May 2003
NSF 03/05/EPADWCTR
Environmental Technology
Verification Report
Physical Removal of Particulate
Contaminants in Drinking Water
Polymem
Polymem UF120 S2 Ultrafiltration
Membrane Module
Luxemburg, Wisconsin
Prepared by
NSF International
Under a Cooperative Agreement with
U.S. Environmental Protection Agency
-------
THE ENVIRONMENTAL TECHNOLOGY VERIFICATION
PROGRAM ^
//
©ERA
E1V
U.S. Environmental Protection Agency
NSF International
ETV Joint Verification Statement
TECHNOLOGY TYPE: MEMBRANE FILTRATION USED IN DRINKING WATER
TREATMENT SYSTEMS
APPLICATION: PHYSICAL REMOVAL OF PARTICULATE
CONTAMINANTS IN DRINKING WATER
TECHNOLOGY NAME: POLYMEM UF120 S2 ULTRAFILTRATION MEMBRANE
MODULE
COMPANY: POLYMEM
ADDRESS: ROUTE DE REVEL F-31450 PHONE: 011.33.5.53.71.79.89
FOURQUEVAUX, FRANCE FAX: 011.33.5.62.71.79.80
EMAIL: polymem@wanadoo.fr
The U.S. Environmental Protection Agency (EPA) supports the Environmental Technology Verification
(ETV) Program to facilitate the deployment of innovative or improved environmental technologies
through performance verification and dissemination of information. The goal of the ETV program is to
further environmental protection by substantially accelerating the acceptance and use of improved and
more cost-effective technologies. ETV seeks to achieve this goal by providing high quality, peer reviewed
data on technology performance to those involved in the design, distribution, permitting, purchase, and
use of environmental technologies.
ETV works in partnership with recognized standards and testing organizations, stakeholder groups
(consisting of buyers, vendor organizations, and permitters), and with the full participation of individual
technology developers. The program evaluates the performance of innovative technologies by developing
test plans that are responsive to the needs of stakeholders, conducting field or laboratory tests (as
appropriate), collecting and analyzing data, and preparing peer reviewed reports. All evaluations are
conducted in accordance with rigorous quality assurance protocols to ensure that data of known and
adequate quality are generated and that the results are defensible.
NSF International (NSF) in cooperation with the EPA operates the Drinking Water Systems (DWS)
Center, one of seven ETV Centers under ETV. The DWS Center recently evaluated the performance of an
ultrafiltration membrane used in drinking water treatment system applications. This verification statement
provides a summary of the test results for the Polymem UF120 S2 Ultrafiltration Membrane Module.
Carollo Engineers, P.C., an NSF-qualified field testing organization (FTO), performed the verification
testing. NSF provided technical and quality assurance oversight of the verification testing described in
this ETV report.
03/05/EPADWCTR The accompanying notice is an integral part of this verification statement. May 2003
VS-i
-------
ABSTRACT
Verification testing of the Polymem UF120 S2 Ultrafiltration Membrane Module was conducted over a
46-day period at the Green Bay Water Utility Filtration Plant, Luxemburg, Wisconsin. The ETV testing
described herein was funded in conjunction with a 12-month membrane pilot study funded by the Energy
Center of Wisconsin. The Energy Center of Wisconsin chose to participate because the overall scope of
the ETV testing fit into the scope of the longer, energy focused study. The testing was performed from
March 11, 2002 through April 26, 2002, representing winter/spring conditions when, historically, feed
water quality was most difficult to treat. The feed water was Lake Michigan. Verification testing was
conducted at optimized conditions based on pilot testing conducted during the 12 months proceeding the
verification test period. The testing was performed using a "generic" custom membrane pilot plant
(CMPP) capable of operating with a variety of membrane modules that are housed in pressure vessels.
Therefore, this ETV testing verified the operation of the membrane module itself, not membrane-specific
process equipment. The membrane unit was operated in dead-end mode during two test runs, each at a
constant specific flux of 40 and 30 1/h-m2 (24 and 18 gfd), respectively. Feed water recoveries ranged
from 89-96 percent. The two test runs were operated for approximately 12.5 and 32.7 days, respectively.
The UF module was chemically cleaned using a "proof of concept" effort based on procedures
recommended by the manufacturer. The cleaning procedures were effective in restoring membrane
productivity. The membrane module achieved significant removal of particulate contaminants and
bacteria, producing an average filtrate turbidity of 0.05 NTU and an average of 4.2 log removal of total
particles (>2 im in size). Average feed turbidity and total particle counts were 1.3 NTU and 4,281
particles/ml, respectively.
TECHNOLOGY DESCRIPTION
The Polymem UF120 S2 Ultrafiltration Module is comprised of 19 individual polysulfone hollow-fiber
membrane bundles housed in a PVC pressure vessel. The bundles are potted on the effluent side of the
module, forming a U-shaped configuration and provide a total of 114 m2 (1227 ft2) of active membrane
surface area. The membrane, classified as an Ultrafiltration membrane, has a nominal pore size of 0.01 (im
as specified by Polymem and was not verified in this verification test. This pore size should provide a
physical barrier to particulate matter, bacteria, protozoans, and viruses when membrane fibers are intact
and operated within the recommended operating ranges.
The membrane module is designed for operation in a dead-end mode, reducing power consumption over
traditional cross flow membrane products, as recirculation pumps are not required. The flow
configuration is outside to inside. This forces the accumulation of particulate matter, pathogens, and
suspended solids on the outside of the membrane fiber. The recommended backwash procedure includes
simultaneous hydraulic backwash, air scour, and chlorine injection. Backwash is accomplished by
pumping filtrate water from the inside to the outside of the fiber. This water is then discharged to waste.
An inlet for air scour is provided at the level of the potting resin via air diffusers located inside the
module. This design makes minimum chemical cleaning intervals of 30 days possible without exceeding
the maximum allowable transmembrane pressure (terminal transmembrane pressure) of 2 bar (29 psi).
The membrane system and operating strategy (flux, recovery, and backwash intervals) are typically
designed for a 30-day chemical cleaning interval. However, significant changes in water quality will
effect membrane performance. Temperature fluctuation, increases in natural organic matter, turbidity, and
pH changes may have the potential to increase membrane fouling rates.
Some fraction of the particulate matter and dissolved constituents in the feed water can accumulate on the
membrane surface and cannot be removed by hydraulic backwash and air scour. This leads to rise in
transmembrane pressure during normal operation. Once the terminal transmembrane pressure has been
reached (29 psi), the membrane must be taken off-line to remove this matter from the membrane with a
03/05/EPADWCTR The accompanying notice is an integral part of this verification statement. May 2003
VS-ii
-------
chemical clean. The membrane polymer is designed to be tolerant to a variety of chemicals, including
chlorine, acids, bases, and chelating agents commonly used for chemical cleaning.
Critical to this testing was the use of a "generic" CMPP. The CMPP was not provided by Polymem. The
CMPP used has the capacity to feed, backwash, and clean a variety of pressure vessel-type MF/UF
modules. Therefore, this testing verified the operation of the membrane module under a given set of
operational parameters, not membrane-specific process equipment.
VERIFICATION TESTING DESCRIPTION
Test Site
The testing site was the Green Bay Water Utility Filtration Plant located at 6183 Finger Road in
Luxemberg, Wisconsin. The Green Bay Water Utility Filtration Plant is fed by one or both of two raw
water intakes located on the western shore of Lake Michigan in Kewaunee, Wisconsin. The raw water is
pumped to the filtration plant in Luxemberg, Wisconsin. A small amount of chlorine (<0.30 mg/L) is
added at each intake to prevent growth of zebra mussels during transmission from intake to the treatment
facility. The CMPP used for this testing was located approximately 200 feet from the raw water channel
at the filtration plant. A submersible pump located 3 feet below the free water surface fed the CMPP via
2-inch schedule 80 PVC pipe, and 1.5-inch PVC tubing.
Methods and Procedures
Onsite bench-top analyses including turbidity, pH, chlorine, and temperature were conducted daily at the
test site according to Standard Methods for the Examination of Water and Waste-water, 20th Edition
(APF£A, 1998) and by Methods for Chemical Analysis of Water and Wastes (EPA, 1979), where
applicable. Standard Methods for the Examination of Water and Wastewater, 20th Edition (APF£A, 1998)
was followed for total coliform analyses conducted at Northern Lake Service, Inc. (NLS), Crandon,
Wisconsin and MWH Laboratories, Pasadena, California. Other analyses conducted by NLS were
conducted using Standard Methods for the Examination of Water and Wastewater, 18th Edition (APF£A,
1992) and by Methods for Chemical Analysis of Water and Wastes (EPA, Revision 1983), where
applicable. Laboratory analyses included alkalinity, total and calcium hardness, total dissolved solids
(TDS), total suspended solids (TSS), total organic carbon (TOC), ultraviolet absorbance at 254
nanometers (UVA), total coliform and heterotrophic plate count (HPC). Alkalinity and total and calcium
hardness analyses were conducted once per month. TDS analyses were conducted every other week. TOC
and UVA analyses were conducted twice per week. TSS, total coliforms, and HPC analyses were
conducted five days per week. Online particle counters and turbidimeters continuously monitored both the
feed and membrane filtrate waters. The particle counters were set up to enumerate particle counts in the
following size ranges: total (>2 (im), 2-3 (im, 3-5 (im, 5-15 (im, and >15 (im. Data from the online
particle counters were stored at 5-minute intervals on a dedicated computer. Online turbidity
measurements were recorded at 10-minute intervals. Challenge testing, microbial or otherwise, was not
performed as part of this study; particle removal was quantified based on turbidity and particle counter
data.
VERIFICATION OF PERFORMANCE
System Operation
Verification testing conditions were established based on pilot study optimization results conducted from
May 2001 to March 2002. The membrane unit was operated at a constant specific flux of 40 L/h-m2 (24
gfd) for the first 12.5 days of operation (Run 1) and 30 L/h-m2 (18 gfd) during the remaining 32.7 days of
operation (Run 2). Production backwashes were performed at 50-minute intervals using an average
volume of 39 and 30 gallons for Runs 1 and 2, respectively. System recoveries ranged from 89-96 percent
03/05/EPADWCTR The accompanying notice is an integral part of this verification statement. May 2003
VS-iii
-------
throughout the testing. The backwash chlorine concentration was set at 5 mg/L for the duration of the
testing.
Test Runs 1 and 2 yielded normalized specific flux decline rates of 7.2 L/h-m2/bar/day (0.29 gfd/psi/d)
and 1.7 L/h-m2/bar/day (0.069 gfd/psi/d), respectively. The improvement in fouling control during Run 2
is likely due to the lower target normalized flux. It should be noted that the 25 percent decrease in specific
flux led to a 260 percent increase in run time before a required chemical cleaning (12.5 vs. 32.7 days).
A total of three membrane cleanings were performed based on the manufacturer's recommended
procedure. A high pH (11-12) chlorine solution (200 mg/L) was injected into the membrane module and
was allowed to soak for at least 4 hours. Flux data was collected after each chemical cleaning to evaluate
specific flux recovery. The first cleaning was performed prior to membrane operation. Therefore,
recovery information was not available for this cleaning. The recovery of specific normalized flux for
Chemical Cleaning #'s 2 and 3 was 62 and 73 percent, respectively. Cleaning #2 was performed at
ambient water temperature, [14-18.6°C (57-65.5°F)], pH = 12.2, and an average total chlorine
concentration of 164 mg/L, for 8 hours. Because recovery of specific flux following Cleaning # 2 was
low, Cleaning # 3 was performed with a similar cleaning solution but at elevated solution temperature
[22-31°C (72-88°F)], for an extended soaking period. Despite these changes, the specific flux recovery
was marginal (73 percent). This may be explained in part by the lack of chemical recirculation. This is
because the CMPP was not equipped with heating and recirculation equipment typically used to perform
clean-in-place (CIP) procedures on this membrane.
Membrane integrity monitoring was conducted prior-to and after this testing. Air pressure-hold tests were
conducted by opening the feed side of the membrane to the atmosphere and applying approximately 10
psi to the filtrate side of the membrane. Once pressurized, the loss of filtrate side pressure was recorded
over a two-minute period. The first membrane integrity test yielded a zero pressure loss with time. The
test at the end of system operation yielded a pressure loss of 0.35 psi/min, which was within the
manufacturers recommended feed side pressure loss (<0.36 psi/min). However, during this test, visual
observations showed a steady stream of air bubbles released to the feed side of the membrane. This
suggested that a membrane fiber (or fibers) and membrane integrity may have been compromised.
Following ETV testing, the membrane module filtrate end cap was removed to further investigate the
bubbles noted during the final integrity test. This investigation followed the integrity test/repair
procedures outlined in the Polymem UF120 S2 Operations and Maintenance (O&M) Manual. One broken
fiber was identified and repaired. One subsequent pressure decay test, performed as described above,
yielded a zero loss in pressure and no visual indicators of a loss of membrane integrity (no bubbles were
detected).
Water Quality Results
The equipment verification testing described in this report was executed using raw Lake Michigan water
obtained from the Green Bay Water Utility Filtration Plant. Water used for CMPP operation was drawn
from the process prior to any treatment (other than C12 addition for zebra muscle control) at the water
facility and was pumped approximately 200 feet to the skid mounted CMPP located inside a module
trailer unit. Table VS-1 below presents the results of the general water quality characterization for both
feed and filtrate waters throughout the ETV verification test. The feed water had the following average
water quality during this evaluation: C12 residual 0.05 mg/L, alkalinity 110 mg/L as CaCO3, total
hardness 130 mg/L as CaCO3, calcium hardness 88 mg/L as CaCO3, TSS 1.3 mg/L, TDS 187 mg/L,
TOC 2.3 mg/L, UVA 0.024 cm'1, algae 34 #/ml, temperature 3°C (37°F), and pH 7.8. As expected, there
was no notable change in alkalinity, total hardness, calcium hardness, or total dissolved solids across the
membrane module. However, there was a small reduction in TOC in the filtrate.
03/05/EPADWCTR The accompanying notice is an integral part of this verification statement. May 2003
VS-iv
-------
Total suspended solids were measured throughout the testing as an indication of particle removal
potential. Filtrate TSS was typically below the detection limit with 32 out of 37 samples reported at or
below the level of detection. Like HPC data, some of the filtrate samples were detected at higher than
expected levels. These results are likely due to the fact that feed and filtrate samples were so near the
detection limit of the analysis. Due to the length of time the equipment was in use prior to the ETV
testing, it is also possible that material had built up in the portion of sample piping permanently fixed to
the CMPP skid. Although the sample ports were allowed to flush prior to sample collection, accumulated
material may have sloughed off during some of the sampling periods.
As presented in Table VS-1, average feed and filtrate bench top turbidities were 1.3 and 0.05 NTU,
respectively. Continuously monitored filtrate turbidity was 0.035 NTU or less 90 percent of the time.
Average feed and filtrate total particle counts were 4,281 and 4 particles/ml, respectively. Table VS-2
summarizes the particulate log removal data. Average particle log removals of 4.2, 4.1, 4.1, 3.4, 3.3, 2.9,
and 2.2 were achieved for particle size ranges of >2 um, 2-3, 3-5, 5-7, 7-10, 10-15, and >15 um,
respectively. The 90th percentile for feed and filtrate total particle counts (>2 #/ml) was approximately
9,911 and 2 particles/ml, respectively. The membrane system removed 3.1 logs of total particles 90
percent of the time. A few of the filtrate particle count data were recorded by the data logger as 0.00
particle/ml (below the detection limit of the instrument). Since these data were recorded as zero values,
log removal data could not be calculated for these data points and were not included in the statistical
analyses. Because the membrane system produced relatively consistent filtrate particle counts, log
removals increased during periods when feed water particle counts were higher and decreased during
periods when feed water particle counts were lower. Relatively higher particle counts were measured in
the filtrate immediately following a backwash due in part to hydraulic and air bubble turbulence. As a
result, particle removals were decreased during these events.
A sensitivity analysis was performed on the data collected from one 24-hour period to determine the
potential effects of backwash events on calculated log removals. Data from March 14, 2002 were chosen
for this analysis due to the clusters of relatively lower log removal data during that time period, thereby
representing a worse case scenario. Log removals calculated for the raw data set (data including backwash
events) were 3.2 logs or greater, 90 percent of the time. Log removals calculated for the data set
excluding data obviously collected during backwash events, increased to 3.6 logs or greater, 90 percent of
the time.
Table VS-3 summarizes total coliform and HPC data. Total coliform enumeration results showed feed
concentration ranging from < 1.1-23 MPN/100 ml. Filtrate results for total coliform enumeration were
reported below the detection limit of <1 MPN/lOOml. HPC were significantly reduced. Feed water HPC
ranged up to 330 CFU/ml. 33 of 38 filtrate HPC samples were at or below the method detection limit of 2
CFU/ml.
03/05/EPADWCTR The accompanying notice is an integral part of this verification statement. May 2003
VS-v
-------
Table VS-1 General Water Quality for Both Feed
Parameter Units
C12 -Residual0 } mg/L
Alkalinity mg/L as CaCO3
Total Hardness mg/L as CaCO3
Calcium Hardness mg/L as CaCO3
TSS(2) mg/L
TDS mg/L
TOC mg/L
UVA cm"1
Algae #/ml
pH Units
Temperature °C (°F)
Bench Top Turbidity NTU
Particles >2 um #/ml
and Filtrate Waters
Feed Water Filtrate
0.05
110 110
130 130
88 87
1.3 1.2
187 203
2.3 2.0
0.024 0.019
34.4
7.80
3.4 (38)
1.3 0.05
4281 4
(1) Measured as part of the daily sampling activities of the Green Bay Water Utility Filtration Plant (GBWUFP).
(2) Limit of detection = 1 mg/L
Table VS-2 Particulate Log Removal
Particle Size Average Feed Count, #/ml Average
>2um 4,281
2-3 um 1,602
3-5 um 1,880
5-7 um 325
7-10um 305
10-15 um 127
>15um 41
Filtrate Count #/ml Average Log Removal
4 4.2
1 4.1
1 4.1
0 3.4
0 3.3
1 2.9
2 2.2
Table VS-3 Average Microbial Water Quality
Parameter Units Feed Water
Total Coliforms (1) MPN/100 ml 6.2
HPC (2) CFU/ml 17
Filtrate Backwash Water
<1.1 <1.1
2 24
( 1 ) Limit of detection =1.1 MPN/1 00 ml
(2) Limit of detection = 2 CFU/ml
03/05/EPADWCTR The accompanying notice is an integral part of this verification statement. May 2003
VS-vi
-------
Operation an d Mainten an ce Results
Operating conditions were established in a Programmable Logic Controller (PLC) prior to beginning the
test. These conditions included flux rate, production dwell time, backwash procedures (interval and
duration), alarm condition settings, chemical feed doses, and data logging intervals. A notable exception
to the logged parameters is air scour flow rate. With the exception of backwash duration, these
parameters were not adjusted during operation. Backwash duration was adjusted as needed to maintain a
recovery of at least 90 percent and ranged from 60-120 seconds. Backwash chlorine was set to a dose of 5
mg/L and was checked daily through onsite analyses.
Operation of the membrane consumed approximately 0.05 and 0.03 Ibs/day of sodium hypochlorite
during test Runs 1 and 2, respectively. Chemical cleanings each consumed 0.06 Ibs of sodium
hypochlorite and approximately 1.5-2 Ibs of sodium hydroxide.
Original Signed by Clyde R. Dempsey Original Signed by
for Hugh W. McKinnon 06/10/03 Gordon Bellen 06/13/03
Hugh W. McKinnon Date Gordon Bellen Date
Director Vice President
National Risk Management Research Laboratory Research
Office of Research and Development NSF International
United States Environmental Protection Agency
NOTICE: Verifications are based on an evaluation of technology performance under specific,
predetermined criteria and the appropriate quality assurance procedures. EPA and NSF make no
expressed or implied warranties as to the performance of the technology and do not certify that a
technology will always operate as verified. The end user is solely responsible for complying with
any and all applicable federal, state, and local requirements. Mention of corporate names, trade
names, or commercial products does not constitute endorsement or recommendation for use of
specific products. This report is not a NSF Certification of the specific product mentioned herein.
03/05/EPADWCTR The accompanying notice is an integral part of this verification statement. May 2003
VS-vii
-------
Availability of Supporting Documents
Copies of the ETV Protocol for Equipment Verification Testing for Physical Removal of
Microbiological and Paniculate Contaminants, dated May 14, 1999, the Verification
Statement, and the Verification Report (NSF Report # NSF 02/05/EPADWCTR) are available
from the following sources:
(NOTE: Appendices are not included in the Verification Report. Appendices are available
from NSF upon request.)
1. ETV Drinking Water Systems Center Manager (order hard copy)
NSF International
P.O. Box 130140
Ann Arbor, Michigan 48113-0140
2. NSF web site: http://www.nsf.org/etv/dws/dws_reports.html and from
http://www.nsf.org/etv/dws/dwsjroject documents.html (electronic copy)
3. EPA web site: http://www.epa.gov/etv (electronic copy)
03/05/EPADWCTR The accompanying notice is an integral part of this verification statement. May 2003
VS-viii
-------
May 2003
Environmental Technology Verification Report
Physical Removal of Particulate Contaminants in Drinking Water
Polymem UF120 S2 Ultrafiltration Membrane Module
Luxemburg, Wisconsin
Prepared for:
NSF International
Ann Arbor, MI 48105
Prepared by:
Carollo Engineers, P.C
Boise, ID 83713
Under a cooperative agreement with the U.S. Environmental Protection Agency
Jeffrey Q. Adams, Project Officer
National Risk Management Research Laboratory
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
Office of Research and Development
U.S. Environmental Protection Agency
Washington, DC
-------
Notice
The U.S. Environmental Protection Agency (EPA) through its Office of Research and
Development has financially supported and collaborated with NSF International (NSF) under
Cooperative Agreement No. R-82833301. This verification effort was supported by the Drinking
Water Systems (DWS) Center operating under the Environmental Technology Verification
(ETV) Program. This document has been peer reviewed and reviewed by NSF and EPA and
recommended for public release.
-------
Foreword
The U.S. Environmental Protection Agency (EPA) is charged by Congress with protecting the
Nation's land, air, and water resources. 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. To meet this
mandate, EPA's research program is providing data and technical support for solving
environmental problems today and building a science knowledge base necessary to manage our
ecological resources wisely, understand how pollutants affect our health, and prevent or reduce
environmental risks in the future.
The National Risk Management Research Laboratory (NRMRL) is the Agency's center for
investigation of technological and management approaches for preventing and reducing risks
from pollution that threaten human health and the environment. The focus of the Laboratory's
research program is on methods and their cost-effectiveness for prevention and control of
pollution to air, land, water, and subsurface resources; protection of water quality in public water
systems; remediation of contaminated sites, sediments and ground water; prevention and control
of indoor air pollution; and restoration of ecosystems. NRMRL collaborates with both public
and private sector partners to foster technologies that reduce the cost of compliance and to
anticipate emerging problems. NRMRL's research provides solutions to environmental problems
by: developing and promoting technologies that protect and improve the environment; advancing
scientific and engineering information to support regulatory and policy decisions; and providing
the technical support and information transfer to ensure implementation of environmental
regulations and strategies at the national, state, and community levels.
This publication has been produced as part of the Laboratory's strategic long-term research plan.
It is published and made available by EPA's Office of Research and Development to assist the
user community and to link researchers with their clients.
Hugh W. McKinnon, Director
National Risk Management Research Laboratory
in
-------
Table of Contents
Section Page
Verification Statement VS-i
Title Page i
Notice ii
Foreword iii
Table of Contents iv
Abbreviations and Acronyms viii
Acknowledgements x
Chapter 1 - Introduction 1
1.1 ETV Purpose and Program Operation 1
1.2 Testing Participants 1
1.3 Definition of Roles and Responsibilities of Project Participants 2
1.3.1 NSF Responsibilities 2
1.3.2 Field Testing Organization Responsibilities 2
1.3.3 Manufacturer Responsibilities 3
1.3.4 Operator and Test Site Staff Responsibilities 3
1.3.5 Water Quality Analyst Responsibilities 4
1.3.6 EPA Responsibilities 4
1.3.7 Funding Source 4
1.4 Verification Testing Site 5
1.4.1 Site Description 5
1.4.2 Source/Feed Water 5
1.4.3 Pilot Effluent Discharge 6
1.5 Background 6
Chapter 2 - Equipment Description and Operating Processes 7
2.1 General Equipment Description 7
2.2 UF120 S2 Ultrafiltration Membrane Module Description 7
2.2.1 General Description 7
2.2.2 Environmental Requirements of UF120 S2 Membrane 8
2.2.3 Materials of Construction of UF120 S2 Membrane 8
2.3 Operating Process 8
2.3.1 Process Instrumentation 9
2.3.2 Consumables 9
2.3.3 Product Performance Capabilities 10
Chapter 3 - Methods and Procedures 11
3.1 Environmental Technology Verification Testing Plan 11
3.1.1 Task 1: Membrane Flux and Recovery 11
3.1.2 Task 2: Chemical Cleaning Efficiency 12
iv
-------
Table of Contents (continued)
3.1.3 Task 3: Finished Water Quality 13
3.1.4 Task 4: Reporting of Membrane Pore Size 13
3.1.5 Task 5: Membrane Integrity Testing 14
3.1.6 Task 6: Data Handling Protocol 15
3.1.7 Task 7: Quality Assurance/Quality Control 15
3.1.8 Task 8: Operational Conditions and Maintenance 21
3.2 Calculation of Operating Parameters 21
3.2.1 Flux 21
3.2.2 Specific Flux 21
3.2.3 Normalized Specific Flux 21
3.2.4 Feedwater System Recovery 22
3.3 Calculation of Data Quality Indicators 22
3.3.1 Representativeness 22
3.3.2 Statistical Uncertainty 24
3.3.3 Accuracy 24
3.3.4 Precision 25
3.4 Safety Measures 25
3.5 Testing Schedule 26
Chapter 4 - Results and Discussion 27
4.1 Task 1: Characterization of Membrane Flux and Recovery 27
4.2 Task 2: Evaluation of Cleaning Efficacy 28
4.3 Task 3: Evaluation of Finished Water Quality 29
4.3.1 Turbidity, Particle Counts, and Particle Removal 29
4.3.2 Microbial Removal 30
4.3.3 Other Water Quality Parameters 31
4.4 Task 4: Reporting Membrane Pore Size 32
4.5 Task 5: Membrane Integrity Testing 32
4.6 Task 6: Data Management 33
4.6.1 Data Recording 33
4.6.2 Data Entry, Validation ,and Reduction 33
4.7 Task 7: Quality Assurance/Quality Control (QA/QC) 34
4.7.1 Data Correctness 34
4.8 Operational Conditions and Maintenance 35
4.8.1 Overall Operation and Maintenance 35
4.8.2 System Chemical Consumption 36
4.8.3 Review of Operations Manual 36
4.8.4 Equipment Deficiencies Experienced During the ETV Program 36
Chapter 5-References 37
Chapter 6 - Vendor Comments 38
Table and Figures 39
-------
Table of Contents (continued)
Appendices
Appendix A. Polymem UF120 S2 Ultrafiltration Module Operation & Maintenance Manual
and Photographs
Appendix B. Custom Membrane Pilot Plant (CMPP) Operating Manual
Appendix C. NLS and MWH Laboratory Certifications
Appendix D. Copy of Field Notebook and Data Log Sheets
Appendix E. Instrument Calibration Certificates
Appendix F. Material Safety Data Sheets
Appendix G. Laboratory Data Sheets
Appendix H. Summary of Pore Size Characterization (By Polymem)
Appendix I. NLS and MWH QA/QC Results
Appendix J. Data Completeness Tables
List of Tables
Table 1-1. Lake Michigan Feed Water Quality Data 41
Table 3-1 Verification Conditions 42
Table 3-2 Operational Data 42
Table 3-3 Chemical Cleaning Procedures and Conditions 43
Table 3-4 Cleaning Verification Data 43
Table 3-5 Water Quality Sampling Schedule 44
Table 3-6 Data Management 45
Table 3-7 Methods for Measuring Precision and Accuracy 46
Table 3-8 Sample Methodology 47
Table 3-9 Corrective Action Plan 48
Table 4-1 Operational Conditions 49
Table 4-2 Summary of Operational Data 50
Table 4-3 Summary of Chemical Cleaning Water Quality Analyses 51
Table 4-4 Summary of Chemical Cleaning Hydraulic Analyses 52
Table 4-5 Summary of Chemical Cleaning Efficacy 53
Table 4-6 Summary of Onsite Bench-top Turbidity Data 53
Table 4-7 Summary of Online Turbidity and Particle Count Data 54
Table 4-8 Summary of Microbial Water Quality 55
Table 4-9 General Water Quality Parameters 56
Table 4-10 TSS Mass Balance 57
Table 4-11 Summary of Verification Data for the Feed Water Particle Counter 58
Table 4-12 Summary of Verification Data for the Filtrate Water Particle Counter 59
Table 4-13 Chemical Consumption Analysis 60
VI
-------
Table of Contents (continued)
List of Figures
Figure 2-1 Process and Instrumentation Diagram 61
Figure 4-1 TMP, Flux, and System Recovery Profiles 62
Figure 4-2 Operational Flux and Temperature Profiles 63
Figure 4-3 Online and Bench-top Turbidity Profiles 64
Figure 4-4 Feed and Filtrate Count Profiles for Particles >2 um 65
Figure 4-5 Feed and Filtrate Count Profiles for Particles 2-3 um 66
Figure 4-6 Feed and Filtrate Count Profiles for Particles 3-5 um 67
Figure 4-7 Feed and Filtrate Count Profiles for Particles 5-7 um 68
Figure 4-8 Feed and Filtrate Count Profiles for Particles 7-10 um 69
Figure 4-9 Feed and Filtrate Count Profiles for Particles 10-15 um 70
Figure 4-10 Feed and Filtrate Count Profiles for Particles >15 um 71
Figure 4-11 Log Removal Profile for Particles >2 um in Size 72
Figure 4-12 Log Removal Profile for Particles 2-3 um in Size 73
Figure 4-13 Log Removal Profile for Particles 3-5 um in Size 74
Figure 4-14 Log Removal Profile for Particles 5-7 um in Size 75
Figure 4-15 Log Removal Profile for Particles 7-10 um in Size 76
Figure 4-16 Log Removal Profile for Particles 10-15 um in Size 77
Figure 4-17 Log Removal Profile for Particles >15 um in Size 78
Figure 4-18 A Frequency Distribution of Filtrate Turbidity 79
Figure 4-19 A Frequency Distribution of Feed Water Turbidity 80
Figure 4-20 A Frequency Distribution for Filtrate Particles >2 um 81
Figure 4-21 A Frequency Distribution for Feed Water Particles >2 um 82
Figure 4-22 A Frequency Distribution of Log Removal for Particles >2 um 83
Figure 4-23 A Frequency Distribution of Log Removal for Particles 2-3 um 84
Figure 4-24 A Frequency Distribution of Log Removal for Particles 3-5 um 85
Figure 4-25 A Frequency Distribution of Log Removal for Particles 5-7 um 86
Figure 4-26 A Frequency Distribution of Log Removal for Particles 7-10 um 87
Figure 4-27 A Frequency Distribution of Log Removal for Particles 10-15 um 88
Figure 4-28 A Frequency Distribution of Log Removal for Particles >15 um 89
Figure 4-29 Feed and Filtrate Particle Count Profiles for Particles >2 um from 3/14/02 00:00
to 3/14/02 20:44 Using Raw Data 90
Figure 4-30 Feed and Filtrate Particle Count Profiles for Particles >2 um from 3/14/02 00:00
to 3/14/02 20:44 Excluding Data Collected During Backwash Events 91
Figure 4-31 A Frequency Distribution of Log Removal for Particles >2 um from 3/14/02
00:00 to 3/14/02 20:44 92
vn
-------
Abbreviations and Acronyms
b
°C
cfm
CPU
cn>
C12
CMPP
CPU
DWS
EPA
°F
Fe
ft2
FTO
GBWUFP
gfd
gpm
HMI
HPC
hr
IESWTR
kWh
L
LT1ESWTR
LT2ESWTR
m2
MF/UF
mg/L
mL
Bar
Degrees Celsius
Cubic feet per minute
Colony Forming Units
Clean-in-place
Chlorine
Custom Membrane Pilot Plant
Central Processing Unit
Drinking Water System
United States Environmental Protection Agency
Degrees Fahrenheit
Iron
Feet squared
Field Testing Organization
Green Bay Water Utility Filtration Plant
Gallons per square foot per day
Gallons per minute
Human-Machine Interface
Heterotrophic Plate Count
Hour
Interim Enhanced Surface Water Treatment Rule
Kilowatt hour
Liters
Long Term 1 Enhanced Surface Water Treatment
Rule
Long Term 2 Enhanced Surface Water Treatment
Rule
Meters squared
Mi crofiltrati on/Ultrafiltrati on
Milligrams per liter
Milliliter
Vlll
-------
jim Micrometers
MPN Most Probable Number
MSDS Material Safety Data Sheets
NaOCl Sodium Hypochlorite
NaOH Sodium Hydroxide
NLS Northern Lake Service, Inc.
NSF NSF International
NTU Nephelometric Turbidity Units
P Pump
P/A Presence/Absence
PLC Programmable Logic Controller
ppm Parts per million
psi Pounds per square inch
PSTP Product Specific Test Plan
QA/QC Quality Assurance/Quality Control
SCFM Standard cubic feet per minute
SDWA Safe Drinking Water Act
STP Standard Temperature and Pressure
SUVA Specific Ultraviolet Absorbance
TDS Total Dissolved Solids
TMP Transmembrane Pressure
TOC Total Organic Carbon
TSS Total Suspended Solids
VFD Variable Frequency Drive
-------
ACKNOWLEDGMENTS
The Field Testing Organization, Carollo Engineers, P.C., was responsible for elements in the
testing sequence, including collection of samples, calibration and verification of instruments,
data collection and analysis, data management, data interpretation and the preparation of this
report.
Carollo Engineers, P.C.
12592 W. Explorer Dr., Suite 200
Boise, ID 83713
Contact Person: Daniel A. Hugaboom, P.E.
The laboratory selected for microbiological analyses and non-microbiological analytical work of
this study was:
Northern Lake Service, Inc.
400 North Lake Avenue
Crandon, WI 54520
Contact Person: R.T. Krueger
The laboratory selected for microbiological analytical work of this study was:
MWH Laboratories
555 E. Walnut Street
Pasadena, CA 91101
Contact Person: Jim Hein
The Manufacturer of the Equipment was:
Polymem
Route de Revel F-31450
Fourquevaux, France
Contact Person: Jean-Michel Espenan
Carollo Engineers, P.C. and Polymem wishes to thank the Energy Center of Wisconsin for
supporting this study.
Carollo Engineers, P.C. wishes to thank Patricia Terry, Ph.D. - University of Wisconsin Green
Bay and the Green Bay Water Utility for their help during this study.
-------
Chapter 1
Introduction
1.1 ETV Purpose and Program Operation
The U.S. Environmental Protection Agency (EPA) has created the Environmental Technology
Verification (ETV) Program to facilitate the deployment of innovative or improved
environmental technologies through performance verification and dissemination of information.
The goal of the ETV program is to further environmental protection by substantially accelerating
the acceptance and use of improved and more cost-effective technologies. ETV seeks to achieve
this goal by providing high quality, peer reviewed data on technology performance to those
involved in the design, distribution, permitting, purchase, and use of environmental technologies.
ETV works in partnership with recognized standards and testing organizations; stakeholder
groups which consist of buyers, vendor organizations, and permitters; and with the full
participation of individual technology developers. The program evaluates the performance of
innovative technologies by developing test plans that are responsive to the needs of stakeholders,
conducting field or laboratory tests (as appropriate), collecting and analyzing data, and preparing
peer reviewed reports. All evaluations are conducted in accordance with rigorous quality
assurance protocols to ensure that data of known and adequate quality are generated and that the
results are defensible.
The EPA has partnered with NSF International (NSF), an independent, not-for-profit testing and
certification organization dedicated to public health, safety and protection of the environment, to
verify performance of small drinking water systems that serve small communities under the ETV
Drinking Water Systems (DWS) Center. A goal of verification testing is to enhance and
facilitate the acceptance of small drinking water treatment equipment by state drinking water
regulatory officials and consulting engineers while reducing the need for testing of equipment at
each location where the equipment's use is contemplated. NSF will meet this goal by working
with manufacturers and NSF-qualified Field Testing Organizations (FTO) to conduct verification
testing under the approved protocols. It is important to note that verification of the equipment
does not mean that the equipment is "certified" by NSF or "accepted" by EPA. Rather, it
recognizes that the performance of the equipment has been determined and verified by these
organizations for those conditions tested by the FTO.
The DWS Center evaluated the performance the Polymem UF120 S2 Ultrafiltration Membrane
Module, which is a membrane filtration technology used in drinking water treatment system
applications. The test evaluated the membrane module's ability to physically remove microbial
and particulate contaminants. This document provides the verification test results for the
Polymem UF120 S2 Ultrafiltration Membrane Module.
1.2 Testing Participants
The FTO was Carollo Engineers, P.C., which provided the overall management of the ETV. The
Ultrafiltration membrane manufacturer for the ETV was Polymem. The operations management
and staff were from the University of Wisconsin-Green Bay and Carollo Engineers. Laboratory
analyses were performed by Northern Lake Service, Inc. (NLS), Crandon, Wisconsin and
-------
Montgomery Watson Laboratories, Pasadena, California (total coliform enumeration only). Data
management and final report preparation were performed by the FTO, Carollo Engineers, P.C.
1.3 Definition of Roles and Responsibilities of Project Participants
The following is a brief description of each ETV participant and their roles and responsibilities.
1.3.1 NSF Responsibilities
NSF is a not-for-profit testing and certification organization dedicated to public health safety and
the protection of the environment. Founded in 1946 and located in Ann Arbor, Michigan, NSF
has been instrumental in the development of consensus standards for the protection of public
health and the environment. NSF also provides testing and certification services to ensure that
products bearing the NSF Name, Logo and/or Mark meet those standards. The EPA partnered
with the NSF to verify the performance of drinking water treatment systems through the EPA's
ETV Program.
NSF provided technical oversight of the verification testing. An audit of the field analytical and
data gathering and recording procedures was conducted. NSF also provided review of the
Product Specific Test Plan (PSTP) and this report.
Contact Information:
NSF International
789 N. Dixboro Rd.
Ann Arbor, MI 48105
Phone: 734-769-8010
Fax: 734-769-0109
Contact: Bruce Bartley, Project Manager
Email: bartley@nsf.org
1.3.2 Field Testing Organization Responsibilities
Carollo Engineers, P.C. conducted the verification testing of the Polymem UF120 S2
Ultrafiltration Membrane Module. Carollo Engineers, P.C. is an NSF-qualified FTO for the
ETV Drinking Water Systems Center.
The FTO was responsible for conducting the verification testing for 30 calendar days. FTO
employees conducted the onsite analyses and data recording during the testing. The specific
responsibilities of the FTO, Carollo Engineers, P.C., were to:
• Provide the overall management of the ETV through the proj ect manager and the proj ect
engineers.
• Provide needed logistical support, the project communication network, and scheduling and
coordination of the activities of participants.
• Manage, evaluate, interpret and report on data generated in the ETV.
-------
• Evaluate the performance of the ultrafiltration membrane technology according to the PSTP
and the testing, operations, quality assurance/quality control (QA/QC), data management and
safety protocols contained therein.
• Provide quality control (QC) information in the ETV report.
• Provide data generated during the ETV in hard copy and electronic form in a common
spreadsheet or database form.
Contact Information:
Carollo Engineers, P.C.
12592 W. Explorer Dr., Suite 200
Boise, ID 83713
Phone: (208) 376-2288
Contact Person: Daniel A. Hugaboom, P.E.
Email: dhugaboom@carollo.com
1.3.3 Manufacturer Responsibilities
The specific responsibilities of the ultrafiltration membrane manufacturer, Polymem, were to:
• Provide complete, field-ready equipment for the ETV at the testing site.
• Provide logistical and technical support as required throughout the ETV.
• Provide partial funding for the project.
• Attend project meetings as necessary.
Contact Information:
Polymem
Route de Revel F-31450
Fourquevaux, France
Phone: Oil.33.5.65.71.79.89
Contact Person: Jean-Michel Espenan
Email: polymem@wanadoo.fr
1.3.4 Operator and Test Site Staff Responsibilities
The specific responsibilities of operations and test site staff from the University of Wisconsin at
Green Bay, Carollo Engineers, P.C., and Green Bay Water Utility, were to:
• Provide set-up, shakedown, operations, maintenance and on-site analytical services according
to the PSTP and the testing, operations, QA/QC, data management and safety protocols.
• Provide the necessary and appropriate space for the equipment to be tested in the ETV.
• Provide necessary electrical power, feedwater and other utilities as required for the ETV.
• Provide necessary drains from the test site.
-------
Contact Information:
University of Wisconsin-Green Bay
24200 Nicolet Dr.
Green Bay, WI 54311-7001
Phone: (920) 465-2278
Contact Person: Patricia Terry, Ph.D.
Email: terryp@uwgb.edu
1.3.5 Water Quality Analyst Responsibilities
The specific responsibilities of the water quality analytical staff from the NLS and MWH
Laboratories were to:
• Provide off-site water quality analyses prescribed in the PSTP according to the QA/QC
protocols contained therein.
• Provide reports with the analytical results to the data manager.
• Provide detailed information on the analytical procedures implemented.
Contact Information:
Northern Lake Service, Inc.
400 North Lake Avenue
Cranon, WI 54520
Phone:(715)478-2777
Contact Person: R.T. Krueger
E-Mail: Krueger@northernlakeservice.com
MWH Laboratories
555 E. Walnut Street
Pasadena, CA 91101
Phone: (626) 568-6400
Contact Person: Jim Hein
1.3.6 EPA Responsibilities
The EPA through its Office of Research and Development has financially supported and
collaborated with NSF under Cooperative Agreement No. R-82833301. This verification effort
was supported by the DWS Center operating under the ETV Program. This document has been
peer reviewed and reviewed by NSF and EPA and recommended for public release.
1.3.7 Funding Source
The ETV testing described herein was funded in conjunction with a 12-month membrane pilot
study funded by the Energy Center of Wisconsin. The 12-month study, focused not only on the
power requirements of membrane filtration plants, but also on promoting the use of more
effective and less energy intensive water treatment technologies. The Energy Center of
-------
Wisconsin chose to participate because the overall scope of the ETV testing fit into of the scope
of the longer, energy focused study.
1.4 Verification Testing Site
1.4.1 Site Description
The testing site was the Green Bay Water Utility Filtration Plant (GBWUFP) located at
6183 Finger Road in Luxemberg, Wisconsin. The plant utilizes pre-ozonation, coagulation,
flocculation, sedimentation, granular media filtration, and chlorination prior to distribution to the
City of Green Bay. The plant also has facilities for recycling filter washwater, which include
settling lagoons and washwater recycle basins.
The GBWUFP is fed by two raw water intakes located on the western shore of Lake Michigan in
Kewaunee, Wisconsin. The primary intake structure is located approximately one mile off shore
at a depth of 50 feet and is operated year round; a second intake structure located 3000 feet off
shore at a depth of 30 feet is operated during peak demand periods to supplement the primary
intake. The raw water is pumped to the filtration plant in Luxemberg, Wisconsin. Chlorine is
added at each intake to prevent growth of zebra mussels during transmission from intake to the
treatment facility. Residual chlorine concentrations (as Cb) in the feed water at the treatment
facility were collected during the ETV test and tested by GBWUFP staff and are included in
Table 1-1. The chlorine residual during the test averaged 0.05 mg/L. Because chlorine is added at
the intake, "raw" Lake Michigan water was not available at the site. However, the low reported
chlorine levels were not expected to significantly effect membrane performance.
The pilot plant was located approximately 200 feet from the raw water channel at the filtration
plant and was fed by either or both of the GBWUFP raw water intakes. A submersible pump
located 3 feet below the free water surface fed the pilot plant via 2-inch schedule 80 PVC pipe,
and 1.5-inch PVC tubing. A tee fitting divided flow to serve two pilot plant units and was located
approximately 20 feet from each pilot plant trailer feed connection. Ball valves located on each
branch provided flow-splitting control. Pictures of the project site are included in Appendix A.
1.4.2 Source/Feed Water
The equipment verification testing described in this report was executed using raw Lake
Michigan water obtained from the GBWUFP. Water used for pilot plant operation was drawn
from the process prior to any treatment at the water facility (other than chlorine addition for
zebra mussel control) and was pumped approximately 200 feet to the skid mounted pilot plant
located inside a module trailer unit.
GBWUFP staff monitor Lake Michigan water quality as part of normal plant operations. In
general, Lake Michigan can be characterized as a high quality raw water. Based on historical
data, the raw water can be characterized by low average turbidity. However, fall and winter
storm events can significantly increase turbidity for several days at a time. Seasonal variations
may also cause significant variation on water temperature and correlate to algae blooms.
Alkalinity, pH, and TOC show less seasonal variation than turbidity and temperature. Specific
feed water quality data collected throughout this verification testing are reported in Table 1-1.
The feed water had the following average water quality during this evaluation: Cb residual 0.05
-------
mg/L, alkalinity 110 mg/L as CaCCb, total hardness 130 mg/L as CaCCb, calcium hardness 88
mg/L as CaCOs, total dissolved solids (IDS) 187 mg/L, total suspended solids (TSS) 1.3 mg/L,
total organic carbon (TOC), 2.3 mg/L, UVA 0.024 per cm"1, algae 34/ml, pH 7.8, temperature
3°C (37°F), total coliform 5.0 MPN/lOOml, HPC 17 CFU/ml, and turbidity 1.3 NTU.
1.4.3 Pilot Effluent Discharge
Sanitary sewers do not exist at the testing site. Therefore, the effluent (filtrate and backwash) of
the pilot treatment unit was routed to the beginning of the GBWUFP facilities for recycling with
filter washwater. Chemical cleaning wastes were neutralized and disposed of off site.
1.5 Background
The Interim Enhanced Surface Water Treatment Rule (IESWTR), promulgated in 1998,
established stringent filtered water quality standards for systems serving over 10,000 people.
Promulgated at the same time as the IESTWR, was the Stage 1 Disinfectant and Disinfection
Byproduct Rule (Stage 1 D/DBPR) to regulate levels of disinfectant residuals and byproducts in
the distribution system of water systems of all sizes. Emerging regulations due in the calendar
year 2001, include the Long Term 1 Enhanced Surface Water Treatment Rule (LTIESWTR) that
will extend the requirements of the IESWTR to systems serving less than 10,000 people. Stage 2
LT2ESWTR will be applicable to systems of all sizes treating surface water and groundwater
under the influence of surface water. The LT2ESWTR "Agreement in Principal" includes
regulation of Cryptosporidium, and technologies that utilities must utilize to remove it from their
water source. These emerging regulations require greater removal of pathogens, lower filter
effluent turbidity standards, and lower disinfectant byproduct concentrations in the distribution
system.
In order to meet these more stringent regulations, water utilities are evaluating numerous
treatment options including low-pressure membrane filtration such as microfiltration (MF) and
ultrafiltration (UF). Although not directly verified in this ETV test, these types of membrane
systems are designed to remove bacteria, protozoa, and in the case of UF membranes, viruses. In
the coming years, most water systems using surface water or groundwater under the influence of
surface water will need to evaluate their capacity to meet these objectives. These systems
constitute the majority of the market for the product being evaluated in this report.
At the time this report was written, amendments to the Surface Water treatment Rule (SWTR), as
mandated by the Safe Drinking Water Act (SOWA) were actively being negotiated. While final
rules and amendments were being negotiated, the implications of these new rules were well
known in the industry, and many water utilities were focusing their plant upgrade and new
construction on these anticipated standards.
-------
Chapter 2
Equipment Description and Operating Processes
2.1 General Equipment Description
Critical to the verification testing was the use a "generic" custom membrane pilot plant (CMPP).
The CMPP was not provided by the membrane manufacturer, Polymem. The CMPP used has the
capacity to feed, backwash and clean a variety of MF/UF modules that are housed in pressure
vessels. Therefore, this ETV testing verified the operation of the membrane module under a
given set of operational parameters, not membrane-specific process equipment. A description of
the equipment used in the process is provided in this section.
2.2 UF120 S2 Ultrafiltration Membrane Module Description
2.2.1 General Description
The Polymem UF120 S2 Ultrafiltration Membrane Module is comprised of 19 individual
polysulfone hollow-fiber membrane bundles housed in a PVC pressure vessel. The bundles are
potted on the effluent side of the module, forming a U-shaped configuration and provide a total
of 114 m2 (1227 ft2) of active membrane surface area. The nominal pore size, as specified by
Polymem and not verified in this ETV test, is 0.01 jim, and is classified as an Ultrafiltration
membrane. This pore size should provide a physical barrier to bacteria, protozoans, and viruses
when membrane fibers are intact and operated within the recommended operating ranges.
The membrane module is designed to be operated in a dead-end mode, reducing power
consumption over traditional cross flow membrane products, as recirculation pumps are not
required. The flow configuration is outside to inside. This forces the accumulation of particulate
matter, pathogens, and suspended solids on the outside of the membrane fiber. The
recommended backwash procedure includes simultaneous hydraulic backwash, air scour, and
chlorine injection. Backwash is accomplished by pumping filtrate water from the inside to the
outside of the fiber. This water is then discharged to waste. An inlet for air scour is provided at
the level of the potting resin via air diffusers located inside the module.
Some fraction of the particulate matter and dissolved constituents in the feed water can
accumulate on the membrane surface and cannot be removed by hydraulic backwash and air
scour. This leads to a rise in transmembrane pressure during normal operation. Once the terminal
transmembrane pressure has been reached (29 psi), the membrane must be taken off-line to
remove this matter from the membrane with a chemical clean. The membrane system and
operating strategy (flux, recovery, and backwash intervals) are typically designed for a 30-day
chemical cleaning interval. However, significant changes in water quality will effect membrane
performance. Temperature fluctuation, increases in natural organic matter, turbidity, pH changes,
and high flux rates may have the potential to increase membrane fouling rates.
The membrane polymer is tolerant to a variety of chemicals, including chlorine, acids, bases, and
chelating agents commonly used for chemical cleaning. Photographs of the module, detailed
specifications, and typical operating parameters are included in Appendix A. Backwash and
chemical cleaning chlorine dose tolerances and pH ranges are also included.
-------
2.2.2 Environmental Requirements of UF120 S2 Membrane
The membrane module should be placed in a location that is protected from high winds, freezing
conditions, direct sunlight, and precipitation that may damage process piping and electrical
equipment.
2.2.3 Materials of Construction of UF120 S2 Membrane
The membrane module's (and CMPP skid) influent and effluent connections are standard-sized
schedule 80 PVC allowing for adaptability to custom designed membrane systems. The
membrane module itself is housed in a PVC vessel. Connection to the feed side of the module is
made via 3-inch schedule 80 PVC threaded into the inlet of the module. Connection on the
filtrate side of the module is made with a 12-inch stainless steel mechanical coupling and a 12-
inch by 3-inch PVC reducer. As noted, the membrane is polysulfone, potted on the filtrate side of
the module body with polyeurothane. A H-inch air scour diffuser inlet is provided at the potting
level to aid in cleaning efficacy. The membrane module weighs 99 pounds when empty and 187
pounds when full of water. The length and diameter of the module are 37 and 12 inches,
respectively.
2.3 Operating Process
The typical UF process consists of pumps, piping, and control systems capable of performing
basic functions necessary for membrane filtration. The process consists of a feed water pump,
membrane module, backwash pump, chemical feed systems, filtrate and chemical storage. Feed
pumping is used during production cycles to provide flow at the necessary pressure for driving
water across the membrane surface. At regular intervals (30-120 minutes), the feed pump is
turned off, and the backwash pump moves water into the effluent side of the membrane at an
elevated pressure for short periods of time (60 seconds) to remove accumulated solids from the
membrane surface. The hydraulic backwash is used in unison with air distributed through
diffusers located inside the membrane module. A schematic of the CMPP process including
pumps, piping schematic, sample ports, chemical feeds, and UF120 S2 membrane location is
included in the process and instrumentation diagram (Figure 2-1).
Because production backwashes do not remove all particulate matter from the membrane
surface, periodic chemical cleanings must be performed to restore membrane permeability.
Chemical cleanings are performed by soaking the membrane fibers in an appropriate solution,
followed by a rinsing cycle. This procedure takes approximately 8 hours to perform, and must be
manually initiated through the Programmable Logic Control (PLC) interface.
Periodic direct integrity monitoring checks to ensure that membrane fibers are not compromised
(therefore allowing passage of particulate contaminants larger than the membrane pore size) are
performed at regular intervals and generally prescribed by the regulatory agency responsible for
the water system. On the CMPP used in the verification, this is a manual operation. The process,
in general, is to pressurize the inside of the membrane fibers with air, and measure the pressure
decay rate. For the UF120 S2 membrane, the acceptable decay rate limit specified by Polymem
is 0.36 psi/min. Decay rates in excess of this limit indicate that fiber integrity may be
compromised.
-------
CMPP equipment requires a 3-phase/480 V power supply to run pumps and variable frequency
drives (VFD). A transformer located in the VFD panel converts the power to 120 V single phase
for the PLC panel.
2.3.1 Process Instrumentation
The CMPP skid is equipped with a PLC unit that provides a significant degree of automation.
The PLC maintains flow rates, pressures, control valve positions, and flow during backwashes
and chemical cleans. The PLC also has the ability to shut down the system in case of a high level
alarm (high transmembrane pressure, low liquid levels, low chemical tank levels, etc.). The
process can be configured for the specific application through a Windows-based interface.
Backwash and cleaning procedures, flow, pressure, alarms, and dose set points can be
customized via the PLC panel. This greatly reduces the required operator training and time
requirements. Functions can be adjusted, initiated, and monitored remotely. The system can be
operated without a full-time operator; however, operators are necessary for routine system
maintenance (e.g., analytical sampling, compressor maintenance, and to maintain acceptable
volumes in chemical storage tanks), maintaining water quality monitoring devices, performing
integrity tests, and trouble-shooting.
Critical data including flow rate, transmembrane pressure, flux, specific flux, temperature, and
turbidity can be downloaded remotely for analysis. However, maintenance issues including
cleaning pre-filters, filling feed tanks, and required sampling must be performed by on-site
personnel. Data collection with the exception of particle count data and manual gage readings
can be done remotely with remote control software over a standard telephone line.
2.3.2 Consumables
2.3.2.1 Chemicals
Chemicals used by the membrane are readily available commodity products [chlorine, sodium
hydroxide (NaOH), or acid if necessary]. Typical chemical consumption includes low doses of
chlorine [approximately 2 pounds per module per month as sodium hypochlorite (NaOCl)]
during production backwashes (occurring every 30 to 120 minutes) and during chemical cleaning
(0.05 pounds per module per month) if performed once each month. NaOH is also typically used
for removal of organic compounds during monthly chemical cleaning (2.4 pounds per module
per month). Chemical cleaning wastes can be disposed of in sanitary sewers as allowed by local
code. If necessary, these chemicals can be quenched (chlorine) or neutralized (acid or caustic)
prior to discharge. Sanitary sewers do not exist at the site. Therefore, chemical cleaning residuals
were neutralized and disposed of off-site.
2.3.2.2 Power
Power requirements for MF/UF membrane systems are driven primarily by feed water pumping
requirements. Operating pressures vary significantly over the course of chemical cleaning
intervals, depending on operating strategies and feed water quality. Typical operating pressures
(estimated from previous studies with the Polymem membrane on Lake Michigan water) range
from 5 to 15 psi. Based on a constant operating pressure of 15 psi, feed pumping power
requirements are approximately 0.157 kWh per day per gpm of feed water flow.
-------
2.3.3 Product Performance Capabilities
The UF120 S2 membrane is capable of removal of solids primarily through a physical sieving
mechanism. The UF120 S2 membrane will remove coliforms, HPC, and other participate matter
from the feed water. The manufacturer's stated performance capabilities were used to shape the
data quality objectives (DQO) and testing plan used for this ETV test.
Note: Challenge testing, microbial or otherwise, was not performed as part of this study; particle
removal was quantified based on turbidity and particle counter data.
10
-------
Chapter 3
Methods and Procedures
This chapter includes a detailed discussion of the ETV experimental plan, testing conditions,
methods, and sampling parameters and frequency. In addition, this chapter details field
operational and maintenance procedures, quality assurance and quality control, and analytical
methods used throughout the ETV testing.
3.1 Environmental Technology Verification Testing Plan
3.1.1 Task 1: Membrane Flux and Recovery
The objective of this task was to demonstrate: 1) appropriate operational conditions for the
membrane module; 2) the product water recovery achieved by the membrane module; and 3) the
rate of specific flux decline observed over extended membrane filtration operation.
3.1.1.1 Work Plan
Fouling rates of a membrane are functions of both water quality and operational conditions.
Several months of pilot plant operation were completed using the test water source prior to the
verification testing. Optimized operational conditions were estimated based on these data, which
were collected prior to March 1, 2002. The verification testing took place over a continuous
46-day period beginning March 11, 2002 and ending April 26, 2002. This time frame was
established to test the ability of the membrane to operate with cold water and during a period of
time when there have historically been turbidity spikes in the GBWUFP raw water. The
operational data used in the testing are shown in Table 3-1 and include flux, recovery, backwash
interval, data logging interval, production backwash chlorine concentration, air scour flow rate,
and terminal transmembrane pressure. The first run was set at a constant normalized flux of 40
1/h-m2 (24 gfd) at 20°C (68°F). The second run of testing was set at a constant normalized flux of
30 1/h-m2 (18 gfd) at 20°C (68°F). The system was operated at constant normalized flux to
minimize temperature related viscosity effects on transmembrane pressure and to facilitate
interruption of flux loss rate.
Operating conditions were set through the PLC human-machine interface (HMI) prior to
beginning the test. These conditions included flux rate, production dwell time, backwash
procedures (interval and duration), alarm condition settings, chemical feed doses, and data
logging intervals. With the exception of backwash duration, these parameters were not adjusted
during operation. Backwash duration was adjusted as needed to maintain a recovery of at least
90 percent. Recovery was determined daily by visually observing the total backwash volume
used during a production backwash cycle as indicated on the filtrate tank site glass. Total
production volumes were determined using the filtrate flow rate at the end of a production cycle
and multiplying by the total production cycle time (i.e. 50 minutes).
During the test run, specific operational data were collected to quantify fouling rates and
hydraulic performance. The operational data and the collection frequency are shown in Table 3-2
and include: feed/filtrate flow rate, feed/filtrate pressure, feed temperature, transmembrane
pressure, flux and specific flux at 20°C (68°F), and recovery.
11
-------
3.1.2 Task 2: Chemical Cleaning Efficiency
The objective of this task was to demonstrate the effectiveness of chemical cleaning for restoring
the specific flux of the membrane module. The cleaning procedures used were selected from the
manufacturer's recommended cleaning strategies, as indicated in the Operations and
Maintenance (O&M) Manual. The task was intended to serve as a "proof of concept" effort.
3.1.2.1 Work Plan
Prior to beginning the ETV testing, a membrane chemical cleaning procedure was performed to
ensure that the module was clean and that the storage solution had been purged. In addition,
chemical cleanings were performed once terminal transmembrane pressure was reached in order
to restore membrane permeability. Lake Michigan water has low concentrations of dissolved
minerals. Therefore, after membrane fouling, a chlorine solution at a pH of 11-12 was used to
oxidize and dissolve organic foulants, and to remove bacterial growth. A total of three chemical
cleanings were performed for this testing as a "proof of concept" effort: 1) prior to beginning the
verification test, 2) after approximately 12.5 days, just prior to when the terminal transmembrane
pressure was reached, and 3) at the end of the 46 days of run time.
The membrane module was chemically cleaned, per the manufacturers recommended cleaning
strategies, with approximately 200 mg/L chlorine and sufficient caustic soda to maintain a pH of
11-12 (5000-7500 mg/L NaOH). The chemicals were dosed automatically with chemical
metering pumps and set points entered into the PLC. Doses were verified by measuring total
chlorine and pH on site immediately following chemical injection. The first two cleanings were
performed at ambient temperature while the final cleaning was performed at elevated
temperatures of approximately 27°C (81°F). Prior to beginning the verification test, the new
membrane was allowed to soak in the chemical cleaning solution for a period of 4 hours. The
two subsequent chemical cleans were allowed to soak for 8 and 48 hours, after production
periods 1 and 2, respectively. Following the chemical cleaning soak, the module was
backwashed with sufficient volume of filtrate to purge the system of the chemical solution (>30
gallons) and was placed in filter-to-waste mode for 90 seconds. A baseline transmembrane
pressure versus flux trend was established immediately following each chemical clean. Chemical
cleaning wastes were neutralized and disposed of off site.
Table 3-3 provides information on the chemicals used and conditions of the chemical cleaning
procedure including chlorine consumption, chemical cleaning steps, pH and Cb of cleaning
residuals, and the volume of water required for chemical cleaning. Detailed procedures are
included in the membrane and CMPP O&M Manuals in Appendices A and B. Table 3-4 shows
the timing and location of samples collected during each chemical cleaning. The samples were
tested for parameters including temperature, pH, TDS, and turbidity. In addition, specific flux,
flux, transmembrane pressure, and flow rate was recorded. Collected data were used to verify
cleaning conditions and assess cleaning effectiveness.
Estimates of specific flux recovery, cleaning efficacy and assessment of irreversible loss of
membrane specific flux were reported for estimating the usable life of the membrane. A
calculation of specific flux recovery and original specific flux loss, as defined in the
12
-------
ETV Protocol for Equipment Verification Testing for Physical Removal of Microbial and
Particulate Contaminants (May 1999) were calculated as follows:
Jsf
% Recovery of Specific Flux = 100 *
Loss of Original Specific Flux = 100 *
1-
Jst_
Js,
Jsf= specific flux at end of current run
Jsj = specific flux at beginning of subsequent run
Jsio = specific flux at beginning of membrane test run
3.1.3 Task 3: Finished Water Quality
The objective of this task was to assess the ability of the membrane module to meet the water
quality objectives outlined by Polymem.
3.1.3.1 Work Plan
To adequately define ETV conditions and to assess the contaminant rejection capacity of the
UF120 S2, feed, filtrate, and backwash water quality were monitored over the course of the
testing. Water quality monitoring was performed per the requirements of the EPA/NSF ETV
Protocol for Equipment Verification Testing for Physical Removal of Microbial and P articulate
Contaminants (May 1999). The parameters, sampling frequency, and specific sampling locations
are presented in Table 3-5. On-site analysis was performed by the CMPP operator. Off-site
analysis was performed by NLS, Crandon, WI. Total coliform enumeration was performed by
MWH Laboratories, Pasadena, California. NLS and MWH state certification documents are
provided in Appendix C. Sampling procedures are detailed in the Section 3.1.7, Task 7: QA/QC.
3.1.4 Task 4: Reporting of Membrane Pore Size
The objective of this task was to report the 90 percent and maximum pore size for the Polymem
UF120 S2 ultrafiltration membrane.
3.1.4.1 Work Plan
The nominal pore size of the UF120 S2 membrane module used in this ETV study was
determined by the manufacturer through the use of multiple methods including scanning electron
microscopy (SEM), flow porometry, and particle retention testing. The UF120 S2 membrane is
an asymmetric hollow fiber with a dense layer containing small pores on the inside and outside,
and a less dense structure containing relatively larger pores. SEM images were used at multiple
resolutions to estimate the size of the pores in the membrane. Flow porometry (Lee, et al., 1997;
Hernandez, et al., 1999; Mietton and Courtois, 1997) was used to determine the mean size of the
pores in the membrane (including the less dense middle layer).
13
-------
3.1.5 Task 5: Membrane Integrity Testing
The objective of this test was to demonstrate the methodology to be employed for monitoring
membrane integrity, and to verify the integrity of the membrane fibers.
3.1.5.1 Work Plan
Following the initial and final chemical cleanings, an integrity test was performed consisting of
direct measurements including a pressure decay and visual bubble test. Indirect measurements of
membrane integrity were made throughout the ETV testing by means of on-line particle count
and turbidity data. The pressure decay and bubble test methodology is included in the membrane
O&M Manual in Appendix A and involved the steps described below.
3.1.5.2 Direct Methods of Membrane Integrity Testing
Pressure Decay Methodology
• Operating the module without air scour for several shortened backwash cycles to remove air
trapped between membrane fibers.
• Pressurizing the inside of the membrane fibers with air through a valve located on the CMPP
filtrate piping to 10 psi.
• Opening the feed sample valve so the outside of the membrane fibers is at atmospheric
pressure during the test.
• Recording measurements from the filtrate pressure transmitter on the HMI every
15 seconds until the 2-minute test has been completed.
• The membrane technical specifications state that a pressure decay rate of greater than
50 mbar over a 2-minute period (0.36 psi/min) could indicate a compromise in membrane
integrity.
Bubble check
• If bubbles are noted in the vertical, transparent section of the CMPP feed water piping during
the pressure decay test, this indicates that fibers have broken.
3.1.5.3 Indirect Methods of Membrane Integrity Testing
In-line particle counting and turbidity data were used as an indirect method of monitoring
membrane integrity. Increases in particle counts or turbidity in the filtrate serve as indicators of
potential fiber breakage. The advantage of this method was that it could be used during normal
operation and did not require the module to be taken out of service as with the direct methods
described above.
14
-------
3.1.6 Task 6: Data Handling Protocol
The objective of this task was to establish a viable structure for the recording and transmission of
field data such that the operational data collected was sufficient and reliable for execution of the
PSTP.
3.1.6.1 Work Plan
Water quality and hydraulic data was maintained on site. Critical process parameters including
pressures, temperature, turbidity, flow rates, and chemical doses were automatically stored on the
membrane skid mounted PLC Central Processing Unit (CPU). This data was also backed up on a
separate personal computer. This data was recorded at a maximum of 10-minute intervals
throughout the ETV testing. In addition, online feed and filtrate turbidities were verified once per
day with bench-top measurements. Backwash turbidity was measured twice per day with bench-
top analysis and was not monitored by online instrumentation.
Feed and filtrate particle count data were recorded at 5-minute intervals for the following size
ranges: 2-3 jim, 3-5 jim, 5-7 jim, 7-10 jim, 10-15 jim, and >15 jim. The instrument
communicates via a RS485 protocol, therefore a RS485/RS232 adapter was required in order to
communicate with the computer used to collect the data. Particle count data was collected on a
dedicated computer located adjacent to the CMPP skid. Both feed and filtrate particle counter
calibration were field verified with microspheres. Further details on the field verification
procedures are provided in Section 3.1.7.
pH, alkalinity, total hardness, calcium hardness, TDS, TSS, TOC, and UVA data were recorded
in tabular format as results were received from the analytical testing laboratory. MWH
Laboratories, Pasadena, California performed total coliform enumeration. Remaining laboratory
analyses were performed by NLS, Crandon, Wisconsin. Likewise, results from biological
samples were recorded in a tabular format. Specifically designed daily log sheets were filled out
to include other details of the test such as field tests, maintenance activities, water type,
backwash procedures, operating cycles, names of visitors, problems, etc. and were logged by the
operator in a bound notebook. Additionally, QA/QC information including manual checks of
pressure, flow and temperature transmitters, results of on-site analyses, instrumentation flow
rates, and other pertinent information were logged in the same manner. Chain of custody records
for off-site laboratory analysis, on-site calibration, and verification of on-line instrumentation
were also recorded. Copies of the laboratory notebooks and data log sheets are included in
Appendix D. Table 3-6 summarizes the information management for this testing including how
the information was collected and stored and the reported format.
3.1.7 Task 7: Quality Assurance/Quality Control
The objective of this task was to maintain strict QA/QC methods and procedures during the test
run.
15
-------
3.1.7.1 Work Plan
The quality assurance project plan (QAPP) for this verification testing specifies procedures that
were used to ensure data quality and integrity. Careful adherence to these procedures ensures that
data generated from the verification testing provided sound analytical results that serve as the
basis for this performance evaluation. The components of the QAPP for this ETV include:
• Routine field procedures;
• Analytical methods;
• Water quality precision and accuracy;
• Critical equipment precision and accuracy;
• Methodology for use of blanks;
• Description of procedures for performance evaluation samples;
• Outline for duplicate sampling;
• Procedures used to ensure data correctness;
• Procedures for calculating indicators of data quality;
• Outline for data reporting; and
• Development of corrective action plan.
3.1.7.2 Routine Procedures
The following procedures were performed prior to the test run:
• In-line turbidimeter reservoirs were cleaned out and calibrated against a 20 NTU standard.
• Particle counters were field-verified using microspheres.
• In-line flow meter was cleaned and the meter output was verified with a bucket and
stopwatch method.
• Sample tubing was checked.
The following procedures were performed daily:
• Routine daily walkthroughs were conducted to verify that each piece of equipment or
instrument was operating properly.
• In-line turbidimeter flow rates were verified.
• In-line turbidimeter readings were checked against a properly calibrated bench model.
• In-line particle counter flow rates were verified.
• Chemical feed pump flow rates were verified.
• Pressure transmitters were checked against pressure gauges that were calibrated against
NIST-traceable standards.
16
-------
The following procedure was performed at the start of the run, and every 2 weeks thereafter:
• In-line flow meter readings were verified with a bucket and stopwatch method.
3.1.7.3 Analytical Methods
Onsite bench-top analyses including turbidity, pH, chlorine, and temperature were conducted
daily at the test site according to Standard Methods for the Examination of Water and
Wastewater, 20( Edition (APHA, 1998) and by Methods for Chemical Analysis of Water and
Wastes (EPA, 1979), where applicable. Standard Methods for the Examination of Water and
Wastewater, 20th Edition (APHA, 1998) was followed for total coliform analyses conducted at
NLS, Crandon, Wisconsin and MWH Laboratories, Pasadena, California. Other analyses
conducted by NLS were conducted using Standard Methods for the Examination of Water and
Wastewater, 18th Edition (APHA, 1992) and by Methods for Chemical Analysis of Water and
Wastes (EPA, Revision 1983), where applicable. The analytical methods utilized for this test to
monitor feed, filtrate, and backwash water quality are further described below:
pH - Analysis for pH was performed according to Standard Method 4500-H+. A 3-point
calibration of the pH meter used in this study was performed daily.
Temperature - Measurement of temperature was conducted in accordance with Standard Method
2550 B. This was done daily to check the in-line temperature transmitter on the CMPP.
In-Line Turbidity - Feed and filtrate water in-line turbidity measurements were logged
continuously on the CMPP. The feed and filtrate turbidimeters were calibrated weekly with
primary calibration standards purchased from the turbidimeter manufacturer. In addition, lenses
were cleaned weekly according to the manufacturer's instructions. The data logging readout was
checked daily against the local turbidimeter display value. Turbidimeter flow rates were checked
daily to ensure that they were within the range required by the manufacturer.
Bench-top Turbidity - Turbidity analysis was performed daily according to Standard Methods
2130 with a bench-top turbidimeter for feed, filtrate, and backwash samples. These results were
used to verify in-line measurements. The bench-top turbidimeter (Hach 2100 AN) was checked
daily against primary standards (0.061, 20, 200 NTU).
The method for collecting grab samples consisted of running a slow, steady stream from the
sample tap, triple-rinsing a dedicated sample beaker in this stream, allowing the sample to flow
down the side of the beaker to minimize bubble entrainment. The sample vial was double-rinsed
with the sample, carefully pouring from the beaker down the side of the sample vial, wiping the
sample vial clean, inserting the sample vial into the turbidimeter, and recording the measured
turbidity upon a stable readout.
Particle Counting - In-line particle counting was performed on both feed and filtrate waters.
Prior to the study, the instrument calibration was field-verified using microspheres as described
in Section 3.1.7.5. Current instrument and particle standard calibration certificates and methods
for demonstration of coincidence error are provided in Appendix E.
17
-------
TOC and W/254 Absorbance - Water samples were collected by Carollo Engineers. Samples
were collected in containers supplied by NLS. Samples were held, preserved, and shipped in a
cooler at approximately 4°C (39°F) to NLS in accordance with Standard Method 5010B. TOC
analysis followed Standard Method 5310B, UV Absorbance followed Standard Method 5910B.
Microbial Parameters - Water samples were collected by Carollo Engineers. Samples were
collected in containers supplied by the testing laboratory and sent to NLS (or MWH) in a cooler
at approximately 4°C (39°F). Total coliform densities were reported in MPN/100 mL, following
Standard Method 9221. Total coliform was analyzed as "present" or "absent" using Standard
Method 9223. HPC densities were reported as colony forming units (cfu) per mL, per Standard
Method 9215B.
Total Alkalinity - Water samples were collected by Carollo Engineers. Samples were collected in
containers provided by NLS, and shipped in a cooler at approximately 4°C (39°F). Samples were
collected in accordance with Standard Method 3010B. Analysis was conducted per Standard
Method 2320B.
Total and Calcium Hardness - Water samples were collected by Carollo Engineers. Samples
were collected in containers provided by NLS, and shipped in a cooler at approximately 4°C
(39°F). Analysis was conducted per EPA Method 200.7.
Total Dissolved Solids - Water samples were collected by Carollo Engineers. Samples were
collected in containers provided by NLS, and shipped in a cooler at approximately 4°C (39°F).
Analysis was conducted per EPA Method 160.1.
Total Suspended Solids - Water samples were collected by Carollo Engineers. Samples were
collected in containers provided by NLS, and shipped in a cooler at approximately 4°C (39°F).
Analysis was conducted per EPA Method 160.2.
Total Chlorine - Total chlorine was measured daily by the Green Bay Water Utility, using
spectrophotometer methods, following HACK Method 8167 (equivalent to Standard Method
4500-C1 G). Sample results were provided by Green Bay Water Utility. In addition, one chlorine
measurement was taken from the CMPP feed water following HACK Method 8167, as noted
above, and served as verification of the readings reported by the Green Bay Water Utility.
3.1.7.4 Water Quality Precision and Accuracy
Table 3-7 describes the methodology used in this ETV for the measurement of precision and
accuracy for each water quality analysis performed during the test run. Duplicate samples for
analysis are also shown for on-site grab sample analysis. The sampling location for each
duplicate grab sample was the filtrate water.
3.1.7.5 Critical Equipment Precision and Accuracy
Flow Meter - Water flow rates were verified prior to the start of the testing and every
2 weeks thereafter. The verification was performed by bucket and stopwatch methods as follows.
The filtrate reservoir on the CMPP skid contains a sight glass. This sight glass contains five-
18
-------
gallon graduations. Following a backwash (when the reservoir is drawn down) a constant flow
rate was established and the operator calculated the amount of time that it takes 50 gallons of
filtrate water to accumulate in the filtrate reservoir. This data was used to calculate the average
flow rate and was checked against the flow rate as indicated on the data acquisition system of the
CMPP.
Pressure Transmitters - Pressure transmitters are located in both the feed and filtrate side of the
membrane module. These pressure transmitters output an analog signal to the PLC indicating the
gauge pressure measured as a function of time. Both of the pressure transmitters were checked
against an independent measurement in the form a pressure gauge (calibrated to NIST traceable
standards) mounted on both feed and filtrate sides of the membrane module. Pressure as
indicated on the data acquisition system of the CMPP was checked against the pressure gauge
readings 7 times per week.
In-Line Particle Counters - Factory calibration of particle counters was performed prior to the
start of the study. However, prior to the study, the instrument calibration was field-verified using
microspheres. Polystyrene sphere suspensions were purchased from EPS Analytical Standards
with concentration certifications provided (Appendix E). The testing was performed with spheres
with diameters of 2, 5 and 10 jim. The particle counter (MetOne) manufacturer's recommended
field procedures were used. Particle suspension concentrations of 1000 particles per ml were
used. The 2 jim particle suspension was supplied at a concentration of 1x109 particles per ml and
required field dilution with a micropipette. Particle suspensions for 5 and 10 jim were supplied at
a concentration of 1000 particles per ml from the manufacturer and did not require in-field
dilution. The guidelines used to perform the verification are listed below. Results of this
verification are presented in Chapter 4.
• The particle counter was thoroughly cleaned using the manufacturer's supplied cleaning
solution.
• The particle counter was then flushed using purified reverse osmosis water for a period of at
least 30 minutes.
• Microsphere suspensions were continuously stirred gently, to prevent entraining air bubbles
into the fluid.
• A peristaltic pump was used to transfer the suspension to the instrument at the required
steady flow rate of 100 ml/min for a period of at least 5 minutes to obtain stable readings.
• Measurements were electronically logged on a data acquisition system every 2 minutes, and
the results were checked against the concentration and size of microspheres in the stock
solution.
In-line Turbidimeters - In-line turbidimeters were calibrated according to manufacturer's
methods with a 20 NTU standard, provided by the manufacturer. Calibration of turbidimeters
was performed in advance of the verification period and once per week throughout the testing.
In-line turbidity measurements were checked daily against a properly calibrated bench-top
instrument.
19
-------
3.1.7.6 Methodology for Use of Blanks
The methodology for using method blanks is summarized in Table 3-8.
Spiked Samples
Spike samples were utilized by NLS for the water quality parameters listed in Table 3-8. Spike
samples were not used for analyses performed in the field. NLS utilized one spike sample per
QA/QC batch. Each batch consisted of approximately 10 samples analyzed for a single
parameter.
Travel Blanks
In order to assess potential travel related contamination, a total of four travel blanks were to be
submitted to the laboratory during the verification testing. The laboratory was to analyze half of
the travel blanks for total organic carbon and half for total coliforms. However, these travel
blanks were not performed and are a deficiency of this report.
3.1.7.7 Performance Evaluation Sample
An evaluation sample was analyzed in accordance with the procedures of NLS. The performance
evaluation sample submitted was a 20 NTU turbidity standard. This performance evaluation
sample was submitted at the start of the verification study by Carollo Engineers to NLS and was
returned with a reported value of 21 NTU.
3.1.7.8 Outline for Duplicate Sampling
Duplicate sampling was performed according to the outline presented in Table 3-7
3.1.7.9 Procedures Used to Ensure Data Correctness
The procedures used to ensure data correctness are detailed in Section 3.3 "Calculation of Data
Quality Indicators."
3.1.7.10 Procedures Used to Calculate Indicators of Data Quality
The procedures used to calculate indicators of data quality are detailed in Section 3.3
"Calculation of Data Quality Indicators."
3.1.7.11 Outline for Data Reporting
Reports were prepared by Carollo Engineers, P.C. Data was reported in a draft and final ETV
report, submitted by Carollo Engineers, P.C. Status reports were not submitted due to the short
duration of the tests.
3.1.7.12 Corrective Action Plan
The corrective action plan for each of the tested water quality parameters is summarized in
Table 3-9.
20
-------
3.1.8 Task 8: Operational Conditions and Maintenance
The objective of this task was to maintain appropriate operational conditions and maintenance
throughout the verification testing.
3.1.8.1 Work Plan
Technical specifications for the Polymem module were provided for this testing and include
normal operating ranges for temperature, pressure, flow, chemical compatibility, instructions for
conditioning, storage, integrity testing, fiber repair, and chemical cleaning. Carollo evaluated the
instructions and procedures in the membrane O&M Manual for their applicability during the
verification test. The CMPP O&M Manual is included in Appendix B. It contains procedures for
filtration, backwashing, chemical cleaning, and routine maintenance. Troubleshooting guidance
is also included.
3.2 Calculation of Operating Parameters
3.2.1 Flux
Membrane flux was calculated as follows:
^ifiltrate
Flux =•
Area
Where:
Qfiitrate = filtrate flow rate
Area = total active membrane surface area
3.2.2 Specific Flux
Specific Flux was calculated as follows:
Flux
Specific Flux =
IMP
Where:
Flux = the result of the flux calculation as shown above
IMP = transmembrane pressure
3.2.3 Normalized Specific Flux
Normalized Specific Flux was calculated as follows:
Stte cificFlux
Normalized Specific Flux = O0239.(r_20)
Where:
Specific Flux = the result of the specific flux calculation as shown above
T= Temperature in Celsius
21
-------
3.2.4 Feedwater System Recovery
Feed Water Recovery was calculated as follows:
' z?rj/
Recovery =
1-
BW
v
total
*100%
Where:
Vtotai = total volume of filtrate produced in a production cycle based on average flow rate and a
production cycle duration of 50 minutes.
VBW = total volume of filtrate used during a production backwash cycle
3.3 Calculation of Data Quality Indicators
Data quality parameters as specified in the ETV protocol include representativeness, statistical
uncertainty, accuracy, and precision. This section details how each of these parameters was
considered throughout this testing.
3.3.1 Representativeness
As specified by the ETV Protocol (EPA/NSF 1999), representativeness of operational parameters
entails collecting a sufficient quantity of data during operation to be able to detect a change in
operational parameters. As specified, detecting a plus or minus 10 percent change in operating
parameter is sufficient for proper QA/QC. Operational parameters including flow rate,
membrane feed, and filtrate pressures were recorded a minimum of five days a week, which NSF
specifies as sufficient for tracking changes in operational conditions that exceed this 10 percent
range.
In addition to ensuring representativeness of operational parameters, representativeness of water
quality samples was ensured by executing consistent sample collection procedures. These
procedures considered sample location, timing of sample collection, sample procedures, sample
preservation, sample packaging, and sample shipping as detailed below.
Sample Locations - The water quality monitoring matrix was presented in Table 3-5, which
presents the water quality monitoring plan for feed, filtrate, and backwash streams of the CMPP.
Further guidance on sampling locations is included in the O&M Manual/Standard Operating
Procedures for the CMPP presented in Appendix B.
Timing of Sample Collection - Feed water quality sampling was performed within one hour of
filtrate water quality sampling to ensure that the filtrate water sample was representative of the
membrane feed water quality. Filtrate water quality was relatively consistent throughout the
duration of the production cycle. However, it is not unusual for turbidity or particle counts to be
slightly higher at the beginning of a production cycle. Since this represents the worst case water
quality, filtrate water samples were collected within the first 15 minutes of a production cycle.
The PLC on the CMPP has a timer, which clearly indicated the duration of a particular
production cycle.
22
-------
The backwash procedure consisted of a three-step procedure. Step 1 included hydraulic
backwashing, air scour, and chlorine addition followed by a short rest period of approximately
five seconds. Step 2 repeated Step 1. Step 3 of the backwash procedure was only hydraulic; air
scour and chlorine addition were not utilized. Backwash samples were collected at the beginning
of the second backwash cycle. The backwash sample tap was opened throughout the duration of
the backwash procedure to ensure that a representative sample was collected.
Sampling Procedures, Preservation, Packaging, and Transport - Prior to the collection of each
individual water quality sample, the sample tap was allowed to run a minimum of 30 seconds in
order to purge the sample tap and sample line of stagnant water. Samples were then collected.
Additional considerations and procedures for individual water quality parameters are included
below:
pH - pH samples were collected at the sample tap in polypropylene beakers and immediately
tested for pH. The temperature at which the pH reading is made was also recorded.
Temperature - In addition to temperature transmitter readings (recorded continuously on the
PLC), temperature gauge readings were manually recorded daily. Temperature was also recorded
while measuring pH. Special preparation or sampling procedures were not necessary for this
measurement.
Turbidity (Bench-top) - The method for collecting bench-top turbidity samples followed the
procedure recommended in the testing protocol developed by the NSF. The procedure was as
follows:
• The slow steady stream was run from the sample tap.
• A dedicated sample cell was triple rinsed with the sample.
• The sample was allowed to flow down the side of the cell to minimize bubble entrainment.
• The sample cell was wiped clean.
• The sample cell was immediately inserted into the turbidimeter.
• The measured turbidity was recorded upon reading stabilization.
Alkalinity - Samples were collected in a polyethylene or borosilicate bottle provided by the
analytical laboratory. The sample was closed tightly and immediately placed into the sample
cooler for transport to the analytical laboratory. Sample agitation and prolonged exposure to the
air was avoided.
Total Hardness - Procedures for sampling total hardness followed alkalinity sampling
procedures.
Calcium Hardness - Samples were collected by the procedures for alkalinity and total hardness.
Total Dissolved Solids - Resistant glass or plastic sample bottles were used as provided by the
analytical laboratory. Samples were collected and immediately placed into the sample cooler for
23
-------
transport to the analytical laboratory. Analysis began as soon possible as specified by EPA
Method 160.1.
Total Suspended Solids - Sampling for total suspended solids followed the same procedure as
specified by total dissolved solids.
Total Coliform and HPC - Sample containers were provided by the analytical laboratory.
Aseptic sampling techniques were used as follows:
• Sample bottles were kept closed until they were filled.
• Sample taps were removed, allowed to soak in a chlorine solution for a minimum of two
minutes, rinsed, and reconnected to the sample valve. Water was then allowed to run through
the tap for a minimum of two minutes. The sample tap was flamed prior to sampling.
• The cap of the sample container was removed without touching the surface of the cap or neck
of the bottle.
• The sample container was filled without rinsing and the cap was replaced immediately.
• Samples were refrigerated immediately after collection and transported to the laboratory in
coolers with frozen blue ice.
• Samples were refrigerated upon receipt at the laboratory and analyzed within holding times
specified in their standard method.
3. 3.2 Statistical Uncertainty
For data sets of eight or more, statistical uncertainty was calculated for grab sample analyses
including TOC, turbidity, HPC, UV254, and TSS. This was done by calculating the 95 percent
confidence interval in the following manner:
- ( e
95% Confidence Interval = X+t , J •
Where:
X = sample mean
t = student t-test with n-1 degrees of freedom
S = sample standard deviation
n = number of independent measurements
3.3.3 Accuracy
Accuracy was quantified as the percent recovery of a parameter in a sample to which a known
quantity of that parameter was added. For this testing an example of accuracy determination in
the ETV was the analysis of a turbidity proficiency sample in comparison of the measured
turbidity of the known level of turbidity of the sample.
Accuracy = Percent Recovery = 100 * [(*_„ - x_e d) + Xknown ]
24
-------
Where:
x-known = known concentration of analyte added to the sample
= measured concentration of parameter
3.3.4 Precision
Precision refers to the degree of mutual agreement among individual measurements that provides
an estimate of random error. The standard deviation and relative standard deviation recorded
from sample analysis were reported as a means to quantify sample precision. The percent relative
standard deviation was calculated in the following manner. The standard deviation is as follows:
Precision Deviation =
-|l/2
n-\
Where:
x = sample mean
X[ = /'th data point in the data set
n = number of data points in the set
As specified in the ETV protocol provided by the NSF, the percent relative standard deviation
for drinking water samples must be less than 30 percent or acceptable precision under the
verification testing program.
Relative Percent Deviation = 1 - 2 * 100%
x
Where:
x = sample mean
x\ = first data point of the set of two duplicate data points
X2 = second data point of the set of two duplicate data points
3.4 Safety Measures
Membrane - The membrane was inspected for visual damage prior to installation, and operated
within the manufacturer's pressure and flow ranges.
Electrical - The electrical work for the CMPP was performed prior to the start of this study by a
licensed electrical contractor in accordance with local codes.
Chemicals - Non-compatible chemicals (acids/bases/chlorine/preservation solutions) were stored
separately from one another at the pilot site. Material Safety Data Sheets were stored on site and
are provided in Appendix F.
25
-------
3.5 Testing Schedule
Field operations specific to the PSTP began February 28, 2002. Testing specific to the PSTP
began March 11, 2002. The verification testing was terminated on April 26, 2002 after 46 days
of operation. Field activities related to this testing were finalized May 3, 2002.
The following key events took place during the test run.
• Conduct initial chemical clean to establish baseline conditions - Day 0;
• Conduct integrity test - Day 0;
• Start continuous membrane filtration pilot testing including sample collection, and hydraulic
performance monitoring as described in Chapter 4 of this report - Day 1;
• First chemical Clean - Day 12;
• End filtration operations - Day 46;
• Second chemical cleaning - Day 51; and
• Conduct integrity testing - Day 54.
26
-------
Chapter 4
Results and Discussion
4.1 Task 1: Characterization of Membrane Flux and Recovery
The operating conditions for the Polymem UF120 S2 module are provided in Table 4-1. These
test conditions were established based on previous pilot study optimization results conducted
from May 2001 to March 2002. The membrane system ran at a constant normalized flux of 40
L/hr-m2 (24 gfd) for the first 12.5 days of operation during the ETV test (Run 1). At the end of
Run 1, the membrane was chemically cleaned to restore flux and to reduce the required
transmembrane pressure. The remaining run time was operated at a constant specific flux of 30
L/hr-m2 (18 gfd) (Run 2).
The backwash interval, or production cycle, was set at 50 minutes of operational run time
throughout the testing and was followed by a total production backwash time ranging from 60-
120 seconds. The backwash cycle duration was adjusted throughout the testing to maintain a
system recovery of at least 90 percent. Backwash water was chlorinated at a target dose of 5
mg/L for the entire duration of the testing and was verified with daily sampling.
Figure 4-1 shows a time series plot of TMP, normalized flux, normalized specific flux, and
system recovery data throughout the test. Operational flux and temperature data are presented as
a time series plot in Figure 4-2. Table 4-2 summarizes operational data collected throughout the
testing. System recovery was calculated daily based on average flow rates and total backwash
volumes as measured on a daily basis by plant operators. Recovery ranged from 89-96 percent
throughout the 46-days of testing. The figures show a plant shut down on 4/18/02 resulting from
a power outage at the CMPP. When the plant was restarted, the default operational setpoints
(different from the target operational data) caused discontinuity in the data. When these set
points were changed back to the target values, system operations returned to normal.
At the start of Run 1 (clean membrane), the TMP began at approximately 0.34 bar (5 psi).
During Run 1, there was a nearly linear rise in TMP at a rate of approximately 0.087 bar per day
(1.5 psi/day). After a chemical clean, the TMP was restored to approximately 0.45 bar (6.5 psi)
and decreased to approximately 0.41 bar (6 psi) within one hour of start up. During Run 2, the
TMP remained stable near 6 psi for approximately 24 hours. The remainder of Run 2 produced a
nearly linear rise in TMP of approximately 0.039 bar per day (0.57 psi/day).
Normalized specific flux at the start of Run 1 was approximately 118 L/hr-m2-b (4.79 gfd/psi).
Due to fouling, normalized specific flux decreased to approximately 28 L/hr-m2-b (1.1 gfd/psi)
by the end of Run 1 with a TMP of 1.43 bar (21 psi). Chemical cleaning restored the normalized
specific flux to approximately 74 L/hr-m -b (3.0 gfd/psi). Run 2 lasted nearly 33 additional days
without another chemical clean. At the time the testing was terminated (at the end of Run 2), the
normalized specific flux was approximately 17 L/hr-m2-b (0.70 gfd/psi) with a TMP of 1.7 bar
(25 psi). The improvement in specific flux decline trends during Run 2 is likely due to the lower
target normalized flux. It should be noted that the 25 percent decrease in normalized flux led to a
260 percent increase in run time before a required chemical cleaning (12.5 vs. 32.7 days).
27
-------
Following a review of the data, the manufacturer suggested that the following changes may have
adversely impact the specific flux decline rate:
• Low Air Scour Flow Rate: Polymem suggested that flow rates are typically 5 m3/hr (2.9
cfm) at standard temperature and pressure (STP). This information is not included in the
O&M Manual provided by Polymem.
• The Run 1 flux rate was selected based on a recovery of 90 percent; however, the membrane
was operated at a median recovery of 94 percent.
4.2 Task 2: Evaluation of Cleaning Efficacy
A total of three chemical cleanings were performed for this testing as a "proof of concept" effort:
1) Prior to beginning the verification test. This initial chemical cleaning was performed on 3/5/02
instead of the first day of the testing (3/11/02) because CMPP equipment verification and the
testing of membrane baseline flux conditions required a few days of work prior to the official
ETV starting date of 3/11/02; 2) following Run 1 when the transmembrane pressure reached 1.4
bar (21 psi). Although the terminal transmembrane pressure of the membrane module was 29 psi,
during Run #1 of the ETV testing, the rate of TMP rise was such that the terminal pressure
would have been exceeded during the weekend when no operator would have been present to
stop the unit. This run was ended somewhat prematurely to avoid possible fiber and module
damage; and 3) following Run 2 at the end of ETV testing, when the transmembrane pressure
reached 1.7 bar (25 psi). Tables 4-3 and 4-4 summarize collected chemical cleaning water quality
and hydraulic data, respectively. It should be noted that, for Chemical Cleaning #3, the TDS
samples were analyzed exceeding the EPA sampling holding times. Table 4-5 summaries the
collected chemical cleaning efficacy data.
The recovery of specific normalized flux for Chemical Cleaning #s 2 and 3 was 62 and 73
percent, respectively. Chemical cleaning conditions (chlorine concentration, pH) were selected
from the manufacturer's recommended procedures. Cleaning # 2 was performed at ambient
water temperature, [14-18.6°C (57-65.5°F)], pH = 12.2, and an average total chlorine
concentration of 164 mg/L, for 8 hours.
Because recovery of specific flux was low, Cleaning # 3 was performed with a similar cleaning
solution but at elevated solution temperature [22 - 31°C (72-88°F)], for an extended soaking
period. Despite these changes the specific flux recovery was marginal (73 percent).
Following a review of the data, the manufacturer suggested that, due to the observed calcium
hardness (average of 88 mg/L during ETV test), some CaCOs fouling could have occurred and
an acid clean would be necessary to restore the specific flux. Additionally, recirculation of the
cleaning solution during cleaning has been shown to produce higher specific flux recoveries than
those observed on other source waters (Hugaboom, et. al, 2001). Further optimization would be
required to improve these recoveries, however, this was not the goal of the study. Despite the
low specific flux recoveries, the goal of providing "proof of concept" was achieved. Ideal
conditions would likely produce much greater cleaning efficiency. Therefore, an accurate
account of the usable membrane life cannot be estimated. The manufacturer's recommended
cleaning procedure is included in Appendix A.
28
-------
4.3 Task 3: Evaluation of Finished Water Quality
Result summary sheets for off-site analyses are provided in Appendix G.
4.3.1 Turbidity, Particle Counts, and Particle Removal
Figure 4-3 presents the on-line and bench-top turbidity profiles recorded throughout the testing.
The figure shows data for feed, filtrate, and backwash water. The bench-top turbidity analysis
results are summarized in Table 4-6 including feed, filtrate, and backwash water analyses. On-
line turbidity data are summarized in Table 4-7. Overall, there was close mutual agreement
between bench-top and online measurements. As shown in Table 4-7, feed water turbidity ranged
from 0.2 to over 19 NTU with an average of 1.3 NTU. Bench-top backwash water turbidity
averaged about 11 NTU. Bench-top filtrate turbidity was typically below 0.05 NTU. Online
turbidity data showed two anomalous filtrate readings dated 3-19-02 and 3-22-02 with readings
of 0.002 and 0.45 NTU, respectively. These data were not included in statistical analyses. Due to
the size of the database for online turbidity and particle counts, this data is not included in the
appendix. However, this data is on file with the NSF.
Table 4-7 also shows online feed and filtrate particle count data. Total feed particle
concentrations (>2 jim) averaged about 4,300 particles/ml. Filtrate particle counts averaged 4
particles/ml. The 90th percentile for feed and filtrate total particle counts (>2 #/ml) was
approximately 9,911 and 2 particles/ml, respectively. Average particle log removals of 4.2, 4.1,
4.1, 3.4, 3.3, 2.9, and 2.2 were achieved for particle size ranges of >2 urn, 2-3, 3-5, 5-7, 7-10, 10-
15, and >15 um, respectively. The membrane system removed 3.1 logs of total particles 90
percent of the time. It should be noted that the particle count instruments collected sample
volumes of 100 ml per each data point. If no particles were detected in that sample volume in the
filtrate, the filtrate particle count data was recorded as 0.00 particle/ml (below the detection limit
for the instrument of 1 particle). Since these data were recorded as zero values, log removal data
could not be calculated for these data points and were not included in the statistical analyses.
Because the membrane system produced relatively consistent filtrate particle counts, log
removals increased during periods when feed water particle counts were higher and decreased
during periods when feed water particle counts were lower.
Figures 4-4 through 4-10 present the time series particle count profile data (>2 jim, 2-3 jim, 3-5
|im, 5-7 um, 7-10 jim, 10-15 jim, and >15 jim) that was collected throughout the testing. Figures
4-11 through 4-17 show log removal profiles for each particle size. The presented data was
collected at 5-minute intervals and includes data collected during backwashes. Evaluation of the
time series data showed relatively higher particle counts during, and immediately following, a
backwash. As a result, particle removals were decreased during these times. These consistently
brief occurrences caused log removal data to decrease or even become negative for short periods.
The reasons for this are described below.
During backwash cycles, the feed pump was turned off thereby halting flow to the feed particle
counter. Because the particle counter output depends on flow rate, this decrease in flow to the
particle counter resulted in lower particle counts.
29
-------
The filtrate particle counter data was affected during backwash cycles as well. The surge in flow
during backwash cycles can cause hydraulic and air bubble turbulence in the particle counter
feed weir. The stirring up accumulated particles within the weir itself and introduction of air
bubbles lead to higher particle counts. Furthermore, the piping used for backwashes was fed
from the bottom of the filtrate tank that may have accumulated particles and air bubbles (caused
by increases in temperature and subsequent degassing) during the stagnant 50-minutes of a
production cycle.
The figures also show higher filtrate particle count data immediately following chemical
cleanings for the same reasons are described above. After system stabilization (within
approximately 4 hours), filtrate particle counts reached typical low-level readings.
For evaluation purposes the data for turbidity, particle counts, and log removal were plotted as
frequency distribution curves. Frequency distribution curves are shown in Figures 4-18 through
4-28. The 90th percentile for feed and filtrate turbidity was 3.5 and 0.035 NTU, respectively. The
90th percentile for feed and filtrate total particle counts (>2 #/ml) was approximately 9,911 and 2
particles/ml, respectively. It should be noted that filtrate particle count average (as shown in
Table 4-7) was higher than the 90th percentile. This is due to the fact that statistical averaging is
largely effected by higher values, especially when the majority of the data is near zero as was the
case in the filtrate particle count data set.
Figures 4-29 and 4-30 show a sensitivity analysis performed on particle count data collected
from one 24-hour period performed to determine the potential effects of backwash events on
calculated log removals. Data from March 14, 2002 was chosen for this analysis due to the
clusters of relatively lower log removal data, thereby representing a worse case scenario. Figure
4-29 presents the "raw" time series of feed and filtrate particle count data (>2 jim) that was
collected throughout the 24-hour period and includes markers indicating likely backwash events.
The presented particle count data was collected at 5-minute intervals and includes data collected
during backwashes. Figure 4-30 presents the time series of feed and filtrate particle count data
(>2 |im) collected throughout the same 24-hour period, with the particle count data removed that
were likely collected during backwash events. It should be noted that Figures 4-29 and 4-30 are
plotted on a log scale. As such, differences in filtrate particle counts are more apparent than with
the feed count data. As shown in Figure 4-30, filtrate particle counts did not exceed 10 #/ml
when the filtrate particle counts recorded during the backwash events were excluded from the
data set.
Figure 4-31 shows the log removal frequency distribution curves calculated for the raw data set
(data including backwash events) and for the data excluding particle count data likely collected
during backwash events. Log removals (>2 jim) for the raw data set and for the data set without
particle counts likely collected during backwash events were 3.2 or greater 90 percent of the time
and 3.6 or greater 90 percent of the time, respectively.
4.3.2 Microbial Removal
The removal of naturally occurring bacteria was also monitored throughout the ETV study. A
summary of this data is shown in Table 4-8. Total coliform bacteria were analyzed through two
means: 1) through a "presence/absence" (P/A) test and 2) through total coliform enumeration.
30
-------
One of the P/A test results, from the sample collected on 3/18/02, confirmed the presence of total
coliforms in the feed water. However, total coliforms were not detected in the corresponding
filtrate sample. Total coliform bacteria was detected in one P/A filtrate sample on 4/12/02 and
one P/A backwash sample on 4/22/02. Total coliform enumeration results showed feed
concentrations ranging from
-------
However, each filtrate sample with detectable TSS was reported as greater than its corresponding
feed water sample. These results are likely due to the fact that feed and filtrate samples were so
near the detection limit of the analysis. Due to the length of time the equipment was in use prior
to the ETV testing, it is also possible that material had built up in the portion of sample piping
permanently fixed to the CMPP skid. Although the sample ports were allowed to flush prior to
sample collection, accumulated material may have sloughed off during some of the sampling
periods. Specifically, TSS data collected on April 5, 2002 showed strong evidence of a sample
mix-up. Feed, filtrate, and backwash TSS values were reported as <1, 22, and 1 mg/L,
respectively. This data was excluded from statistical analysis.
Table 4-10 presents the mass balance conducted on the TSS results obtained throughout the
testing. In the cases where the measured TSS was below the detection limit (<1 mg/L), half of
the detection limit was used for the calculations. Six of the 36 calculated results showed positive
correlation (<50% relative percent deviation) between calculated and measured backwash water
TSS. These relatively poor correlations may in part be due to the feed water TSS being close to
the method detection limit. In addition, backwash water samples were collected during the
second cycle of the backwash. It is likely that the majority of TSS were flushed out during the
first cycle of the backwash. Therefore, the reported backwash water TSS concentration may not
accurately represent the entire backwash water volume consumed.
4.4 Task 4: Reporting Membrane Pore Size
The nominal pore size of the UF120 S2 membrane module used in this ETV study was
determined through the use of multiple methods; however, the manufacturer has not determined
a pore size distribution. Methods used in determining nominal pore size include scanning
electron microscopy (SEM), flow porometry, and particle retention testing. The first two
methods are described here.
The membrane is an asymmetric hollow fiber with dense layer containing small pores on the
inside and outside, and a less dense structure containing larger pores. SEM images were used at
multiple resolutions to estimate the size of the pores in the membrane. Based on these images,
the estimated nominal pore diameter is less than 100 nanometers (nm).
Flow porometry (Lee, et al., 1997; Hernandez, et al., 1999; Mietton and Courtois, 1997) was
used to determine the mean size of the pores in the membrane (including the less dense middle
layer). The mean pore diameter used in this procedure was determined to be less than 50 nm. The
above data are taken from a letter supplied by the manufacturer that is included in Appendix H.
This data is provided for informational purposes only and the results were not verified during the
ETV testing.
4.5 Task 5: Membrane Integrity Testing
A total of two membrane integrity tests were performed for this testing: 1) prior to the start of the
test and 2) following the final chemical cleaning. The first integrity test was performed with an
applied air pressure of 10 psi. Measurements of air pressure were taken every 15 seconds for a
period of 2 minutes. Over the course of the 2 minutes there was zero loss in pressure and no
visual indicators of a loss of membrane integrity (no bubbles were detected). The second
32
-------
integrity test was again performed with an applied pressure of 10 psi. Measurements were taken
every 15 seconds for a period of 2 minutes. The test was repeated 4 times yielding an average
pressure drop of 0.35 psi/min. This was within the allowable pressure drop as specified by the
manufacturer (<0.36 psi/min). However, during the visual inspection a small amount of air
bubbles were noted in the vertical, transparent section of the CMPP feed water piping. This
suggests that a fiber (or fibers) may have broken.
In addition to the visual air bubble test, in-line particle counting and turbidity data was used as an
indirect method of monitoring membrane integrity. Increases in particle counts or turbidity in the
filtrate serve as indicators of potential fiber breakage. During the ETV testing, there were no
significant increases in filtrate particle counts or turbidity that would have indicated a loss in
membrane integrity. However, following ETV testing (but prior to any further operation of the
membrane module) on 6/5/02, the membrane module filtrate end cap was removed to further
investigate the bubbles noted during the final integrity test. This investigation followed the
integrity test/repair procedures outlined in the Polymem UF120 S2 O&M Manual. One broken
fiber was identified and repaired. One subsequent pressure decay test, performed as described
above, yielded a zero loss in pressure and no visual indicators of a loss of membrane integrity
(no bubbles were detected).
4.6 Task 6: Data Management
4.6.1 Data Recording
Water quality and hydraulic data was maintained on site. Critical process parameters including
pressures, temperature, turbidity, flow rates, and chemical doses were automatically stored on the
membrane skid mounted PLC CPU. This data was also backed up on a separate personal
computer. This data was recorded at a maximum of 10-minute intervals throughout the ETV
testing. Feed and filtrate count data was recorded at 5-minute intervals for the following size
ranges: 2-3 |im, 3-5 jim, 5-7 jim, 7-10 jim, 10-15 jim, and >15 jim.
Data were manually recorded in tabular format for water quality parameters received from NLS
including alkalinity, total hardness, calcium hardness, TDS, TSS, TOC, and UVA. Likewise,
results from biological samples were recorded in a tabular format.
Onsite bench-top analyses were recorded in specifically designed daily log sheets. Other details
of the ETV testing such as field analyses, maintenance activities, water type, backwash
procedures, operating cycles, names of visitors, problems, etc. were logged by the operator in a
bound notebook. Additionally, QA/QC information including manual checks of pressure, flow
and temperature transmitters, results of on-site analyses, instrumentation flow rates, and other
pertinent information were logged in the same manner. Chain of custody records for off-site
laboratory analysis, on-site calibration, and verification of on-line instrumentation were also
recorded. Copies of the laboratory notebooks and data log sheets are included in Appendix D.
4.6.2 Data Entry, Validation, and Reduction
Data were entered from data sheets into similarly designed data entry forms in electronic
spreadsheet format. Following data entry, the spreadsheets were printed and checked against
33
-------
handwritten datasheets. All corrections were noted on the hard copies and were then corrected on
the electronic copy. The hard copies of the electronic data are included in Appendix D.
4.7 Task 7: Quality Assurance/Quality Control (QA/QC)
QA/QC results and chain of custody records reported by NLS and MWH are included in
Appendix I.
4.7.1 Data Correctness
There are five indicators of data correctness including representativeness, statistical uncertainty,
completeness, accuracy, and precision. This section includes the summary of analyses conducted
to ensure correctness of the data. The methods used for data analysis are outlined in Chapter 3.
Sampling and testing protocols were conducted per the requirements of the ETV Protocol for
Equipment Verification Testing for Physical Removal of Microbial and Particulate
Contaminants (May 1999).
4.7.1.1 Representativeness
Representativeness of the data was ensured through strict adherence to sampling and testing
methods outlined in Chapter 3. In addition, sampling efforts were coordinated so that sampling
was conducted at a consistent time and location throughout the testing.
4.7.1.2 Statistical Uncertainty
For data sets of eight or more, statistical uncertainty was calculated for grab sample analysis
including TOC, turbidity, HPC, UV254, and TSS. In addition, statistical uncertainty was
calculated for online measurements such as flux, flow rate, temperature, system recovery, etc.
Statistical uncertainty was evaluated by calculating the 95 percent confidence interval. The
results were presented in the previous sections.
4.7.1.3 Completeness
Data completeness refers to the amount of data collected during the ETV testing compared to the
amount of data proposed in the PSTP. Data completeness was determined for onsite water
quality measurement, laboratory water quality measurement, and operational data recording.
Completeness tables can be found in Appendix J. Nearly 100 percent of the parameters were
complete. However, travel blank samples were not collected during this testing. CMPP feed
water chlorine residual was scheduled for testing once per week to verify the daily measurements
collected as a part of the GBWUFP. Only one CMPP sample was collected during the first day of
testing to verify the readings recorded by the utility. It should also be noted that the backwash
water total coliform sample collected on 4/17/02 was not analyzed due to NLS error.
4.7.1.4 Accuracy
Accuracy was quantified as the percent recovery of a parameter in a sample to which a known
quantity of that parameter was added. Accuracy determination in this ETV testing was performed
by the analysis of a turbidity proficiency sample and onsite bench-top turbidimeter standards. A
34
-------
comparison was made between the measured turbidity and the known turbidity level of the
standard. Bench-top turbidity accuracy ranged from 84-113 percent with an average of 98
percent. The accuracy of the turbidity proficiency sample analyzed by NLS from a sample
collected on 3/12/02 was 95 percent. This data is included in Appendix I. Accuracy was also
ensured by calibration of bench-top and online turbidimeters as well as particle counter and flow
meter verification as outlined in Chapter 3. Results for the online particle counter calibration
verification are shown in Tables 4-11 and 4-12 for the feed and filtrate particle counters,
respectively. Standards for 2, 5 and 10 um particle sizes were used for the verification. Each
standard solution had a total particle concentration of 1000 particles/ml. The data show good
correlation between the tested standard and particle counter output. A minor amount of particles
were detected outside of the range of each standard. This is likely due to the standard solutions,
which contain a distribution of particle sizes rather than a pure concentration of one particle size.
Accuracy data is presented in Appendix D.
4.7.1.5 Precision and Relative Percent Deviation
Duplicate measurements were taken throughout the ETV testing as outlined in Chapter 3. Each
duplicate measurement was analyzed to determine the consistency of sampling and analysis
using relative percent deviation. The relative percent deviations averaged within 9 percent for
onsite bench-top turbidity measurements. General water quality parameters (hardness, calcium
hardness, alkalinity, algae, TDS, TOC, and UVA) sampled for the waters (feed, filtrate, and
backwash) were as high as 40 percent with an average of 9 percent relative deviation. TSS
analyses were as high as 189 percent (backwash) with an average of 50 percent relative
deviation. HPC analyses for the feed and filtrate were as high as 150 percent with an average of
30 percent relative deviation. These data are included in Appendix G.
Relative percent deviation is greatly effected when analytical measurements are close to the
lower detection limit. Specifically filtrate HPC and TSS were largely effected in this manner. 33
out of 38 filtrate HPC results were at or below the detection limit. 32 of 37 TSS results were at or
below the detection limit.
4.8 Operational Conditions and Maintenance
4.8.1 Overall Operation and Maintenance
There were no major reoccurring problems experienced during this ETV testing program.
However, plant operators noticed a few occasions of air bubble entrapment in the particle
counter feed lines signaled by a decrease in particle counter flow. Influent waters were
approximately 4°C (39°F) while the CMPP trailer was kept near 20°C (68°F). This temperature
difference could have caused degassing of dissolved species and subsequent entrapment of air
bubbles; however, air bubble introduction during backwash flow surges was the likely cause of
entrapment of air bubbles. Particle counter feed flow rates were verified daily and corrective
action was taken if flow restrictions were found. A small tubing brush was used to remove the air
bubbles.
35
-------
4.8.2 System Chemical Consumption
Chemical consumption was calculated for the system based on data collected throughout the
testing. Daily measurements of total backwash volume and chlorine dose were used to estimate
daily consumption of chlorine for production backwashes. Chemical cleaning chemical
consumption was based on the input parameters for sodium hydroxide and chlorine. Chemical
dosing was verified through onsite pH and chlorine measurements taken immediately following
chemical injection.
Table 4-13 provides a summary of chemical consumption during the ETV test. For Run 1, a total
of 0.61 pounds of chlorine (as NaOCl) were consumed for production backwashes. The chemical
cleaning following Run 1 consumed 0.06 pounds of chlorine (as NaOCl) and 2.04 pounds of
NaOH (as NaOH). For Run 2, a total of 1.06 pounds of chlorine (as NaOCl) were consumed for
production backwashes. The chemical cleaning following Run 2 used 0.06 pounds of chlorine (as
NaOCl) and 1.47 pounds of NaOH (as NaOH).
4.8.3 Review of Operations Manual
After reviewing the O&M Manual provided by Polymem, the following points were noted:
• The O&M Manual should be reviewed for grammar and word usage;
• The air scour flow rate during backwash should be specified;
• Chemical cleaning information should contain the recommendation that heated water be used
if cleaning is inefficient at ambient temperatures;
• Pages 4 and 8 recommend procedures for integrity testing and for identifying a compromised
fiber. The O&M Manual should be clarified to explicitly state the purpose of these two
procedures.
4.8.4 Equipment Deficiencies Experienced During the ETV Program
4.8.4.1 Online Particle Counting
Throughout the testing there were no mechanical problems with the particle counter equipment.
However, there was one period of approximately three days from 3/29/02 3:48 to 3/31/02 17:10
in which collected particle count data was unavailable for retrieval from the data logging
software. The operators experienced similar problems prior to the start of the ETV testing. The
data logging software technical support staff were unsuccessful in diagnosing this problem. In
addition, plant operators noticed a few occasions of air bubble entrapment in the particle counter
feed lines as described in section 4.8.1.
4.8.4.2 Membrane Equipment and Online Turbidimeters
One CMPP shut down was experienced during the testing during the night of 4/18/02 due to
lightning. The online turbidimeters experienced damage resulting from an electrical overload and
were out of commission until replacement parts were available during the first week of May. The
CMPP shut down lasted approximately 10 hours. Operational staff successfully restarted the
CMPP with no concerns outside of the online turbidimeters.
36
-------
Chapter 5
References
EPA/NSF International. ETVProtocol for Equipment Verification Testing for Physical Removal
of Microbiological and Particulate Contaminants, May 1999.
U.S. Environmental Protection Agency. Methods for Chemical Analysis of Water and Wastes,
EPA 600/479-020, March 1979.
U.S. Environmental Protection Agency. Methods for Chemical Analysis of Water and Wastes,
EPA 600/479-020, Revision March 1983.
Hugaboom, D. A., Roquebert, V.R., Crozes, G.F., Seacord, T.F., and Mahady, J.J., Retrofit of
Existing Granular Media Filters With Ultrafiltration Membrane Proceedings, AWWA Water
Quality Technology Conference, 2001, Nashville, TN.
Hernandez, J.I., Calco, P., Pradanos, P., and Palacio, L. A Multidisciplinary Approach Towards
Pore Size Distributions of Microporous and Mesoporous Membranes In Surface Chemistry and
Electroschemistry of Membranes. Ed. T.S. Sorensen, 1999.
Jacangelo, J.G., S.S. Adham, and J-M. Laine. Mechanism of Cryptosporidium, Giardia, and
MS2 Virus Removal by MF and UF, Journal AWWA, 87(9)107-121, 1995.
Lee, Y., Jeong, J., Young, I.J., and Lee, W.H. Modified Liquid Displacement Method for
Determination of Pore Size Distribution in Porous Membranes, Journal of Membrane Science,
130(1+2) 149-156, 1997.
Mietton Puechot, M. and Courtois, T. (1997) The Performance Limits of Bi-Liquid Porometry in
Determining the Distribution of Membrane Pore Diameters, Euromembrane '97, June 2-3, 1997,
held at University ofTwente (Netherlands).
Standard Methods for the Examination of Water and Wastewater, 1992, 18* Edition, APHA,
AWWA, and WEF, Edited by L.S. Clesceri, A.E. Greenberg.
Standard Methods for the Examination of Water and Wastewater, 1998, 20th Edition, APHA,
AWWA, and WEF, Edited by L.S. Clesceri, A.E. Greenberg.
37
-------
Chapter 6
Vendor Comments
Polymem submitted the following comments on the DRAFT report to the NSF. These comments
were not included in the body of the text.
1. The pilot plant [CMPP] was not equipped with a dedicated air compressor for providing air for
production backwashes. As a result, the airflow rate dropped below the specified flow rate
during backwashing.
2. Concerning airflow rate during backwashing, the manufacturer typically recommends a
constant flow rate of 3 to 5 m3/h at standard temperature and pressure. Due to the limited
capacity of the compressor, this recommendation was not strictly followed.
3. An acid cleaning was not performed following Run 1, which may have increased the specific
flux.
4. It is believed that the membrane fibers were uncompromised during the test run as evidenced
by the turbidity and particle count profiles. If membrane integrity had been compromised, it
would be expected to find a notable increase in particle concentrations in the filtrate.
38
-------
Tables and Figures
39
-------
40
-------
Table 1-1. Lake Michigan Feed
Parameter Units Count
Cl2-Residual(2)
Alkalinity
Total Hardness
Calcium
Hardness
TDS
TSS
Total Coliforms
Total Coliforms
HPC
TOC
UVA
Algae
pH
Temperature
Turbidity
(Bench-top)
mg/L
mg/L as
CaCOs
mg/L as
CaC03
mg/L as
CaC03
mg/L
mg/L
P/A
MPN/ 100
ml
CFU/ml
mg/L
cm"1
#/ml
-
°C (°F)
NTU
42
2
2
2
o
J
37
38
5
38
13
13
3
40
76
39
Water Quality Data
Median'1' Range
0.05
N/A
N/A
N/A
200
<1
N/A
<1.1
3
2.2
0.022
32.8
7.79
2.9 (37)
1.0
0.03-0.06
110-110
130-130
87-88
150-210
<1-8.0
A_p(3)
-------
Table 3-1 Verification Conditions
Operating Parameter Verification Test Value
Flux 30-40 1/h-m2 @ 20°C (1) (18-24 gfd)
Recovery >90%(2)
Backwash Interval 50 min(3)
Data Logging Interval 10 minutes (4 times per backwash cycle)
Production Backwash Chlorine Concentration 5 mg/L
Air Scour Flow Rate Approximately 2 cfm
Terminal Transmembrane Pressure 29 psi
1) Test flux was determined based on results of pilot testing prior to the verification study.
2) Recovery was at least 90%, however the exact values were calculated daily during verification testing and are reported in
Chapter 4.
3) Backwash duration was set to achieve minimum recovery criteria.
Table 3-2 Operational Data
Operation Parameter Frequency
Feed/Filtrate Water Flow(1) Continuous
Feed Pressure Continuous
Feed Temperature Continuous
Filtrate Pressure Continuous
Transmembrane Pressure Continuous
Flux @ 20°C Continuous
Specific Flux @ 20°C Continuous
Recovery Daily
1) Feed flow equals filtrate flow as measured on the filtrate side.
42
-------
Table 3-3 Chemical Cleaning Procedures and Conditions
Chemical
Approximate Quantity
Sodium Hypochlorite
Sodium Hydroxide
0.05 Ib (as NaOCl) /module/month
2.4 Ib (as NaOH)/module/month
Cleaning Step
Hydraulic Condition
Duration
Step 1 - Chemical backwash
Step 2 - Cleaning soak
Step 3 - Cleaning rinse
Backwash to drain
Soak at elevated temperatures not to exceed 90°F (32°C)
Backwash to drain
60 seconds or approx.
35 gal
4, 8, and 44 hours0)
90 seconds
Residual Characteristic
Approximate Value
pH
Residual Chlorine
Total Volume of Residual
10-12
150 mg/L as C12
150 gallons/module/cleaning
1) Three chemical cleans were included as part of this testing. The first clean took place prior to membrane operation and was
allowed to soak for 4 hours. The second and third chemical cleans were performed after elevated TMPs were reached and
were allowed to soak for 8 and 44 hours, respectively.
Table 3-4 Cleaning Verification Data
Analytical Parameter
Temperature
pH
TDS
Turbidity
Visual observations
Operational Data
Specific Flux @ 20°C
Flux @ 20°C
Transmembrane pressure
Flow rate
Sample Timing
2,3,4,5
2,3,4,5
2,3,4,5
2,3,4,5
2,3,4,5
Sample Timing
1,5
1,5
1,5
1,5
1 ) Prior to stopping production step.
2) Immediately following chemical dosing backwash for chemical cleaning.
3) 120 minutes following chemical dosing backwash for chemical cleaning.
4) During cleaning rinse backwash at the end of the chemical soaking period.
5) During production immediately following rinsing.
43
-------
Table 3-5 Water Quality Sampling Schedule
Sampling
Parameter Frequency Feed Filtrate Backwash
On-Site Analysis
pH I/day X
Temperature Continuous X
Turbidity (in-line)(1)(2) Continuous X X 2/day
Particle Counts (in-line)(3) Continuous X X
Total Chlorine(4) I/week X
Off-Site Laboratory Analysis
Alkalinity I/month X X
Total Hardness I/month X X
Calcium Hardness I/month X X
TDS 2/month X X
TSS 5/week XX X
Total Coliforms 5/week XX X
HPC 5/week X X
TOC 2/week X X
UVA 2/week X X
1) Turbidity data was recorded at 10-minute intervals.
2) Online feed and filtrate turbidities were verified once per day with bench-top measurements. Backwash turbidity was
measured twice per day with bench-top analysis and was not monitored by online instrumentation.
3) Particle count data was recorded at 10-minute intervals.
4) Backwash samples were analyzed once per day for total chlorine to verify the target dose of 5 mg/L. Feed water chlorine
was monitored daily by the GBWLTFP and was verified by means of a pilot plant grab sample prior to ETV testing.
44
-------
Table 3-6 Data Management
Data Type
Documentation Format
Reporting Format
Membrane Hydraulic Data
Chemical Cleaning
Observations/Measurements
Production Backwash
Observations/Measurements
In-line Feed/Filtrate Turbidity
In-line Feed/Filtrate Particle Counts
In-line instrument flow measurements
Daily Walkthrough Observations
Monthly QA\QC Observations
Off-site analyses
On-site analyses
pH, temperature, turbidity,
conductivity
Photographs
SCADA
Notebook
Notebook
SCADA
Dedicated Computer
Notebook
Notebook
Notebook
Feed and validated
data from NLS
Notebook
Notebook
Plots
Table
Table
Plots
Plots
Table
Table
Table
Table/Plots
Table/Plots
45
-------
Table 3-7 Methods for Measuring Precision and Accuracy
Parameter
Precision
Accuracy
pH
Temperature
Total Chlorine0 >
Bench-Top Turbidity
Alkalinity
Total Hardness
Calcium Hardness
Total Dissolved Solids
Total Suspended Solids
Total Organic Carbon
UV Absorbance
Total Coliform/HPC
7 measurements per week with
duplicates (100%)
7 measurements per week with
duplicates (100%)
7 measurements per week
14 measurements per week with
weekly duplicates (14%)
One duplicate (100%)
One duplicate (100%)
One duplicate (100%)
4 measurements with 1 duplicate
(25%)
60 measurements with 12
duplicates (20%)
4 measurements per week with 1
duplicate (25%)
4 measurements per week with 1
duplicate (25%)
15 each measurements per week
with 3 duplicates (20%)
Daily 3-point calibration with pH
buffers at 4, 7, and 10
Initial and weekly testing against NIST
thermometer
Instrument "Zeroed" daily
Daily checks against primary standards
(0, 20, 200 NTU)
NLS standard procedures
NLS standard procedures
NLS standard procedures
NLS standard procedures
NLS standard procedures
NLS standard procedures
NLS standard procedures
NLS standard procedures
1) Samples were collected and analyzed daily by GBWLTFP as part of normal operating procedures. For verification, pilot
plant feed waters were measured for chlorine at the beginning of the verification testing.
46
-------
Table 3-8 Sample Methodology
Parameter
Methodology for All Samples
Method
pH
Temperature
Turbidity (Bench-top)
Alkalinity
Total Hardness
Calcium Hardness
Total Dissolved Solids
Total Suspended Solids
Total Organic Carbon
UV Absorbance
Total Coliform/HPC
Certified pH buffers
None
Ultra pure water as necessary
NLS standard procedures
NLS standard procedures
NLS standard procedures
NLS standard procedures
NLS standard procedures
NLS standard procedures
NLS standard procedures
NLS and MWH Laboratories standard
procedures
Standard Method 4500-H+
Standard Method 2550B
Standard Method 2130
Standard Methods 3010B, 2320B
EPA Method 200.7
EPA Method 200.7
Standard Method 2540C
EPA Method 160.2
Standard Method 5310B
Standard Methods 5910B
Standard Methods 9223/9215B (NLS)
Standard Method 9221 (MWH)
47
-------
Table 3-9 Corrective Action Plan
Parameter
Acceptance Criteria
Sequence of Steps for
Corrective Action
Any Duplicate
Any Method Blank
Any Performance Evaluation
(PE) or Proficiency Sample
pH
Temperature
Turbidity (Bench-top)
Alkalinity, Total Hardness,
Calcium Hardness, Total
Dissolved Solids
Total Suspended Solids
(filtrate)
< 10% apart
See Table 3.8; criteria set by
EPA-certified laboratory
performing the analysis
Within recovery specified for
each PE or proficiency sample
< 10% difference from previous
day; < 1 pH unit difference from
previous measurements
< 20% difference from previous
day
No increasing or decreasing
trend indicated by results of bi-
weekly proficiency samples.
Measurement deviates less than
±10% from standard.
< 20% difference from previous
reading
Assuming very low TSS
concentrations, < 100%
difference from previous reading
Re-sample duplicates
Check instrument calibration;
recalibrate instrument
See Table 3.8; perform procedures
specified to each analysis as
determined by the state-certified,
third-party, or EPA-accredited
laboratory performing the analysis
Check and verify all steps in sample
collection and analysis
Re-do PE or proficiency sampling and
analysis
Resample
Check instrument calibration
Re-calibrate instrument
Check for change in feed water source
to supply
Verify turbidimeter operation and
status of sample tap
Perform routine maintenance/cleaning
of instrument
Verify calibration using secondary
standards
Re-calibrate using primary standards
Verify change in feed water source or
supply
Verify corresponding increase in
turbidity
Re-sample
Check membrane integrity
48
-------
Table 4-1 Operational
Operating Parameter
Run Period
Start Date and Time
End Date and Time
Run Length
Termination Condition
Normalized Flux (20 °C)(1)
Recovery(2)
Backwash Interval
Backwash Duration(3)
Conditions
Units
-
-
-
day:hrs
-
L/h-m2 (gfd)
%
minutes
seconds
Hydraulic Data Logging Interval minutes
Particle Count Logging Interval
Production Backwash Chlorine
Dose Setpoint
Air Scour Flow Rate
Terminal Transmembrane
Pressure
minutes
mg/L
cfm
Bar (psi)
1
3/11/0208:30
3/23/0222:17
12 days 14 hrs
Fouling
40 (24)
>90
50
60-120
10 (4 per backwash
cycle)
5
5
Approximately 2
1.4(21)
2
3/24/02 19:36
4/26/02 12:22
32 days 17 hours
Time
30 (18)
>90
50
60-120
10 (4 per backwash
cycle)
5
5
Approximately 2
1.7 (25)
1) Test flux was determined based on results of pilot testing prior to the verification study.
2) Recovery was at least 90%. The exact values were calculated during verification testing and are reported in Chapter 4.
3) Backwash duration was set to achieve minimum recovery criteria.
49
-------
Table 4-2 Summary of Operational Data
Parameter
Runl
Flux
Flow Rate
Temperature
Recovery
Specific Flux
Decline at
20°C(1)
Run 2
Flux
Flow Rate
Temperature
Recovery
Specific Flux
Decline at
20°C(2)
1) Average daily
2) Average daily
Units
L/h-m2
(gfd)
gpm
°C (°F)
%
L/h - m2-
bar-d
(gfd/psi-d)
L/h-m2
(gfd)
gpm
°C (°F)
%
L/h -m2-
bar-d
(gfd/psi-d)
Count
1396
1396
1397
13
N/A
2738
2738
2738
27
N/A
Median
23.8
(14.0)
12.0
2.2 (36)
94
N/A
18.5
(10.9)
9.3
3.3 (38)
94
N/A
Range
17.9-25.4
(10.5-15.0)
9.0-12.8
1.5-3.2 (35-38)
89-96
N/A
16.1-24.2
(9.48-14.3)
8.1-12.2
1.6-6.8 (35-44)
89-95
N/A
Average
23.8
(14.0)
12.0
2.2 (36)
93
7.2
(0.29)
18.7
(11.0)
9.4
3.5 (38)
94
1.7
(0.069)
Standard
Deviation
0.5
0.3
0.3
1.3
N/A
0.98
0.49
1.31
1.2
N/A
95%
Confidence
Interval
23.8-23.8
(14.0-14.0)
12.0-12.0
2.2-2.2 (36-36)
93-94
N/A
18.8-18.8
(11.1-11.1)
9.4-9.4
3.5-3.5 (38-38)
93-94
N/A
specific flux decline for the duration of the Period 1 .
specific flux decline for the duration of the Period 2.
50
-------
Table 4-3
Parameter
Summary of Chemical
Units
Prior to
Clean
Cleaning
Start of
Clean
Water Quality Analyses
Sample Collection Time
During Clean
Backwash
Rinse
Production
after Clean
Chemical Cleaning 1(1) March 5, 2002
Time
Temperature
pH
TDS
Turbidity
Total C12
(hours)
°C (°F)
mg/L
NTU
mg/L
t<0
N/A
N/A
N/A
N/A
N/A
t=0
8.3 (47)
12.5
-
1.1
215
t=2
7.8 (46)
12.6
-
0.5
192
t=4
3.3 (38)
8.6
-
-
0.46
t= next run
3.3 (38)
7.9
-
0.1
-
Chemical Cleaning 2 <2) March 24, 2002
Time
Temperature
pH
TDS
Turbidity
Total C12
(hours)
°C (°F)
mg/L
NTU
mg/L
Chemical Cleaning 3 <3) May
Time
Temperature
pH
TDS(4)
Turbidity
Total C12
(hours)
°C (°F)
mg/L
NTU
mg/L
t<0
2(36)
7.8
-
0.032
5.1
1, 2002
to
6.3 (43)
-
-
0.037
-
t=0
14 (57)
12.2
4100
28.8
160
t=0
22 (72)
12.0
3900
31.1
186
t=2 t=8
15.8 (60.4)
12.2
4700 5900
10.9
168
t=2 t=22 t=44
25(77) 28(82) 31(88)
11.9 11.9 11.9
4500 5600 6500
17.6 17.2 15.7
180 176 180
t=8
18.6 (65.5)
12.18
130
12.3
72
t=44
16 (61)
11.9
3600
10.9
174
t=next run
7.5 (46)
8.1
120
0.121
0.7
t=next run
6.1(43)
-
-
-
-
1) Chemical Cleaning 1 was conducted prior to membrane operation. The total soaking duration for chemical cleaning 1 was four
hours.
2) Chemical Cleaning 2 was allowed to soak for eight hours.
3) Chemical Cleaning 3 was allowed to soak for 44 hours.
4) These samples were analyzed exceeding the EPA holding times for TDS.
51
-------
Table 4-4 Summary of Chemical Cleaning Hydraulic Analyses
Value Prior to Chemical Value Following
Parameter Units Clean Chemical Clean
Chemical Cleaning 1(1) March 5, 2002
Flux
Normalized Flux
Specific Flux
Normalized Specific Flux
TMP
Flow Rate
Chemical Cleaning 2 March
Flux
Normalized Flux
Specific Flux
Normalized Specific Flux
TMP
Flow Rate
Chemical Cleaning 3 May 1,
Flux
Normalized Flux
Specific Flux
Normalized Specific Flux
TMP
Flow Rate
1/h-m2 (gfd)
1/h-m2 (gfd)
l/h-m2-b (gfd/psi)
l/h-m2-b (gfd/psi)
psi
gpm
24, 2002
1/h-m2 (gfd)
1/h-m2 (gfd)
l/h-m2-b (gfd/psi)
l/h-m2-b (gfd/psi)
psi
gpm
2002
1/h-m2 (gfd)
1/h-m2 (gfd)
l/h-m2-b (gfd/psi)
l/h-m2-b (gfd/psi)
psi
gpm
N/A(1)
N/A
N/A
N/A
N/A
N/A
23.5(13.8)
39.4 (23.2)
16.5 (0.670)
27.7(1.12)
20.7
11.8
20.3 (12.0)
29.7(17.5)
11.8(0.479)
17.2 (0.698)
25.1
10.2
24 (14)
39 (23)
68 (2.8)
118(4.79)
5
12
19.2(11.3)
30.4(17.9)
45(1.8)
74.0(3.01)
6.1
10
21 (12)
32(19)
42(1.7)
63 (2.6)
7.2
10.6
1) Chemical Cleaning 1 was performed prior to membrane operation. Therefore, this information was not available.
52
-------
Table 4-5 Summary of Chemical Cleaning Efficacy
Normalized
Specific Flux
Prior to
Cleaning
L/h-m2-b
Cleaning Number Date (gfd/psi)
1 3-5-02 -(1)
2 (end of period 1)(2) 3-23-02 28 (1.1)
3 (end of period 2)(5) 5-1-02 17 (0.70)
Normalized
Specific Flux
Following
Cleaning
L/h-m2-b
(gfd/psi)
118(4.79)
74(3) (3.0)
63 (2.6)
Recovery of
Specific Flux
(%)
-J1)
62(4)
73
Loss of
Original
Specific Flux
(%)
-J1)
37(4)
47
1) Testing began with a new module. Therefore, this information was not available.
2) The end of period 1 was following approximately 12.5 days of non-stop operation at a constant specific flux of 40 L/hr-m
(24 gfd).
3) The readings were 67 L/h-m2-b immediately after chemical cleaning and stabilized after an increase to 74 L/h-m2-b.
4) The manufacturer's recommendations include chemical recirculation at elevated temperature [90-95T (32-35°C)]. However,
no recirculation loop or heating coil was used for this cleaning.
5) The end of period 2 was folio wing approximately 33 days of non-stop operational a constant specific flux of 30 L/h-m2 (18
gfd).
Table 4-6 Summary of Onsite Bench-top Turbidity Data
95%
Standard Confidence
Parameter Units Count Median Range Average Deviation Interval
Feed Water
pH 40
Temperature °C (°F) 76
Turbidity NTU 39
Filtrate
Turbidity NTU 41
Backwash
Water
Turbidity NTU 78
7.79 7.67-8.01 7.80 0.07 7.73-7.88
2.9(37) 1.9-6.5(35-44) 3.4(38) 1.6 3.4-3.4(38-38)
1.0 0.4-4.5 1.3 0.9 1.3-1.3
0.05 0.05-0.10 0.05 0.01 0.05-0.05
11 1.7-21 11 3.6 11-11
53
-------
Table 4-7 Summary of Online Turbidity and Particle Count
Parameter Units Count Median Range Average
Feed Water
Turbidity
>2 um Particles
2-3 um Particles
3-5 um Particles
5-7 um Particles
7-10 um Particles
10-15 um Particles
> 15 um Particles
Filtrate
Turbidity
>2 um Particles
2-3 um Particles
3-5 um Particles
5-7 um Particles
7-10 um Particles
10-15 um Particles
> 15 um Particles
Log Removal of
Particles'1*
>2 um Particles
2-3 um Particles
3-5 um Particles
5-7 um Particles
7-10 um Particles
10-15 um Particles
> 15 um Particles
NTU
#/ml
#/ml
#/ml
#/ml
#/ml
#/ml
#/ml
NTU
#/ml
#/ml
#/ml
#/ml
#/ml
#/ml
#/ml
4,124
11,175
11,175
11,175
11,175
11,175
11,175
11,175
4,124
11,175
11,175
11,175
11,175
11,175
11,175
11,175
8,941
6,016
6,261
3,567
4,032
3,651
6,185
0.85
2,835
1,201
1,208
178
153
60
19
0.0532
0.00
0.00
0.00
0.00
0.00
0.00
0.00
4.4
4.3
4.3
3.6
3.4
3.0
2.4
0.25-19
0-21,529
0-4,676
0-8,772
0-3,289
0-5,590
0-3,628
0-1,336
0.00-0.05
0-6,125
0-2,862
0-2,850
0-249
0-135
0-140
0-313
-1.9-5.9
-1.9-5.3
-1.9-5.5
-1.7-5.1
-1.4-5.2
-1.5-5.0
-2.8-4.7
1.3
4,281
1,602
1,880
325
305
127
41
0.05
4
1
1
0
0
1
2
4.2
4.1
4.1
3.4
3.3
2.9
2.2
Data
Standard
Deviation
1.4
3731
1,095
1,691
394
430
194
69
0.007
65
28
28
3
3
4
16
0.9
0.7
0.8
0.8
0.8
0.9
0.9
95%
Confidence 90th
Interval Percentile
1.3-1.3 2.9
4,278-4,283 9911
1,601-1,603
1,879-1,881
325-326
305-306
127-127
41-41
0.05-0.05 0.05
4-4 2
1-1
1-1
0-0
0-0
1-1
2-2
10th
Percentile
4.2-4.2 3.1
4.1-4.1 3.3
4.1-4.1 3.2
3.4-3.5 2.6
3.3-3.3 2.4
2.9-2.9 2.0
2.2-2.2 1.1
1) Evaluation of the time series data showed relatively higher particle counts during, and immediately following, a backwash. As a
result, particle removals were decreased during these times. Some of these consistently brief occurrences caused log removal data
to become negative for short periods. Negative data points were not included in the statistical analysis for log removals.
54
-------
Table 4-8
Parameter
Feed Water
Total
Coliforms
HPC
Filtrate
Total
Coliforms
HPC
Backwash
Water(4)
Total
Coliforms
HPC
Summary of Microbial Water Quality
Units Count Median'1' Range Average
P/A
MPN/100 ml
CFU/ml
P/A
MPN/100 ml
CFU/ml
P/A
MPN/100 ml
CFU/ml
38 N/A(2) A-P(2) A
4 <1.1
-------
Table 4-9 General Water Quality Parameters
Parameter Units Count Median'1' Range
Feed Water
C12-
Residual(2)
Alkalinity
Total Hardness
Calcium
Hardness
TSS
TDS
TOC
UVA
Algae
Filtrate
Alkalinity
Total Hardness
Calcium
Hardness
TSS
TDS
TOC
UVA
Backwash
Water(3)
C12
TSS
mg/L
mg/L as
CaC03
mg/L as
CaC03
mg/L as
CaC03
mg/L
mg/L
mg/L
cm"1
#/ml
mg/L as
CaCO3
mg/L as
CaC03
mg/L as
CaC03
mg/L
mg/L
mg/L
cm"1
mg/L
mg/L
42
2
2
2
37
3
13
13
o
3
2
2
2
37
o
J
19
19
40
36
0.05
N/A
N/A
N/A
<1
200
2.2
0.022
32.8
N/A
N/A
N/A
<1
210
2.0
0.019
4.8
20.
0.03-0.06
110-110
130-130
87-88
<1-8.0
150-210
1.6-3.4
0.017-0.043
32.6-37.8
110-110
130-130
87-87
<1-9.0
190-210
1.6-3.0
0.015-0.027
2.0-11.7
<1-41
Average
0.05
110
130
88
1.3
187
2.3
0.024
34.4
110
130
87
1.2
203
2.0
0.019
4.7
20
Standard
Deviation
0.01
N/A
N/A
N/A
1.8
N/A
0.5
0.007
N/A
N/A
N/A
N/A
1.9
N/A
0.3
0.003
1.5
10
95%
Confidence
Interval
0.05-0.05
N/A
N/A
N/A
1.3-1.3
N/A
2.3-2.3
0.024-0.024
N/A
N/A
N/A
N/A
1.2-1.2
N/A
2.0-2.0
0.019-0.019
4.7-4.7
20-20
1) Values reported as non-detect were assumed to be one-half of the detection limit for the purposes of statistical evaluation.
2) Measured as part of the daily sampling activities of the GBWUFP.
3) Sampled during the second cycle of the production backwash.
56
-------
Table 4-10 TSS Mass Balance
Filtrate Fi"nfm
Date Flow _<*?.
Length
(gpm) (min)
Test Period 1
3/11/2002
3/11/2002
3/12/2002
3/13/2002
3/14/2002
3/15/2002
3/18/2002
3/19/2002
3/20/2002
3/21/2002
3/21/2002
3/22/2002
Test Period 2
3/25/2002
3/26/2002
3/27/2002
3/28/2002
3/28/2002
3/29/2002
4/1/2002
4/2/2002
4/3/2002
4/4/2002
4/4/2002
4/8/2002
4/9/2002
4/10/2002
4/11/2002
4/11/2002
4/15/2002
4/16/2002
4/17/2002
4/18/2002
4/18/2002
4/19/2002
4/22/2002
4/23/2002
12
12
12
12
12
12
12
12
12
12
12
12
9.4
9.4
9.4
9.4
9.4
9.4
9.4
9.4
9.4
9.4
9.4
9.4
9.4
9.4
9.4
9.4
9.4
9.4
9.4
9.4
9.4
9.4
9.4
9.4
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
Volume
Filtered
(gal)
600
600
600
600
600
600
600
600
600
600
600
600
470
470
470
470
470
470
470
470
470
470
470
470
470
470
470
470
470
470
470
470
470
470
470
470
Backwash
Volume
(gal)
50
50
40
55
50
32
45
40
37
35
35
34
50
42
38
36
36
30
33
32
30
30
30
26
25
32
27
27
24
28
28
28
28
30
24
24
Measured
Feed
TSSd)
(mg/L)
1.0
0.5
8.0
6.0
1.0
0.5
1.0
0.5
0.5
1.0
2.0
0.5
4.0
0.5
0.5
0.5
4.0
1.0
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
Measured
Backwash
TSS(1)(2)
(mg/L)
0.5
4
26
3
23
40
41
25
31
24
24
19
14
20
14
10
13
27
14
2
20
0.5
17
29
26
23
15
20
22
12
30
20
21
36
27
16
Calculated
Backwash
TSS
(mg/L)
12
6
120
65
12
9
13
8
8
17
34
9
38
6
6
7
52
16
7
7
8
8
8
9
9
7
9
9
10
8
8
8
8
8
10
10
RPD(3)
%
184
40
129
182
63
124
102
108
117
33
35
73
91
113
77
42
120
53
65
114
87
176
74
105
94
103
53
79
77
35
113
82
86
129
94
48
1) Values reported as non-detect (<1 mg/1) were assumed to be one-half of the detection limit (0.5 mg/1) for the purposes of this
evaluation.
2) Sampled during the second cycle of the production backwash.
3) Relative Percent Deviation.
57
-------
Table 4-11 Summary of Verification Data for the Feed Water Particle Counter
Standard
Parameter Units Count Median Range Average Deviation'1'
2 um Particle Standard at a
>2 um Particles
2-3 um Particles
3-5 um Particles
5-7 um Particles
7-10 um Particles
10-15 um Particles
> 15 um Particles
#/ml
#/ml
#/ml
#/ml
#/ml
#/ml
#/ml
5 um Particle Standard at a
>2 um Particles
2-3 um Particles
3-5 um Particles
5-7 um Particles
7-10 um Particles
10-15 um Particles
> 15 um Particles
#/ml
#/ml
#/ml
#/ml
#/ml
#/ml
#/ml
10 um Particle Standard at
>2 um Particles
2-3 um Particles
3-5 um Particles
5-7 um Particles
7-10 um Particles
10-15 um Particles
> 15 um Particles
#/ml
#/ml
#/ml
#/ml
#/ml
#/ml
#/ml
95%
Confidence
Interval'1'
Concentration of 1000 #/ml
3
3
3
3
3
3
3
1143.7
720.0
303.2
43.5
41.5
27.4
13.2
1141.7-1332.2
714.8-753.7
302.2-406.8
41.9-63.4
39-60.1
25.9-33.2
12.9-15.1
1205.9
729.5
337.4
49.6
46.9
28.8
13.7
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
Concentration of 1000 #/ml
3
3
3
3
3
3
3
1121 A
129.9
780.5
149.6
37.5
17.9
8.3
a Concentration of
3
3
3
3
3
3
3
952.1
43.8
70.5
26.8
585.1
212.9
11.3
1108.4-1130.9
129.5-132.3
774.8-780.9
147.5-150.9
35.8-39.5
14.8-18
6-13
1000 #/ml
944.954.6
36.5-47.3
60.6-70.7
25.5-27.2
579.2-587.2
212.4-220.1
6.1-27.7
1122.2
130.6
778.7
149.3
37.6
16.9
9.1
950.2
42.5
67.2
26.5
583.8
215.1
15.0
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
1) Less than eight data points exist for this data. Therefore, this statistical analysis was not performed.
58
-------
Table 4-12 Summary of Verification Data for the Filtrate Water Particle Counter
95%
Standard Confidence
Parameter Units Count Median Range Average Deviation'1' Interval'1'
2 um Particle Standard at a
>2 um Particles
2-3 um Particles
3-5 um Particles
5-7 um Particles
7-10 um Particles
10-15 um Particles
> 15 um Particles
#/ml
#/ml
#/ml
#/ml
#/ml
#/ml
#/ml
5 um Particle Standard at a
>2 um Particles
2-3 um Particles
3-5 um Particles
5-7 um Particles
7-10 um Particles
10-15 um Particles
> 15 um Particles
#/ml
#/ml
#/ml
#/ml
#/ml
#/ml
#/ml
10 um Particle Standard at
>2 um Particles
2-3 um Particles
3-5 um Particles
5-7 um Particles
7-10 um Particles
10-15 um Particles
> 15 um Particles
#/ml
#/ml
#/ml
#/ml
#/ml
#/ml
#/ml
Concentration of 1000 #/ml
2
2
2
2
2
2
2
N/A(2)
N/A(2)
N/A(2)
N/A(2)
N/A(2)
N/A(2)
N/A(2)
1089.7-1126.4
639.1-647.7
333.7-348.3
42.6-47.3
39.2-43.4
23.5-28.5
11.3-11.5
1108.0
643.4
341.0
45.0
41.3
26.0
11.4
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
Concentration of 1000 #/ml
5
5
5
5
5
5
5
727.9
68.3
564.0
81.5
5.9
1.2
5.9
a Concentration of
4
4
4
4
4
4
4
840.4
49.7
113.3
46.6
476.0
145.9
6.4
719.4-732.8
67.1-69
557-571.6
75.4-87.4
4.7-6.3
0.7-2.5
3-8.9
1000 #/ml
830.6-849.6
47-56.1
108.9-117.4
45.1-50.3
471.1-484.6
144.4-147.1
4.4-9.2
727.4
68.1
564.3
81.9
5.7
1.5
6.0
840.3
50.6
113.2
47.2
476.9
145.8
6.6
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
1) Less than eight data points exist for this data. Therefore, this statistical analysis was not performed.
2) Less than three data points exist for this data. Therefore, this statistical analysis was not performed.
59
-------
Table 4-13 Chemical Consumption Analysis
Approximate Total
Product Volume
Produced
Operation (gal)(1)
Run 1 197,463
Chemical x T .
™ ~ NA
Clean 2
Run 2 402,388
Chemical x T .
™ , NA
Clean 3
Approximate Total
Backwash Volume
Used
(gal)<2)
13,822
35
Total for Period 1
24,143
32
Total for Period 2
Chlorine Consumed
(Ibs as NaOCl)<3)
0.61
0.06
0.67
1.06
0.06
1.12
NaOH
Consumed
(Ibs as NaOH)(4)
NA
2.04
2.04
NA
1.47
1.47
1) Based on average flow data from Table 4.2 and a complete production and backwash cycle time of 55 minutes.
2) Based on daily observations of production backwash volumes and observation during each chemical clean.
3) Based on a production cycle backwash chlorine dose of 5 mg/L and chemical cleaning dose of 200 mg/L as verified
through onsite analysis.
4) Based on the programmed dose of 7,000 and 5,500 mg/L for Chemical Cleanings 2 and 3, respectively.
60
-------
10
12
13
15
16
17
18
19
20
22
23
24
SODIUM
HYPO-
CHLORITE
PUMPP310
(0.75GPH®
80PSIG)
RAW WATER
PUMP P100
(1.5HP,45GPM@28PSIG)
BACKWASH WATER
PUMP P600
(7.5HP, 150GPM@55PSIG)
Figure 2.1
Intuitechs.
THIS DRAWING IS THE INTELLECTUAL PROPERTY OF INTUITECH AND
MAY NOT BE REPRODUCED IN FULL OR IN PART FOR ANY PURPOSE
ASIDE FROM THE PROJECT AS SPECIFIED ON THIS DOCUMENT
10
11
12
13
DIMENSIONS ARE IN INCHES
UNLESS OTHERWISE NOTED
TOLERANCES ARE AS
FOLLOWS, UNLESS
OTHERWISE NOTED
1/X
1/XX
.OX
.OOX: +.005
ANGLES : +1°
14 is
REFERENCE
DRAWINGS
TITLE:
UF PILOT
PROCESS & INSTRUMENTATION DIAGRAM
CLIENT: CAROLLO ENGINEERS - BOISE
DESIGN BY: EJH
DRAWN BY: EJH
CHECK BY:
RELEASED FOR: FABRICATION
16
17
FILE NAME: P&ID
is
PROJECT: UF PILOT #1
DESIGN DATE: 6-18-99
DRAWN DATE: 6-18-99
CHECK DATE:
P.O.:
REVISION: 3
19
20
21
22
23
24
-------
0.0 4*
3/11
3/16
3/21
3/26 3/31
Transmembrane Pressure (bar)
A System Recovery
• Flux@20C
I Specific Flux @ 20 C
4/5 4/10 4/15
Time
4/20
160
140
120
Transmembrane Pressure
4 A A A/A A A A A A A A A A A A A A A A. AAA
/ \
o
o
o
100 o ~
O
CM
80 N~
o
0)
Q.
V)
20
0
4/25 4/30
Note: The membrane unit was off-line for 21 hours and 10 minutes following
Run 1 for chemical cleaning. In addition, on 4/18/02, the membrane unit was
shut down for approximately 10 hours due to lightning.
Figure 4-1
IMP, Flux, and System Recovery Profiles for the
Polymem UF Module UF120 S2 Membrane Module
o
es
.0.
S
--3 •=
Q.
(0
-------
O>
CO
28
26
24
22
20
18
16
14
12
10
8
6
4
2
0
Temperature
3/11
3/16
• Temperature
l Operational Flux
3/21 3/26 3/31 4/5 4/10
Run Time (Days)
4/15
4/20
4/25
4/30
Note: The membrane unit was off-line for 21 hours and 10 minutes following
Run 1 for chemical cleaning. In addition, on 4/18/02, the membrane unit was
shut down for approximately 10 hours due to lightning.
Figure 4-2
Operational Flux and Temperature Profiles for the
Polymem UF Module UF120 S2 Membrane Module
-------
100
O>
Raw Online Turbidity
0.001
0.01
Filtrate Online Turbidity
3/11
3/16
3/21
3/26
3/31
Raw Online Turbidity
D Raw Benchtop Turbidity
Filtrate Online Turbidity
A F&trateate Benchtop Turbidity
O Backwash Benchtop Turbidity
Note: The membrane unit was off-line for 21 hours and 10 minutes following
Run 1 for chemical cleaning. In addition, on 4/18/02, the membrane unit was
shut down for approximately 10 hours due to lightning.
4/5 4/10 4/15 4/20 4/25 4/30
Time
Figure 4-3
Online and Benchtop Turbidity Profiles for the
Polymem UF Module UF120 S2 Membrane Module
-------
CD
cn
100000
10000
c
3
o
o
0)
o
s.
1000
=• 100
Feed Particle Counts
Filtrate Particle Counts
0.01
3/11
3/16
3/21
3/26
3/31
I Filtrate Particle Counts
• Feed Particle Counts
4/5
Time
4/10
4/15
4/20
4/25
4/30
Note: The membrane unit was off-line for 21 hours and 10 minutes following
Run 1 for chemical cleaning. In addition, on 4/18/02, the membrane unit was
shut down for approximately 10 hours due to lightning.
Figure 4-4
Feed and Filtrate Count Profiles for Particles >2 um for the
Polymem UF Module UF120 S2 Membrane Module
-------
CD
CD
100000
c
3
o
o
0)
o
t
re
Q.
10000
1000
~ 100
0.01
3/11
3/16
3/21
3/26
3/31
Filtrate Particle Counts
l Feed Particle Counts
4/5
Time
4/10
4/15
4/20
4/25
4/30
Note: The membrane unit was off-line for 21 hours and 10 minutes following
Run 1 for chemical cleaning. In addition, on 4/18/02, the membrane unit was
shut down for approximately 10 hours due to lightning.
Figure 4-5
Feed and Filtrate Count Profiles for Particles 2-3 um for the
Polymem UF Module UF120 S2 Membrane Module
-------
O>
10000
1000
100
3
o
o
0)
o
t
re
Q.
y.M'M
Filtrate Particle
I/,1,11 Counts
IB •IBMHIMMIIi
0.01
3/11
3/16
3/21
3/26
3/31
4/10
4/15
4/20
4/25
I Filtrate Particle Counts
• Feed Particle Counts
Note: The membrane unit was off-line for 21 hours and 10 minutes following
Run 1 for chemical cleaning. In addition, on 4/18/02, the membrane unit was
shut down for approximately 10 hours due to lightning.
4/5
Time
Figure 4-6
Feed and Filtrate Count Profiles for Particles 3-5 um for the
Polymem UF Module UF120 S2 Membrane Module
4/30
-------
CD
00
10000
1000
100 -
5
4-1
c
o
o
s.
0.01
3/11
3/16
3/21
3/26
3/31
4/10
4/15
4/20
4/25
4/30
I Filtrate Particle Counts
• Feed Particle Counts
Note: The membrane unit was off-line for 21 hours and 10 minutes following
Run 1 for chemical cleaning. In addition, on 4/18/02, the membrane unit was
shut down for approximately 10 hours due to lightning.
4/5
Time
Figure 4-7
Feed and Filtrate Count Profiles for Particles 5-7 um for the
Polymem UF Module UF120 S2 Membrane Module
-------
CD
10000
1000
100
3
o
o
S.
0.01
III III II
•III
Mill III!
Ill
iiiiniiHi IIIIHIIIIII mm 111 mi in
HUH i iiiiiiiniMiiiiiniiini
• mini i nun
IMIII in mi
3/11
3/16
3/21
3/26
3/31
Filtrate Particle Counts
l Feed Particle Counts
4/5 4/10
Time
4/15
4/20
4/25
4/30
Note: The membrane unit was off-line for 21 hours and 10 minutes following
Run 1 for chemical cleaning. In addition, on 4/18/02, the membrane unit was
shut down for approximately 10 hours due to lightning.
Figure 4-8
Feed and Filtrate Count Profiles for Particles 7-10 um for the
Polymem UF Module UF120 S2 Membrane Module
-------
10000
1000
o
O
re
Q.
Vj'V taflU
0.1
0.01
3/11
3/16
3/21
I Filtrate Particle Counts
• Feed Particle Counts
3/26
3/31
4/5
Time
4/10
4/15
4/20
4/25
4/30
Note: The membrane unit was off-line for 21 hours and 10 minutes following
Run 1 for chemical cleaning. In addition, on 4/18/02, the membrane unit was
shut down for approximately 10 hours due to lightning.
Figure 4-9
Feed and Filtrate Count Profiles for Particles 10-15 um for the
Polymem UF Module UF120 S2 Membrane Module
-------
10000
1000
o
o
-J 7;
re
Q.
Filtrate Particle
Counts
0.01
3/11
3/16
3/21
3/26
3/31
I Filtrate Particle Counts
• Feed Particle Counts
4/5
Time
4/10
4/15
4/20
4/25
4/30
Note: The membrane unit was off-line for 21 hours and 10 minutes following
Run 1 for chemical cleaning. In addition, on 4/18/02, the membrane unit was
shut down for approximately 10 hours due to lightning.
Figure 4-10
Feed and Filtrate Count Profiles for Particles >15 um for the
Polymem UF Module UF120 S2 Membrane Module
-------
_ 4
re
§
0)
re
Q.
Backwash
Events
I1',IN,'!.!
9
'.''
.....
.••
' i'll .
I.1 11
<> f
1 ' ,'
III I
'I
-I-
1 it'
3/11
3/16
3/21
3/26
3/31
4/5
Time
4/10
4/15
4/20
4/25
Note: The membrane unit was off-line for 21 hours and 10 minutes following
Run 1 for chemical cleaning. In addition, on 4/18/02, the membrane unit was
shut down for approximately 10 hours due to lightning.
4/30
Figure 4-11
Log Removal Profile for Particles >2 um in Size for the
Polymem UF Module UF120 S2 Membrane Module
-------
CO
Backwash 1
Events
3/11
3/16
4/25
Note: The membrane unit was off-line for 21 hours and 10 minutes following
Run 1 for chemical cleaning. In addition, on 4/18/02, the membrane unit was
shut down for approximately 10 hours due to lightning.
4/30
Figure 4-12
Log Removal Profile for Particles 2-3 um in Size for the
Polymem UF Module UF120 S2 Membrane Module
-------
Backwash
Events
3/11
3/16
4/25
4/30
Note: The membrane unit was off-line for 21 hours and 10 minutes following
Run 1 for chemical cleaning. In addition, on 4/18/02, the membrane unit was
shut down for approximately 10 hours due to lightning.
Figure 4-13
Log Removal Profile for Particles 3-5 um in Size for the
Polymem UF Module UF120 S2 Membrane Module
-------
en
Backwash
Events
3/11 3/16
4/25 4/30
Note: The membrane unit was off-line for 21 hours and 10 minutes following
Run 1 for chemical cleaning. In addition, on 4/18/02, the membrane unit was
shut down for approximately 10 hours due to lightning.
Figure 4-14
Log Removal Profile for Particles 5-7 um in Size for the
Polymem UF Module UF120 S2 Membrane Module
-------
CD
Backwash
Events
3/11 3/16
4/25 4/30
Note: The membrane unit was off-line for 21 hours and 10 minutes following
Run 1 for chemical cleaning. In addition, on 4/18/02, the membrane unit was
shut down for approximately 10 hours due to lightning.
Figure 4-15
Log Removal Profile for Particles 7-10 um in Size for the
Polymem UF Module UF120 S2 Membrane Module
-------
Backwash
Events
3/11 3/16
4/25 4/30
Note: The membrane unit was off-line for 21 hours and 10 minutes following
Run 1 for chemical cleaning. In addition, on 4/18/02, the membrane unit was
shut down for approximately 10 hours due to lightning.
Figure 4-16
Log Removal Profile for Particles 10-15 um in Size for the
Polymem UF Module UF120 S2 Membrane Module
-------
00
Backwash
Events
Note: The membrane unit was off-line for 21 hours and 10 minutes following
Run 1 for chemical cleaning. In addition, on 4/18/02, the membrane unit was
shut down for approximately 10 hours due to lightning.
4/25
4/30
Figure 4-17
Log Removal Profile for Particles >15 um in Size for the
Polymem UF Module UF120 S2 Membrane Module
-------
CD
O)
c
'•B
o>
o>
o
X
LLI
+j
O
z
4-i
0)
o
0)
Q.
Q0%
80%
40%
1 0%
i
p
1
• t *
••
0.01
0.02 0.03 0.035
Turbidity (NTU)
0.04
0.05
Figure 4-18
A Frequency Distribution of Filtrate Turbidity for the
Polymem UF Module UF120 S2 Membrane Module
-------
00
O
O)
'•&
-------
00
O)
c
O
X
LU
0)
O
0)
Q.
10%
0%
456
Particle Count (#/ml)
10
Figure 4-20
A Frequency Distribution for Filtrate Partilces >2 um for the
Polymem UF Module UF120 S2 Membrane Module
-------
00
O)
O
X
LU
£
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
5000
10000 15000
Particle Count (#/ml)
20000
25000
30000
Figure 4-21
A Frequency Distribution for Feed Water Partilces >2 um for the
Polymem UF Module UF120 S2 Membrane Module
-------
00
CO
O)
O
X
LU
+J
0)
O
L.
0)
Q.
100%
90%
Log Removal
Figure 4-22
A Frequency Distribution of Log Removal for Particles >2 um for the
Polymem UF Module UF120 S2 Membrane Module
-------
00
O)
c
0)
o
X,
LU
0)
o
0)
Q.
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0% +
234
Log Removal
Figure 4-23
A Frequency Distribution of Log Removal for Particles 2-3 um for the
Polymem UF Module UF120 S2 Membrane Module
-------
oo
en
O)
c
0)
-------
00
CD
TO
c
TJ
0)
8
X
LU
•4-1
0)
H
0)
Q.
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
0
234
Log Removal
Figure 4-25
A Frequency Distribution of Log Removal for Particles 5-7 um for the
Polymem UF Module UF120 S2 Membrane Module
-------
00
O)
0)
0)
u
X
HI
0)
£
0)
0.
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
\
0%
234
Log Removal
Figure 4-26
A Frequency Distribution of Log Removal for Particles 7-10 um for the
Polymem UF Module UF120 S2 Membrane Module
-------
00
00
O)
c
0)
-------
oo
VO
O)
O
X
LU
+J
0)
O
0)
Q.
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
3 4
Log Removal
Figure 4-28
A Frequency Distribution of Log Removal for Particles >15 um for the
Polymem UF Module UF120 S2 Membrane Module
-------
mnnn
mnn
E
5
' — ' mn
c
3
o
o
03 m
o IU
t
re
Q.
-\
D
D
0.1
001
ii, i'
'i ' , '
i
I I I I I II II I I I II I II II I
• /
• | • • Time of Backwash Events
" f • ' i • •• • • •"
• Ml •• • • • • MB
oooooooooooooo
oooooooooooooo
c»
c
c
c
c*
T-
?
} ^3 ^"™ C^ CO ^f LO CO 1*^ CO O5 ^3 ^~™ C^
JCN|CN|CN|CN|CN|CN|CN|CN|CN|CN|T~T~T~
JOOOOOOOOOOCNICNICNI
JT^T^T^T^T^T^T^T^T^T^^^^
"cocococococococococo^^^
•) CO CO CO
.''',
1, '"l, '
'
1 1 1 1
•
H'
> ' \'
• 4 •
MB • •
o o o o
o o o o
CN| CN| CN| CN|
''l,
' ^"'
\
Feeql Particle Counts
II 1 ^ 1 II
1 Filtrate Particle Counts
- /'.
• m m
mm m • •
• • •• •
o o o o o
o o o o o
CN| CN| CN| CN| CN|
c
CN
CN
CN
^
T-
COCOCOCOCOCOCOCOCOC1"
l Feed Particle Counts T\me
Figure 4-29
• Filtrate Particle Counts Feed and FMtrate particle Count Profiles for Particles >2 urn
from 3/14/02 00:00 to 3/14/02 20:44 Using Raw Data for the
I Markers Indicating the Time of Likely Polymem UF Module UF120 S2 Membrane Module
-------
IUUUUU
mnnn
mnn
E
' — ' mn
3
o
o
03 m
o IU
t
-\
^o
0.1
Om
\
''
''' 1 \ I1
'' ' ' ' '""^.
\
Feed Particle Counts
I I I I I II II I I I II I
/
• | • Time of Backwash Events
• • _•
• f-«" •• (*"" • i " • i"
" f • ' i • •• i i~
• •• • •• M • M M
M1H •• ^ • • • 1H •
ooooooooooooo
ooooooooooooo
COO-*— OJCO^LOCQI^COOOT—
CN|CN1CN1CN1CN1CN1CN1CN1CN1CN1CN1T~T~
CNIOOOOOOOOOOCNICNI
COT^T^T^T^T^T^T^T^T^T^^^
CO CO CO
1 II 1 1 1 1 1 II 1 1 1 II
Filtrate Particle Counts
- / .
• • •
~S* % mP i f i
mm mm mm m mmmm mm mm
mmm mm •••••
ooooooooooc
ooooooooooc
T-T-T-T-T-T-T-T-CNCsicN
CNICNICNICNICNICNICNICNICNICNICN
COCOCOCOCOCOCOCOCOCOf
I Feed Particle Counts Time Figure 4-30
Feed and Filtrate Particle Count Profiles for Particles >2 um
• Filtrate Particle Counts , _ .. . .__ __ __ . _ .. . .__ __ .,.,,_ •_•- r* r*. *. .1
from 3/14/02 00:00 to 3/14/02 20:44 Excluding Data Collected
During Backwash Events for the
I Markers Indicating the Time of Likely _ . ,,.-..,, » «« . ....
Backwash Event Polymem UF Module UF120 S2 Membrane Module
-------
VO
to
O)
£
0)
0)
o
X
LU
+j
0)
o
L.
0)
Q.
100%
90%
70%
l% - -
0%
^r
^^
/^ .
Curve Developed with Raw Data Set /
/
/
/
/
Curve Developed Excluding Data Collected
,
\
j
A
7
\ ,
L *L
'•\.\
\
\
\
\
"ll
\,
\
\
\
I
\
V
\
\
ll.
Log Removal
3.2 3.6
Figure 4-31
A Frequency Distribution of Log Removal for Particles >2 um
from 3/14/02 00:00 to 3/14/02 20:44 for the
Polymem UF Module UF120 S2 Membrane Module
------- |