Read Microsoft Word - CA - Marin, Final Report.doc text version

Pilot Study Report for the ZeeWeed® 1000 Ultrafiltration System provided to Marin Municipal Water District

Submitted to: Kennedy/Jenks Consulting, Inc. 622 Folsom Street San Francisco, CA 94107 Attn: Todd Reynolds P.E.

Submitted by: ZENON Membrane Solutions 3239 Dundas Street West Oakville, Ontario L6M 4B2 June 2006

All patents, trade secrets and other intellectual property in this document shall be the property of ZENON. The Customer shall retain all technical information or other trade secrets developed or applied by ZENON and learned by the Customer through this document in confidence until such time as the information has become wholly disclosed to the public (otherwise than by default of the Customer) or until disclosure is authorized in writing by ZENON. One or more ZENON Environmental Inc. (ZENON) patents or patent applications may cover the technology disclosed herein. Any disclosure in this document does not hereby grant, and nothing contained in this document shall obligate ZENON to grant an option to obtain a license to any technology or any other rights under any patent now or hereafter owned or controlled by ZENON.

ZEEWEED® 1000 Pilot Study Marin, CA

June 2006

ACKNOWLEDGEMENTS

ZENON would like to thank the staff of the Marin Municipal Water District for their dedication over the course of the pilot study. The hard work and dedication of Paul Sellier was instrumental in the success of the pilot study. ZENON would also like to thank Todd Reynolds and Patrick Treanor of Kennedy/Jenks Consultants for their continued involvement throughout this study. Thank you to all involved.

II

ZENON Private

ZEEWEED® 1000 Pilot Study Marin, CA

June 2006

EXECUTIVE SUMMARY

ZENON Environmental Inc., in conjunction with Kennedy/Jenks Consultants, Inc., conducted an 11 month pilot study using the ZeeWeed® 1000 ultrafiltration membrane system. The study was conducted for the Marin Municipal Water District at the Marin Rod and Gun Club using sea water from San Pablo Bay. The primary piloting objective was to demonstrate stable membrane performance for both the dry and wet seasons while meeting applicable standards for RO pretreatment.

ZeeWeed UF pretreatment to a seawater RO desalination system removes the risks associated with RO pretreatment. A Zenon system can produce high quality permeate regardless of the raw sea water variability while occupying a 30-40% smaller footprint and 2-8% lower life cycle costs than conventional systems. The technical and the economic analyses favor a ZeeWeed UF system over conventional granular media pretreatment. Zenon can provide the best solution for RO pretreatment.

This document has been written to provide a summary of the operational, analytical, membrane integrity and cleaning results obtained throughout the pilot study. The following sections

highlight the conclusions that can be drawn from the piloting for the Marin Municipal Water District.

Membrane Performance · During the dry season, from May to November, stable performance was observed at a flux of 35 gfd and 95% recovery. At these conditions and while performing daily 100 mg/L sodium hypochlorite maintenance cleans a predicted cleaning interval of greater than 50 days was achieved when corrected to a temperature of 20oC. · During the wet season, from December to April, stable performance was observed at a flux of 35 gfd and 95% recovery. The maintenance clean strategy utilized an extended aeration procedure as part of the daily heated 100 mg/L sodium hypochlorite maintenance clean. The maintenance clean combined with a 120 second backwash drain time resulted in a predicted cleaning interval of greater than 60 days when corrected to a temperature of 20oC.

III

ZENON Private

ZEEWEED® 1000 Pilot Study Marin, CA Water Quality ·

June 2006

The ZeeWeed® Ultrafiltration System consistently produced excellent water quality. Permeate turbidity averaged 0.05 NTU throughout the entire duration of the study even with feed water turbidity averaging 14 NTU and spikes up to 90 NTU during the wet season.

Membrane Integrity · · Throughout the course of piloting no leaks or breaks occurred in the membrane. Membrane integrity tests proved valuable in detecting leaks in the piping of the system.

Cleaning · Soaking the membranes in a 10 g/L citric solution heated to 95oF followed by a soak in 1000 mg/L of a sodium hypochlorite solution heated to 95oF restored the membrane permeability to approximately 6 gfd/psi.

IV

ZENON Private

ZEEWEED® 1000 Pilot Study Marin, CA

June 2006

Table of Contents

ACKNOWLEDGEMENTS ......................................................................................................... II EXECUTIVE SUMMARY .........................................................................................................III 1. INTRODUCTION .................................................................................................................. 1 2. PILOT OBJECTIVES............................................................................................................ 2 3. MEMBRANE, PROCESS, AND PILOT DESCRIPTIONS ............................................... 3 3.1 ZEEWEED® 1000 MEMBRANE............................................................................................. 3 3.2 ZEEWEED® 1000 WATER TREATMENT PROCESS ............................................................... 6 3.3 ZEEWEED® 1000 PILOT ...................................................................................................... 8 4. OPERATIONAL TERMINOLOGY .................................................................................... 9 4.1 FLUX ................................................................................................................................... 9 4.2 TRANSMEMBRANE PRESSURE............................................................................................. 9 4.3 PERMEABILITY.................................................................................................................. 10 4.4 RECOVERY ........................................................................................................................ 10 4.5 BACKWASH PROPERTIES .................................................................................................. 10 4.6 MAINTENANCE CLEANS AND AERATION ......................................................................... 11 5. DISCUSSION - MEMBRANE PERFORMANCE ............................................................ 13 5.1 PHASE 1: FLUX OPTIMIZATION ­ DRY SEASON (MAY 26 TO NOV 30, 2005) ................... 13 5.2 PHASE 2: FLUX OPTIMIZATION ­ WET SEASON (DEC 5 TO APR 25)................................. 15 6. DISCUSSION - WATER QUALITY.................................................................................... 23 6.1 TURBIDITY ........................................................................................................................ 23 7. DISCUSSION - MEMBRANE INTEGRITY RESULTS .................................................. 24 7.1 MEMBRANE INTEGRITY TESTS ......................................................................................... 24 8. DISCUSSION - CLEANING RESULTS.............................................................................. 25 9. CONCLUSIONS..................................................................................................................... 27 APPENDIX A. MEMBRANE PERFORMANCE RESULTS................................................. 28 V ZENON Private

ZEEWEED® 1000 Pilot Study Marin, CA

June 2006

APPENDIX B. WATER QUALITY RESULTS....................................................................... 35 APPENDIX C. MEMBRANE INTEGRITY RESULTS ......................................................... 39

Index of Tables and Figures

Table 1. ZeeWeed® 1000 Membrane Module Characteristics ....................................................... 4 Table 2: Pilot Performance Summary, Phase 1 ­ Dry Season....................................................... 14 Table 3: Pilot Performance Summary, Phase 2 ­ Wet Season ...................................................... 16

Figure A. Membrane Fiber Cross-Section....................................................................................... 3 Figure B. ZeeWeed® 1000 Module.................................................................................................. 4 Figure C. ZeeWeed® 1000 Cassette ................................................................................................ 5 Figure D. ZeeWeed® 1000 Train..................................................................................................... 6 Figure E. Typical ZeeWeed® 1000 Ultrafiltration Pilot System ..................................................... 8 Figure F. Photograph of Solids Accumulation in Module prior to Extended Aeration Clean ...... 20 Figure G. Photograph of Module following the Extended Aeration Clean ................................... 20

VI

ZENON Private

ZEEWEED® 1000 Pilot Study Marin, CA

June 2006

1. INTRODUCTION

The increasing demand for water resources in California combined with the decreasing availability of fresh surface and ground water is driving the development of large seawater desalination plants in the region. Marin Municipal Water District (MMWD) is investing in their long term future by investigating the feasibility in constructing one of these plants and is conducting a long-term pilot study in San Rafael, CA to determine the optimal process by which to produce high quality and cost effective desalinated water.

ZENON Environmental's ultrafiltration membrane, the ZeeWeed® 1000, has been chosen to evaluate UF pretreatment to a reverse osmosis system. The ZeeWeed® 1000 uses a nominal pore size of 0.02 microns to produce high quality permeate ideal for feed to a reverse osmosis system. Other advantages, including reduced footprint, ease of operation and Zenon's reputation as a world leader in membrane technology, also benefit the MMWD project.

This document has been written to provide a summary of all operational, analytical, and cleaning results obtained throughout the Marin pilot study. The pilot objectives are stated in Section 2. The basic operating principles of the ZeeWeed® 1000 ultrafiltration membrane water treatment process and the ZENON pilot are presented in Section 3. Section 4 outlines the operational terminology that is used to characterize the membrane performance. Detailed discussions of membrane performance, water quality, membrane integrity testing, and cleaning results are presented in Sections 5 through 8.

1

ZENON Private

ZEEWEED® 1000 Pilot Study Marin, CA

June 2006

2.

PILOT OBJECTIVES

The following were the specific objectives of the Marin County pilot study:

1) Determine the optimal design parameters that will generate stable membrane performance

with a recovery clean (CIP) frequency of no more often than 45 days for the dry season, from May to November, and no more frequent than 30 days for the wet season, from December to April.

2) Demonstrate a minimum system recovery of 93%.

3) Demonstrate that the ZeeWeed® Ultrafiltration Membrane Water Treatment System will

produce treated water that will meet all of the applicable standards.

4) Develop an effective cleaning regime, including chemical types and maximum time between

cleanings.

2

ZENON Private

ZEEWEED® 1000 Pilot Study Marin, CA

June 2006

3. MEMBRANE, PROCESS, AND PILOT DESCRIPTIONS

The ZeeWeed® based drinking water treatment is a low energy immersed ultrafiltration membrane process that consists of outside-in, hollow-fiber modules immersed directly in the feed-water (Figure A). The small pore size of the ultrafiltration membranes ensures that no particulate matter, including Cryptosporidium oocysts, Giardia cysts, suspended solids or other contaminants of concern, will pass into the treated water stream.

Figure A. Membrane Fiber Cross-Section

3.1

ZeeWeed® 1000 Membrane

Figure B illustrates an individual ZeeWeed® 1000 membrane module, which is the building block and smallest replaceable unit within a ZeeWeed® 1000 filtration system. One ZeeWeed® 1000 membrane module consists of thousands of horizontally oriented hollow fibers mounted between two vertical plastic headers. Shrouds enclose the fibers, leaving only the bottom and top open to create a vertical flow upwards through the fiber bundles. The membrane module characteristics are summarized in Table 1.

3

ZENON Private

ZEEWEED® 1000 Pilot Study Marin, CA

June 2006

Figure B. ZeeWeed® 1000 Module . Table 1. ZeeWeed® 1000 Membrane Module Characteristics

Size of module used in study Configuration Nominal membrane area Nominal membrane pore size Membrane material / construction Membrane surface properties Maximum dry shipping weight Maximum weight of module Typical operating transmembrane pressure Maximum operating temperature Operating pH range 27.2 x 26.9 x 4.1" (LxHxW) Outside-in hollow fiber 450 ft² and 600 ft2 0.02 m PVDF Non-ionic and Hydrophilic 12.9 kg (25.5 lb) 17.4 kg (38.4 lb) -1 to -13 psi 35oC (95oF) 5 - 9.5

4

ZENON Private

ZEEWEED® 1000 Pilot Study Marin, CA

June 2006

Cassettes are built by assembling modules in the vertical and horizontal dimensions. In the vertical dimension, a standard stack has three modules, but stacks of two modules can be easily assembled to fit in shallow tanks. This modular structure allows for the modules to be assembled into cassettes that can be designed to fit into virtually any size tank, allowing for efficient retrofitting of existing tanks. The cassette illustrated in Figure 3 is a 3-stack cassette containing 60 modules. A module can be inserted into, or removed from, the cassette by sliding it like a book into a bookcase. Each stack of modules is connected via an isolation valve to a permeate manifold that runs horizontally above the cassette.

Figure C. ZeeWeed® 1000 Cassette A ZeeWeed® 1000 train is a production unit containing a number of cassettes immersed in an open tank. Figure 4 shows two trains containing six cassettes each. Feed enters each train at the bottom of the tank from a feed channel that runs along one side of the train. Permeate is collected through a common header.

5

ZENON Private

ZEEWEED® 1000 Pilot Study Marin, CA

June 2006

Figure D. ZeeWeed® 1000 Train

3.2

ZeeWeed® 1000 Water Treatment Process

The ZeeWeed® 1000 system is operated as a simple semi-batch process where filtration and backwash alternate in sequence. During the filtration cycle, permeate is withdrawn through the membranes by applying vacuum to the permeate piping. The water removed by permeation is replaced with feed water to maintain a constant level in the tank. No aeration is used while in filtration mode. At the end of each filtration cycle (typically 15 ­ 60 minutes), a backwash is performed (typically for 30 seconds). During the backwash, the membranes are simultaneously aerated and backpulsed to dislodge solids. Solids are loosened from the surface of the

membranes and suspended in the process tank due to the aeration. Once the backwash is complete, the process tank is completely drained and aerated, which rids the tank of any accumulated solids. The process tank is then refilled with feed water and production resumes. Other processes used as part of the ZeeWeed® 1000 operation include maintenance cleans, recovery cleans, and pressure decay tests. The properties of these processes are explained in Section 4.

6

ZENON Private

ZEEWEED® 1000 Pilot Study Marin, CA 3.2.1 Coagulation Pretreatment

June 2006

Enhanced coagulation pretreatment generally consists of adding an aluminum or iron based coagulant to the feed stream. The pH may also be adjusted with an acid or base to enhance organics removal efficiency.

In the case of this pilot, Kennedy/Jenks provided equipment for coagulant dosing and mixing. Coagulant was only dosed from February 23, 2006 to March 01, 2006. Doses of 5 to 20 mg/L of ferric chloride were dosed during this period.

3.2.2

UF Pretreatment for RO

Long life and efficient operation of reverse osmosis (RO) membranes largely depends on the quality of the feedwater that a pretreatment system can provide. Whether RO membranes are desalting seawater, producing ultrapure water for industrial processes, or removing impurities from drinking water, pretreatment systems must provide a consistent supply of high quality feedwater to ensure trouble-free RO operation. Ineffective or unreliable pretreatment can adversely affect the RO system with problems such as high rates of membrane fouling, excessive cleaning requirements, lower recovery rates, high operating pressure, reduced membrane life, and poor quality product water. Each of these factors contributes to higher operational costs and lower RO plant productivity. ZeeWeed® immersed hollow fiber ultrafiltration (UF) membranes are increasingly being selected over conventional granular media systems for their ability to provide superior protection of valuable RO systems from particulate or biological fouling. UF membranes form a physical barrier that effectively blocks virtually all suspended particles from entering the feedwater stream regardless of the turbidity of the raw water. This is particularly important for desalination plants where turbidity can vary greatly during sea storms. Conventional systems may not be able to consistently produce high quality water under such conditions. UF membrane fibers are inherently insensitive to upsets caused by high turbidity or variable raw water quality. ZeeWeed® systems can consistently deliver RO feedwater with a turbidity of less than 0.1 NTU and a low silt density index (SDI).

7

ZENON Private

ZEEWEED® 1000 Pilot Study Marin, CA

June 2006

3.3

ZeeWeed® 1000 Pilot

ZENON supplied a pilot-scale ZeeWeed® 1000 system for the evaluation study in Marin County, including three ZeeWeed® 1000 membrane modules manufactured by ZENON Membrane Products in Ontario, Canada. A photograph of the typical equipment used in a pilot set-up is shown in Figure 5.

Figure E. Typical ZeeWeed® 1000 Ultrafiltration Pilot System

8

ZENON Private

ZEEWEED® 1000 Pilot Study Marin, CA

June 2006

4. OPERATIONAL TERMINOLOGY

Flux, transmembrane pressure, permeability, recovery, backwash properties (frequency, duration, flux, pressure, permeability, and airflow) and maintenance clean properties (frequency, chemical type, and chemical dose) are operating parameters used to evaluate the performance of the ZeeWeed® 1000 membrane. These terms are described in the following paragraphs, along with remarks about the values employed or achieved during piloting. Detailed discussions of pilot results follow in Sections 5 ­ 8. Operational pilot data has been compiled as charts in Appendix A.

4.1

Flux

Flux is a measure of the rate at which the product (or permeate) passes through the membrane per unit of outside surface area of membrane. It is reported in units of gal/ft2/day (gfd). The net flux is a calculation that takes into account the frequency and duration of backwashing, accounting for the lost production time as well as the actual volume of permeate lost during the backwash. In addition, losses associated with maintenance cleans and recovery cleans are taken into account. The instantaneous flux does not account for the backpulse volume that is used during backwashing or the volume of permeate used during the maintenance cleans, and is therefore a higher volume. The fluxes stated in this report are instantaneous values.

Figure A1 illustrates the instantaneous flux throughout the study. The pilot was operated at 25 to 40 gfd over the course of this study.

4.2

Transmembrane Pressure

Transmembrane pressure (TMP) refers to the vacuum required to pull clean water through the membrane. The ZeeWeed® 1000 system is designed to maintain a constant flux. Therefore, as the membrane becomes fouled, the transmembrane pressure increases. A cleaning is typically required once the transmembrane pressure reaches approximately 13 psi.

TMP profiles for the entire study are shown in Figures A2 and A3. A detailed discussion of TMP data is presented in Section 5.

9

ZENON Private

ZEEWEED® 1000 Pilot Study Marin, CA

June 2006

4.3

Permeability

Permeability is a calculated parameter of flux normalized against transmembrane pressure. It is reported in units of gfd/psi. Permeability is the parameter that is typically corrected to account for temperature variations. Adjusting the permeability for temperature allows the influence of fouling to be isolated from those variations caused by temperature. The formula used to calculate permeability at x°C is shown below:

Permeability @ x °C = Permeability @ T *1.025 ( x -T )

Membrane permeability throughout the different phases in the study is shown in Figures A4 and A5, along with the temperature profile. The temperature varied between 9°C and 24°C over the duration of the study.

4.4

Recovery

The recovery is the percent of the raw water passing through the ultrafiltration membrane as permeate. It is primarily controlled by flux and length of the filtration cycle. The only waste created through the ZeeWeed® 1000 process is that which is rejected and disposed of either during the backwash procedure or during maintenance cleaning.

The pilot was operated at recoveries of 92 and 95% over the duration of the study.

4.5

Backwash Properties

The backwash is used as a method of cleaning in the ZeeWeed® 1000 process. Every 15-60 minutes, flow is reversed through the membrane (backpulsed) for 30 seconds, pushing clean water from the inside of the membrane lumen to the outside. The water used for the backpulse is permeate that has been collected in the backpulse tank. Following the membrane backpulse, the contents of the membrane tank are completely drained to rid the tank of any accumulated solids. The backwash frequency is varied in order to achieve the desired recovery set point. The backwash frequency (or cycle length) varied from 24 to 38 minutes.

10

ZENON Private

ZEEWEED® 1000 Pilot Study Marin, CA

June 2006

Backpulse flux refers to the rate at which the backpulse water (permeate from the backpulse tank) passes through the membrane per unit of surface area of membrane. The backpulse flux is typically set to equal the permeate flux. Backpulse pressure is the transmembrane pressure required to push clean water from the inside to outside of the membrane during a backpulse. Backpulse permeability is a calculated parameter that represents the permeability of the membrane observed when the flow through the membrane is reversed. Air is applied during the backwash and maintenance clean procedures for the ZeeWeed® 1000 process. The air is supplied to the bottom of the membrane module. As it travels through the membrane stack to the surface of the process tank, it scours the outside of the membrane fibers and removes any larger particles that have adhered to the surface of the fibers. In addition, the airflow creates an airlift effect within the membrane tank to hold in suspension any solids from the membrane surface so that they are easily flushed out of the membrane tank during the drain at the end of the backwash and maintenance clean procedure. The air remains on during the tank drain portion of the backwash, while a VFD on the blower ensures the airflow remains consistent throughout the drain cycle. The duration of the drain cycle can be altered in order maximize the solids removal efficiency. During the piloting, the airflow was maintained at 3 dcfm while the drain time ranged between 45 and 180 seconds.

4.6

Maintenance Cleans and Aeration

A maintenance clean is another operational strategy used to control membrane fouling. Maintenance cleans are typically initiated once per day, but the frequency can be increased or decreased as needed.

During maintenance cleans the process tank is drained and then filled with permeate. Chemicals such as sodium hypochlorite or citric acid are added to achieve a desired concentration in the process tank. Once the tank is full, the membranes are recirculated or soaked with the chemical solution for 10 - 60 minutes and then the solution is drained from the tank. Standard sodium hypochlorite doses range from 50 to 250 mg/L. Citric acid is typically used at 0.5 to 2 g/L. Before resuming production, chemical residuals are flushed from the process tank. In the case of

11

ZENON Private

ZEEWEED® 1000 Pilot Study Marin, CA

June 2006

this pilot, sodium hypochlorite maintenance cleans ranged from 100 to 200 mg/L. The cleaning solution was recirculated at a flux of 15 gfd every 24 hours beginning on July 4, 2005.

Another tool that can be utilized with ZeeWeed® technology to extend the recovery cleaning interval is "extended aeration" procedures. This aeration procedure can be initiated

independently or coupled with a chemical maintenance clean and is typically used in high solids or high turbidity feed waters. During this chemical-free clean, the membrane tank is aerated for a total time of 24 minutes before the membrane tank is completely drained. The tank is aerated at eight different level set points, starting with a full tank, and intermittently drained down to the next lower set point until the tank has been aerated at every level and has been fully drained. In the case of this pilot, after the solids accumulation observed in the module during the first two weeks of January 2006, extended aerations were performed every 24 hours as part of the daily chemical maintenance clean.

12

ZENON Private

ZEEWEED® 1000 Pilot Study Marin, CA

June 2006

5. DISCUSSION - MEMBRANE PERFORMANCE

The Marin County, CA pilot study was divided into two primary phases. These were the Flux Optimization Phase for the dry season and Flux Optimization Phase for the wet season. During each phase different maintenance clean strategies were utilized to extend the recovery cleaning interval. Performance graphs can be found in Appendix A.

5.1

Phase 1: Flux Optimization ­ Dry Season (May 26 to Nov 30, 2005)

The objective of this phase was to determine optimal pilot parameters for the full scale design during the dry season in Marin County. During the course of the piloting the dry season occurred between May and November and is characterized by lower raw water turbidities than during the remainder of the year. A performance summary is found in Table 2 below.

13

ZENON Private

ZEEWEED® 1000 Pilot Study Marin, CA

June 2006

Table 2: Pilot Performance Summary, Phase 1 ­ Dry Season

Date May 26 to Jun 22, 2005 Jun 22 to Jun 24, 2005 Jun 24 to Jul 4, 2005 Jul 4 to Jul 12, 2005 Jul 14 to Jul 24, 2005 Jul 28 to Sep 12, 2005 Sep 15 to Nov 30, 2005

Flux Recovery Run Time Change in Change in Vacuum Cleaning Average Feed (gfd) (%) (days) Vacuum (psi) per day (psi/day) Interval Turbidity (NTU)

Maintenance Clean Strategy

Notes

25

92

26.82

1.75

0.07

> 60 days

5.2

none

Feed turbidity spikes of up to 20 NTU

27

92

1.97

0.44

0.22

37 days

3.4

none

27

95

10.19

2.95

0.29

28 days

3

none

27

95

7.97

-1.35

-0.17

n/a

3.6

Daily 100 mg/L NaOCl

30

95

8.35

0.97

0.12

> 60 days

6.9

Daily 100 mg/L NaOCl

New modules installed (450 ft2). Feed turbidity spikes of up to 10 NTU

35

95

29.91

3.81

0.13

53 days

4.2

Daily 100 mg/L NaOCl Feed turbidity spikes of up to 35 NTU

35

95

52.02

2.02

0.04

> 60 days

6.7

Daily 100 mg/L NaOCl

Recovery clean performed prior to this run. Feed turbidity spikes of up to 30 NTU

Data presented is corrected to 20oC

14

ZENON Private

ZEEWEED® 1000 Pilot Study Marin, CA

June 2006

Excellent performance was observed throughout the dry weather trials from May to November. Throughout the duration of testing, fluxes were tested at 25 to 35 gfd with a range of recoveries from 92 to 95%. Very stable performance was observed at the beginning of the piloting with a cleaning interval of greater than 60 days for a flux of 25 gfd and 92% recovery. As such, the flux was increased to 27 gfd on June 22 and the recovery was subsequently increased to 95% on June 24. These changes resulted in cleaning intervals of 37 and 28 days respectively.

From July 4 to November 30, the end of Phase 1, the flux ranged from 27 to 35 gfd while performing daily 100 mg/L sodium hypochlorite recirculation maintenance cleans. The cleaning interval was not significantly impacted by the range of fluxes tested, and was in fact greater than 50 days for each of the fluxes tested under these conditions. Important dates to note during this period are July 14, September 12, and November 30. In early July the membrane modules in Marin were noted as coming from a batch where possible manufacturing defects may have occurred. Although no issues had been detected with the initial modules during the first few months of operation, as a precaution, the modules were replaced with new modules on July 14. On September 12 and November 30 a CIP was performed. More detail on these cleans is found in Section 8. It should be noted that it was discovered that the temperature transmitter was faulty throughout this Phase. Temperatures obtained from grab samples were used to calculate cleaning intervals corrected to 20oC.

For the duration of Phase 1, the turbidity ranged from 3.0 to 6.9 NTU. Under these conditions a flux of 35 gfd can be sustained when 100 mg/L sodium hypochlorite maintenance cleans are performed daily.

5.2

Phase 2: Flux Optimization ­ Wet Season (Dec 5 to Apr 25)

The objective of this phase was to determine optimal pilot parameters for the full scale design during the wet season in Marin County. The wet season is characterized by higher turbidities and was tested between December 5, 2005 and April 25, 2006. A performance summary is found in Table 3 below.

15

ZENON Private

ZEEWEED® 1000 Pilot Study Marin, CA

June 2006

Table 3: Pilot Performance Summary, Phase 2 ­ Wet Season

Date

Flux Recovery Run Time Change in Change in Vacuum Cleaning Average Feed (gfd) (%) (days) Vacuum (psi) per day (psi/day) Interval Turbidity (NTU)

Maintenance Clean Strategy

Notes

Dec 5, 2005 to Jan 3, 2006

35

95

23.53

3.45

0.15

46 days

14.5

Daily 100 mg/L NaOCl Feed turbidity spikes of up to 60 NTU

Jan 12 to Jan 14, 2006

40

95

1.71

4.70

2.74

2 days

36.4

New modules installed (600 ft ). Solids accumulation in the modules due to elevated turbidity caused the Daily 100 mg/L NaOCl pilot to shut down. Series of aeration maintenance cleans restored permeability. Feed turbidity spikes of up to 60 NTU. Daily extended aeration prior to daily 100 mg/L NaOCl. Backpulse drain time set to 3 minutes.

2

Jan 24 to Jan 30, 2006

25

95

5.94

0.47

0.08

> 60 days

22.4

Feed turbidity spikes of up to 80 NTU

Jan 30 to Feb 7, 2006

28

95

6.72

0.27

0.04

> 60 days

30.6

Daily extended aeration prior to daily 100 mg/L Feed interuption caused infrequent shut downs. Feed NaOCl. Backpulse drain turbidity spikes of up to 75 NTU time set to 3 minutes

Feb 8 to Feb 14, 2006

32

95

5.81

0.47

0.08

> 60 days

18.5

Daily extended aeration prior to daily 100 mg/L Feed turbidity spikes of up to 55 NTU NaOCl. Backpulse drain time set to 3 minutes

16

ZENON Private

ZEEWEED® 1000 Pilot Study Marin, CA

June 2006

Date

Flux Recovery Run Time Change in Change in Vacuum Cleaning Average Feed (gfd) (%) (days) Vacuum (psi) per day (psi/day) Interval Turbidity (NTU)

Maintenance Clean Strategy

Notes

Feb 14 to Feb 16, 2006

32

95

1.96

0.20

0.10

> 60 days

25.5

Daily extended aeration prior to daily 100 mg/L Feed turbidity spikes of up to 65 NTU NaOCl. Backpulse drain time reduced to 45 seconds.

Feb 16 to Feb 17, 2006

36

95

1.09

0.34

0.31

21 days

17.9

Daily extended aeration prior to daily 100 mg/L Feed turbidity spikes of up to 25 NTU NaOCl. Backpulse drain time set to 45 seconds

Feb 17 to Feb 21, 2006

36

95

3.93

0.43

0.11

> 60 days

16.3

Daily extended aeration prior to daily 200 mg/L Feed turbidity spikes of up to 30 NTU NaOCl. Backpulse drain time set to 45 seconds

Feb 23 to Mar 1, 2006

36

95

5.96

0.09

0.02

> 60 days

28.4

Daily extended aeration prior todaily 200 mg/L 5 to 8 mg/L of ferric chloride dosed for inline NaOCl. Backpulse drain coagulation. Feed turbidity spikes of up to 70 NTU time set to 45 seconds

Mar 1, 2006

36

95

0.13

negative

negative

n/a

25.5

Daily extended aeration prior to daily 200 mg/L 10 to 20 mg/L of ferric chloride dosed for inline NaOCl. Backpulse drain coagulation. time set to 45 seconds

Mar 7 to Mar 14, 2006

36

95

6.89

1.28

0.19

35 days

34.7

Daily extended aeration prior to daily heated Feed turbidity spikes of up to 80 NTU 100 mg/L NaOCl. Backpulse drain time set to 45 seconds

17

ZENON Private

ZEEWEED® 1000 Pilot Study Marin, CA

June 2006

Date

Flux Recovery Run Time Change in Change in Vacuum Cleaning Average Feed (gfd) (%) (days) Vacuum (psi) per day (psi/day) Interval Turbidity (NTU)

Maintenance Clean Strategy

Notes

Mar 14 to Apr 2, 2006

36

95

15.37

0.49

0.03

> 60 days

30.7

Daily extended aeration prior to daily heated 100 Feed turbidity spikes of up to 80 NTU mg/L NaOCl. Backpulse drain time increased to 2 minutes. Daily extended aeration prior to daily heated 100 Compressor failure caused extended shutdown mg/L NaOCl. prior to starting this run. Feed turbidity spikes of up Backpulse drain time set to 90 NTU to 2 minutes

Apr 11 to Apr 17, 2006

36

95

6.40

0.40

0.06

> 60 days

30.2

Apr 21 to Apr 25, 2006

36

95

4.32

0.16

0.04

> 60 days

22.8

Daily aeration MCs with daily heated 100 mg/L Lack of feed water caused extended shutdown prior NaOCl. Backpulse drain to this run. Feed turbidity spikes of up to 80 NTU time set to 2 minutes

Data presented is corrected to 20oC

18

ZENON Private

ZEEWEED® 1000 Pilot Study Marin, CA

June 2006

The ZeeWeed® 1000 with 600 ft2 modules was able to meet the recovery cleaning interval objective while seeing feed turbidity averaging 25 NTU and spikes up to 90 NTU. Various features or tools incorporated into the ZeeWeed® process toolbox to combat high feed water TSS or high feed water turbidity including flexible maintenance clean procedures and various aeration methods. Throughout the duration of Phase 2, maintenance cleans and backwash drain times were modified to target solids removal. The concentration of sodium hypochlorite maintenance cleans was varied between 100 and 200 mg/L. Aeration cleans were also implemented during this phase and coupled with the maintenance clean. This feature improves solids removal from the

membranes and also enhances the performance of the chemical portion of the clean by removing solids that would otherwise decrease the efficiency of the chemicals. Backwash drain times varied between 45 and 180 seconds and included constant aeration at 3 scfm during the drain. Increasing the time to drain the membrane tank increases the amount of time the membranes are aerated and therefore increases the amount of solids that are removed. With the proper

combination of solids removal techniques and with the optimal maintenance clean strategy stable performance was once again observed.

The first run in Phase 2 occurred from December 5, 2005 to January 3, 2006. Feed turbidity during this time averaged 14.5 NTU, which was approximately double the feed turbidity observed during the dry season. The pilot ran at 35 gfd and 95% recovery, performing 100 mg/L sodium hypochlorite cleans daily. Under these conditions a cleaning interval of 46 days was achieved. Due to the good performance to date ZENON saw an opportunity at this point in time to test modules with a larger surface area (600 ft2 as opposed to the 450 ft2 modules tested to date). One major advantage offered to MMWD through the use of a higher surface area module, is that a smaller footprint can be observed for the full scale plant or more spare space can be available for future expansion, depending on the design of the full scale plant. On January 12, 600 ft2 modules were installed in the pilot. The pilot was restarted at a flux of 40 gfd and 95% recovery with 100 mg/L sodium hypochlorite cleans being performed daily. The combination of the higher surface area modules, the increased flux, and the elevated feed turbidity levels of 36 NTU resulted in the pilot alarming on high TMP after 3 days of operation. In the following week, attempts were made to restart the pilot but as this was done without attempting to clean the modules first, the pilot would shut off on high TMP shortly after start up. On January 24, the modules were removed from the membrane tank for inspection. As shown in 19 ZENON Private

ZEEWEED® 1000 Pilot Study Marin, CA

June 2006

Figure 4 below, solids accumulation in the membrane fiber bundle was observed. At this point a series of three extended aeration cleans were performed using the procedure described in Section 4.6. Following the third extended aeration the modules were inspected again and were deemed to be relatively free of solids. Photographs of the module before and after the aeration clean are found in Figures F and G. This event demonstrated that sludging of the modules could be reversed quickly, without chemical, permanently, and with standard tools available in the ZeeWeed® process toolbox.

Figure F. Photograph of Solids Accumulation in Module prior to Extended Aeration Clean

Figure G. Photograph of Module following the Extended Aeration Clean

20

ZENON Private

ZEEWEED® 1000 Pilot Study Marin, CA

June 2006

Following the extended aeration clean, the pilot was restarted at a more conservative flux of 25 gfd. The recovery remained at 95% and daily 100 mg/L sodium hypochlorite maintenance cleans were scheduled. Due to the previous solids accumulation observed from January 12 to 14, a daily extended aeration was scheduled to replace the normal rapid draining of the membrane tank that was occurring prior to the sodium hypochlorite maintenance clean. To further aid in solids removal the backwash drain time was increased from 45 to 180 seconds, allowing more time for aeration during the drain. These changes resulted in a cleaning interval of greater than 60 days until January 30 when the flux was increased to 28 gfd. From January 24 to January 30 the feed turbidity averaged 22.4 NTU with spikes up to 80 NTU.

From January 30 to February 14, the flux was increased to 28 gfd and again to 32 gfd on February 8 with no other changes to the operating parameters. The average feed turbidity during these runs was 30.6 and 18.5 NTU respectively. Feed turbidity spikes up to 75 NTU were observed during this period. The cleaning interval was not impacted and was greater than 60 days for both conditions tested.

From February 14 to 21, several short runs were performed in order to try to optimize more quickly. On February 14 the backwash drain time was reduced back to 45 seconds and all other parameters remained the same. Minimal impact on cleaning interval was observed, therefore, the flux was increased to 36 gfd on February 16. With the change in flux the estimated cleaning interval over this short period of time was reduced to 21 days when compared to greater than 60 days at 32 gfd. Since the target cleaning interval was not being achieved, the concentration of the daily sodium hypochlorite maintenance clean was increased to 200 mg/L. As a result of this change, a cleaning interval of greater than 60 days was observed. During this period the feed turbidities ranged between 16.3 and 25.5 NTU with spikes up to 65 NTU regularly observed.

At this point, Kennedy/Jenks wished to observe the impact of using coagulant in the feed. Doses of 5 to 8 mg/L of ferric chloride were tested from February 23 to March 1. No changes were made to the pilot operating parameters during this time. Excellent performance was observed with a cleaning interval of greater than 60 days. When the coagulant dose was increased up to 20 mg/L of ferric chloride on March 1, no negative impact was observed, and in fact the TMP started to decrease. Feed turbidity during the coagulant testing averaged 27 NTU with spikes up to 70 NTU observed. 21 ZENON Private

ZEEWEED® 1000 Pilot Study Marin, CA

June 2006

With the end of the piloting approaching three runs were tested from March 7 to April 25. The pilot was tested at 36 gfd and 95% recovery. Daily heated maintenance cleans were performed at a concentration of 100 mg/L of sodium hypochlorite, with a target temperature of 95oF. The daily extended aeration prior to the sodium hypochlorite maintenance clean also continued. The first run during this period used a 45 second backwash drain time. At these conditions a 35 day cleaning interval was observed. An increase in the drain time to 120 seconds on March 14, resulted in cleaning intervals of greater than 60 days for the remainder of the testing. From March 7 to April 25 the average feed turbidity ranged between 22.8 and 34.7 NTU with spikes up to 90 NTU. A compressor failure was the cause for the shutdown between April 2 and 11. Two cleaning intervals were calculated for the period of April 12 to April 25 due to a shutdown caused by lack of feed water April 17 and 21.

The higher feed turbidity experienced during the course of Phase 2 posed a challenge for the membranes as evidenced by the lower cleaning intervals at the start of the phase. With the proper combination of solids removal techniques combined with the optimal maintenance clean strategy, stable performance was once again observed with cleaning intervals of greater than 60 days. A CIP was performed prior to switching the modules on January 12 with another occurring on March 2 following the coagulant testing. More detail on the CIP can be found in Section 8. The faulty temperature transmitter was replaced when the 600 ft2 modules were installed on January 12.

22

ZENON Private

ZEEWEED® 1000 Pilot Study Marin, CA

June 2006

6. DISCUSSION - WATER QUALITY

6.1 Turbidity

Turbidity is a measure of the clarity of water and is commonly expressed in nephelometric turbidity units (NTU). Suspended solids and colloidal matter (e.g. clay, silt and microscopic organisms) cause turbidity. Both feed and permeate turbidity were monitored on-line

continuously with a HACH 1720D turbidimeters.

Figure B1 shows the feed and permeate turbidity for the dry season while Figure B2 shows the feed and permeate turbidity for the wet season. During the dry season the feed turbidity averaged 4.71 NTU, while the permeate turbidity averaged 0.05 NTU. During the wet season the feed turbidity averaged 30 NTU while the permeate turbidity remained unaffected and averaged 0.05 NTU. Permeate turbidity spikes observed from the middle of February to the start of March and at the end of March were due to incorrect flow to the instrument.

23

ZENON Private

ZEEWEED® 1000 Pilot Study Marin, CA

June 2006

7. DISCUSSION - MEMBRANE INTEGRITY RESULTS

7.1 Membrane Integrity Tests

A Membrane Integrity Test (MIT) is a periodic means of confirming membrane integrity and consists of pressuring the inside of the membrane fibers with air to 10 psi. This pressure is allowed to stabilize to ensure that all of the water inside the membrane fibers has been displaced by air. The flow of air to the membrane is then turned off, and the decay in pressure is measured over time. A pressure decay of less than 0.65 psi over two minutes is the criteria used to verify membrane integrity. In this pilot study, pressure decay tests were performed daily beginning on July 17.

Figure C1 shows the membrane integrity test results for the tests performed. From February 23 to March 3 the pilot failed the integrity test. The cause of the failure was due to a leak in the piping on the pilot and the membranes were confirmed to be intact. All other membrane integrity tests during the study were considered a pass and over the course of the 11 month study no broken fibers occurred.

24

ZENON Private

ZEEWEED® 1000 Pilot Study Marin, CA

June 2006

8. DISCUSSION - CLEANING RESULTS

A recovery clean is required to restore the permeability of the membrane once the membrane becomes fouled. A fouled membrane condition occurs if the TMP approaches and does not stabilize at values of approximately 12 to 13 psi (terminal vacuum). The cleaning chemicals that are typically used are sodium hypochlorite for the removal of organic foulants and citric acid for the removal of inorganic contaminants. The typical pilot procedure for cleaning ZeeWeed® 1000 membranes consists of soaking them in a 500 mg/L sodium hypochlorite solution for 4-6 hours. This procedure is then repeated with a citric acid solution, for which the concentration is usually 2 g/L. Hydrochloric acid is sometimes added to further reduce the pH of the citric acid solution. Variations upon this practice can include only using one of the cleaning chemicals, changing chemical concentrations and/or durations and heating the cleaning solution.

Recovery cleans were performed on September 12, November 30, January 11 and March 1. The recovery cleans on September 12 and November 30 consisted of a 500 mg/L sodium hypochlorite solution and soaked overnight. This was followed by a six hour soak in 2 g/L of citric acid with the pH reduced to approximately 2.5 using HCl. On these occasions the temperature corrected permeability was restored to 5.2 and 5.5 gfd/psi for the cleans conducted on September 12 and November 30 respectively.

Different cleaning strategies were targeted on the remaining two cleans to optimize the cleaning performance. For the clean that occurred on January 11, the membranes were first soaked in hot water, approximately 95oF but not maintained, containing 5 g/L of citric acid to reduce the pH to 2.2. After 3 hours, more citric acid was added to obtain a concentration of 10 g/L and the membranes were allowed to soak for an additional one and a half hours. The citric soak was followed by a 500 mg/L sodium hypochlorite hot water soak. Again the target temperature of the clean was 95oF. After 16 hours the hypochlorite solution was replaced with a 2000 mg/L sodium hypochlorite solution and the membranes soaked for half an hour in this new solution. The clean on January 11 restored the temperature corrected permeability to 6.2 gfd/psi.

25

ZENON Private

ZEEWEED® 1000 Pilot Study Marin, CA On March 1 a recovery clean was performed on the 600 ft2 modules.

June 2006 For this clean the

membranes soaked for 15 hours in a 1000 mg/L sodium hypochlorite solution that was originally heated to 95oF. This soak was followed by a citric acid soak in which 10 g/L of citric acid were added to the heated solution to reduce the pH to 2.5. The target temperature for this soak was again 95oF. The citric acid soak was followed up by another sodium hypochlorite soak at a concentration of a 1000 mg/L and a temperature of 95oF. This procedure restored the temperature corrected permeability to 5.7 gfd/psi, however, no permeability was recovered from the additional sodium hypochlorite soak.

26

ZENON Private

ZEEWEED® 1000 Pilot Study Marin, CA

June 2006

9. CONCLUSIONS

The following sections highlight the conclusions that can be drawn from the Marin County pilot study.

Membrane Performance · During the dry season, from May to November, stable performance was observed at a flux of 35 gfd and 95% recovery. At these conditions and while performing daily 100 mg/L sodium hypochlorite maintenance cleans a predicted cleaning interval of greater than 50 days was achieved when corrected to a temperature of 20oC. · During the wet season, from December to April, stable performance was observed at a flux of 36 gfd and 95% recovery. The maintenance clean strategy utilized an extended aeration procedure as part of the daily heated 100 mg/L sodium hypochlorite maintenance clean. The maintenance clean combined with a 120 second backwash drain time resulted in a predicted cleaning interval of greater than 90 days when corrected to a temperature of 20oC.

Water Quality · The ZeeWeed® Ultrafiltration System consistently produced excellent water quality. Permeate turbidity averaged 0.05 NTU throughout the entire duration of the study even with feed water turbidity spikes of up to 90 NTU during the wet season.

Membrane Integrity · · Throughout the course of the piloting no leaks or breaks occurred in the membrane. Membrane integrity tests proved valuable in detecting leaks in the piping of the system.

Cleaning · Soaking the membranes in a 10 g/L citric acid solution heated to 95oF followed by a soak in 1000 mg/L of a sodium hypochlorite solution heated to 95oF restored the membrane permeability to approximately 6 gfd/psi.

27

ZENON Private

ZEEWEED® 1000 Pilot Study Marin, CA

June 2006

APPENDIX A. MEMBRANE PERFORMANCE RESULTS

28

ZENON Private

ZEEWEED® 1000 Pilot Study Marin, CA

June 2006

List of Figures

Figure A1: Flux for the Entire Pilot Study. Figure A2: Transmembrane Pressure for the Dry Season. Figure A3: Transmembrane Pressure for the Wet Season. Figure A4: Permeability and Temperature Profiles for the Dry Season. Figure A5: Permeability and Temperature Profiles for the Wet Season.

29

ZENON Private

ZEEWEED® 1000 Pilot Study Marin, CA

June 2006

Figure A1: Flux for the Entire Pilot Study.

30

ZENON Private

ZEEWEED® 1000 Pilot Study Marin, CA

June 2006

Figure A2: Transmembrane Pressure for the Dry Season.

31

ZENON Private

ZEEWEED® 1000 Pilot Study Marin, CA

June 2006

Figure A3: Transmembrane Pressure for the Wet Season.

32

ZENON Private

ZEEWEED® 1000 Pilot Study Marin, CA

June 2006

Figure A4: Permeability and Temperature Profiles for the Dry Season.

33

ZENON Private

ZEEWEED® 1000 Pilot Study Marin, CA

June 2006

Figure A5: Permeability and Temperature Profiles for the Wet Season.

34

ZENON Private

ZEEWEED® 1000 Pilot Study Marin, CA

June 2006

APPENDIX B. WATER QUALITY RESULTS

35

ZENON Private

ZEEWEED® 1000 Pilot Study Marin, CA

June 2006

List of Figures and Tables

Figure B1: Feed and Permeate Turbidity for the Dry Season. Figure B2: Feed and Permeate Turbidity for the Wet Season.

36

ZENON Private

ZEEWEED® 1000 Pilot Study Marin, CA

June 2006

Figure B1: Feed and Permeate Turbidity for the Dry Season.

37

ZENON Private

ZEEWEED® 1000 Pilot Study Marin, CA

June 2006

Figure B2: Feed and Permeate Turbidity for the Wet Season.

38

ZENON Private

ZEEWEED® 1000 Pilot Study Marin, CA

June 2006

APPENDIX C. MEMBRANE INTEGRITY RESULTS

39

ZENON Private

ZEEWEED® 1000 Pilot Study Marin, CA

June 2006

List of Figures and Tables

Figure C1: Membrane Integrity Test results for the Entire Study

40

ZENON Private

ZEEWEED® 1000 Pilot Study Marin, CA

June 2006

Figure C1: Membrane Integrity Test Results for the Entire Study

41

ZENON Private

Information

Microsoft Word - CA - Marin, Final Report.doc

48 pages

Report File (DMCA)

Our content is added by our users. We aim to remove reported files within 1 working day. Please use this link to notify us:

Report this file as copyright or inappropriate

839338