEMICAL ANALYSIS OF FEEDWATER (CHWA POTABLE WATER)
Parameter
Alkaliuity(as CaCOj). 16 mg/L
Calcium . 33 og/L
Chloride 200 ng/L
Conductivity(as ag/L NaCl) 470 ag/L
Fluoride 0.3 ag/L
Magnesium 23 ag/L
Potassium 3.9 mg/L
Silicon 2.6 mg/L
Strontium 7.1 mg/L
Sulfate 80 ag/L
Total Hardness(as CaCOj) 185 ag/L
Total Slllca(S102) 5.7 mg/L
The raw water for the CHWA water is drawn from the upper Hawthorn aqui-
fer located approximately 1.5 miles northeast of the CHWA treatment facili-
ties and pretreated with sodium hexametaphosphate and sulfuric acid before
entering three two-stage reverse osmosis units. The RO product water is
blended with raw water, degasified, chlorinated and stabilized with soda ash
before distribution. At this point, the test water was drawn for the research
project. CHWA finished water was used in almost all cases except for the
radium and uranium tests. The test water for the radium experiments was CHWA
raw water containing natural radium. Well water containing naturally occurr-
ing uranium was obtained for the uranium tests from a small community in
southern Florida.
Feedwater Pretreatment
As shown in Figure 1, the pilot plant test system utilized a recircula-
tion flow pattern with permeate and concentrate flows blended together and
returned to the feedwater holding tank. As a result of this, the water
required an Initial pH adjustment to conform to the operating specifications
of the particular membrane element being tested. The proper pH was accom-
plished by the adding of small amounts of sulfuric acid. Frequently, however,
the target pH was exceeded and soda ash was added In order to compensate.
Throughout the testing period, the pH tended to drift upward and consequently
very small dosages of sulfuric acid were added to maintain the pH goal. IXie
to the low concentrations of the natural chemical constituents the use of a
sequestering agent was not necessary because the solubility products were not
exceeded.
Exceptions to the above procedures occurred with two natural waters,
containing uranium and radium. The CHWA water containing radium was pumped
directly into the feed water holding tank via a tap in the CHWA Influent
piping and then, in order to remove hydrogen sulfide, was degasified through
vigorous recirculation which bypassed the RO module. The well water with
4
-------
natural uranium was collected In a 378.5 L (100 gal) plastic storage tank and
transferred to the feedwater holding tank. Both of the waters were then sub-
jected to arid pretreatnent for pH adjustment.
Initially, the feedwater was filtered through a 5 micrometer cartridge,
upflow filtration unit. Upon spiking the water with mercury, a rapid decline
In feed concentration occurred which was thought to be caused by an adsorp-
tion of the mercury on the filter cartridges. The filter cartridges were
removed from their housing after which tine the feedwater concentrations of
mercury were considerably more stable. Use of the filters was then discon-
tinued for the remaining test period.
Contaminant Addition
Spiking of the test water with the test contaminants was achieved by
weighing out an amount of source material based upon the desired feedwater
concentration, mixing it in distilled water and adding the solution to the
feedwater holding tank. Mixing was accomplished by direcc recirculation
of the test water for 30-45 minutes using the R.0 feed punp (by passing the R0
module).
Monitoring Instrumentation
Process and control instrumentation consisted of continuous monitoring
of feedwater pH, product flow, concentrate flow, feedwater pressure, product
water pressure, and concentrate pressure (Figure 1).
Sampling Ports
Three stapling locations were utilized: one'port each for feed, product
and concentrate waters (Figure 1).
5
-------
I Low Pressure Shu tuff Switch
H High Pressure Shutoff Switch
P Pressure valve
S Sample valve
F Flowmeter
pH_pH meter
PERMEATE
HEAT
EXCHANGER
038m 3
(100 gal)
ST AIM_ESS
Lqj, STEEL
* TATvK
CONCENTRATE
5 MICROMETER
CARTRIDGE
FILTER
BOOSTER
PUMP
ou
SHUTOFF
VALVE
FEED PRESSURE
CONTROL VALVE
®"
©-
©-
HIGH PRESSURE
PUT-P
f
1
1
CONCENTRATE,
FLOW
CONTROL VALVE'
Figure FkMdiagwmof CHWA 19m /day reverse oamoaia research units.
-------
SECTION 4
PILOT PLANT EXPERIMENTS AND DATA COLLECTION
SYSTEM OPERATION
Initially, the test water was spiked with one conramlnant and Che system
was operated continuously for 6 to 8 hours per day during the regular 5 day
work week. Because this schedule required an Inordinate amount of time to
complete a full series of tests with all the inorganic contaminants, the
decision was made to combine 2 or 3 contaminants and to shorten the run tine
to permit 3 to 4 test runs dally. Thus, the time required to study a single
membrane element was significantly reduced. Each test ran approximately 2.5
hours with continuous cooling of the feedwater. Usually, one hour elapsed
between test runs although this time varied according to ambient conditions.
After each contaninant group test run, the feedwater holding tank was
emptied. Vigorous flushing with fresh water of both the holding tank and
the KO module ensued. This water was then pumped to waste and the flushing
procedure repeaced two more times to ensure complete removal of the contami-
nants.
Each membrane element was run according to the manufacturer's specifica-
tions for the cestlng of that particular aeobrane element. This, the
product water flow rates, feedwater pressures and the specific recoveries
differed between membrane elements.
SYSTEM PERFORMANCE DATA
The operation of the RO pilot system was monitored by direct and contin-
uous measurement of feedwater pH and product water and concentrate flows.
Pressure guages Installed on the feed water, product water and concentrate
streams were referenced on an hourly basis during each test run. Performance
data were collected immediately prior to test water sample collection. Due
to feed water temperature fluctuations, the desired system recovery tended
to drift and therefore very frequent fine-tuning of the feed and concentrate
flows was necessary.
WATER SAMPLE 00LIE CIION
Test water samples consisted of feedwater, product water and concentrate;
all three were collected at each sampling In the order of product water, concen-
trate and then feedwater so as not to disturb the system recovery by lowering
the feedwater pressure prior to product water and concentrate sampling.
7
-------
The amount of water sample collected varied according to the analyses to
be performed but was generally around one liter. This one liter sample was
then split to provide for both ln-house analytical work and those analytical
procedures conducted by Environmental Quality Laboratory or the U.S. Environ-
mental Protection Agency. EQL provided wide mouth 250 ml polyethylene bottles
for their water samples and all samples were preserved utilizing the proce-
dures ricanmended by the USEPA (6).
CHEMICAL ANALYSES
Routine ln-house chemical analyses were petfonned immediately following
sample collection and consisted of pH (Corning pH meter 125), TDS (by con-
ductivity, tyron-L meter), total hardness (EDTA titrimetrlc method) and
chlorides (argentometrlc method).
Analyses of the spiked Inorganic contaminants, e.g. fluoride, nitrate,
arsenic, selenium, cadoium, mercury, chromium and lead, were conducted by EQL
and or by the USEPA, using USEPA approved procedures (8). Radium, uranium,
molybdenum, copper and nitrite nitrogen determinations were exclusively
conducted by the USEPA. A list of analyses and analytical methods is shown in
Table 2.
QUALITY CONTROL
Both EQL and U.S. EPA analytical laboratories followed standard labora-
tory QC procedures In conducting analyses of water samples. Quality control
samples were run with unknown samples and each laboratory participated In the
U.S. EPA performance evaluation studies twice per year.
TEST SCHEDULE
At the beginning, only one contaminant was investigated at a time. This
proved too lengthy considering the time available and therefore, the decision
was made to test 2 or 3 contaminants concurrently. This arrangement is de-
tailed in Table 3.
Testing of each contaminant or contaminant group was generally conducted
for a time to permit the collection of 12 or more samples per test run. Oc-
casionally a test was repeated to verify unrealistic or Inconsistent data par-
ticularly In the case of mercury where analytical results were quite varied.
During the last study with the Hydranautlcs membrane, several short term tests
were added. These tests provided limited data of only 4-8 samples.
MEMBRANE ELEMENT OPERATING SPECIFICATIONS
As a general rule the manufacturer's operating specifications were ad-
hered to quite strictly. As a result of this, some variations exist in oper-
ating parameters between the membrane elements tested; a listing of operating
specifications for each element Is shown in Tables 6 to 8.
8
-------
TABLE 2. LIST OF CHKMICAL ANALYSES AND ANALYTICAL METHODS
Laboratory* Parameter Analytical method
CHWA pH Corning pH meter 125
CHWA TDS .'.Myron L TDS meter (conductivity)
CHWA Total hardness Tltriraetric, EDTA
CHWA Chi orlde Argentoraetrlc
EQL Fluoride Potentlooetric, Ion selective
electrode
EQL Nitrate Coloriraetric, automated cadmium
reduction
EQL Arsenic AA, graphite furnace
EQL Selenium AA, graphite furnace
EQL Cadoium AA, flaoe photometric
EQL Mercury AA, manual cold vapor technique
EQL Chromium AA, graphite furnace
EQL Lead AA, graphite furnace
EPA WERL Fluoride Technlcon-Allzarln fluorine blue
EPA WERL Arsenic AA, graphite furnace
EPA WERL Selenium AA, graphite furnace
EPA WERL Cadmium AA, graphite furnace
>0.2 irg/L: AA, flaae photometric
EPA WEKL Mercury AA, manual cold vapor technique
EPA WEKL Chromium AA, graphite furnace
>0.2 mg/L: AA, flame photooetrlc
EPA WERL Lead AA, graphite furnace
EPA WERL .....Uranium Laser Induced fluorometry, EPA
Method 909.2
EPA WEKL Radium Radon em'.natlon technique, EPA
Method '/03.1
EPA WERL Molybdenum AA, graphite furnace
EPA WEKL Copper AA, fiame photometric
EPA WERL Nitrite Color imetrlc, automated cadmium
reduction
* CHWA : On-site laboratory at Charlotte Harbor Water Association
Reverse Osmosis Water Treatment Plant
EQL : Environnental Quality Laboratory, Port Charlotte, Florida
EPA WEKL : United States Environmental Protection Agency, Water
Engineering Research Laboratory, Cincinnati, Ohio
9
-------
TABLE 3. OONTAMINANT CROUPS
Feedwater
Concentration
Group Contaminant (nig/L) Source
1
Fluoride
8-12
sodium fluoride
2
Araenlc(+3)
Selenlum(+4)
1-2
1.5-3
sodium arsenlte
sodium selenlte
3
Arsenic(+5)
Selenium(+6)
Chromium(+6)
1.5-3
1.5-3
1.5-3
sodium arsenate
sodium selenate
sodium dlchrornate
4
Lead
Nltrate(as X)
1-2
15-25
lead nitrate
sodium nitrate
5
Cadmium
Mercury
Chromlu:n(+3)
3-4
0.6
3-4
cadmium chloride
mercuric chloride
chromic chloride
6
Uranium
natural
nar ural
7
Radium
natural
nftural
8
Molybdenua
3-5
molybdenum trloxide
9
Copper
3-5
copper sulfate
10
Nltrlte(as N)
3-5
sodium nitrite
10
-------
TABLE 4. TORAY SC 3100 TECHNICAL SPECIFICATIONS
Membrane type modified cellulose acetate
Membrane configuration.. spiral wound
Maximum feedwater pressure....*..... 600 pslg
Standard feedwater pressure........... 428 pslg
pH range 4 - 7.5
Maximum feedwater temperature 40s C
Standard feedwater temperature 30' C
Maximum chloride concentration 1 ppts
Maximum feed flow rate.. 11.9 gptn
Maximum recovery . .25Z
TABLE 5. FILrfTEC BW 30-4021 TECHNICAL SPECIFICATIONS
Membrane type non-celluloslc
Membrane configuration spiral wound
Maximum operating pressure ..200 pslg
Recommended initial operating pressure 160 - 180 pslg*
Maximum recommended feed flow rate per element ...........5 gpm
Maximum pressure drop per element 8 pslci
Maximum feedwater turbidity 1 ntu
Maximum feedwater temperature. 30° C
Recommended feedwater pH range 4 - 10
Antltelescoplng device ....bonded to element
Dry weight 4 lbs
Nominal diameter 4 in
* This assumes a feedwater temperature of less than 30° C.
The recommended operating pressure for temperatures of 30s C to
50° C will be approximately 10 - 20 pslg lower.
TABLE 6. 00W R0-5K TECHNICAL SPECIFICATIONS
Membrane type ...cellulose triacetate
Membrane configuration hollow fiber
Maximum operating pressures - feed and brine 450 pslg
Maximum feedwater turbidity.............. I jtu
Maximum feedwater chlorine concentration 1.0 ng/L
Maximum feedwater temperature 30° C
Recommended feedwater pH range . 4 - 7.5
Dimensions - case length 48 In (121.9 cm)
case diameter...........................6.25 in (15.9 cm)
Shipping weight 86 lbs (39.1 kg)
11
-------
TABLE 7. DUFONT B-9 MODEL 0440 TECHNICAL SPECIFICATIONS
Membrane type B-9 arasid
Membrane configuration hollow fiter
Initial product water capacity* 4200 gpd nominal +152, -10X
Salt rejection as shipped* > 90Z
Rated operating pressure 400 pslg
Temperature range 0 - 35° C
pH range, continuous exposure 4 - 11
Minimum brine rate ...3200 gpd
Maximum brine rate 9600 gpd
Shell dimensions - outer diameter 5-1/4 in
inner diameter 4-5/8 in
length 47 in
Shell material filament wound fiberglass epoxy
End plates fiberglass epoxy
Snap rings SAE 1075 carbon steel, cadmium plated
Connections - feed and product 1/2 " female, KPT
brine 3/8 " female, KPT
brine sample 1/8 " female, KPT
Operating position horizontal or vertical
Permeator weight, filled with water 50 lbs
* Based on operation with a feed of 1500 og/L sodium chloride at
400 pslg, 25° C and 75Z conversion, standard test conditions.
TABLE 8. HYDRANAUTICS K)DEL P/N 4040-LSY-IFC1 TECHNICAL SPECIFICATIONS
Membrane type
Membrane configuration spiral wound
Rated initial chloride ion rejection*... average 97.52
minimum 95.0%
Rated initial permeate productivity* 1600 gpd
Maximum feed flow to element 18 gpm
Maximum applied pressure 600 pslg
Minimum concentrate flow @ rated permeate output 4.9 gpo
Maximum operating temperature..... 45s C
Feed pH range 4 - 9
Oxidant tolerance 0.0 ppm
* Above ratings are based on a test solution of 2000 ppm sodium
chloride at a temperature of 25° C. Under an applied pressure of
270 pslg, a water recovery of 10Z and a pH of 5 - 6.
12
-------
SECTION 5
RESULTS
INTRODUCTION
Pilot studies for the removal of the Inorganic contaminants were conduct-
ed with five different RO modules. The modules are considered state-of-the-
art membranes with all being practically applied to treat drinking waters.
The primary objective of the study was to obtain RO rejection data for most
of the EPA regulated Inorganic contaminants and not to compare one membrane
against another. Initially this report was planned to be written on a contam-
inant basis, but because of the rather significant variation in operating
conditions, a decision was made to present the data by Individual membrane.
Although membrane comparison will naturally be made, the reader should
be aware that the membranes were not operated under similar conditions and
that caparison of rejection values between the membranes Is not totally
valid. The purpose of the study therefore was to provide rejection data for
comparison of specific contaminant rejection and comparison of these values
to the rejection of the more comoon naturally occurring substances such as
sodium, chloride, calcium, etc.
System Operation Performance
The parameters used to evaluate RO system performance are pressure,
flow, and water quality. The monitoring Instrumentation provided continuous
readouts of pressures (feed, product and concentrate) and flows (product and
concentrate). The reading from these monitors along with system elapsed
operating time, feed water pH and feedwater temperature were recorded each
time a sec of water samples were collected. Additionally, the TDS of the feed,
product and concentrate water samples was measured and recorded.
The system performance information for each membrane element Is pre-
sented in the discussion of each membrane. Each membrane element was operat-
ed according to the manufacturer's specifications and, therefore, significant
variation exists between the operating pressures, flows and system recoveries
of the membranes.
Natural Constituents
The typical method for evaluating a reverse osmosis membrane Is to deter-
mine the desalinating capacity, i.e. the ability to reject salts as measured
13
-------
by «he difference In TDS in the feed and product-screams. Concurrent with
Inorganic contaminant testing, analyses for several of the common natural
constituents in the feedwater were performed to'establish baseline system
performance criteria by which any problems ,C£uld be detected as evidenced by
a decline in rejection capacity. This testing consisted of TDS, total hard-
ness, and chloride Ion determinations for all membrane elements. Sodium and
calcium were also analyzed during the testing of the Dupont and Hydranautlcs
elements. A review of these data reveals no major differences in rejection
efficiencies for any of the membrane elements tested with percent rejections
for all elements varying as follows: TDS (93.8 - 97.6), total hardness (97.7 -
99.3) and chloride lo-> (91.4 - 94.5). For the Dupont and Hydranautlcs elements
only, the sodium (96 percent) and calcium (98-99 percent) removals were also
reported.
The removal data for the natural substances showed In some c&^es a
decline in removal with time. Although che decreases were relatively small,
2-5 percent, these changes were observed and noted. The most significant
change occurred in the early stages of the Fllmtec test program when TDS
rejection declined from about 97 percent to about 87 percent and returned
to the original 97 percent.
TORAY MEMBRANE
The first series of tests were conducted with a Toray membrane whose
general characteristics are given in Table 4. The system was run for 104
days (620 hrs) at an average feed water pressure of 1960 kPa (284 psl^).
During the test period of day 1 to day 57, CHWA raw well water (TDS 2000
mg/L) was used as the test water. Because of the high sulfate and some
precipitation problems, Che test water source was changed to CHUA finished
water starting on test, day 58. This change was the reason for the decrease
in TDS of the feed water from around 2000 mg/L for the first 57 days to 500 -
700 mg/L for the remaining tests.
A summary of the operational data collected is shown in Table 9. A
summary of the removal results of the natural occurring substances that were
measured, TDS, chloride and hardness, are also shown in Table 10 and the TDS
data Is plotted in Figures 2 and 3. For the 104 day test period, removals
averaged 95 percent for TDS, 98 percent for hardness and 93 percent for
chloride.
A summary of the removal values for the spiked contaminants is shown
In Table 11. Because of either testing problems or analytical problems,
removal data is lacking for Cr+3, Hg, and U.
The test data shows that best removals (97-99 percenc) were achieved
on cadmium, selenium >4 and +6, arsenic -*-5, lead and chromium +6. Lower
removals (44 - 94 percent) were achieved on fluoride, nitrate and arsenic
III. The low removals for arsenic III were verified by repeating the tests
several times. A wide variation in removals (44-79 percent) was observed
for arsenic +3. Partial oxidation of arsenic *-3 to arsenic +5 may have
been the reason for this variation.
14
-------
TABLE 9. SUMMARY OK TOKAY MKMBRANK OPERATIONAL DATA
RUN DAYS 1-13 16-13
FEEO WATER SOURCE A A
CONTAMINANTS P NO 3
SAMPLES/READINGS 2b 35
34-35
A
As+1
36-57
A
Cd
Hg
Cr+3
46
67-71
B
As+3
Se+4
10
72-77
B
As+3
II
76-89 92-97
B C
Pb U
20
14
100-104
8
Cr+6
As+5
Se+6
11
AVG
DAYS
1-104
HEKUWATER pH (units)
—Average 5.8 5.8 5.8 5.6 5.6 5.6 5.7
—Minimum 5.5 5.4 5.8 5.1 5.4 5.4 5.2
•-Maximum 6.3 6.3 5.9 6.0 6.0 6.0 6.1
5.7
5.4
6.1
5.7
5.4
6.1
FKEUWATER TEMP (C*)
—Average 39 38
--Minimum 32 35
--Maximum 46 47
37
24
45
32
27
39
32
25
38
34
28
39
32
20
41
33
23
41
35
27
42
PEEUWATER PRESSURE (PSIC)
—Average 261 260
—Minimum 250 255
• -Maximum 285 270
271
250
295
237
275
295
295
290
300
292
275
300
299
280
320
309
290
350
284
271
302
FEEUWATER PLOW (CPM)
--Averago 6.9 6.8
—Minimum 6.3 6.6
--Maximum 7.0 7.0
6.8
7.1
6.1
6.8
6.6
7.0
6.6
6.5
6.7
6.5
6.3
6.7
6.7
6.5
7.0
6.6
5.9
6.9
6.7
6.5
6.8
RECOVERY (percent)
—Average 10.6 9.5
--Minimum 9.3 8.2
•-Maximum 12.9 11.5
9.7
8.0
11.5
9.6
8.3
11.1
10.0
8.8
11.4
10.3
9.0
J2.4
9.2
7.3
10.9
8.8
6.5
10.2
9.8
8.3
11.5
A - CHUA Kaw Water
B - CIIUA Treated Water
C - FL Ground Water With Natural U
-------
TABLK 10. SUMMARY OK TURAY MEMBRANE TEST UATA
KUN DAYS
1-15
16-33
34-35
J6-57
67-/1
72 -77
78-89
92-97
100 - 104
AVI
SOURCE OF WATER
A
A
A
A
B
B
H
C
B
UAV
CONTMINANTS
V
NO 3
AbO
Cd
A9+-3
As+3
Pb
U
Cr*6
1-11
Se+4
As *-5
Cr+3
Se+6
SAMPLES/READINGS
26
35
4
46
10
11
20
—
11
—
FEEDWA1ER OONC (rag/L)
TUS - AVG
J988
2059
2063
1913
672
877
920
—
438
—
TUS - M1N
1800
1875
1975
1700
525
825
675
410
—
TUS - 11AX
2225
2250
2100
2100
850
950
1150
—
460
HARDNESS -
AVG
609
024
645
625
198
—
285
115
.
HARDNESS -
MIN
58C
550
610
540
180
—
J 70
—
100
—
HARDNESS -
MAX
660
720
660
760
210
—
390
—
120
—
CHLOKI1X1 -
AVG
651
661
680
613
225
279
287
—
176
CHLORIDE -
MIN
600
570
655
550
214
261)
228
—
160
CHLORIDE -
MAX
730
720
690
685
246
300
330
—
200
—
PERCENT REMOVAL
TUS - AVC
95
94
94
94
96
96
97
—
96
95
TUS - MIN
94
94
94
94
95
96
96
—
95
95
TUS - MAX
96
95
94
95
97
97
97
—
97
96
HARDNESS -
AVG
99
98
98
98
98
—
99
99
99
HARDNESS -
MIN
98
97
98
98
97
—
99
—
99
98
HARDNESS -
MAX
99
99
98
99
90
--
99
—
99
99
CHLORIDE -
AVG
93
93
93
93
92
94
94
92
93
CHLORIDE -
MIN
92
92
92
92
90
93
93
—
89
92
CHLORIDE -
MAX
94
94
93
94
93
94
94
—
94
94
A - CI1HA Raw Water
B - CHWA Treated Hater
C - KL Ground Water with Natural U
-------
oooo
i
fZ
Si
REJECT WATER
wto-.
^ c
8 too
CO
e
FEED WATER
PRODUCT WATER
fan
I I—I—I—I—l—i—l—I—I—I—I—I—I—i—i—I—i—i—i—i
0 5 10 15202530354045 50 55 60 65 70 75 80 83 90 95100105
RUNDAYS
Figure 2. Removal of TDS with Toray membrane.
-------
100
95
90
85
80
75
70
65
60
55
50
kbrftfWwai,S^gpssssp
-i—i—i—i—i—i—i—j—i—i—i—i—i—i—i—i—i—i—i—i—i
5 tO 15202530354045 50 5560657075 80 8590 95tOO%&
RUNDAYS
gure 3. Rejection (percent) of TDS with Toray membrane.
-------
TABLE ] 1. SUMMARY OF CONTAMINANT REMOVAL WITH TORAY MEMBRANE
Run Saaples Feedwater Concentratlon-ng/L Percent Rejection
/s Contaminant No. Mln Max Avg Mln
1-13
F
26
3.0
10.0
6.1
73
94
90
16-32
N03(N)
35
1.7
25.3
11.8
35
82
69
34-35
As(+3)
4
0.03
0.34
0.14
58
70
63
36-57
Cd
46
0.02
0.54
0.23
95
99
99
Ug
0
-
-
-
-
-
-
Cr(+3)
0
-
-
—
•
•
—
67-71
As(+3)
10
0.03
0.68
0.30
44
79
66
Se(+4)
10
0.12
0.74
0.33
96
99
97
72-77
As(+3)
11
0.15
0.68
0.30
46
76
64
78-89
Pb
12
0.24
1.3
0.55
97
99
98
92-97
U
—
—
—
—
—
--
LOO-104
Cr(+6)
6
0.31
0.96
0.60
97
98
97
As(+5)
12
0.12
0.74
0.35
97
>99
99
Se(+6)
12
0.26
1.0
0.61
99
>99
>99
19
-------
FILMTEC (CMBRANE
The second test series was conducted with a Fllmtec membrane whose
description Is given In Table 5. The test period lasted for 74 days (929
hrs) and the feed water pressure averaged 1318 kl"a (191 pslg). Thus, the
operating pressure averaged about 950 kPa (100 pslg) less than the average
for the To ray membrane test.
A summary of the operational data for a 74 day test period is shown In
Table 12. The rejection results of the natural substances measured (TDS,
chloride, hardness) are given In Table 13 and the rejection results of TDS
are plotted in Figures 4 and 5. Removal averaged 95 percent for TDS, 98
percent for hardness and 92 percent for chloi'lde.
The TDS data in Figure 5 shows a steady decrease in TDS rejection from
about day 9 (97%) through day 37 (84Z) and then a return to the initial
rejection value (97%) day 39. The reason for this decrease is not known, but
this decline suggests some type of an operation problem.
A summary of the rejection data for the spiked contaminants and along
with natural uranium Is shown In Table 14. The results were somewhat similar
to the Toray membrane results with highest removals (95 - 982) achieved on
arsenic +5, selenium +4 and +6, chromium +3 and +6, lead, cadmium, and uran-
ium. Lower removals were obtained with fluoride, nitrate, arsenic +3 and
mercury. The widest variation between minimum and maximum removals were
experienced with arsenic +3 as had also occurred with the Toray tests.
Oxidation of some arsenic +3 to arsenic +5 is again suggested as a possible
cause for this wide variation in removals.
DOW MEMBRANE
The Dow membrane was the third membrane tested. The test period lasted
72 days (760 hrs) and the average feed pressure was 1911 kPa (277 pslg).
A summary of the operational data collected Is shown in Table 15. This
membrane had the highest percent recovery (55 - 60) of the five membranes
tested. The rejection data from the natural elements (TDS, chloride, hard-
ness) are shown in Table 16 and the rejection data for TDS is plotted in
Figures 6 and 7. TDS rejection averaged 96 percent through the 72 days.
Figure 7 shows, however, that for the period, day 13 - 25, that TDS rejection
decreased from about 97 percent to 94 percent and then returned to around 96
percent for the duration of the test period. Why this slight decrease occur-
red is not known. Removals for hardness was 98 percent and for chloride, 93
percent.
A summary of removal of spiked contaminants and for uranium and radium
is shown in Table 17. The pattern of removals was similar to that of the
first two membranes. Best removals (95 - 99 percent) were achieved on lead,
cadmium, chromium +3 and +5, arsenic +5, selenium +4 and +6, uranium and
radium. Lower removals were achieved with fluoride, nitrate, arsenic +3 and
mercury. Some questions exist on mercury removals because of analytical
problems and adsorption within the system. However, two different test
20
-------
TABLE 12. SUMMARY OF PILMTKC MEMBRANE OPERATIONAL DATA
RUN DAYS 1-10
SOURCE Of WATER B
CONTAMINANTS P
SAMPLES/READINGS 22
PEEDWATER pH (UNITS)
-AVG 6.8
-MIN 5.7
-MAX 7.1
PEEDWATEK TEMP (C )
-AVC 27
-MIN 17
-MAX 36
PEEDWATER PRESSURE (PSIti)
-AVG 180
-MIN 165
-MAX 210
PEEDWATER n.OW (GPM)
-AVC 3.7
-MIN 3.5
-MAX 1.8
RECOVERY
-------
tOOOi
REJECT WATER
100-
10
FEED WATER
da?PQqJHPPo
~ rnUl
¦«/*¦ »SSu ,F
PRODUCT WATER ~ n rfff "
~3 ~
m
t 1 1 1 1 1 1 r^h 1 1 r-om—i 1
0 5 *> 15 202530354045505560657075
RLJNDAYS
Figure 4. Removal of TDS with Fiimtec membrane.
-------
100
95
SO
85
BO
75
70
65
60
56
50
Q
%£°aMPa,0'
o
—I 1 1 1 1 I—I 1 1 1 1 1 1 1 1
5 10 6202530354045 50 55 60 65 70 75
RUNDAYS
ire 5. Rejection (percent) of TDS with Fllmtec membrane.
-------
TABLE 13. SUMMARY OF KILMTKC MEMBRANE TEST DATA
RUN DAYS
1-10
11-20
21-37
38-49
50-62
63-66
67-70
71-74
AVC
DAYS
PEED WATER SOURCE
B
B
B
B
B
B
B
C
—
CONTAMINANTS
f
As+3
N03
Aa+5
Cd
N03
Cd
U
1-74
Se+4
Pb
Se+6
»g
Cr+3
Cr+6
Cr+3
—
SAMPLES/HEADINGS
22
21
32
16
21
8
9
7
~
FEED CONC. (ng/L)
669
648
810
664
702
393
420
420
IDS - AVC
600
575
500
600
650
380
410
400
TDS - MIN
800
700
900
725
775
420
735
470
—
T US - MAX
HARDNESS-AVC
327
250
155
155
148
91
102
145
,
MIN
310
230
130
140
130
65
85
130
MAX
340
260
260
180
160 .
100
110
170
CHLORIDE-AVC
224
324
293
234
223
164
185
133
MIN
200
280
230
225
200
160
175
125
—
MAX
255
360
320
245
240
170
200
140
PERCENT REMOVAL
TDS - AVG
97
95
88
96
98
98
95
95
95
TDS - MIN
96
92
85
84
97
97
94
92
92
TDS - MAX
98
96
93
99
98
98
96
96
97
HARDNESS - AVG
99
98
94
99
99
99
99
99
98
MIN
98
97
86
94
99
99
99
99
96
MAX
99
99
98
99
99
99
99
99
99
CHLORIDE - AVC
93
91
88
93
94
94
92
89
92
MAX
90
87
85
81
92
93
90
85
88
MIN
95
94
91
96
96
96
94
91
94
A - CljWA Raw Water
8 - CHWA Treated Water
C - 1'L Ground Water with Natural U
-------
RUN
DAYS
TABU 14.
SUMMARY OF
CONTAMINANT
removal
WITH PILMTEC
MEMDRAlfE
ction
CONTAfflNA-ST
SAMPLES
(NO.)
FEEDUATER
CONCENTRATION - or/L
Percent Relei
Ml n
Max
Avr
MI n
Max
Avr
1-10
f
22
8.4
10.2
8.9
72
92
83
11-20
Aa(+3)
7
0.04
0.18
0.10
55
83
69
Se(+4)*
21
0.02
0.08
0.04
>83
>96
—
21-37
N03(N)
20
12.8
14.3
13.7
71
73
75
Pb
32
0.04
0.13
0.07
65
94
89
39-49
Aa(+5)
5
0.10
0.47
0.26
98
>99
99
Se(+6)
16
0.58
2.6
1.2
96
>59
99
Cr(+6)
9
0.04
1.3
0.73
87
>99
97
50-62
Cd
11
0.28
0.36
0.32
>99
>99
>99
Hg
10
0.002
0.109
0.040
60
89
78
Cr<+3)
0
—
—
—
—
--
63-66
N03CN)
0
--
—
—
—
—
--
Pb
8
0.19
1.32
0.41
78
>99
97
67-70
Cd
Ui*
9
A
2.5
2.6
2.6
99
>99
99
Cr(+3)
u
9
0.05
0.29
0.12
94
98
96
71-74
U
7
0.533
0.879
0.682
99
99
99
•Product water concentrations all less than detectable Holt of 0.005 mg/L
25
-------
TAHLK 15. SUHMAKY OK OUW KKrtHKANK UPKWAriONAL DATA
RUN MAYS 1-17
SOURCE OF WATER B
CONTAMINANTS F
18-26
B
NO 3
Pb
27-J 4
B
Cd
I'R
Cr+3
34-41
U
As+5
Se(+b)
Crf6
42-50
B
A9+3
Se+4
5J-63
C
64-67
B
"g
68-69
tl
Ra
70-71
B
»g
72-7 J
8
Aa«-3
AVG
DAYS
1-7 J
SAMPLKS/READINGS 29
19
21
16
24
22
FEEO WATER
pH (UNITS)
--AVU
—MIN
--MAX.
6.3
4.0
7.4
6.2
5.6
6.4
6.0
5.5
6.5
5.6
S.O
6.4
4.9
4.5
5.1
7.6
7.2
6.0
6.7
6.6
6.8
7.5
7.4
7.5
6.3
6.2
6.3
6.1
6.1
6.1
6.3
5.8
6.7
FEED WATER
TKNP (C)
—AVIJ
—MIN
- -MAX
23
17
27
25
23
27
25
20
30
25
20
35
26
21
28
25
22
35
34
31
36
19
16
20
19
19
19
19
19
19
24
21
28
FEED rfATER
PRESSURE (PSIC)
—AVG 261 261 266 262 263 272 253 295 320 320 277
—MIN 250 250 240 225 245 230 240 290 320 320 261
- -MAX 27 5 270 290 28 5 305 28 0 260 300 320 320 290
FEEO WATER
FLOW (GHM)
—AVG
—MIN
--MAX
6.6
6.2
7.0
6.9
6.8
7.0
5.8
5.6
6.0
5.9
5.7
6.2
5.7
5.3
6.6
6.1
5.9
6.3
6.9
6.8
7.0
6.0
5.9
6.0
6.8
6.8
6.8
6.8
6.8
6.8
6.3
6.2
6.6
RECOVERY
—AVG
—MIN
--MAX
<*)
55.3
53.7
59.6
56.2
54.4
57.9
61.8
58.6
65.0
60.5
57.8
64.9
64.1
58.7
65.4
60.4
58.3
65.0
61.B
59.4
66.1
55.4
54.2
56.6
55.8
55.8
55.8
55.5
55.5
55.5
58.7
56.6
61.2
A - CIIWA Raw Water
B - CHWA Treated Water
C - KL Ground Water with Natural U
-------
TABLE
16. SUMMARY OK DOW
tKMBKANK
TEST
DATA
RUN DAYS
1-17
18-2 b
27-34
34-41
42-50
51-63
64-67
68-69
70-7J
•n
1
N
AVC
DA *5
1-7:
SOURCE UF WATER
B
B
B
B
B
C
B
A
B
B
CONTAMINANTS
f
NO3
Cd
As+5
Ab + 3
U
Ra
Hg
As+3
„
--
Pb
"g
Se(+6)
Se+4
—
...
—•
—
Cr+3
Crf6
--
--
—
SAMPLES/HEADINGS
29
19
21
16
24
22
6
2
2
2
—
FEED WATER CUNC
(mg/L)
TUS - AVC
627
6.1
630
625
702
443
625
1763
800
800
TDS - MIN
550
575
620
600
650
350
625
1750
800
800
TUS - MAX
750
775
640
650
750
600
625
1775
800
800
—
HARDNESS - AVC
J 39
126
178
161
172
160
148
480
209
205
__
HARDNESS - MIN
120
120
165
150
150
100
140
480
208
200
HARDNESS • MAX
160
150
185
180
200
270
150
480
210
210
--
CHLORIDE - AVU
219
203
233
225
263
131
235
535
250
225
.
CHLORIDE - MIN
205
185
220
215
240
J25
235
530
250
225
CHLORIDE - MAX
225
220
245
233
275
135
235
540
250
225
¦—
PERCENT REMOVAL
TOS - AVC
97
95
97
97
96
97
95
96
97
97
96
TOS - MIN
95
94
96
96
96
96
95
96
97
97
96
rus - MAX
98
96
97
97
96
97
95
95
97
97
96
HARDNESS - AVU
98
96
96
98
98
98
97
98
99
99
98
HARDNESS - MIN
96
94
96
98
97
98
97
98
99
99
97
HARDNESS - MAX
99
98
98
99
98
99
97
98
98
99
98
CHLORIDE - AVU
94
9J
93
93
94
91
92
94
94
94
93
CHLORIDE - MIN
93
92
92
91
92
90
92
94
94
94
92
CHLORIDE - MAX
95
94
94
95
94
92
93
95
94
94
94
A - CIIWA Raw Water
B - CIIWA Treated Water
- t'L Cround Water wl tli Natural U
-------
10000
fO
00
c
J2
c
8
tn
&
tooo
09
REJECT WATER
oocxxxcAxoP^bccx^
d»
too
FEED WATER
%
J3^
O
~r
5
PRODUCT WATER
T 1 1 1 1 1 1 1 1 1 1 1
10 15 20 253035404550556065
RUNDAYS
Figure 6. Removal of TDS with Dow membrane.
-------
100
95
90
85
80
75
70
65
60
55
50
aoooccfcoooocm
1 1 1 1 1 1 1 1 I 1 1 1 1
5 10 15 20253035404550556065
RUNDAYS
gure 7. Rejection (percent) of TDS with Dow membrane.
-------
TABLE 17. SUMMARY OF CONTAMINANT REMOVAL WITH DOW MEMBRANE
RUN
DAYS
CONTAMINANT
SAMPLE
SOS.
FiEDVATER
CONCENTRATION - mrf/L
PERCENT
'ECTION
Mia
Max
Avg
Mln
!"JX
Avg
1-17
F
29
5.5
l't.5
8.4
56
97
91
18-26
no3(n)
17
12.0
41.4
28.0
82
36
85
Pb
18
0.09
1.0
0.60
94
93
96
27-34
Cd
21
1.1
3.7
2.2
98
93
98
Hg
9
0.508
0.636
0.557
12
17
14
Cr(+3)
21
0.06
1.3
0.49
95
93
97
34-41
As(+5)
16
0.47
1.9
1.1
98
99
98
Se(+6)
16
1.4
3.3
2.1
99
99
98
Cr(+6)
16
1.1
3.6
2.0
95
97
96
42-50
As(+3)
11
0.36
0.41
0.39
97
99
98
Set+4)
11
0.51
0.65
0.56
98
99
99
As(+3)
13
1.1
1.3
1.2
73
73
75
Se(+4)
13
1.4
1.9
1.7
97
99
98
51-f3
U
22
0.330
1.650
0.670
98
99
99
64-67
Hg
6
0.0002
0.010
0.003
52
81
64
68-69
Ra(pCi/L)
1
5.05
5.05
5.05
97
97
97
70-71
Hg
2
0.071
0.081
0.076
10
22
16
72-73
As(+3)
2
0.73
0.85
0.79
82
84
83
30
-------
periods shoved very low removals of 10 - 20 percent, the lowest of all the
rejection values.
DUWNT MEMBKVJE
The fourth membrane evaluated was a Dupont membrane. The test period
was 43 days (327 hrs) and the average feed water pressure was 2650 kPa (384
psig), the highest of all the tests conducted. The percent recovery for this
membrane was about 50 percent.
A summary of the operational data is shown in Table 18. A summary of
the removals of the natural elements (TDS, chloride, hardness, calcium and
sodium) is given in Table 19. Rejection data for TDS for the test period is
also plotted in Figures 8 and 9. Although TDS rejection was high, 96-99
percent, a slight decline was observed during the test period from the ini-
tial high rejection of 99 percent to the 96 percent rejection during the last
few days. Again no reason, except for membrane usage, Is offered to explain
the slight decline. Hardness removals averaged 99 percent and chloride 94
pe rcent.
A summary of the reaoval for tt*.e spiked contaminants and uranium and
radium is shc^n in Table 2'J. For the most part, the same pattern of removal
results exisit-i. Highest removals were achieved on lead, cadnium, chromium
+3 and +6, selenium +¦'. and +6, arseaic +5, uranium and radium. Lower remov-
als were obtained on fluoride, nitrate, arsenic +3 and mercury. The or
major difference existed for mercury where removal ranged from 65 - 98 per-
cent which was significantly higher Chan the 15 - 20 percent removal with the
Dow membrane. Because of some analytical uncertainty and observed adsorption
with the system, some doubts exist oa the validity of all the mercury data.
HYDRANAUTICS MEMBRANE
The last membrane to be studied was the Hydranautics membrane. This
membrane was tested for the shortest period of time, only 29 days (303 hrs).
The feed water pressure averaged 1953 kPa (283 pslg) and percent recovery vas
around 11 percent.
A summary of the operational data is shown in Table 21. A summary of
the removals for the natural substances in the feed water (TDS, chloride,
hardness, calcium, sodium) is also shown in Table 22. The TDS rejection data
for the test period is plotted in Figures 10 and 11. Once again the TDS
rejection data showed a slight decrease with time as the average rejection
went from about 99 percent (days 1-5) to around 95 percent (days 26-29).
General usage again Is suggested as the only explanation for the decrease.
L'lrdness removal averaged 96 percent and chloride 95 percent.
A summary of the removal data for the spiked contaminants and for uranium
and radium is ^hown in Table 23. As reported with all other membranes, high-
est removals (95 - 99 percent) were achieved with lead, cadraiun, chromium
31
-------
+3 and +6, selenium +4 and +6, arsenic +5, uranium and radium. Lower removals
were obtained on arsenic +3. However, for the firsc and only time high
removals were achieved on fluoride (98 percent) and for nitrate (97 percent).
Why these removals vere significantly different from the other membrane
results Is noc known.
For the first time, tests were conducted for the removal of nitrite
copper, and t»lybdemuau Although the tests were short, 2 - 3 days, the re-
moval data (average) showed high removals for all three substances; greater
than 92 percent for nitrite, 97 percent for copper, and greater than 97 per-
cent for molybdemum.
32
-------
TABLE 18. SUMMARY OK UUHJNT MF.MHItANK OPERATIONAL DATA
RUN UAYS 1-5
SOURCE OF WATER B
CONTAMINANTS F
SAMPLES/READINGS 14
FRED WATER pll
(UN ITS)
—AVC 6.2
—M1N 6.1
--MAX 6.6
FEEU WATER TEMP(C)
--AVC 2 3
—M1N 16
- -MAX 33
FKKI) WATER PRESSURE
(PS LC)
- -AVC 379
—M1N 330
- -MAX 400
FEED WATER FLOW
(CPU)
—AVC 4.5
—MIN 4.2
-MAX 4.7
RECOVERY (X)
—AVC 50.0
—MIN 47.6
--MAX 52.1
6-10
B
Cd
MR
Cr+3
14
5.9
5.2
7.5
24
23
27
37 7
350
395
4.5
4.4
4.9
48.8
44.4
50.0
11-16
B
As+5
Se+6
Crf6
17
6.0
5.7
6.2
27
18
35
383
285
400
4.6
4.4
5.2
50.2
48.8
52.1
17-22
B
Pb
15
5.3
5.1
5.4
24
21
32
395
370
400
4.5
4.3
5.1
49.7
47.7
51.1
23-28
B
NO3
13
5.2
5.0
5.4
27
23
29
378
360
400
4.5
5.2
5-0
50.8
50.0
52.3
29-34
B
As+3
Se*4
16
5.8
5.6
6.4
34-35
A
Ita
24
21
27
385
375
395
4.5
4.4
4.6
50.5
50.0
52.1
5.5
5.5
5.5
25
24
25
383
380
385
4.4.
4.4
4.4
50.0
50.0
50.0
36-41
C
U
15
8.8
7.9
9.3
42-43
B
N03
AVC
DAYS
1-43
26
22
29
393
385
405
4.4
4.4
4.4
50.0
50.0
50.0
5.5
5.5
5.5
27
26
27
390
390
390
4.4
4.4
4.4
50.0
50.0
50.0
5.8
5.8
6.0
25
21
30
384
360
395
4.5
4.3
4.6
50.0
48.8
51.0
A - CIIWA Raw Water
H - CIIWA Treated Water
C - IX Cr-Hiiul Water with Natural U
-------
TABLE 19. SUMMARY OF OUFUNT MEMBRANE TEST DATA
KUM JHYS
1-5
6-10
11-16
17-22
23-23
29-34
34-35
36-41
42-4 3
AVG
DAYS
souxce or mm
1
B
8
B
8
8
A
C
8
—
CWfLUU MUTTS
t
Cd
Al+5
Pb
"3
Aa*3
Ka
a
HO,
—
Og
Sa*6
—
SaM
—
—
--
—
Cr»3
Cr*6
—
—
--
--
--
--
—
SARFUS/KEADIWS
14
1*
17
13
13
16
2
15
2
ruo turn cotc
(a«/L)
TOS-AVG
793
946
765
813
810
(67
1700
760
800
—
TDS-KIN
775
925
750
800
800
850
1700
7 SO
800
—
TBS-Mil
800
975
775
825
825
67}
1700
774
800
—
HMONESS-AVG
160
222
160
204
200
215
480
274
200
NAISNESS-HIN
160
210
160
200
190
20}
480
260
200
—
BAXSHESS-iM
16S
240
160
210
205
220
480
285
200
—
aaouoe-Avc
233
289
230
249
247
267
533
149
238
aaoutSB^MN
230
260
220
240
240
255
320
140
23C
—
obuxik-hax
240
305
235
265
255
27}
}45
160
245
_
CALCTDH-AVC
27
38
26
31
14
88
60
__
CALCIUM-KIN
27
37
25
28
—
31
86
57
—
—
CAUaOt-KAI
28
39
28
32
—
33
89
62
—
—
SOOIUM-AVG
It 4
US
109
130
_
113
262
68
_
—
sootm-taii
106
109
103
122
—
105
260
36
—
—
SOOIUH-HAX
120
U6
113
136
—
129
263
76
—
—
Ktaicr KQfOVAL
TU6-AVC
99
98
98
98
98
98
96
97
98
98
TDS-KIN
99
98
97
98
98
97
96
96
98
98
TDS-HAX
99
99
99
98
98
98
97
97
98
98
HUOKESS-AVC
99
99
99
99
99
99
99
99
99
99
BABDNESS-HIN
99
89
99
99
99
99
99
99
99
99
HA8MESS-MU
99
100
99
100
99
100
99
100
100
99
aauuoe-AVG
96
96
95
93
93
94
9}
93
93
94
QOOUOS-NIH
96
96
94
95
92
93
95
92
93
94
aUUUOK-MX
97
97
96
95
95
94
9}
94
93
95
CALC10M-AVC
98
99
98
98
—
99
99
99
—
99
CALCIUft-MIR
98
99
98
98
—
99
99
99
—
99
CAuamMux
98
99
98
98
-—
99
99
99
•—
99
SOOIIM-AVC
98
97
96
96
—
96
95
94
—
96
sooimncM
97
88
95
86
—
94
94
89
92
SOOIUM-tMX
98
98
98
97
—
97
95
96
—•
97
A - CBW Urn Watar
B - OWL Traatad Ifctar
C - ft Crouod Uacar «lth Natural U
34
-------
u
u>
OOOOi
S SCO
£
c
J2
8
CO
e
too
REJECT WATER
0000000000000000c<»00c00c0000000c»® ®000000°°
©B&D&SS©
FEED WATER
~r
5
cnmn
DP
PRODUCT WATER
T~
10
—r-
15
I 1—
20 25
RUNDAYS
30
i
35
l
40
"I
45
Figure 8. Removal of TDS with Dupont membrane.
-------
100
95
90
85
80
75
70
65
60
55
50
OO
oooooooo
1 1 1 1 1 1 1 1 1
5 10 15 20 25 30 35 40 45
RUNDAYS
jure 9. Rejection (percent) of TDS with Dupont membrane.
-------
TABLE 20.
SUMMARY
OF CONTAMINANT REMOVAL
WlTH D(JPONT MEM3RANE
XUN
DAYS
CONTAMINANT
SAMPLES
(NO.)
FEED'.'ATEX
CCNCENTXATIO
>'! - zz/l
PERCE
NT REJECTION
>11N
MAX
AVG
MIN
"AX
AVG
1- 5
F
12
5.2
5.4
5.3
88
96
92
6-10
Cd
14
0.66
1.79
1.22
99
99
99
Hg
14
0.0027
0.064
0.026
65
>98
—
Cr(+3)
14
0.15
0.39
0.26
96
99
99
11-16
As(+5)
17
0.70
1.4
1.03
98
99
99
Se(+6)
17
1.2
2.0
1.6
98
99
99
Cr(+6)
17
1.59
1.94
1.76
98
99
98
17-22
Pb
15
0.12
0.7
0.33
>96
>99
>98
23-28
N03(N)
13
12.4
13.2
12.7
93
95
94
29-34
As(+3)
16
0.38
1.05
0.61
46
84
71
Se(+4)
16
0.37
1.75
0.88
97
99
98
34-35
Ra
2
1.83
2.19
2.01
96
97
96
36-41
U
15
0.103
0.182
0.154
96
99
98
42-43
N03(N)
2
13.5
13.8
13.6
95
95
95
37
-------
TABLW 21. SUMMARY OK IIYIJKANAUTICS MEMBRANE OPERATIONAL IJATA
A VU
HUN DAYS
i-3
4-6
7-9
10- n
14-16
17-20
21-23
24-25
26-27
28-29
OAY
SOURCK OF HATER
B
U
B
e
B
C
B
A
B
B
1-29
CONTAMINANTS
Cd
P
As+5
Ho
Pb
U
A9+3
Ra
Cu
NO^
"8
—
Sef6
—
NO 3
—
Sev4
--
--
—
Cr+3
—
Cr+6
—
—
—
--
S.VMPLKS/RKADINtS
12
12
12
15
12
12
12
6
6
4
—
FEEOWATEK pll
UNITS
—AVG
5.2
5.9
5.8
5.6
5.7
7.5
5.8
6.5
5.9
6.4
6.0
—M1N
5.1
5.6
5.6
5.5
5.4
6.6
5.5
6.2
5.7
6.4
5.8
—.'VAX
5.2
6.2
5.9
5.7
6.6
8.3
5.2
6.7
6.1
6.5
6.3
FEEUWATER
TEMP (C)
—AVC
26
31
32
33
34
33
34
32
31
31
32
—MIN
19
27
26
26
25
27
29
26
28
23
25
—MAX
32
34
40
38
39
40
40
37
34
40
37
FEEUWATER
PRESS (PS10)
—AVG
292
286
268
284
298
294
299
280
275
254
283
—MIN
265
275
260
260
290
265
295
260
270
250
269
- -MAX
315
295
275
300
310
315
305
300
280
260
296
FEEMWATER
FLOW (C'PM)
--AVG
7.J
6.6
6.6
6.3
6.3
6.1
5.9
5.9
5.5
5.1
6.2
—MIN
7.0
6.5
6.4
6.0
6.0
6.0
5.4
5.6
5.5
4.9
5.9
- -MAX
7.3
7.0
6.7
7.1
6.5
6.4
5.1
6.1
5.7
5.2
6.3
RECOVBRY(X>
--AVU
10.0
11.4
10.6
10.3
11.1
10.8
11.)
10.4
11.0
10.4
10.7
—MIN
8.9
10.2
8.9
7.1
9.2
7.8
JO.O
8.4
9.7
8.7
8.9
--HAX
12.2
12.a
12.5
11.6
12.0
12.6
12.6
12.6
11.9
13.2
12.4
A - CHWA Raw Water
B - CHWA Treated Wat«c
C - KL Ground Water with Natural U
-------
TABU! 22. SUMMARY CK UYOHANAUTICS tCMBKANE TEST DATA
RUM OUTS
1-3
4-6
7-9
10-13
14-16
17-20
21-23
24-25
26-27
16-29
AVC
souttce >it u*ren
1
8
•
1
8
C
e
A
8
B
OATS
O0HT.VU.1MT1
Cd
r
U«3
n
U
A»*3
¦l
Cu
HOj
—
us
—
S««6
—
NO]
--
'•
—
--
—
--
Cr+1
—
Cr«6
—
--
—
-•
SAXPUS/HKAS1M3
U
12
12
15
12
12
12
6
6
4
—
KBO WICK 00(C
(«g/L>
-
TCS-AVG
629
747
390
571
685
575
311
1379
518
517
—
TU9HUH
620
680
570
560
590
560
500
1550
510
510
—
TOS-tUX
640
790
600
590
710
590
520
1600
520
52C
—
tUUUXKSS-AVG
178
170
143
140
144
354
;»
315
130
130
—
HAKONESS-NIM
170
16a
iso
140
110
150
105
310
130
130
—
UUUM8SS-4UK
190
180
150
140
155
360
130
520
130
130
*-
CHLOUUS-AVS
236
21}
197
199
196
88
190
334
175
183
—
CHUHLIDC-taK
22}
200
190
190
190
80
175
525
170
170
CHLUUUC-MUC
245
225
205
210
205
95
200
54}
185
193
—
calciun-avc
28
28
25
24
26
95
21
85
21
19
—
CALCIUH-K1N
25
28
22
21
21
89
18
77
19
18
—
CALClUH-tUX
29
29
27
29
33
98
22
88
22
20
*-
SUUtUM-AVC
90
12}
79
88
126
?7
88
230
95
9)
—i
SOOIUH-mN
83
124
69
72
104
25
74
203
94
97
—
SOUlUH-tttlC
98
14)
84
98
141
29
94
238
97
98
—
renojirr kouval
TOS-MC
97
98
97
97
97
97
96
95
95
95
96
TDS-HU
96
98
95
9C
96
97
l>
9}
95
93
96
TOS-NAX
98
99
97
97
97
97
96
95
96
95
97
autmss-tNG
99
99
99
99
99
99
98
98
98
98
99
tUHMKSS-MM
99
99
99
99
98
98
98
98
98
98
98
IIUUMBSS-Ntt
99
99
99
99
99
99
99
98
98
98
99
CHLOMM-AVC
96
97
96
96
96
92
94
93
93
94
95
CKLCBI US-tan
9)
96
95
95
95
91
93
92
93
93
94
aaoain-Muc
97
97
97
97
97
93
96
94
94
95
96
CALCtm-AVC
>98
>98
>98
>98
>98
98
>99
98
97
97
>98
CALCtUN-MIM
798
>98
>98
>98
>98
98
>99
97
97
96
>98
CJUXXOHUI
>98
>98
>98
>98
>98
98
>99
98
97
97
>98
SOOIUM-AVC
99
99
98
98
98
94
95
94
95
92
96
SOOim-MM
99
98
98
98
96
92
94
93
95
92
96
SODIUM-ttUC
99
99
98
99
98
95
96
93
95
93
97
A - Qfltt In Ifccar
• - am Trtmtad Iktit
C • fl Craaad bur with Instil II
39
-------
0000*3
tooo
100
88
ee688@88888888838888g§s 8088
HEUECT^fflER
FEED WATER
~ ~
PDDnDDDDDDDDDD°nDnDODDDn aaD|
a PRODUCT WATER
I i i i i i i i I i i i i i i i i i i i i i I l » I I i i I
012345678 9 tO 111213141516f718t9202122232425262728293O
RUNDAYS
Figure 10. Removal of TDS with Hydronautlcs membrane.
-------
95
90
85
80
75
70
65
60
56
SO
00«000000000©0000000000000000
I I I I I I I I I I I I I I I I I I I I I I I I I I I I 1 I
1 2 3 4 S 6 7 8 9101112151415161718192021222324252627282930
RUNDAYS
Rejection (percent) of TDS with Hydronautlcs membrane.
-------
TABLE 23. SUMMARY OF CONTAMIMANT REMOVAL WITH HYDRANAUTICS
RUN
SAMPLES
FEEDWATER CONCENTRATION-os/L
PERCENT
REJECTION
DAYS
CONTAMINANT
(NO.)
KIN
MAX
AVG
MIN
MAX
AVG
1- 3
Cd
Cr(+3)
12
1.17
1.36
1.31
99
99
99
12
0.86
1.46
1.23
99
99
99
4- 6
F
12
14.0
16.0
14.5
98
ve
98
7- 9
As(+5)
Se(+6)
Cr(+6)
12
12
12
1.3
2.0
4.16
2.0
3.2
5.96
1.7
2.7
4.46
96
99
97
99
99
98
9#
99
98
10-13
Mo
15
1.6
4.3
2.4
88
>98
>97
14-16
NOj(N)
Pb
12
12
18.1
1.7
43.1
4.8
27.1
2.6
96
98
98
99
97
99
17-10
U
12
0.252
0.310
0.277
99
99
99
21-23
As(+3)
Se(+4)
11
12
0.8
1.0
1.1
2.4
0.92
1.5
5
93
75
98
46
95
24-25
Ra(pCl/L)
6
7.86
9.83
8.91
96
98
97
26-27
Cu
6
4.8
5.9
5.1
97
98
97
28-29
KO2
4
4.8
4.8
4.8
90
92
92
42
-------
SECTION 6
SUMMARY AND CONCLUSIONS
GENERAL
A soall recirculating RO pilot plant system was operated with five
different state-of-the-art membranes to determine the rejection values for 13
inorganic contaminants from ground water. Because of various factors, the
operating test conditions were not identical for all the membrane tests. tr>r
example, operating pressures varied from a low average of 1318 kPa (190 pslg)
(Filmtec) to a high average of 2650 kPa (384 psig) (DuPont). Recovery varied
from around 10 percent for the Filmtec, Toray and Hydranautics membranes to
50 - 60 percent for the Dow and DuPont hollo* fiber membranes. Some diffi-
culties also occurred In maintaining constant operating conditions of pressure
and temperature duriry each individual (one-three day) test run. Consequent-
ly, comparison of the performance of the membranes is not considered totally
valid. Nevertheless, the results of the study, as summarized in Table 24,
show a general pattern of removals for the contaminants studied for the five
membranes. Furthermore, the results can provide general guidance for estima-
ting the approximate removals that can be achieved by reverse osraosis treat-
cent. A short summary and general u'scusslon for each contaminant follows.
NATURAL SUBSTANCES
Three different test waters were used eacn having a different background
concentration of the natural substances monitored, TDS, hardness, chloride,
calcium, sodium. For all five membrane tests, TDS, hardness and chloride
analyses were conducted on raw, product, rejects waters. Only during the
last two series of tests for the Pupont and Hydranautics membranes were the
calcium and sodium tests performed in addition to TDS, hardness and chloride.
Because extensive data exists on the remo"al of the natural occurring
substances measured, the primary reason for the monitoring of these substances
was to evaluate the general performance of the membranes during the testing
period.
All of the membranes averaged above 95 percent removal of TDS with some
averaging 98 percent. One jmbrane (Filmtec) showed a noticeable decline in
TDS rejection during the first 40 days of test run from around 98 percent to
85 percent. After the 40th day, TDS rejection returned to the initial level
of around 97-98 percent and remained constant for the last 30 plus days. The
reaoon for the decline is not known, but It suggested that some problem
existed and thus the removal results of specific contaminants tested during
this period may be lower than that achleveable under proper membrane perfor-
mance.
43
-------
TABLE
24. SUMMARY OF R£VERSE
OSMOSIS PILOT
PLANT TESTS
MEMBRANE
INFORMATION
DOW
DUPONT
FILMTEC
HYDRANAUTICS
TORAY
Hate rial
CIA
ARAMID
Non-C
MCA
CA
Conflguration
HF
H?
SU
SW
SU
Model No.
5K
B9 0440-0-42
3W30-4021
P/N4040
SCJ100
CHEMICAL
REMOVAL DATA-Z
Arsenic -*-3
75
71
69
46
65
Arsenic +5
98
99
99
98
99
Cadmium
98
99
99
99
98
Calcium
NA
99
NA
>98
NA
Chloride
93
94
92
95
93
Chromium +3
97
99
99
99
SA
Chromium +6
96
98
97
98
98
Copper
XA
KA
NA
97
NA
Fluoride(pH)
91(6.3)
92(6.2)
83(6.8)
98(5.9)
90(5.8)
Hardness
96
99
98
99
99
Lead
96
>98
97
97
98
Mercury (I)
14
NR
78
NA
NA
Molybdenlum
XA
NA
NA
>97
NA
Nitrate
83
94
75
99
67
N1trite
NA
KA
NA
92
NA
Radium
97
96
NA
97
NA
Seleniua +4
93
98
NR.
95
97
Selenium +6
99
99
98
99
99
Sodium
KA
96
NA
96
NA
TDS
96
96
95
96
95
Uranium
99
98
99
99
NA
TEST CONDITIONS
RUN DAYS
73
43
74
29
104
AVERAGE:
X recovery
59.0
50.0
10.4
10.7
9.8
Feed pressure
277
384
191
283
282
Influent pH
6.3
5.8
6.7
6.0
5.7
Influent temp
24
25
29
32
35
Flow rate (CPM)
6.3
4.5
4.0
6.2
6.7
NA - Not available
NR. - Not reportable
44
-------
All of Che membranes removed above 98 percent of total hardness and
93-95 percent of the chloride. Data from the last two series of tests for
the Dupont and Hydranautics membranes showed average removals of around 98
percent for calcium and 96 percent for sodium. Because calcium is the pri-
mary constitutent of total hardness, calcium removal results should be simi-
lar to hardness removal as was found. All results for the reaoval of Che
naturally occurring substances were consistent with manufacturer's guidelines
for performance.
SPECIFIC CONTAMINANTS
Arsenic
Arsenic ca.i cccur in four oxidation states; however, it is normally
found as an anion in only the trivalent (arsenite) ana pentavalent (arsenate)
forms. Each of the two oxidation states forms several species in natural
waters. The soluble arsenate species are H^AsO^, and HAsO^ with
the most predoainant one (in che pll 4-10 range) being the neutral species
H^AsOj. The soluble arsenite species are four: HjAsO^, l^AsO^-, HAsO^'^and
AsOj . Of these four, che most significant ones are t^AsO^-'* and HAsO^~~.
The arsenic removal data for all the membranes show excellent removals
(greater than 9)5 percent) for arsenic h5 and low and variable removals (20-95
percent) for arsenic +3. The arseni; +3 removals ^averaged between
-------
chromium occurs as a cation Cr+3 an(j hexavaJent chromium as an anion 45 as
either chromate (HCrO^ /CrC^ or dlchromate (Cr,0^ ). Both anion forms
are very soluble In water and ttie formation of each Is pH dependent. The
chromate ion exists In alkaline water and the dlchroaate Ion In acidic water.
The test data showed excellent removals for both chromium +3 and chromi-
um +6. All membranes achieved better than 96 percent removal of both forms
and several membranes averaged 99 percent removal of chromium +3. Therefore,
chromium is easily removed by RO regardless of the form found In the water
source.
Copper
A copper test was added to the last study with tne Hydranautlcs membrane
because EPA has proposed that It, along with several other new inorganic
contaminant, be considered for regulation (9). To provide some data, a two
day (6 samples) test was conducted with the Hydranautlcs membrane. Copper
being a divalent cation similar to cadmium, lead, and calclua, good removals
were expected. The short test period proved this to be true with the average
removal being 97 percent. The conclusion is, therefore, that if copper is
found in the source water it should be easily removed from drinking water by
reverse osmosis.
Fluoride
The Dupont hnginee ring Design manual states that removal of fluoride and
bicarbonate for their B-IO perraeators are pH dependent. Their data show
about 50Z removal at pH 5.5 increasing to about 95 percent removal at about
pH 7,5. The test data f?r the five membranes showed a range of removal
average from 83 to 98 percent. The tests program was not designed to evalu-
ate the effect of pH, but because the feed water pH did occasionally vary,
some variation in pH did happen. Unfortunately, pressure and temperature also
varied making it difficult to determine the effect of pH alone. A review of
the fluoride data obtained showed only the Dow aeobrane data having a wide
range of pH values (4-7) for feed water and these data do indicate a trend of
increasing removals with Increasing pH (Figure 12).
Lead
Lead is a divalent cation aad forms various carbonate and hydroxide
complexes in natural waters. The test data for the five membranes showed
high removals of above 96 percent removal for all cembanes with two of them
averaging 98 percent. The data thus indicated that lead is easily removed
from ground water by reverse osmosis.
Mercury
Mer-ury has three basic oxidation states in aqueous solutions: (1) the
pure metal, tig; (2) the monovalent Ion (mercurous), H+; and (3) the divalent
ion (mercuric), Hg+^. Besides forming the common inorganic salts, mercury
has the capacity to form organic complexes, the most significant being the
very toxir methyl mercury ion, CH^Hg*. In water having a prt above 5, the most
46
-------
100-1
o
s
UJ
~
or
5 ui
O
£5
Q-
PH
Figure 12. Effect of pH on Fluoride Removal
-------
predominant mercury species is metallic mercury, Hg° with a relatively low
solubility. In high chloride concentration waters, the solubility of mercury
increases with the formation of the uncharged complexes of HgCl2 and Hg(0ri>2«
Tests were not designed to evaluate the effect of pH, but because the pH of
the feed water varied to some degree some variation in pH did occur. Unfor-
tunately, pressure and temperature also varied making it difficult to determine
the effect of pH alone.
Another important characteristic of mercury is the tendency of mercury
to adsorb to various materials. In the Initial RO studies when a prefilter
was in line, a decrease in the total mercury content of the water was observed.
After the filter was removed, this decrease was not as great, thereby sugges-
ting that some of the mercury was adsorbing to the filter.
Because of various reasons, data on mercury removal were reported for
only three membranes and this data varied from a low average of 14 percent to
a high of 80 percent. It is difficult to determine the cause of the variabil-
ity, but based upon the results of other contaminants it is unlikely to be
membrane differences.
Molybdenum
A molybdenum test run was added to the study during the last series of
tests with the Hydranautics membrane because molydbenum appeared on the EPA
inorganic list of possible or proposed regulations (9).
Molybdenum Is a transition metal that can e.'ist in oxidation states froa
2~ to 6+. In aqueous solution, molybdenum will occur in various forms depend-
ing on the water composition and the oxidation-reduction potential of the
water. In most natural waters, the most predominant species is MoO^
Although the test data Is limited to one membrane, the result of 97
percent removal suggests that molydbenum is easily removed by RO treatment.
Nitrate
Nitrate (NO-j ) is a common ground water contaminant and RO information
indicates that it is not highly rejected by most membranes. The test data for
the five membranes showed removal averages from 67 to 9V percent.
Unfortunately, the lack of very tight operating conditions prevents
making any firm conclusion regarding specific membrane rejection capability.
The general conclusion is that nitrate is not as highly rejected as most
contaminants with rejections in the 65-90 percent range.
The genera1, literature suggests that some of the newer RO membanes may
have a greatei capability to remove nitrate than the older membrane type.
Again, because of the lack of very tightly controlled conditions, it is
difficult to draw any firm conclusion from this study.
48
-------
Nitrite
Because nitrite (NO^) is proposed for consideration as an EPA regulated
contaalnant, a nitrite test was added to the last study with the I'ydranautics
membrane. The very limited test results (2 days, 4 samples) Indicated good
removal. The two day test results showed a 90-92 percent removal range with
an average removal of 92 percent. The nitrite average of 92 percent was
slightly less than the removal average of 97 percent for nitrate for this
membrane. The operating pressures differed by 1960 It Pa (284 pslg) (average)
for the nitrate test to 1750 kPa (254) pslg (average) for the nitrite study.
Whether this pressure difference is the reason for the difference In removal
Is not known.
Radium
Radium is a divalent cation that has chemical and physical properties
slallar to the elements in the alkaline earth metals group - calcium,
magnesium, barium and strontium. Because of radium's similarity to calcium
and magnesium (hardness elements), removal of radium by RO should be
similar to the removal of these two elements and, of course, total hardness.
Data exist on the removal of radium from ground water by full scale RO
systans (10). For this reason and also because of the complexity of radium
analyses, only one day tests were comple:ed on each membrane. The test
data confirm the full scale system results. All systems removed around
96-97 percent of the naturally occurring radium in the ground water. Fur-
thermore, these results were very similar to the removal values reported
for hardness and calciua. Thus, RO is considered a good method for radium
removal.
Selenium
Selenium is somewhat similar to arsenic in that selenium has several
oxidation states, but only two are predominant in water: selenium +4 and
selenium +6. Moreover, like arsenic, selenium occurs as an anion in water
and thus has acid characteristics.
- -2
Selenium +4 forms two primary species in water, HSeOj and SeOj .
At pH 7, the predominant one is the divalent SeOj . Selenium +6 fonns only
one species, in water, the divalent SeO^ .
The RO test data showed high removals (95-99 percent) for both selenium
+4 aad selenium +6 by all membranes. , Consequently, removals are not valence
dependent: both foras are easily removed by RO and the valence is not
Important.
Uranium
Uranium occurs as an anion cotnplexer in natural water and the species
that predominatein the pH range of 7-10 are likely to be the carbonates
forms, IK^CO}^ anc' ^02(00^)3 . Because of the high ionic charge, high
49
------- |