Membrane Autopsy Report Completed for: Ashland Chemicals March 2012 WO#021712-2 USA: Avista Technologies, Inc. Phone: 760 744-0536 info@avistatech.com United Kingdom: Avista Technologies (UK) Ltd. Phone: 44 (0) 131 4496677 info@avistatech.co.uk F O C U S E D S O L U T I O N S
Table of Contents Executive Summary 3-4 Initial Tests on Elements as Received Element weight 5 Element Wet Testing Results 5 Element Integrity Test 6 External Inspection Spiral-Wound Membrane Element Construction 7-8 Fiberglass Casing Damage 9 Brine Seal Damage 10 Anti-Telescoping Device (ATD) Damage 10 Permeate Tube Damage 10 Internal Inspection and Testing Scroll End Examination 11 Membrane Surface Visual Contamination 11-12 Feed Spacer Visual Contamination 13 Permeate Spacer Damage 13 Glue Line Integrity 13 Testing for Exposure to Oxidizing Halogens (Chlorine) 14 Cell Testing for Water Passage Flow and Salt Passage 15 Detailed Foulant Analysis Organic Content Testing 16 Foulant Density Measurement 16 Testing for Presence of Carbonates 16 Testing for the Presence of Microbiological Organisms 17 Testing for the Presence of Coagulant 18 Testing for Organic Foulant Constituents 19 Testing to Identify Inorganic Foulant Constituents 20-22 Cell Test & Laboratory Clean-In-Place Study 23-26 Laboratory Certification 27 Autopsy Report 3.0 Avista Technologies, Inc. Page 2 of 27
Executive Summary Background Ashland Chemicals provided one Filmtec BW30-365 reverse osmosis element to Avista Technologies for analysis. Initial Element Testing SN# F4812779 was wet tested and data was normalized to the manufacturer s standard conditions. The element produced higher than normal flow (136% of normal), normal rejection, and high delta pressure (10 psi) upon full element wet testing. External Inspection The fiberglass casing, anti-telescoping devices, brine seal, and permeate tube were in good condition and showed no signs of physical damage. The element also passed the integrity test confirming that there was no damage to the mechanical components of the spiral wound element. Internal Inspection and Testing The feed scroll end (brine seal end) had some black colored plastic shavings on the surface. The active membrane surfaces of SN# F4812779 were unevenly coated with a pink colored foulant material. The foulant material was also observed in the feed spacer which most likely contributed to the higher than normal delta pressure observed during full element wet testing. The permeate spacers and glue lines were in good condition. The membrane tested positive for the presence of halogens (i.e. chlorine) in the membrane structure. Cell Testing Results SN# F4812779 flat sheet cut membrane samples produced 140% of normal water passage and normal salt passage during the baseline cell test. Autopsy Report 3.0 Avista Technologies, Inc. Page 3 of 27
Foulant Analysis Organic content of the foulant material was determined from loss on ignition to be 96.7% indicating that the majority of the foulant material was organic. Microscope analysis of foulant scraped from the membrane surface identified Gram negative bacteria, Gram positive bacteria, algae and colloidal solids in the foulant layer. FT-IR analysis located strong bands associated with organic material (proteins and carbohydrates) and weak band between 1100 and 900 cm -1 indicating silica. EDX analysis identified only trace amounts of silica and calcium on the membrane surface. The high carbon and oxygen weight percents indicate a significant amount of organic material. SEM images showed a smooth foulant layer on the top of the membrane surface with colloidal particles embedded in the foulant layer. The dried foulant layer was peeling from the membrane surface indicating high organic content and the colloidal solids were between 2-20 microns in size. Chromatic elemental imaging (CEI) confirmed that the foulant layer was mainly organic material by showing a high distribution of carbon across the foulant layer. The colloidal solids embedded in the layer were composed of silica and calcium. Cleaning Study Results RoClean P111 (2% by weight, heated, with extended soak time) removed more than 90% of the foulant material from the flat sheet samples and feed spacers. Autopsy Report 3.0 Avista Technologies, Inc. Page 4 of 27
Initial Element Test Results Because element weight is often indicative of the degree of fouling, elements are weighed prior to the autopsy. SN# F4812779 weighed 35 pounds; new elements of this model weigh approximately 32 to 35 pounds. WET TEST The element was wet tested on dechlorinated San Marcos, CA city water. Wet test results were normalized to the manufacturer s published test conditions. Filmtec BW30-365 Flow, gpm Rejection, % DP, psi SN# F4812779 8.99 99.2 10 Manufacturer s Specifications 6.1 to 6.6 99.0 to 99.5 3 to 5 Element Wet Testing Autopsy Report 3.0 Avista Technologies, Inc. Page 5 of 27
Vacuum (in Hg) Percent Loss INTEGRITY TEST To determine if a membrane performance problem is possibly caused by mechanical damage, membranes are tested to check for vacuum decay that may indicate abnormal bypass. In this test a vacuum of about 25 in Hg is applied to the permeate side of the membrane for a duration of 120 seconds. If over 35% of the vacuum is lost within a 120 second period, then the membrane can be said to have severe physical damage. 30 25 20 15 10 5 0 SN #F4812779 100 90 80 70 60 Vacuum 50 Percent loss 40 30 20 10 0 0 15 30 45 60 75 90 105 120 Time (seconds) SN# F4812779 passed the integrity test. Autopsy Report 3.0 Avista Technologies, Inc. Page 6 of 27
External Inspection The external inspection of a membrane element is an important step in the autopsy process. Physical damage to the exterior components can contribute to performance issues in the element or may yield clues as to the operating conditions of the membrane system that led to poor membrane performance. As most of the external components are damaged during the autopsy process, documenting any significant finds before further work is completed is essential. This section covers the fiberglass casing, anti-telescoping devices (ATDs), permeate tube, and brine seal. In addition the scroll ends are also examined for any foulant/scale material that may be interfering with flow and for feed spacer extrusion also known as telescoping and gapping in the scroll end which may cause localized scaling (uneven hydraulics). Spiral-Wound Membrane Element Construction: Autopsy Report 3.0 Avista Technologies, Inc. Page 7 of 27
Autopsy Report 3.0 Avista Technologies, Inc. Page 8 of 27
Fiberglass Casing: The fiberglass casing is an integral part of each element. The purpose of this filament wound wrap is to protect the element from external differential pressure, provide compressive strength to prevent telescoping and to ensure that the various membrane components are held in their correct position for optimum performance. Damage to the fiberglass wrap can be an indication of rough handling or damage from excessive differential pressure across the membrane surface. The fiberglass casing of SN# F4812779 was in good condition and showed no signs of physical damage. External fiberglass casing for SN# F4812779 Image of feed end ATD (left) and concentrate end ATD (right) Autopsy Report 3.0 Avista Technologies, Inc. Page 9 of 27
Brine Seal: The purpose of the brine seal is to seal against the inside diameter of the pressure vessels and the outside diameter of the membrane to ensure that all the feed water passes through the membrane element. Chevron type seals are used to aid in membrane loading and to seal to a variety of pressure vessel inside surfaces. The brine seal for SN #F4812779 was in good condition and showed no signs of damage. Anti-Telescoping Device (ATD): When assembled at the factory, membrane elements are commonly fitted with Anti- Telescoping Devices (ATDs) at each end of the element. These devices are designed to prevent telescoping of the membrane leaves under normal operating conditions that can cause membrane damage. The ATD s were in good condition and showed no signs of physical damage (i.e. gouges, cracks). Permeate tube: At the center of each membrane element is a round section of pipe that is called the permeate tube. Down the length of the tube, holes are drilled through the pipe wall to the tube center. This tube is bonded to the membrane leaves and permit water to flow from the leaves outward at each end of the full element and the through the holes for collection. To function properly, the permeate tube must be free from gouges or damage that can prevent proper o-ring sealing at each end. Poor sealing can result in bypass from the high-pressure feed/concentrate flow into the permeate stream. No gouges were visible on the ends of the permeate tube that could allow by-pass of feed water. Autopsy Report 3.0 Avista Technologies, Inc. Page 10 of 27
Internal Inspection and Testing Scroll End Examination: Once the anti-telescoping devices are removed, the scroll ends of the membrane leaves are examined for presence of colloidal particles, biofouling, feed spacer extrusion and membrane gapping. In addition, each scroll end is examined for the gradual axial shift of the element leaves from outer diameter of the element towards the permeate tube. This type of damage is termed "telescoping" and is caused by the development of high differential pressure (usually greater than 10 psi) across the element. Black colored plastic shavings were observed on the feed scroll end of SN #F4812779. The concentrate scroll end was in good condition and free of foulant debris. Feed scroll end (left) and concentrate scroll end (right) of SN# F4812779 Membrane Surface Visual Contamination: When assembled, the surface of the membrane is a uniform, shiny surface with no visual contamination or impurities. Although the membrane surface contamination can be sometimes hard to detect visually many times contamination is very visible and easy to detect with the naked eye. The exposed membrane surfaces of SN# F4812779 were unevenly coated with a pink colored foulant material. Autopsy Report 3.0 Avista Technologies, Inc. Page 11 of 27
Flow Path Flow Path Exposed membrane surface for SN# F4812779 Image of the feed end of element SN# F4812779 Autopsy Report 3.0 Avista Technologies, Inc. Page 12 of 27
Feed Spacer Visual Contamination: The feed spacer is a plastic net material designed to separate membrane surfaces to form a flow path and to promote turbulence within feed water channel. The feed spacers in SN#F4812779 had foulant material trapped in the plastic netting. Permeate spacer Damage: Permeate spacer provides a path for permeate flow to channel towards the central permeate tube which minimizes permeate-side pressure losses. The permeate spacers were in good condition for element SN# F4812779. Glue line Integrity: Membrane leaves are glued on three sides to separate feed and permeate streams. Glue lines are inspected to ensure that there are no sections of unbounded material referred to as glue flaps that may block the feed channel into the element module. In some worst case situations, glue lines may fail at the feed end of the membrane permitting contamination. The glue lines are also inspected for pouching and delamination which often occur on the concentrate end of last stage elements. This type of physical damage may indicate permeate backpressure caused by positive pressure on the permeate side of the membrane. Glue lines in element SN# F4812779 were in good condition and showed no signs of pouching, delamination, or glue flaps. Autopsy Report 3.0 Avista Technologies, Inc. Page 13 of 27
Testing for the Presence of Oxidizing Halogens (i.e. chlorine, bromine or iodine) The Fujiwara test is used to confirm that a polyamide (PA) thin-film membrane has been exposed to an oxidizing halogen, such as chlorine, bromine, or iodine. This test analyzes whether halogens have become part of the polymer structure through oxidative attack. Please note that the Fujiwara test is a qualitative test and that any color change indicates the presence of a halogen in the membrane structure. However the test does not quantify the amount of exposure or which exact halogen is attached. Fujiwara testing for SN# F4812779 was positive for the presence of halogens (i.e. chlorine) in the membrane structure. Example of the color change observed during a typical Fujiwara test. A negative result is on the left and a positive result on the right. Autopsy Report 3.0 Avista Technologies, Inc. Page 14 of 27
Cell Testing for Permeate Flow and Salt Passage To determine membrane performance characteristics membrane samples are tested in a cell test apparatus (CTA). The water passage constant is expressed as the A value, and the salt passage constant is expressed by a B value. Both constants are functions of the chemical-physical properties of the membrane plus any fouling layer present. A and B value constants are also independent of operating parameters such as pressure, temperature, and salt content of the feed stream. A value units are cm/sec/atm. B value units are cm/sec. The table below shows baseline performance data before cleaning. Comparing cell testing of the membrane material to the original specification for the full spiral membrane element is a useful comparison tool. This data is collected in order to factor out any additional mechanical aspects the element construction may have caused in the spiral configuration. Filmtec BW30-365 SN#F4812779 Water Passage Constant A Value Flow (GFD/PSI) Salt Passage Constant B Value Rejection (%) Flat Sheet Membrane 1.46 E-04 140% of Normal 0.18 1.07 E-05 Normal 99.4 Manufacturer s original specifications 0.77 to 1.04 E-04 Normal Range 0.12 to 0.13 Normal Range 1.06 to 1.44 E-05 Normal Range Note: Testing conducted with dechlorinated city water from San Marcos, CA. Flow (GFD/PSI) Flat Sheet Results 99.0 to 99.5 Normal Range -LLooww - FFl llooww RR aannggee- - -HHi - iigghh FFl llooww RR aannggee- - 0 0.10 MFG. 0.15 0.20 0.25 >0.30 Spec Rejection (%) Flat Sheet Results <94 95 96 97 98 Approx. Rejection of Cell Test MFG. Spec. Autopsy Report 3.0 Avista Technologies, Inc. Page 15 of 27
Foulant Analysis Organic Content Testing: Loss on ignition (LOI) testing gives an approximation of the organic content of the foulant. Values in excess of 35% typically represent the presence of organic content. Organic content for SN# F4812779 is shown in the graph below. Loss on Ignition % for SN# F4812779 Inorganic portion 3.3 Organic portion 96.7 Foulant Density Measurement: Membrane foulant density is the weight of dry foulant per area of membrane surface. Foulant densities determined from past autopsies range from 0.02 to 1.84 mg/cm 2 and average 0.49 mg/cm 2. SN# F4812779 had a foulant density of 0.13 mg/cm 2. Testing for Presence of Carbonates: Acid testing is used to determine the presence of carbonates on the membrane surface. In this test, several drops of dilute hydrochloric acid were placed on the foulant surfaces. Effervescing indicate a positive test result. No effervescing was observed on the membrane surface of SN# F4812779 when acid was applied. Autopsy Report 3.0 Avista Technologies, Inc. Page 16 of 27
Testing for the Presence of Microbiological Organisms: Foulant samples were stained and examined with a light microscope at 1000x using an oil immersion lens. Gram positive bacteria are stained blue while Gram negative bacteria are stained red. Foulant scraped from the membrane surface contained Gram positive bacteria, Gram negative bacteria, algae and some colloidal particles. Light microscope image (1000X) of foulant scraped from SN# F4812779 Autopsy Report 3.0 Avista Technologies, Inc. Page 17 of 27
Testing for the Presence of Coagulant: Zeta potential testing of the membrane surface foulant can determine the presence of excess coagulant by measuring the charge associated with the surface colloids. Most naturally occurring colloids are negatively charged and surrounded by a double layer of counter ions. Zeta potential is the charge that resides at the double layer boundary, which we can conveniently measure with a zeta potential meter. Electrostatic repulsion becomes significant when two colloids approach each other and their charged double layers begin to interfere. Because of this mutual repulsion, coagulation and flocculation are difficult to accomplish and coagulants are often overfed into the RO system resulting in a positive zeta potential. Samples that show a near zero or neutral zeta potential represent the optimum coagulant dosage. Two grams of wet foulant was suspended in 200 ml of deionized H2O + 10 ppm NaCl to provide conductivity. The foulant material of SN#F4812779 had a zeta potential of -20.5 mv indicating there was no coagulant on the membrane surface. Image based on diagram from Particle Characterization Laboratories, Inc. Autopsy Report 3.0 Avista Technologies, Inc. Page 18 of 27
Testing for Organic Foulant Constituents: Fourier Transformed Infrared (FT-IR) analysis identifies organic foulant constituents. FT-IR is a measurement technique whereby spectra are collected based on measurements of the temporal coherence of a radiative source, using time-domain measurements of the electromagnetic radiation or other type of radiation. Spectra are compared against a library of more than 71,000 known constituents. FT-IR spectrum of a foulant sample from SN# F4812779 located strong bands associated with organic material (proteins and carbohydrates) and a weak band between 1100 and 900 cm -1 indicating the presence of silica. FT-IR spectral image of foulant material from SN# F4812779 Autopsy Report 3.0 Avista Technologies, Inc. Page 19 of 27
Testing to Identify Inorganic Foulant Constituents: Energy Dispersive X-ray (EDX) analysis is conducted in conjunction with scanning electron microscopy (SEM) to identify inorganic foulant constituents. The electron beam in the microscope causes specimens to emit x-rays including those from the k, l and m atomic shells. Spectrometer counts of these x-rays, which are said to be characteristic of the elements present in the specimen, can be used to calculate composition for a full qualitative analysis. Autopsy Report 3.0 Avista Technologies, Inc. Page 20 of 27
Inorganic Foulant Constituents Test Results: Elements (wt. %) SN#F4812779 Carbon 52.83 Silicon 1.05 Sulfur 4.59 Calcium 0.94 Oxygen 40.59 Testing Comments and Interpretation: EDX analysis identified only trace amounts of silica and calcium on the membrane surface. The high carbon and oxygen weight percents indicate a significant amount of organic material. The sulfur weight percent is contributed by the membrane itself and not part of the foulant layer. SEM images showed a smooth foulant layer on the membrane surface with colloidal particles embedded in the foulant layer. The dried foulant layer was peeling from the membrane surface indicating high organic content. The colloidal solids embedded in the foulant layer were between 2-20 microns in size. Chromatic elemental imaging (CEI) confirmed the high organic content observed in the EDX by showing a high distribution of carbon across the foulant layer. The colloidal solids embedded in the foulant layer were composed of silica and calcium. Autopsy Report 3.0 Avista Technologies, Inc. Page 21 of 27
Chromatic Elemental Imaging (CEI): Chromatic elemental imaging is an image showing the spatial distribution of elements in a sample. CEI can be useful for displaying element distributions in textural context, particularly for showing compositional zonation. Autopsy Report 3.0 Avista Technologies, Inc. Page 22 of 27
Cell Test and Laboratory Clean-In-Place Study Flat sheet membrane samples harvested from the full element are placed in a cell test apparatus and cleaned with various Avista chemicals to determine the most effective cleaner combinations and the amount of time required for an effective cleaning. The table below shows performance data before and after cleaning. Flat sheet samples and feed spacers were soaked in RoClean P111 (2% by weight in RO/DI water, heated to 35-40 C) over night (approximately 12 hours). The cleaner solution was then circulated for two hours after the extended soak time. Filmtec BW30-365 SN#F4812779 Water Passage Constant A Value Flow (GFD/PSI) Salt Passage Constant B Value Rejection (%) Pre Clean 1.46 E-04 140% of Normal 0.18 1.07 E-05 Normal 99.4 RoClean P111 Post Clean 1.59 E-04 152% of Normal 0.20 1.32 E-05 Normal 99.0 Manufacturer s original specifications 0.77 to 1.04 E-04 Normal Range 0.12 to 0.13 Normal Range Note: Testing conducted with dechlorinated city water from San Marcos, CA. 1.06 to 1.44 E-05 Normal Range 99.0 to 99.5 Normal Range Flow (GFD/PSI) Pre Clean Post Clean -LLooww - FFl llooww RR aannggee- - -HHi - iigghh FFl llooww RR aannggee- - 0 0.10 MFG. 0.15 0.20 0.25 >0.30 Spec. Rejection (%) Pre Clean Post Clean <94 95 96 97 98 Approx. Rejection of Cell Test Autopsy Report 3.0 Avista Technologies, Inc. Page 23 of 27
Pre clean membrane and feed spacer Post clean membrane and feed spacer Autopsy Report 3.0 Avista Technologies, Inc. Page 24 of 27
Stereoscope Image (10x) of pre clean feed spacer Stereoscope Image (10x) of post clean feed spacer Autopsy Report 3.0 Avista Technologies, Inc. Page 25 of 27
Close up stereoscope Image (60x) of pre clean feed spacer Close up stereoscope Image (60x) of post clean feed spacer Autopsy Report 3.0 Avista Technologies, Inc. Page 26 of 27
Certification by Laborataory Report Number WO #021712-2 Report Content Standard Spiral Autopsy Element Serial Number (s) Start Date of Testing Report Date F4812779 February 23, 2012, March 19, 2012 We the undersigned being the Technical Specialists in Membrane Autopsy and related testing procedures and protocol for Avista Technologies certify to the best of our knowledge and belief that the tests listed above have been conducted following Avista standard testing practices and that the results are accurate and complete. By signing this certificate neither the laboratory employees nor their employer makes any warranty, expressed or implied, concerning the cleaning study results. Date: 03/19/2012 Signed: Dan Gibson Laboratory Services Specialist Sara Pietsch Laboratory Services Chemist Autopsy Report 3.0 Avista Technologies, Inc. Page 27 of 27